Searching \ for '[PIC] PIC prototype board comments?' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: 'PIC prototype board comments?'.

Exact match. Not showing close matches.
PICList Thread
'[PIC] PIC prototype board comments?'
2005\05\18@175101 by olin_piclist

face picon face
I may shortly regret ever asking this, but what the heck, here goes anyway.
I'm looking for feedback on the design of a new product before it's too late
to change.

I'm seriously considering offering a family of PIC prototyping boards.  I've
never found the existing offerings to be that useful for just quickly wiring
up and trying a PIC circuit, or for serving as the base for a one-off hand
built device.  They seem to be either aimed as showing off what a PIC can
do, or try to solve too many problems with one board.  The prototype areas
I've seen are also too small and not very useful with just an array of
unconnected holes.

I've created a schematic of my idea what a PIC prototype board should be at
http://www.embedinc.com/temp/qprot1.pdf.  Here is a page by page overview:

Page 1: Power supplies.  The board will run from many cheap AC wall warts.
There will be a 1.3mm socket for the same wall wart I use for the EasyProg
and ProProg, plus screw posts for both AC and DC bare wire inputs.  The
final rectified voltage needs to be about 11-30 volts.  DC supplies from 12
to 30V will work fine.  A basic buck regulator makes 5.5V, which is linearly
regulated to 5.0V with an LDO.  I've done this in another design and it
worked very well.

Page 2: This is the PIC with its immediately surrounding circuitry.  There
are pads for two types of crystals, one thru hole and one surface mount.  Of
course at most one is intended to be used at a time.  Pushbuttons to ground
are connected to the port B interrupt on change pins with internal pullups.

Page 3: RS-232 interface.  The three jumpers make this optional.

Page 4: Diagnostic LEDs.  These are uncommitted so that you can wire them to
any signal you wish.  Each signal has two LEDs, one lit when the signal is
high, the other when low.  This is useful for seeing short pulses when a
single LED would othewise be in the on state.  It also avoids having to
configure LED polarity.

Page 5: Prototype area.  There will be uncommitted strips of 5 pads
surrounded by bus lines, much like the common "protoboards".  The schematic
is roughly laid out like the real board will be, except that there won't be
a space every 10 pins.  That's just for my sanity in working with the
schematic.  The pad strips and busses will be connected on the bottom layer,
with the top being a ground plane flowing around the holes.  The top silk
screen will also indicate how the holes are connected.

Overall I intend most of that pats in the schematic to be pre-installed, and
therefore surface mount.  Neither of the crystals would be installed, and
there would be a DIP socket for the PIC.  All the jumpers would be thru hole
wires that come installed.  They can be easily clipped to undo those
sections of the circuit not used in a particular design.  Yes I know that
means in some cases people will be paying for parts they don't end up using.
This makes sense when you realize that having only one variant to
manufacture allows unit to be cheaper than the price of a few parts bought
in volume.

Instead of having two or three boards trying to cover lots of ground, I'd
rather create more targeted boards that are good fits for what they are
trying to to.  I thought the generic PIC in a 28 pin DIP package would be
the first target.  40 pins, dsPICs, CAN, USB, Ethernet, etc, etc, will all
be different boards.  I'm hoping to have the first one available by Masters.

Comments?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\18@181519 by Charles Craft

picon face
I've had good luck using the internal RC on most recent projects but did use ceramic resonators at one time.

Thoughts on having PTH to mount one (ceramic resonator) instead of the crystal and caps?


{Original Message removed}

2005\05\18@182830 by Richard Prosser

picon face
How about adding a dual opamp to assist analog conditioning.?
Just something cheap & simple in the prototype area that can have
resistors & caps easily added to it.

RP

On 5/19/05, Charles Craft <spam_OUTchuckseaTakeThisOuTspammindspring.com> wrote:
> I've had good luck using the internal RC on most recent projects but did use ceramic resonators at one time.
>
> Thoughts on having PTH to mount one (ceramic resonator) instead of the crystal and caps?
>
>
> {Original Message removed}

2005\05\18@191317 by olin_piclist

face picon face
Charles Craft wrote:
> I've had good luck using the internal RC on most recent projects but
> did use ceramic resonators at one time.
>
> Thoughts on having PTH to mount one (ceramic resonator) instead of the
> crystal and caps?

Is there a reason for using a resonator instead of a crystal other than
cost?  If not, then the extra $.20 or so seems irrelevant for a one-off
unit, especially when your buying a ready-made prototype board.  If there is
a reason for wanting a resonator regardless of cost, what would be required
to support that?  The caps can be lifted easily enough, and the board comes
without crystals so those pads are available.  Would it be good enough to
put a ground pad between the two pads for the thru hole crystal?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\18@192327 by olin_piclist

face picon face
Richard Prosser wrote:
> How about adding a dual opamp to assist analog conditioning.?
> Just something cheap & simple in the prototype area that can have
> resistors & caps easily added to it.

This sounds like it's getting more and more application specific.  Many
projects might not have any analog at all.  Would this not be better served
by providing a generous prototype area that can be used exactly as needed by
any project?  In other words, would you find it objectionable if the answer
is "that's what the prototyping area is for"?

That brings up the question of how much prototype area is enough.  I looked
at the protoboards we have lying around that have been used many times to
quickly implement temporary PIC projects.  The vast majority of the time,
these projects fit on one board.  The ones we use have 47 double columns,
which is a little less than 5 inches wide.  I figured 50 double columns
would be good enough for most cases, especially considering that the PIC,
power supply, RS-232 interface, and debug LEDs don't eat into that budget.
I'm figuring that 10% to 15% of projects are too big for this board is about
right.  Otherwise it would be too expensive and therefore not applicable at
the low end.  The question is how big is about right?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\18@193110 by Charles Craft

picon face
I only asked because I've got a couple bags of resonators left over from when I used the evil white
prototyping boards with the spring contacts. %-)

If the pin spacing of a crystal matches the outer pins of a ceramic resonator then adding the PTH
for the center ground pin on the resonators sounds like a pretty cool solution.

{Original Message removed}

2005\05\18@194322 by olin_piclist

face picon face
Charles Craft wrote:
> If the pin spacing of a crystal matches the outer pins of a ceramic
> resonator then adding the PTH for the center ground pin on the
> resonators sounds like a pretty cool solution.

OK, I'll have to look into that.  I've never used a resonator myself.  I
guess I just haven't run into any high volume designs that needed pretty
good but not really accurate PIC frequency.

By the way, what's a "PTH"?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\18@195656 by Mauricio Jancic

flavicon
face
PTH = Plated Trough Hole

Mauricio Jancic
Janso Desarrollos - Microchip Consultants Program Member
.....infoKILLspamspam@spam@janso.com.ar
http://www.janso.com.ar
(54) 11 - 4542 - 3519

2005\05\18@200210 by Richard Prosser

picon face
Yes - but pretty much whatever you add is not going to be used by
everyone. If I was doing something with analogs, I might not be using
the switches or all the leds, or maybe not the RS232. I just thought
that a few cents for a simple Opamp might be of use ~30% of the time.
Maybe I'm biased toward analog interfacing but I don't generally find
"digital only" projects have the same appeal.
Your call.

Re the  prototype size, you'd probably have as much experiance with
this as anyone. I generally don't use it as I prefer to use an
external circuit to avoid damage or commit to a "valuable" resource as
I may want to try a different project without wanting to strip
everything off.
50 double collumns sounds quite generous particularly, as you say, the
leds & switches & RS232 are already taken care of.

I'd say basing on 10% to 15% projects being too big to fit is quite
reasonable especially as cost is going to be a major factor.

RP


{Quote hidden}

2005\05\18@200215 by Richard Prosser

picon face
Yes - but pretty much whatever you add is not going to be used by
everyone. If I was doing something with analogs, I might not be using
the switches or all the leds, or maybe not the RS232. I just thought
that a few cents for a simple Opamp might be of use ~30% of the time.
Maybe I'm biased toward analog interfacing but I don't generally find
"digital only" projects have the same appeal.
Your call.

Re the  prototype size, you'd probably have as much experiance with
this as anyone. I generally don't use it as I prefer to use an
external circuit to avoid damage or commit to a "valuable" resource as
I may want to try a different project without wanting to strip
everything off.
50 double collumns sounds quite generous particularly, as you say, the
leds & switches & RS232 are already taken care of.

I'd say basing on 10% to 15% projects being too big to fit is quite
reasonable especially as cost is going to be a major factor.

RP


{Quote hidden}

2005\05\18@201123 by Dave Lag

picon face

Love the RS232 section and the LED section.
Whats your target audience- Worried all those power supply parts will
make it expensive.
All that power supply and I can't run it at LF voltages?
D


Olin Lathrop wrote:
> I may shortly regret ever asking this, but what the heck, here goes anyway.
> I'm looking for feedback on the design of a new product before it's too late
> to change.

2005\05\18@210214 by Jan-Erik Soderholm
face picon face
Olin Lathrop wrote :

> Page 1: Power supplies.  The board will run from many
> cheap AC wall warts. There will be a 1.3mm socket for
> the same wall wart I use for the EasyProg and ProProg,
> plus screw posts for both AC and DC bare wire
> inputs.  The final rectified voltage needs to be about
> 11-30 volts.  DC supplies from 12 to 30V will work fine.
> A basic buck regulator makes 5.5V, which is linearly
> regulated to 5.0V with an LDO.  I've done this in another
> design and it worked very well.

Just thinking here...

*Must* a PSU for a rather small board have *that*
many components and beeing so complex. What happend
to the simple DC-in/7805-and-a-couple-of-caps type of PSU ?
Isn't that good enough for this kind of board ?

Otherwise, on the rest of the pages, I can't see any particular
problem right now.

How do you plan the connections to the outside ?
Pin headers like on Wouters Dwarf boards ?
Or maybe miniature "terminal blocks" ? Where you use just
a screw driver to to connect other equipment. Could be
part of the prototyping area, I guess...

Jan-Erik.



2005\05\18@210234 by Peter van Hoof

flavicon
face
how about replacing the rs232 section with a usb chip similar to the olimex usb
pic boards , it's getting hard to find laptos and pc's with serial ports

Peter van Hoof


2005\05\18@212425 by Byron A Jeff

face picon face
On Wed, May 18, 2005 at 09:02:32PM -0400, Peter van Hoof wrote:
> how about replacing the rs232 section with a usb chip similar to the
> olimex usb pic boards , it's getting hard to find laptos and pc's with
> serial ports

But it's easy to get a USB to serial cable.

We can't presume that the only thing that you'd want to connect a PIC to
is a PC with USB.

BAJ

2005\05\18@213550 by Spehro Pefhany

picon face
At 07:43 PM 5/18/2005 -0400, you wrote:
>Charles Craft wrote:
>>If the pin spacing of a crystal matches the outer pins of a ceramic
>>resonator then adding the PTH for the center ground pin on the
>>resonators sounds like a pretty cool solution.
>
>OK, I'll have to look into that.  I've never used a resonator myself.  I
>guess I just haven't run into any high volume designs that needed pretty
>good but not really accurate PIC frequency.

They're great if you need reliable serial communication, and other timing
is okay to within +/-0.5% or so. You can use the same 3-pin pattern for a
HC-49/U crystal or a resonator and you'll have all three options-- crystal,
two-pin resonator (I have not used those ones), three-pin resonator.

Some crystals have a ground pin in the middle anyway (I've ordered them
that way for high EMI industrial situations), so using the three-pin
pattern makes sense (the middle wire is just welded to the housing)

>By the way, what's a "PTH"?

Plated through hole, I'd assume.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
speffspamKILLspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com




2005\05\18@221543 by Phil Keller

flavicon
face
Olin,

 Another approach to prototype area is a daughter board approach.  It
does not answer you original question of how big but allows the real
design to progress with the possibility of various daughter boards, some
even with Analog functions predefined to be developed later.  For my use
it sure makes life easier when I can just peal off an interface and
begin a new project without damaging or loosing previous work.  Cuts
down on the number of mother boards sold but adds greatly to the overall
flexibility.

Phil

Olin Lathrop wrote:

{Quote hidden}

--
===========================================
PK Consulting
Phil Keller
Fremont, Ca

.....PhilKILLspamspam.....PKConsulting.net



2005\05\18@231503 by Matthew Miller

flavicon
face
Hi Olin,

On Wed, May 18, 2005 at 05:50:58PM -0400, Olin Lathrop wrote:
> Page 1: Power supplies.  The board will run from many cheap AC wall warts.
> There will be a 1.3mm socket for the same wall wart I use for the EasyProg
> and ProProg, plus screw posts for both AC and DC bare wire inputs.  The
> final rectified voltage needs to be about 11-30 volts.  DC supplies from 12
> to 30V will work fine.  A basic buck regulator makes 5.5V, which is linearly
> regulated to 5.0V with an LDO.  I've done this in another design and it
> worked very well.

As another poster mentioned, this power supply seems overly complicated. I
like flexability, but I think the prototype area could be used for
this. Most of my power supplies (I'm a hobbiest btw) are simple 78L05's or
LP2951 type chips. I haven't designed any commercial products, so my power
supply opinions might not be valid for anyone else. :)

> Page 4: Diagnostic LEDs.  These are uncommitted so that you can wire them to
> any signal you wish.  Each signal has two LEDs, one lit when the signal is
> high, the other when low.  This is useful for seeing short pulses when a
> single LED would othewise be in the on state.  It also avoids having to
> configure LED polarity.

Wow! This is a great concept! This would be very useful in debugging as
well.

> Page 5: Prototype area.  There will be uncommitted strips of 5 pads
> surrounded by bus lines, much like the common "protoboards".  The schematic
> is roughly laid out like the real board will be, except that there won't be
> a space every 10 pins.  That's just for my sanity in working with the
> schematic.  The pad strips and busses will be connected on the bottom layer,
> with the top being a ground plane flowing around the holes.  The top silk
> screen will also indicate how the holes are connected.

This sounds good. IMO, I think the prototype area should be about 3/4 to 1.5
the size of the board. That estimate is just a guess though.

> Overall I intend most of that pats in the schematic to be pre-installed, and
> therefore surface mount.

All of my PIC circuits, so far, have been made using soldered point-to-point
wiring. Using a board like you have described would be great; my circuits
would look a bit better using a board like this. ;)

I like this idea. Especially, if it comes with nothing but the power supply
components installed. Olin, I like your ideas.

Take care, Matthew.

--
"This is not a scientific issue, it's a political issue," he said.
"There isn't a scientific issue about the validity of evolution. The
only issue is whether schoolchildren will learn real science or not."
         -- Adrian Melott

2005\05\18@233737 by Chen Xiao Fan

face
flavicon
face
Cannot agree more with you on the power supply design. It only
serves the purpose of demonstrating the nice use of PIC 10F
as a SMPS controller but it can not beat a simple SMPS
controller. Since the maxim supply for this board is only
5V, a simple 9V supply with a 7805 will be the cheaper
and more robust design.

If I use a 10F as a power supply controller I will put
some over voltage protection in the output as well just
in case the 10F runs wild. The design already consider
the soft-start (with Q1, R1 nd D3, C5), why not put
a 6V2 zener after the 5V5 output? Maybe I will even put
a fuse (at least a resistor) in front for over current
protection.

Xiaofan

{Original Message removed}

2005\05\19@001400 by Denny Esterline

picon face
> Is there a reason for using a resonator instead of a crystal other than
> cost?  If not, then the extra $.20 or so seems irrelevant for a one-off
> unit, especially when your buying a ready-made prototype board.  If there
is
> a reason for wanting a resonator regardless of cost, what would be
required
> to support that?  The caps can be lifted easily enough, and the board
comes
> without crystals so those pads are available.  Would it be good enough to
> put a ground pad between the two pads for the thru hole crystal?
>
>

I thought the 'normal' justification for a resonator over a crystal is the
shock resistance and cost. But in this case since the added cost is only an
extra hole, I don't see a reason to no preserve the option.

I'm also wondering why you chose to make the power supply this complicated.
Seems most applications can get by with a 7805 and a couple caps - maybe a
couple diodes for protection. So I'm wondering what lead you to that design?
How about cost?

-Denny

2005\05\19@030440 by Wouter van Ooijen

face picon face
> Would it be
> good enough to
> put a ground pad between the two pads for the thru hole crystal?

yes

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\05\19@031039 by Wouter van Ooijen

face picon face
> I'm seriously considering offering a family of PIC
> prototyping boards.  I've
> never found the existing offerings to be that useful for just
> quickly wiring
> up and trying a PIC circuit, or for serving as the base for a
> one-off hand
> built device.

You might check the approach I took with the DwarfBoards:
http://www.voti.nl/dwarf

As Jan-Erik said: one or two 2-pin headers would make your board
compatible with my 'peripheral' boards.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\05\19@033415 by Wouter van Ooijen

face picon face
> As Jan-Erik said: one or two 2-pin headers would make your board
> compatible with my 'peripheral' boards.

sorry, 10-pin headers

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\05\19@035610 by Alan B. Pearce

face picon face
>> If the pin spacing of a crystal matches the outer pins of a ceramic
>> resonator then adding the PTH for the center ground pin on the
>> resonators sounds like a pretty cool solution.
>
>OK, I'll have to look into that.  I've never used a resonator myself.  I
>guess I just haven't run into any high volume designs that needed pretty
>good but not really accurate PIC frequency.
>
>By the way, what's a "PTH"?

Resonators seem to come in 0.1" pin spacing, but I have soldered a surface
mount one onto holes.

And my input on the board - what size do you envisage? My suggestion would
be to make it a Eurocard 100x160mm (4" x 6" nominal) and have provision for
a three row DIN 41619 connector on the prototype area end. This being a 0.1"
pin spacing would fit most other connector types as well (eg IDC pin row for
ribbon cable). And have the 9 pin D connector at the other end for the
serial.

Would you see any advantage in having an FT232 and USB connector on the PCB
for those wanting to do USB without going to a USB PIC? Even if only done as
the pads, without fitting the parts, and having the connection points right
beside the ones going to the MAX232?

2005\05\19@040519 by Lindy Mayfield

flavicon
face
Please forgive me if this is obvious, but would I be able to make (or is it feasible to implement) a connection from my ICD2 to program and debug things?

{Original Message removed}

2005\05\19@043610 by Alan B. Pearce

face picon face
>Please forgive me if this is obvious, but would I be able to make
>(or is it feasible to implement) a connection from my ICD2 to
>program and debug things?

On page 2 he shows an RJ12 connector in the top left as the programming
connector, and it looks as though the connections match the Microchip ones.
As it is Olin designing it, I would be surprised if they didn't match.

2005\05\19@072209 by olin_piclist

face picon face
Richard Prosser wrote:
> Re the  prototype size, you'd probably have as much experiance with
> this as anyone. I generally don't use it as I prefer to use an
> external circuit to avoid damage or commit to a "valuable" resource as
> I may want to try a different project without wanting to strip
> everything off.

But you're supposed to buy a new board for each new project ;-)

If this is really how you'd use it, then I think there isn't a problem here.
You can use the prototype area to add whatever "standard" analog circuitry
you wish the board had, and your per-project stuff would be external.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@082348 by olin_piclist

face picon face
Dave Lag wrote:
> Love the RS232 section and the LED section.

Thanks.

> Whats your target audience-

I think there are several different customer types.  One is guys like me.  I
know I would have used something like this a dozen times already if I had a
few of these in the closet.  It probably wouldn't substitute for the real
customer target circuit most of the time, but occasionally you need a quick
circuit to produce a specific sequence of pulses, measure something, quick
and dirty control something from a PC, be a test interface to the customer
circuit, etc.  The cost of these boards would be low enough that it's a no
brainer to use another board as apposed to wiring what you need on a
protoboard with power supply, RS-232 interface, etc.  In the end, I'd throw
the board in the box for that customer project so it'd be available a year
later when the customer want something changed.  With the protoboard
approach, by that time the protoboard would have gotten re-used, and you'd
have to figure out all over again what the test setup was.  Been there, done
that.

The other target is for hobbyists where the board becomes the project.  The
one off cost will still be cheaper than getting a PC board made, especially
if you really only need one.  I know this is a somewhat tougher sell since
many hobbyists either want to do everything themselves or are willing to
spare no expense to save a few pennies.  Still, there will be some that just
want to get on with their project.  This board allows you to get a simple
PIC circuit up and running with external hardware in an evening, which is
not possible if you've got to send out for board or even etch your own.

> Worried all those power supply parts will
> make it expensive.

Expensive as apposed to what?  You will still need a regulated power supply
somehow.  One way or the other, you're going to pay for a full wave bridge,
caps, and a regulating power supply, whether these are in a wall wart, a
bench supply, or on the board.  The cheapest solution would be a linear
regulator like a 7805.  Let's say pessimistically that the parts cost of the
switcher is $3 more than the 7805 solution parts cost.  On the other hand,
the switcher will run from a wall wart available for $1.70 in single
quantity and is more flexible in its input voltage requirements.  Is it
worth it?  I think for the first type of customer it is, but not sure about
the hobbyist.  What do you think?

> All that power supply and I can't run it at LF voltages?

Good point.  Do you want to run at low voltage to test the circuit or to
save power?  If the latter, this board probably isn't for you.  What about
if I add some jumpers around the 5.5V to 5.0V LDO?  That allows you to
disconnect the PIC power from the LDO output, which allows using your own
5.5V to 3.3V (or whatever) regulator on the prototype section, or you could
replace IC3 with a different voltage.  You can even change R5 and R6 to set
the switcher to a different output voltage.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@082813 by olin_piclist

face picon face
Peter van Hoof wrote:
> how about replacing the rs232 section with a usb chip similar to the
> olimex usb pic boards , it's getting hard to find laptos and pc's with
> serial ports

I'm thinking of a different prototype board that will be for the new 18F USB
PICs and be USB powered.  If you just want RS-232 connectivity on a PC that
doesn't have a COM port, then what about an off the shelf USB to RS-232
converter?  I'm thinking that most of the target audience would either have
a PC with COM port, or already have such a converter.  Do you think this is
a bad assumption?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@083509 by olin_piclist

face picon face
Jan-Erik Soderholm wrote:
> *Must* a PSU for a rather small board have *that*
> many components and beeing so complex.

No, it certainly doesn't.  You're the second person to mention this.  I'll
have to go back and figure out the costs to both alternatives carefully.
Note that the 7805 solution does push some cost to other parts of the
system.

> How do you plan the connections to the outside ?

The power inputs are 1.3mm power jack and screw terminals.  I was leaving
the rest as just plated thru holes .046" in diameter.  For fixed
installation, this allows soldering wires directly.  For more flexibility
screw terminals can be installed in the prototype area.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@083800 by olin_piclist

face picon face
Spehro Pefhany wrote:
> You can use the same 3-pin pattern for
> a HC-49/U crystal or a resonator and you'll have all three options--

Yup, this one's a no brainer.

*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@084550 by olin_piclist

face picon face
Matthew Miller wrote:
> This sounds good. IMO, I think the prototype area should be about 3/4
> to 1.5 the size of the board.

Sorry, but I don't follow this.  How can the prototype area be bigger than
the board it's on?

Given the number of prototype pads I've shown in the schematic, the
prototype area would be about 5 inches wide by 1.5 inches deep.  This will
dictate the board with, which will probably be 6 inches to allow for rubber
feet and mounting holes.  I'm planning on the prototype area being at the
front of the board with the fixed circuitry in as thin a strip as it will
fit in at the back of the board.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@084807 by William Bross

picon face
How about adding a half duplex RS485 chip, like a 75176 with transmit
control?  You could just double up on the same jumpers. Makes for one
less circuit in the kludge area.

The dual LED idea is great.  I know I've had to reverse my indicator
circuits on more than one occasion.

Bill

2005\05\19@085522 by olin_piclist

face picon face
Chen Xiao Fan wrote:
> Cannot agree more with you on the power supply design. It only
> serves the purpose of demonstrating the nice use of PIC 10F
> as a SMPS controller but it can not beat a simple SMPS
> controller.

It does in price.  I've been down this route a few times and the 10F
solution has come out cheaper.  I've got similar circuits in several volume
products, and they work very nicely.

> a simple 9V supply with a 7805 will be the cheaper
> and more robust design.

Yeah, I've been hearing this a lot.  I think the cost differential is less
than people imagine, but I agree it would still be cheaper.  My main
motivation was to have flexible input voltage requirements.  It would be no
problem to specify 9V DC in and use a 7805, but would that really cost less
since the external supply would cost more?

I'm going to carefully price out a few complete solutions (wall wart and
all) and see where things come out.  How much is flexible input power
requirement worth?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@091203 by olin_piclist

face picon face
> You might check the approach I took with the DwarfBoards:
> http://www.voti.nl/dwarf
>
> As Jan-Erik said: one or two 2-pin headers would make your board
> compatible with my 'peripheral' boards.

I've saved this email to look into this later.  I need to look at your bus
spec (electrical and mechanical) before deciding how to address this, and I
can't do that right now.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@091551 by Matthew Miller

flavicon
face
On Thu, May 19, 2005 at 08:45:55AM -0400, Olin Lathrop wrote:
> Matthew Miller wrote:
> >This sounds good. IMO, I think the prototype area should be about 3/4
> >to 1.5 the size of the board.
>
> Sorry, but I don't follow this.  How can the prototype area be bigger than
> the board it's on?

Sorry Olin, I feel a bit silly! I didn't write what I was thinking :) My
thought was that the prototype area should be 3/4 to 1.5 times the area
taken by the soldered componeents. I guess I was trying to give a
rule-of-thumb...

> Given the number of prototype pads I've shown in the schematic, the
> prototype area would be about 5 inches wide by 1.5 inches deep.  This will
> dictate the board with, which will probably be 6 inches to allow for rubber
> feet and mounting holes.  I'm planning on the prototype area being at the
> front of the board with the fixed circuitry in as thin a strip as it will
> fit in at the back of the board.

Anyway, I like your idea for this type of board!

Take care, Matthew.

--
"Few people can be happy unless they hate some other person, nation, or
creed."    -- Bertrand Russell

2005\05\19@092217 by olin_piclist

face picon face
Alan B. Pearce wrote:
> Would you see any advantage in having an FT232 and USB connector on the
> PCB for those wanting to do USB without going to a USB PIC? Even if
> only done as the pads, without fitting the parts, and having the
> connection points right beside the ones going to the MAX232?

But then the users wouldn't be "doing USB", only using USB as the transport
for RS-232.  The PC software would be talking to a COM port.  Given that,
why not just use an off the shelf USB to RS-232 adapter for those PCs that
don't have COM ports?

Putting the pads for FT232 isn't that unreasonable, but this is where other
realities set in.  I've never used a FT232 chip, and would require more time
than I have available for the proper testing to get this done before
Masters.  Yes I know it's supposed to be easy, and it probably is, but I'm
not going to put something in a product I haven't tested carefully.  Sorry,
sometimes other constraints dictate.

I do plan to have a USB offering in this family of prototype boards.  This
will be specifically for the new 18F USB PICs, but this will certainly not
be available by Masters.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@092432 by olin_piclist

face picon face
Lindy Mayfield wrote:
> Please forgive me if this is obvious, but would I be able to make (or
> is it feasible to implement) a connection from my ICD2 to program and
> debug things?

Yes.  That's what P7 in the top left corner of page 2 is.  The target
circuit end of the standard ICD2 cable will plug directly into P7.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@100359 by Lindy Mayfield

flavicon
face
Oh, I get it now.  (-:  I was thinking it was more like a replacement for things like PICDem2 / PICDem4 boards.

What would be the price range?  As a hobby/learning person this sounds great.  

{Quote hidden}

2005\05\19@101015 by Wouter van Ooijen

face picon face
> Putting the pads for FT232 isn't that unreasonable, but this
> is where other realities set in.

IMHO with the arrival of the 18F USB PICs and the all-available USB
converters the FT232 has lost most of its charm for this kind of board.
But an USB connector and what else is needed for the USB PICs would sure
make sense (and add the option to power the board from USB!).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu



2005\05\19@101020 by Wouter van Ooijen

face picon face
> I've saved this email to look into this later.  I need to
> look at your bus
> spec (electrical and mechanical) before deciding how to
> address this, and I
> can't do that right now.

bus spec is simple: gnd, 5V, (up to) 8 PIC I/O pins. Connector is a 10
pin 'shrouded header'. My supplier calls it an ML10. It is about the
cheapest connector I could find. gnd and 5V are swapped as to what you
would expect because some dark-side demo boards do this too.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\05\19@101022 by Wouter van Ooijen

face picon face
> I'm going to carefully price out a few complete solutions
> (wall wart and
> all) and see where things come out.  How much is flexible input power
> requirement worth?

I don't know how things are at other parts of the world, but I can buy a
low-quality wall-wart for E 2 (300mA unregulated DC), a high-quality one
for E 5 (1A switcher). For both an 7805 + some capacitors would do.
Perosnally I always add a series diode and protection dioded for the
7805.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\05\19@102327 by alan smith

picon face
9V battery connection.  Seems like alot of demo boards
are doing this now, makes it nice when you don't want
to mess with the wallwart or other stuff.  Course you
might need a box of batteries as well



               
__________________________________
Yahoo! Mail Mobile
Take Yahoo! Mail with you! Check email on your mobile phone.
http://mobile.yahoo.com/learn/mail

2005\05\19@113546 by olin_piclist

face picon face
Lindy Mayfield wrote:
> What would be the price range?  As a hobby/learning person this sounds
> great.

I don't know yet.  I think the total parts cost will be well under $10, but
the board will cost more than usual due to the large number of holes.
Assembly shouldn't be that bad since that has nothing to do with the holes.
And then of course I've got to make a buck in there somewhere too.  I'll
know better in a few weeks I hope.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@113951 by olin_piclist

face picon face
Wouter van Ooijen wrote:
> bus spec is simple: gnd, 5V, (up to) 8 PIC I/O pins. Connector is a 10
> pin 'shrouded header'. My supplier calls it an ML10. It is about the
> cheapest connector I could find. gnd and 5V are swapped as to what you
> would expect because some dark-side demo boards do this too.

It is a single row of pins, or two?  What is the pin pitch on the PCB side
of the connector?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@125333 by Dave Lag

picon face
Olin Lathrop wrote:
.....snip.......
> Good point.  Do you want to run at low voltage to test the circuit or to
> save power?  If the latter, this board probably isn't for you.  What about
> if I add some jumpers around the 5.5V to 5.0V LDO?  That allows you to
> disconnect the PIC power from the LDO output, which allows using your own
> 5.5V to 3.3V (or whatever) regulator on the prototype section, or you could
> replace IC3 with a different voltage.  You can even change R5 and R6 to set
> the switcher to a different output voltage.

You did say for one-off we should just use this board
(and buy another ;) - so the latter

If you stick to your guns- is there an adjustable version of the LDO so
one just needs to swap resistors/move a strap?

I like the concept of 10F+LDO (would like to try it myself) but as you
indicated if BOM is $10 and regulation is ~$3 or 30% of that and if your
markup scales accordingly...

Success*
D


(* copyright Vasile Surdican ;)


2005\05\19@134843 by Charles Craft

picon face
Trying to get caught up on this thread so maybe already mentioned:

Where's the outlet for the wall wart in my car?  :-p

Yep, flexible power input would be cool for automobile projects.

----------------------------------------------------------------------------------

I'm going to carefully price out a few complete solutions (wall wart and
all) and see where things come out.  How much is flexible input power
requirement worth?


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@144019 by olin_piclist

face picon face
Charles Craft wrote:
> Trying to get caught up on this thread so maybe already mentioned:
>
> Where's the outlet for the wall wart in my car?  :-p
>
> Yep, flexible power input would be cool for automobile projects.

Hmm.  It costs money and board space to protect against the nasty spikes.
I'll look into leaving pads for an alternate 12V automotive power input.
This would have a series diode with either a series inductor or resistor
followed by an active clamp that clips to a bit under 30V since the rest of
the circuit can already handle that.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@144232 by Wouter van Ooijen

face picon face
> It is a single row of pins, or two?  What is the pin pitch on
> the PCB side
> of the connector?

2x5, 2.54mm pitch. Think PC HD connector, but less pins. see
http://www.voti.nl/dwarf/notes/DN001.pdf, page 4, pin assignment on page
5.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu



2005\05\19@150815 by Charles Craft

picon face
You don't have to add extra filtering room for auto stuff - it could be external.

Just wanted to make the point that the wide power supply input would come in handy.
Could be something run off a solar panel and battery for example.


{Original Message removed}

2005\05\19@171539 by Howard Winter

face
flavicon
picon face
Olin,

On Wed, 18 May 2005 19:43:15 -0400, Olin Lathrop wrote:

> Charles Craft wrote:
> > If the pin spacing of a crystal matches the outer pins of a ceramic
> > resonator then adding the PTH for the center ground pin on the
> > resonators sounds like a pretty cool solution.
>
> OK, I'll have to look into that.  I've never used a resonator myself.  I
> guess I just haven't run into any high volume designs that needed pretty
> good but not really accurate PIC frequency.

I agree that this would be a good addition - I've got loads of resonators of various frequencies with the
central 0V pin for the internal capacitors.  One reason to use them is that they are more robust than crystals
(which can actually break if you shock them hard enough) and they seem to be more reliable at starting up if
conditions aren't optimum.  I always use them unless there's a real reason to use a crystal.

Cheers,


Howard Winter
St.Albans, England


2005\05\19@174151 by Mike Hord

picon face
> Page 1: Power supplies.  The board will run from many cheap AC wall warts.
> There will be a 1.3mm socket for the same wall wart I use for the EasyProg
> and ProProg, plus screw posts for both AC and DC bare wire inputs.  The
> final rectified voltage needs to be about 11-30 volts.  DC supplies from 12
> to 30V will work fine.  A basic buck regulator makes 5.5V, which is linearly
> regulated to 5.0V with an LDO.  I've done this in another design and it
> worked very well.

As others have mentioned, allowing for adjustable Vdd and a 9V battery
might be nice.  Also, and this may be a minor point or something for the
user to worry about, but it'd be cool to see some feature which makes
monitoring the current consumption easy- maybe a jumper that can be
removed to provide a shunt?  Although I can see that degenerating into
"Why the bloody #$%^ won't my circuit run?  It only runs when I'm
watching the current consumption!"

I like the thought of being able to "jumper out" the on-board regulation.
Things like that make it easier to use this to prototype low-power
circuits.

> Page 2: This is the PIC with its immediately surrounding circuitry.  There
> are pads for two types of crystals, one thru hole and one surface mount.  Of
> course at most one is intended to be used at a time.  Pushbuttons to ground
> are connected to the port B interrupt on change pins with internal pullups.

Perhaps being certain to include pads for a 32.768 kHz crystal near the PIC,
for easy and accurate Timer1 RTCs?  This could probably be covered by the
planned holes, if a ground hole is provided near by.

{Quote hidden}

A small surface mount prototyping area?  Laid out properly, it'd be fairly easy
to allow the user to put down, say, three 8-pin SOICs, or two 14-pin, or one
28-pin, and provide a couple of holes to tie those components to others in
the prototyping area.

I also want to ask:  what happens as the pads in the prototyping area start
to "wear out" (i.e., my ham-fisted soldering skills cause them to lift, etc.)?
Other than using a solderless breadboard (ick), is there a fix for that?
Perhaps some clever method to make that part of the board replacable?

> Instead of having two or three boards trying to cover lots of ground, I'd
> rather create more targeted boards that are good fits for what they are
> trying to to.  I thought the generic PIC in a 28 pin DIP package would be
> the first target.  40 pins, dsPICs, CAN, USB, Ethernet, etc, etc, will all
> be different boards.  I'm hoping to have the first one available by Masters.

If you get it done soon enough, I'll order a few with the EasyProg I'm
putting off ordering.  ;-)

Mike H.

2005\05\19@183718 by olin_piclist

face picon face
Wouter van Ooijen wrote:
> 2x5, 2.54mm pitch. Think PC HD connector, but less pins. see
> http://www.voti.nl/dwarf/notes/DN001.pdf, page 4, pin assignment on
> page 5.

That looks reasonable enough.  I had figured that is someone wanted to have
I/O signals via a connector, they could solder a single row header to the
prototype area.  However, the double row is more compact and providing a
space for it keeps I/O from eating into the proto area.  I've put your
connector on the board with 10 labeled pads coming out so that it can be
hooked up any way someone wants.  I'm thinking this would not come
installed, but the pads would be there.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\19@195406 by Howard Winter

face
flavicon
picon face
Alan,

On Thu, 19 May 2005 07:23:26 -0700 (PDT), alan smith wrote:

> 9V battery connection.  Seems like alot of demo boards
> are doing this now, makes it nice when you don't want
> to mess with the wallwart or other stuff.  Course you
> might need a box of batteries as well

Well around these parts a couple of 9V alkaline batteries cost as much as a decent wall-wart!  If messing with
a wallwart involves plugging it in, I'll do that, especially if it's the one that fits the Easyprog, which I
already have.

Cheers,



Howard Winter
St.Albans, England


2005\05\19@205229 by Chen Xiao Fan

face
flavicon
face
How much is the total budget power consumption? If it is not so high,
quite some linear regulator will be able to cope with wide input
voltage (say <=30V or even <=45V). It also looks a bit strange
for me to see MCP1700 here. Is it really safe to be used since its
maximum voltage is only 6V (or 6V5)?

I still like the idea of using a SMPS. As for simple SMPS controller,
I think they are comparable to the price of 10F204 and saves you
the cost of programming of 10F204. For example we have an old design
using MC33063AD SO8 and it costs about US$0.3. It may even save you
the external MOSFET Q5 and the associated driving circuits (R3, Q2, Q4,
Q3, R2, R4 and C7)if your budget current consumption from 5.5V
is less than 500mA (I think it should be the case since MCP1700
can only deliver 250mA, unless you have other plan for the 5V5).

Xiaofan

/**** Quoted from On-Semi Website ****/
The MC34063A Series is a monolithic control circuit containing the
primary functions required for DC-to-DC converters. These devices
consist of an internal temperature compensated reference, comparator,
controlled duty cycle oscillator with an active current limit circuit,
driver and high current output switch. This series was specifically
designed to be incorporated in Step-Down and Step-Up and
Voltage-Inverting applications with a minimum number of external
components. Refer to Application Notes AN920A/D and AN954/D
for additional design information.
* Operation from 3.0 V to 40 V Input
* Low Standby Current
* Current Limiting
* Output Switch Current to 1.5 A
* Output Voltage Adjustable
* Frequency Operation to 100 kHz
* Precision 2% Reference


{Original Message removed}

2005\05\19@210936 by James Newtons Massmind

face picon face
> ...but the board will cost more than usual due to the
> large number of holes.


Anyone thought about doing LARGE surface mount as a way to keep bread
boarding cost down? 1206 are REALLY not hard to solder. Neither are any of
the SO... IC packages. And if you have an existing stock of through hole
parts, you can always just bend the pins under or out and solder them down.

Probably not going to happen, but I'd be interested in SMT bread boarding
systems... Not much available in that area.

---
James Newton, massmind.org Knowledge Archiver
@spam@jamesKILLspamspammassmind.org 1-619-652-0593 fax:1-208-279-8767
All the engineering secrets worth knowing:
http://techref.massmind.org What do YOU know?



2005\05\19@215302 by Charles Craft

picon face
www.onepasinc.com/bonnie/smt/smt.html

A web search returned a reference to the onepas stuff on the sxlist.

{Original Message removed}

2005\05\19@222045 by Charles Craft

picon face
I'm guessing SH* is a shunt?

What are SH1, SH2 and SH3 for?


2005\05\19@222734 by Denny Esterline

picon face
> Wouter van Ooijen wrote:
> > 2x5, 2.54mm pitch. Think PC HD connector, but less pins. see
> > http://www.voti.nl/dwarf/notes/DN001.pdf, page 4, pin assignment on
> > page 5.
>
> That looks reasonable enough.  I had figured that is someone wanted to
have
> I/O signals via a connector, they could solder a single row header to the
> prototype area.  However, the double row is more compact and providing a
> space for it keeps I/O from eating into the proto area.  I've put your
> connector on the board with 10 labeled pads coming out so that it can be
> hooked up any way someone wants.  I'm thinking this would not come
> installed, but the pads would be there.
>
>

How about adding a couple extra rows of holes in the normaly blank area
where the ICs would set on the prototype area - i.e. no gap. That would
allow someone to put any size double row header in to suit thier needs.

-Denny

2005\05\19@222959 by Chen Xiao Fan

face
flavicon
face
0207/0204/0102/1206/0805 are all not too bad. 0603 are also
acceptable for those not so old. 0402 is a bit tough though
but still manageable after a while. So anything not smaller
than 0805 will be okay for most people.

Maybe the problem is with the sourcing and cost of low
SMD components in low quantity. The other problem is for
SMD ceramic capacitors (generally no marking). Not many
people will have a capacitance meter.

Xiaofan

{Original Message removed}

2005\05\20@034944 by Alan B. Pearce

face picon face
>I also want to ask:  what happens as the pads in the prototyping
>area start to "wear out" (i.e., my ham-fisted soldering skills
>cause them to lift, etc.)?  Other than using a solderless breadboard
>(ick), is there a fix for that?
>Perhaps some clever method to make that part of the board replacable?

dare I suggest that the fix for this is to put a pin header on and use
ribbon cable to another proto board? Endeavouring to fix lifted pads is a
beggar at the best of times.

2005\05\20@035927 by Alan B. Pearce

face picon face
>I still like the idea of using a SMPS. As for simple SMPS controller,
>I think they are comparable to the price of 10F204 and saves you
>the cost of programming of 10F204. For example we have an old design
>using MC33063AD SO8 and it costs about US$0.3. It may even save you
>the external MOSFET Q5 and the associated driving circuits (R3, Q2, Q4,
>Q3, R2, R4 and C7)if your budget current consumption from 5.5V
>is less than 500mA (I think it should be the case since MCP1700
>can only deliver 250mA, unless you have other plan for the 5V5).

Well I guess it would be advisable to drive all those LEDs of the 5V5
supply, rather than the smoothed 5v or 3v3 if it gets dropped to that.

However to take it back a step, I do wonder just what the purpose of having
the linear regulator is. If one is looking for real smooth supplies for
analogue purposes then I would suggest that one would regulate this
separately anyway to keep the supply noise from the PIC out of the picture.
I cannot see any real advantage in not having the switcher give the 5V
direct. A pad to fit a resistor to pull the output down to 3V3 (or any
voltage from 5V down to 2v5) would then be easy, for those who want the
lower voltage for some reason.

2005\05\20@040407 by Alan B. Pearce

face picon face
>How about adding a couple extra rows of holes in the normaly blank area
>where the ICs would set on the prototype area - i.e. no gap. That would
>allow someone to put any size double row header in to suit thier needs.

Well I did suggest that he put a triple row of holes down the edge to fit a
DIN 41419 style connector, and for those not using these you can fit pin
header strips up to 64 pins long, so with a double row you could have 128
pins.

My suggestion was for the PCB to be Eurocard size, but haven't seen any
comment on this.

2005\05\20@040557 by Alan B. Pearce

face picon face
>Maybe the problem is with the sourcing and cost of low
>SMD components in low quantity. The other problem is for
>SMD ceramic capacitors (generally no marking). Not many
>people will have a capacitance meter.

Yeah, I don't know why these don't get marked. But most DVMs have a
capacitance range now that will do almost anything above about 100pF with
good accuracy.

2005\05\20@075625 by olin_piclist

face picon face
Mike Hord wrote:
> Also, and this may be a minor point or something for the
> user to worry about, but it'd be cool to see some feature which makes
> monitoring the current consumption easy- maybe a jumper that can be
> removed to provide a shunt?

I was thinking of providing pads for automotive power input, which would
include a series resistor so that the voltage clamp has something to work
against.  This could also be used as a current sense.

> Perhaps being certain to include pads for a 32.768 kHz crystal near the
> PIC, for easy and accurate Timer1 RTCs?

Timer 1 oscillator crystal seems so obvious I don't know how it got missed
in the first place.  The prototype area would not be good for that since you
want to keep the leads really short.  I'll add pads for a watch crystal and
caps close to the processor.

> A small surface mount prototyping area?

Hmm.  There are so many different SMD footprints I think it would take up a
lot of board space for something not often used.  I've made a note though
and will consider it after seeing how much board space is left over, if any.
Maybe one 14 pin SOIC commonly used by opamps, many of which aren't
available in DIP.

> I also want to ask:  what happens as the pads in the prototyping area
> start to "wear out"

They're not supposed to wear out, you're supposed to buy a new one for each
project ;-)

Actually I'm not totally joking.  Temporary projects are best done with
those protoboards with the zillion pin sockets.  I was envisioning my boards
to save the tedium of wiring up one off but more premanent projects.  One of
my boards together with a protoboard would be a good fit for
experimentation.  My board takes care of all the mandatory basics like power
supply, oscillator and some nice options like RS-232 and debug LEDs.  You
solder wires to the processor pin pads, power, and ground, then use the
protoboard to connect them to the experimental circuit.

> If you get it done soon enough, I'll order a few with the EasyProg I'm
> putting off ordering.  ;-)

I think I'll just get it done in time for Masters, which is only two months
away.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@082611 by olin_piclist

face picon face
Chen Xiao Fan wrote:
> How much is the total budget power consumption? If it is not so high,
> quite some linear regulator will be able to cope with wide input
> voltage (say <=30V or even <=45V).

The bare circuit doesn't draw that much, but I wanted to provide reasonable
extra power for the custom part.  Yes, linear regulators can handle 30V, but
the power dissipation gets nasty.  Even with only 50mA at 5V, that's 1 1/4
watts dissipated by the regulator at 30V in.  That's more than just a 7805
sticking up is going to handle.  I know the linear regulator sounds so easy
and simple, but there are some messy issues once you look more closely.

> It also looks a bit strange
> for me to see MCP1700 here. Is it really safe to be used since its
> maximum voltage is only 6V (or 6V5)?

Yes, I've done this before in production designs.  Note that the switcher
produces a nominal 5.5V, so the MCP1700 is being run totally within specs.
The Microchip analog parts are very nice when you can live within the
constraints of 6V max.  When I did this in a production design, I looked at
various LDOs and the Mircrochip part had the best specs and either cost the
same or less - again considering the input would be 6V or less.

> I still like the idea of using a SMPS. As for simple SMPS controller,
> I think they are comparable to the price of 10F204 and saves you
> the cost of programming of 10F204. For example we have an old design
> using MC33063AD SO8 and it costs about US$0.3. It may even save you
> the external MOSFET Q5 and the associated driving circuits (R3, Q2, Q4,
> Q3, R2, R4 and C7)if your budget current consumption from 5.5V
> is less than 500mA (I think it should be the case since MCP1700
> can only deliver 250mA, unless you have other plan for the 5V5).

I looked around at this a while ago, and don't specifically remember the
switcher you mention.  In general, they were either too low input voltage,
too inefficient, or too expensive.  I've done the 10F trick a few times now,
and it works very nicely.  Note that the most expensive part in that circuit
is the inductor, which you have to supply yourself in any case.

And yes, I intend the 5.5V to be available beyond what the MCP1700 requires.
If more 5V is needed, additional LDOs can be added.  The 5.5V is also
available for things like relays that require significant current but can
tolerate some ripple.  At that input voltage, just about any 5V LDO can be
run at its maximum current without worrying about heat.  For example, the
MCP1700 will only dissipate 125mW at its maximum output of 250mA.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@082757 by olin_piclist

face picon face
James Newtons Massmind wrote:
> Probably not going to happen, but I'd be interested in SMT bread
> boarding systems... Not much available in that area.

I think one of the problems is that there are so many incompatible
footprints.  And then you have to provide thru hole pads for interconnects
anyway.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@083644 by olin_piclist

face picon face
Charles Craft wrote:
> I'm guessing SH* is a shunt?
>
> What are SH1, SH2 and SH3 for?

I call them "shorts".  They are ways to break up nets in Eagle even though
both sides are connected to each other.  Sometimes this is done to have a
differnet net class on each side, sometimes its to guarantee all the current
between the two subnets flow thru a single point that I can manually place.
This is the case for SH1 on page 1 for example.  It connects the power
section GND-PWR to the main ground for the rest of the board.  The switching
power supply can have large current spikes flowing accross its ground, which
I want to keep off the main ground.  The switcher will connect to the rest
of the circuit only at its ground and output points so all the large
internal currents don't cause ground offsets all over the board.  I may need
another short at its 5.5V output, or just make sure that the output funnels
thru one connection during layout.  I often add shorts once I start working
on routing, especially when using the auto router.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@084115 by olin_piclist

face picon face
Denny Esterline wrote:
> How about adding a couple extra rows of holes in the normaly blank area
> where the ICs would set on the prototype area - i.e. no gap. That would
> allow someone to put any size double row header in to suit thier needs.

Interesting idea I hadn't thought of, but I'm already using the gap to run
5V and ground bus lines.  I think in the end these will have more general
use than the ability to put double row headers in the prototype area.
Wouter's 10 pin double row header and the fact that you can put a lot of
single row headers in the prototype area won't cover all cases, but a good
many I think.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@090457 by olin_piclist

face picon face
Alan B. Pearce wrote:
> Well I guess it would be advisable to drive all those LEDs of the 5V5
> supply, rather than the smoothed 5v or 3v3 if it gets dropped to that.

Yeah, good point.

> However to take it back a step, I do wonder just what the purpose of
> having the linear regulator is. If one is looking for real smooth
> supplies for analogue purposes then I would suggest that one would
> regulate this separately anyway to keep the supply noise from the PIC
> out of the picture.

But what would you regulate it from?  If the switcher produces 5V directly,
then there is no convenient input to a LDO to make the analog supply.  I've
found that the 5V supply stays quite clean with decent bypass caps on the
PIC and an output cap at the LDO.

> I cannot see any real advantage in not having the
> switcher give the 5V direct.

I seriously considered that, but then you are left with no easy way to get a
clean supply.  I've found that the MCP1700 with a good output cap and a
properly bypassed PIC make a quite clean supply that will be good enough for
the vast majority of 10 bit analog applications.  It's all a tradeoff, but
in the end I thought the $.35 or so for the LDO was a relatively cheap way
to cover most of the main stream applications.  The unusual applications can
jumper out the LDO and adjust the switcher feedback resistors to make any
custom voltage they want.  They will "waste" a $.35 part, but I think that
will be a small minority of applications.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@091011 by olin_piclist

face picon face
Alan B. Pearce wrote:
> Well I did suggest that he put a triple row of holes down the edge to
> fit a DIN 41419 style connector, and for those not using these you can
> fit pin header strips up to 64 pins long, so with a double row you
> could have 128 pins.
>
> My suggestion was for the PCB to be Eurocard size, but haven't seen any
> comment on this.

I guess that's because I don't know what a DIN 41419 is, and don't know what
size a "Eurocard" is, nor why that would be a desirable size.

The board is basically built around the prototype area.  I'm currently
thinking the board will be about 6" wide, which accomodates the 50 prototype
columns with room for rubber feet or mounting holes.  Extra connector pads
could fit in the extra width not used up by the mounting.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@094033 by Lindy Mayfield

flavicon
face
I'm really actually enjoying this thread.  It is interesting for me to see the design thoughts that go into developing a hardware idea into a real product.

I am curious about something, but I'm not sure how to explain.  This is a prototype board, meaning I think a way to test an idea.  Say for example I come up with an interesting design.  If I use parts of the schematics of your board with my final product, (which I might, correct, since the entire thing is a part of the new prototype?) would I need to be concerned about that? IOW, would I be stealing part your design?  

Or is the hardware world different from the software one? (-:

-----Original Message-----
From: KILLspampiclist-bouncesKILLspamspammit.edu [RemoveMEpiclist-bouncesTakeThisOuTspammit.edu] On Behalf Of Olin Lathrop
Sent: Friday, May 20, 2005 13:57
To: Microcontroller discussion list - Public.
Subject: Re: [PIC] PIC prototype board comments?

Mike Hord wrote:

> I also want to ask:  what happens as the pads in the prototyping > area start to "wear out"

Olin answered:

They're not supposed to wear out, you're supposed to buy a new one for each
project ;-)



2005\05\20@100207 by Howard Winter

face
flavicon
picon face
Olin,

> Comments?

Looks like a good idea to me!  :-)

As others, I am a tad confused at the complexity of the power supply - I like the basic concept of having the
regulation on-board, and the "idiot-proofing" of using a bridge rectifier, but I'm finding it hard to think of
a reason for a 5V5 supply!  Perhaps I've missed something?

On Wed, 18 May 2005 17:50:58 -0400, Olin Lathrop wrote:

> Page 5: Prototype area.  There will be uncommitted strips of 5 pads
> surrounded by bus lines, much like the common "protoboards".  The schematic
> is roughly laid out like the real board will be, except that there won't be
> a space every 10 pins.  That's just for my sanity in working with the
> schematic.  The pad strips and busses will be connected on the bottom layer,
> with the top being a ground plane flowing around the holes.  The top silk
> screen will also indicate how the holes are connected.

Excellent idea!  I've never understood why prototype areas on (say) Microchip's PICDem boards have just an
individual anulus of copper around each hole - why not have strips?  I said as much to the Microchip reps at
the recent seminar here - they didn't know either but thought it may be to do with the way people in the USA
do prototype wiring!

The idea of having the prototype area at the front with the pre-installed parts behind sounds good, however
I'm not clear what the physical layout of your pad-strips would be.  

My suggestion would be a pair of horizontal bus strips (not connected to power - leave that to the user) with
vertical strips either side, making it easy to install DIL chips and connect them to power, easy to put
decoupling caps close to their ends, and other components around them.  However, I've seen boards where the
strips are not continuous full-length, giving "interwoven" strips which can be handy - basically you have
breaks in the strips every now and then, which do not line up across the strips, for example: ( O=O  are
connected holes,  O*O  is a break in the strip):

O=O=O=O=O=O=O*O=O=O=O=O*O=O=O=0=0=0
O=O=O=O=O*O=O=O=O=O=O=O=O=O*O=O=O=O
O=O=O=O=O=O=O*O=O=O=O=O*O=O=O=O=O=O
O=O=O=O=O*O=O=O=O=O=O=O=O=O*O=O=O=O
O=O=O=O=O=O=O*O=O=O=O=O*O=O=O=O=O=O

I presume you'll have pads at the back of the prototype area connected to the PIC I/O pins, to make patching
to them neat and easy?

I feel it would be useful to have some patterns that are not 0.1" grid - putting a surface mount device could
be allowed-for by having a small area with pads for the most common SMT spacings (0.050" ?), fanned-out to
small strips on 0.1" grid, rather like the central part of the "Surfboard" products such as this:
http://www.capitaladvanced.com/9081.htm

The other thing that I always have trouble with on 0.1" boards is things like D-type and RJ45 sockets, which
have the pins o 0.1" grid but with one row 50% staggered in relation to the other.  So a strip of staggered
holes along one edge (also fanned out to "straight" 0.1" strips) would be really handy!

And finally while I'm here:  how do you get Eagle to create strips of connected holes with no components in
them?

Cheers,


Howard Winter
St.Albans, England


2005\05\20@112715 by Mike Hord

picon face
> > A small surface mount prototyping area?
>
> Hmm.  There are so many different SMD footprints I think it would take up a
> lot of board space for something not often used.  I've made a note though
> and will consider it after seeing how much board space is left over, if any.
> Maybe one 14 pin SOIC commonly used by opamps, many of which aren't
> available in DIP.

That'd be wonderful.  There are a LOT of GOOD chips available only in SOIC
packaging; two columns of .025"x.2" pads at .05" spacing would allow a
user to place any of the most common SOIC package chips.

For example, Atmel's DataFlash ICs aren't available in through hole.  That
could be a good reason right there.

Mike H.

2005\05\20@112856 by olin_piclist

face picon face
Lindy Mayfield wrote:
> I am curious about something, but I'm not sure how to explain.  This is
> a prototype board, meaning I think a way to test an idea.  Say for
> example I come up with an interesting design.  If I use parts of the
> schematics of your board with my final product,
>
> Or is the hardware world different from the software one? (-:

Apparently part of your paragraph got cut off somewhere, probably because
you are sending paragraphs as single long lines.  Compliant SMTP and POP3
servers are allowed to truncate long lines, which is apparently what
happened.  Try sending lines limited to no more than 80 characters.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@114942 by Lindy Mayfield

flavicon
face
Sorry, I had no idea I was doing that.  I'm just using Outlook with Plain Text.

Anyway:


-----Original Message-----
From: spamBeGonepiclist-bouncesspamBeGonespammit.edu [TakeThisOuTpiclist-bouncesEraseMEspamspam_OUTmit.edu] On Behalf Of Olin Lathrop
Sent: Friday, May 20, 2005 17:29
To: Microcontroller discussion list - Public.
Subject: Re: [PIC] PIC prototype board comments?

Lindy Mayfield wrote:
> I am curious about something, but I'm not sure how to explain.  This is
> a prototype board, meaning I think a way to test an idea.  Say for
> example I come up with an interesting design.  If I use parts of the
> schematics of your board with my final product,
>
> Or is the hardware world different from the software one? (-:

Apparently part of your paragraph got cut off somewhere, probably because
you are sending paragraphs as single long lines.  Compliant SMTP and POP3
servers are allowed to truncate long lines, which is apparently what
happened.  Try sending lines limited to no more than 80 characters.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@115040 by Howard Winter

face
flavicon
picon face
Olin,

On Fri, 20 May 2005 09:11:17 -0400, Olin Lathrop wrote:

{Quote hidden}

Eurocard is 100 x 160mm, and the connector Alan is talking about is designed to fit on the end of it with
connectors extending parallel to the board, making it into a plug-in card for rack mounting.  The standard
allows for three rows of 32 pins (called A B and C) and any of them can be implemented - using A and C to give
a 64-way connector is common.  Lots of rack mounting systems are designed to take Eurocards with the DIN
connector to connect them to the backplane.  The first computer I built (Z80) used the technique, but with
rather larger boards.

As for why it's a useful size, it's a nice aspect ratio, it's commonly used in Europe, Eagle's maximum size
for the free and cheap versions is half a Eurocard, and Olimex's standard "single" PCB is one Eurocard.

I was going to post a link to a picture of the connectors, but Google can't seem to find them - I wonder if
that number is right (it's a long time since I used it!).

> The board is basically built around the prototype area.  I'm currently
> thinking the board will be about 6" wide, which accomodates the 50 prototype
> columns with room for rubber feet or mounting holes.  Extra connector pads
> could fit in the extra width not used up by the mounting.

Well 160mm is pretty close to 6.3", and 100mm is just under 4", which sounds like a good size to me!

Cheers,


Howard Winter
St.Albans, England


2005\05\20@115100 by Lindy Mayfield

flavicon
face
Sorry Olin, I had no idea I was doing that.  I'm using Outlook and the type is set to plain text.

Anyway, I asked:

I'm really actually enjoying this thread.  It is interesting for
me to see the design thoughts that go into developing a hardware
idea into a real product.

I am curious about something, but I'm not sure how to explain.  
This is a prototype board, meaning I think a way to test an idea.  
Say for example I come up with an interesting design.  If I use
parts of the schematics of your board with my final product,
(which I might, correct, since the entire thing is a part of the
new prototype?) would I need to be concerned about that? IOW,
would I be stealing part your design?  

Or is the hardware world different from the software one? (-:


{Original Message removed}

2005\05\20@120206 by Alan B. Pearce

face picon face
>I guess that's because I don't know what a DIN 41419 is, and don't
>know what size a "Eurocard" is, nor why that would be a desirable size.

OK, I did state it in my first email on this topic, and I think you at
skimmed that, because I recall comment on fitting an FT232 for USB.

However a Eurocard is 100mm (4") wide, by 160mm (nominally 6") long, and
many rack mount systems are designed for a 3U rack to have these on edge
uncased. The Eurocard size is what many PCB houses are basing their cheap
development sampling costings on (e.g. Olimex).

The DIN connector is a 3 row 0.1" pitch normally 64 pin long, but 32 pin
long versions are also available. This is across one end of the PCB. Checked
the connector, and I put the wrong DIN number sorry, the correct one is
41612. The "Brick" type power supplies for modular rack systems typically
use a version of this connector that has spade type connectors for high
current and mains use. You can also get versions that take coax or circular
high current pins a bit like some D connectors do.

http://www.jrmiller.demon.co.uk/products/mod96.html has a picture of a
typical Eurocard, with the DIN connector pads on the right hand end (the
area marked PL on the silk screen). Note the holes within the silk screen
marking for the connector are for connector mounting screws,
. As these pads are on a 0.1" x 0.1" grid if someone wants to fit a pin
header to connect to ribbon cable IDC connectors, these same holes suit that
purpose. You can also get double and quad height ones. IIRC VME cards are
quad height Eurocards.

a Eurocard would also fit the sort of use you envisage - one off projects
that would be installed somewhere. Having a one off type CPU card that could
go in a development system consisting of a Eurocard cardcage would be real
useful.

>The board is basically built around the prototype area.  I'm currently
>thinking the board will be about 6" wide, which accomodates the 50
prototype
>columns with room for rubber feet or mounting holes.  Extra connector pads
>could fit in the extra width not used up by the mounting.

That would be reasonably commensurate with what I envisage the Eurocard size
having.

2005\05\20@121129 by olin_piclist

face picon face
Lindy Mayfield wrote:

This time the whole message apparently got lost:

> Sorry, I had no idea I was doing that.  I'm just using Outlook with
> Plain Text.
>
> Anyway:
>
>
> {Original Message removed}

2005\05\20@121931 by Alan B. Pearce

face picon face
>I'm really actually enjoying this thread.  It is interesting
>for me to see the design thoughts that go into developing a
>hardware idea into a real product.
>
>I am curious about something, but I'm not sure how to explain.
>This is a prototype board, meaning I think a way to test an idea.
>Say for example I come up with an interesting design.  If I use
>parts of the schematics of your board with my final product,
>(which I might, correct, since the entire thing is a part of
>the new prototype?) would I need to be concerned about that?
>IOW, would I be stealing part your design?
>
>Or is the hardware world different from the software one? (-:

I think there are several issues here.

1. There are bits of the circuit which would be effectively hard to
copyright as such. The circuit around the PIC 16F device is straight out of
the datasheet stuff, so you could get that from anywhere.

2. If you were to go on and produce your own PCB you would not include the
LED drivers, and would probably do the power supply differently to suit a
more specific source of power. These have both been included to make it a
very general purpose device for development.

3. If you did decide to copy the power supply, then you would have to work
out your own program for the 10F device used in the switching supply, or
purchase the rights or devices to use. However because the PCB is designed
for a very wide range generic source of supply volts, I doubt that any
design you did would go this way.

4. If you needed "a dozen" to do a "quick and dirty" system or set of
systems, then it would be worth buying the needed number of boards from Olin
and building on the prototype area. If you were doing more than this, with
them all being the same, then doing your own PCB is probably viable, while
using one of Olins for prototyping and software development while your PCbs
get made.

2005\05\20@122030 by olin_piclist

face picon face
Lindy Mayfield wrote:
> I am curious about something, but I'm not sure how to explain.
> This is a prototype board, meaning I think a way to test an idea.
> Say for example I come up with an interesting design.  If I use
> parts of the schematics of your board with my final product,
> (which I might, correct, since the entire thing is a part of the
> new prototype?) would I need to be concerned about that? IOW,
> would I be stealing part your design?

Not at all.  Most of my "design" is the concept for the board and maybe some
layout details.  The circuit is mostly hooking a PIC to a power supply,
RS-232 interface, and LEDs.  In any case, I've already posted it publicly
without restrictions.  The final schematic for the board, notes on how to
modify it for various purposes, a parts list, board drawing, and parts index
will all be available on the web site without restriction.

I wouldn't be please if someone copied the whole board as is, but there is
probably little I could do about it.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@122824 by olin_piclist

face picon face
I tried to send this with a small attachment that was the Eagle script for
my PROTO library that Howard asked about, but I got this message from the
list server:

> Your message cannot be delivered to the following recipients:
>
>   Recipient address: RemoveMEpiclistspamTakeThisOuTmit.edu
>   Reason: SMTP transmission failure has occurred
>   Diagnostic code: smtp;554 5.7.1 Executable attachments not accepted
>   Remote system: dns;pacific-carrier

Which is strange because I thought I did this with another Eagle library a
few weeks ago without problems.  Is there any way to have the list server
minds its own business and stop trying to protect stupid people from their
stupidity?

Anyway, here is the text only of my reply to Howard:


Howard Winter wrote:
> The idea of having the prototype area at the front with the
> pre-installed parts behind sounds good, however I'm not clear what the
> physical layout of your pad-strips would be.

11-30V O-O-O-O-O

 5.5V O-O-O-O-O

 5.0V O-O-O-O-O

      O O O O O
      | | | | |
      O O O O O
      | | | | |
      O O O O O
      | | | | |
      O O O O O
      | | | | |
      O O O O O

 5.0V O-O-O-O-O

  GND O-O-O-O-O

      O O O O O
      | | | | |
      O O O O O
      | | | | |
      O O O O O
      | | | | |
      O O O O O
      | | | | |
      O O O O O

  GND O-O-O-O-O

Except picture this 50 columns wide instead of the 5 I've shown.  All holes
are .1" apart in both directions.  The vertical strips are each 5 holes
connected together not connected to anything else.  The horizontal strips
are busses.  This is similar to the common "protoboards" that have wire
sockets on .1" grids.

> My suggestion would be a pair of horizontal bus strips (not connected
> to power - leave that to the user)

Most of the time you will want them connected to power.  In fact I don't
think I've ever used a protoboard where I didn't use the two busses as power
and ground.  Still, I could connect the power busses via jumpers that could
be cut in the rare cases where someone wants to use them for something else.
This can't be done with the ground busses since each pad just connects to
the ground plane on the top layer.  There have been a lot of good
suggestions here, but I also don't want to throw in too many kitches sinks.

> I presume you'll have pads at the back of the prototype area connected
> to the PIC I/O pins, to make patching to them neat and easy?

Yes.  These will be double pads for each I/O port line.  These are shown on
page 2 of the schematic.

> I feel it would be useful to have some patterns that are not 0.1" grid
> - putting a surface mount device could be allowed-for by having a small
> area with pads for the most common SMT spacings (0.050" ?), fanned-out
> to small strips on 0.1" grid, rather like the central part of the
> "Surfboard" products such as this:

This is now the second or third time this has been suggested, so I'll
consider it.  I'm a bit worried about board space but I'll consider this
once I see where things are at.  Maybe the center busses could stop a little
short of one end.  I could then add a 14 pin SOIC pattern there and hook it
up to the vertical strips at that end.  It would require losing the inner
most pad on the vertical strips, so there would be 4 instead of 5 for those
7 columns.  I'll see how/if this might fit later.

> The other thing that I always have trouble with on 0.1" boards is
> things like D-type and RJ45 sockets, which have the pins o 0.1" grid

Actually they're not even .1" within a row either.

> but with one row 50% staggered in relation to the other.  So a strip of
> staggered holes along one edge (also fanned out to "straight" 0.1"
> strips) would be really handy!

Hmm.  This would certainly be useful in just the right circumstance, but it
sounds like too much of a kitchen sink to me.

> And finally while I'm here:  how do you get Eagle to create strips of
> connected holes with no components in them?

I've attached the script exported from my PROTO library.  It's only 9Kb so I
think it's OK.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@123215 by Howard Winter

face
flavicon
picon face
On Fri, 20 May 2005 16:50:36 +0100 (BST), I wrote:

> I was going to post a link to a picture of the connectors, but Google can't seem to find them - I wonder if
> that number is right (it's a long time since I used it!).

I was right (above!) it was the wrong number - it's DIN 41612.  Here's a link to Mouser's catalogue page which
shows the range:

http://www.mouser.com/catalog/622/892.pdf

Cheers,


Howard Winter
St.Albans, England


2005\05\20@141133 by olin_piclist

face picon face
Alan B. Pearce wrote:
> However a Eurocard is 100mm (4") wide, by 160mm (nominally 6") long,

That's about the size I was envisioning anyway.

> The DIN connector is a 3 row 0.1" pitch normally 64 pin long, but 32 pin
> long versions are also available. This is across one end of the PCB.
> Checked the connector, and I put the wrong DIN number sorry, the
> correct one is 41612.

I typed this into DigiKey and found only one DIN connector.  It was straight
out from the PCB and had two rows of 32 pins.  That doesn't sound like what
you're talking about since you want a right angle connector if I understand
it correctly.  Can you give me a pointer to the datasheet for the exact
connector you would like this board to take?

I guess this would take up 6 columns of pads on the right side of the board.
Three for the connector itself, and another three to wire to from on the
board.

> http://www.jrmiller.demon.co.uk/products/mod96.html has a picture of a
> typical Eurocard, with the DIN connector pads on the right hand end (the
> area marked PL on the silk screen).

I noticed that doesn't show the connector itself.  I assume its right angle,
but can't tell from the picture whether it would be male or female.

> As these pads are on a 0.1" x 0.1" grid if someone wants to fit a pin
> header to connect to ribbon cable IDC connectors, these same holes suit
> that purpose.

That's a good point which helps justify the rather large board area for this
feature.  Wouter could fit 2 or 3 of his connectors onto the same pad area,
plus many other connectors with .1" pin pitch become possible.

How important is it to have all 3 rows or would 2 rows of 32 be sufficient
for this purpose?  I can't see this board requiring that much I/O.  After
all, there are only 100 uncommitted nets on the prototype area and the PIC
only has 28 pins total.

> a Eurocard would also fit the sort of use you envisage - one off
> projects that would be installed somewhere. Having a one off type CPU
> card that could go in a development system consisting of a Eurocard
> cardcage would be real useful.

I see what you're saying.  I guess these things are more popular in Europe.
They might be more popular here if they weren't called "Eurocard".  Whoever
came up with that failed marketing 101.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@152719 by Denny Esterline

picon face
> > > A small surface mount prototyping area?
> >
> > Hmm.  There are so many different SMD footprints I think it would take
up a
> > lot of board space for something not often used.  I've made a note
though
> > and will consider it after seeing how much board space is left over, if
any.
{Quote hidden}

And Ramtron's FRAM parts...

Sounds like a general purpose SOIC breakout might be a good idea. If there's
enough space I'd think pads (unpopulated) for the more common PITA
connectors would be a good idea- maybe an extra (uncommitted) DB9 and RJ45?
Or maybe run the existing serial and ICSP connections through a set of
jumpers that could be cut to allow other uses to the connectors. (I'm
thinking about things like CAN and LIN)


You have to be a little careful trying to please everyone, or pretty soon
you'll end up with a Stuka dive bomber.
-Denny

2005\05\20@170328 by olin_piclist

face picon face
Denny Esterline wrote:
> Sounds like a general purpose SOIC breakout might be a good idea. If
> there's enough space I'd think pads (unpopulated) for the more common
> PITA connectors would be a good idea- maybe an extra (uncommitted)
> DB9 and RJ45? Or maybe run the existing serial and ICSP connections
> through a set of jumpers that could be cut to allow other uses to the
> connectors. (I'm thinking about things like CAN and LIN)

I think this is going too far.  CAN will be a different board.  Adding much
more to what is already on this board will make it so big that nobody wants
it because they will only use 10% of it.

I've gotten some useful ideas from this thread, and some changes will
definitely be made from my original design.  However, there is no way to
please all the people all the time, and it would be a mistake to try.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@174338 by Bob Ammerman

picon face
>> Your message cannot be delivered to the following recipients:
>>
>>   Recipient address: piclistEraseMEspam.....mit.edu
>>   Reason: SMTP transmission failure has occurred
>>   Diagnostic code: smtp;554 5.7.1 Executable attachments not accepted
>>   Remote system: dns;pacific-carrier

Zip up your file.

Bob Ammerman
RAm Systems

2005\05\20@175108 by David Minkler

flavicon
face

Olin Lathrop wrote:

>I've gotten some useful ideas from this thread, and some changes will
>definitely be made from my original design.  However, there is no way to
>please all the people all the time, and it would be a mistake to try.
>

Agreed.   It will be far more important that you actually produce
something useful than that you design the perfect product.  I'll be
waiting for your dsPic version.

Incidentally, a little sheet with suggestions on a suitable tool chain
would add a lot to the value of such boards.  Unlike the hobbyist who
"will spare no expense to save a few pennies", I don't mind spending a
little to have a solid (reliable) toolchain adequate to the task at
hand.  On the other hand I don't care to waste a lot on tools that are
inappropriate or unreliable (or far beyond what I really need for what
I'm trying to do).  Clues are appreciated.

Regards,

Dave


2005\05\20@200149 by olin_piclist

face picon face
I've been looking around the web to try and understand the Eurocard and DIN
41612 connector thing.  Unfortunately there seem to be many incompatible
flavors of this type of connector.  It's not even totally clear which side
goes onto the cards, male or female.  I'm gathering that the normal thing to
put on a card are right angle male connectors, with the backplane having
sockets.  There are also a lot of different sizes.  The one that seems to
pop up more than the others is 32 x 2 pins.  32 x 3 also comes up a lot, but
that seems overkill for a board of this complexity.

So, I'm thinking I'll leave pads on the right board edge for such a "type B"
DIN 41612 connector, which means two rows of 32 pins each on a .1" grid.
Can someone confirm that the Eurocard folks would indeed consider this
"normal", or at least one of the "normal" options?

If so, I've found detailed drawings of such connectors.  I think I have
enough to make the Eagle package definition.  However, how far out (or back)
should the end of the pins sticking into the backplane connector be from the
end of the board?  There must be a standard for that, but I haven't been
able to find it.  The same goes for the vertical placement.  The connector
appears centered in all the pictures I could find, but I haven't seen
anything actually specifying this.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\20@211342 by Dave Tweed

face
flavicon
face
olin_piclist@embedinc.com (Olin Lathrop) wrote:
> So, I'm thinking I'll leave pads on the right board edge for such a
> "type B" DIN 41612 connector, which means two rows of 32 pins each
> on a .1" grid.

Double-check this. IIRC, the VME bus that we used to use at Apollo was
based on the Eurocard standard, and I believe that a "type B" connector
depopulated the center row of a 3-row connector, leaving two rows of 32
pins each (on 0.100" centers) with 0.200" between the two rows.

The only other comment I have about your project is that I see no need
for a separate connector and diode for a "DC" power input. There's no
problem with feeding DC through a bridge rectifier. If the power source
is negative-ground, then one of the bridge diodes is harmlessly shorted
out. If it's positive-ground, then you've got problems that even your
separate input doesn't solve. :-)

Any extra protections you were thinking of applying to the DC input for
automotive sources could just as well be included in the AC input path.

-- Dave Tweed

2005\05\21@091500 by olin_piclist

face picon face
Dave Tweed wrote:
> Double-check this. IIRC, the VME bus that we used to use at Apollo was
> based on the Eurocard standard, and I believe that a "type B"
> connector depopulated the center row of a 3-row connector, leaving
> two rows of 32 pins each (on 0.100" centers) with 0.200" between the
> two rows.

What I've been running into most is .1" between the rows of pins.  A
representative drawing is at http://www.embedinc.com/temp/din41612b_1.pdf.

> The only other comment I have about your project is that I see no need
> for a separate connector and diode for a "DC" power input. There's no
> problem with feeding DC through a bridge rectifier. If the power
> source is negative-ground, then one of the bridge diodes is
> harmlessly shorted out. If it's positive-ground, then you've got
> problems that even your separate input doesn't solve. :-)

Good point.  I wanted the possibility of a hard ground connection without a
diode drop for use with DC supplies, but that can be solved just as well by
providing a jumper that shorts one of the AC inputs to ground.  Without the
jumper you have AC inputs, with the jumper you have a + and GND inputs.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\23@041312 by Alan B. Pearce

face picon face
>I typed this into DigiKey and found only one DIN connector.  It
>was straight out from the PCB and had two rows of 32 pins.  That
>doesn't sound like what you're talking about since you want a right
>angle connector if I understand it correctly.  Can you give me a
>pointer to the datasheet for the exact
>connector you would like this board to take?

Howard has already done this, providing a link to the Mouser page. It sounds
like you found the half length version at Digikey.


>I guess this would take up 6 columns of pads on the right side of
>the board. Three for the connector itself, and another three to
>wire to from on the board.

Could do it that way, but I always do it on the reverse of the PCB, using
the pin end to wrap the wire around.


>> http://www.jrmiller.demon.co.uk/products/mod96.html has a picture of a
>> typical Eurocard, with the DIN connector pads on the right hand end (the
>> area marked PL on the silk screen).
>
>I noticed that doesn't show the connector itself.  I assume its right
angle,
>but can't tell from the picture whether it would be male or female.

Yes, the card mount one is always a right angle with pins, and the backplane
always a straight with sockets. The connector body is keyed by its shape so
they will not mate in the wrong orientation.


>> As these pads are on a 0.1" x 0.1" grid if someone wants to fit a pin
>> header to connect to ribbon cable IDC connectors, these same holes suit
>> that purpose.
>
>That's a good point which helps justify the rather large board area for
this
>feature.  Wouter could fit 2 or 3 of his connectors onto the same pad area,
>plus many other connectors with .1" pin pitch become possible.
>
>How important is it to have all 3 rows or would 2 rows of 32 be sufficient
>for this purpose?  I can't see this board requiring that much I/O.  After
>all, there are only 100 uncommitted nets on the prototype area and the PIC
>only has 28 pins total.

You really need to do all three rows if you wish to provide holes on 0.1" x
0.1" pitch, as there are two variations of the connector which have two rows
of pins. One version is the 3 row connector body with the centre row
missing, and the other is a narrower one with only two rows. I would be
tempted to provide three rows of holes, but only bring out the two nearest
the edge of the PCB to pads as that is the form of the second connector
above, but if people insist on using the other one they can fit it, or the
three row version.


>> a Eurocard would also fit the sort of use you envisage - one off
>> projects that would be installed somewhere. Having a one off type CPU
>> card that could go in a development system consisting of a Eurocard
>> cardcage would be real useful.
>
>I see what you're saying.  I guess these things are more popular in Europe.
>They might be more popular here if they weren't called "Eurocard".  Whoever
>came up with that failed marketing 101.

<VBG> I think you may fine that compatible ones are available in certain
marketing areas of the USA, in specific markets of rack mount gear, unless
the VME style card has the market share. However if you wish to market in
Europe then being able to say it is Eurocard form factor and connector
compatible will help you. Guess this depends on how big you see your market.

2005\05\23@081046 by olin_piclist

face picon face
Alan B. Pearce wrote:
> You really need to do all three rows if you wish to provide holes on
> 0.1" x
> 0.1" pitch, as there are two variations of the connector which have two
> rows of pins. One version is the 3 row connector body with the centre
> row missing, and the other is a narrower one with only two rows. I
> would be tempted to provide three rows of holes, but only bring out the
> two nearest the edge of the PCB to pads as that is the form of the
> second connector above, but if people insist on using the other one
> they can fit it, or the three row version.

I asked this on Friday and didn't get an answer, so I went ahead with what
seemed reasonable to me.  I needed to get this mostly done over the weekend.

I ended up leaving pads for the two row "type B" connector.  These were the
most common ones I found in Google searches, and are really bigger than
seems necessary for a board of this type.  I brought the two rows of 32 pins
each out to two rows of pads, and made sure the pads were back far enough so
that simple shrouded headers like Wouter's connectors could be used without
covering the access pads.  This just left room for 50 columns of pads in the
prototype area.  Due to the extra board height, there was room for two 14
pin SOIC footprints such as is common with opamps, many of which are not
available in DIP.  The access pads for the two SOICs are in addition to the
prototype area.

> <VBG> I think you may fine that compatible ones are available in certain
> marketing areas of the USA, in specific markets of rack mount gear,
> unless the VME style card has the market share.

I frankly don't see a lot of card-cage rack mounted equipment anymore.

> However if you wish to
> market in Europe then being able to say it is Eurocard form factor and
> connector compatible will help you. Guess this depends on how big you
> see your market.

Well the board is 160 x 100 mm and it has the footprint for the type B
Eurocard connector on the right side.  I noticed that I didn't exactly get
flooded with answers when asking specifics about Eurocards, so I'm not sure
how important this feature is.  Hopefully I've done enough to make it at
least somewhat usable in that role.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\05\23@085811 by Alan B. Pearce

face picon face
>I asked this on Friday and didn't get an answer, so I went
>ahead with what seemed reasonable to me.

I had gone home by then (one of the things about being on an international
list- you guys are just lighting up the PICList at about the time I go
home). Looking at the resultant comments this morning I figured that you had
come to some sensible looking decisions.

>I needed to get this mostly done over the weekend.

<VBG> I have projects like that too.

2005\05\26@065423 by vasile surducan

picon face
On 5/19/05, Olin Lathrop <EraseMEolin_piclistspamembedinc.com> wrote:
> Matthew Miller wrote:
> > This sounds good. IMO, I think the prototype area should be about 3/4
> > to 1.5 the size of the board.
>
> Sorry, but I don't follow this.  How can the prototype area be bigger than
> the board it's on?

 
 great topic!

 Let say you are prototyping only with SMD parts, did you heard about
like that?
 So, the board have at least two "faces" and an infinite numbers of
layers. As biggest the technological development in the developers
country, as many the numbers of layers on the PCB, no matter if are
necessary indeed or not. :)
  One half of one face is the circuit itself and 1.5 is the prototyping area.

For those which need a bigger prototype area, prototyping also on
layers 2 to 15 will be available at request.

best regards,
Vasile





{Quote hidden}

> -

2005\05\26@094430 by Paul James E.

picon face

What!


{Quote hidden}

>> --

2005\05\26@094440 by Paul James E.

picon face

What? !!!!!!!


{Quote hidden}

>> --


'[PIC] PIC prototype board comments?'
2005\06\12@220336 by Roy Ward
flavicon
face
Olin Lathrop wrote:

> Dave Lag wrote:
>
>> All that power supply and I can't run it at LF voltages?
>
>
> Good point.  Do you want to run at low voltage to test the circuit or to
> save power?  If the latter, this board probably isn't for you.  What
> about
> if I add some jumpers around the 5.5V to 5.0V LDO?  That allows you to
> disconnect the PIC power from the LDO output, which allows using your own
> 5.5V to 3.3V (or whatever) regulator on the prototype section, or you
> could
> replace IC3 with a different voltage.  You can even change R5 and R6
> to set
> the switcher to a different output voltage.

The other reason for lower voltages is to connect to components that
won't handle 5 volts. My last project involved connecting a PIC to a
CPLD (I used a cable to connect a PIC board to a CPLD board, but still
needed to match the voltages).

However, 3.3V only components seem to be always always SMT anyway, so
such uses are probably outside your target market anyway.

Cheers,
Roy Ward.

2005\06\12@220645 by Roy Ward

flavicon
face
Apologies for that last post - I'd accidently switched to view by thread
rather than by date, so didn't realize this was an old thread.

Cheers,
Roy Ward.

More... (looser matching)
- Last day of these posts
- In 2005 , 2006 only
- Today
- New search...