Searching \ for '[PIC] Why the preoccupation with bus powered progr' 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/power.htm?key=power
Search entire site for: 'Why the preoccupation with bus powered progr'.

Exact match. Not showing close matches.
PICList Thread
'[PIC] Why the preoccupation with bus powered progr'
2006\12\02@135746 by tachyon 1

picon face
Of all the microcontroller programmers out there, I see a preoccupation
with making a big deal out of the fact that programmer X needs no
external power supply (wall wart).
Instead a great deal is made about complicated DC-DC converters and
charge pumps, and USB power, and serial port power, and even application
circuit power.

This all makes no sense to me whatsoever. Why bother with all the extra
complexity, expense, and possible power problems when you can simply hook
up a $2 wall wart?
I personally don't give a rip about the supposed inconvenience of a wall
wart. It's not like I'm programming while riding a motorcycle. I'm quite
content to use my desk at home or work.

Can someone please explain to me what, if any, advantage there is to not
using and external power supply? I can't think of any.

--

Search for products and services at:
http://search.mail.com

2006\12\02@144251 by olin piclist

face picon face
tachyon_1@email.com wrote:
> Of all the microcontroller programmers out there, I see a
> preoccupation with making a big deal out of the fact that programmer
> X needs no external power supply (wall wart).

I don't.  Many programmers require a external power supply, some don't.  Of
my programmers, 2 require external power and 1 doesn't.

> Instead a great deal is made about complicated DC-DC converters and
> charge pumps, and USB power, and serial port power, and even
> application circuit power.

Perhaps here on a technical list where people might be interested in the
internal workings of such things.  I don't see a "great deal" being made of
it in other contexts, like when explaining to a end user what the features
are.

> This all makes no sense to me whatsoever. Why bother with all the
> extra complexity, expense, and possible power problems when you can
> simply hook up a $2 wall wart?

You truly can't imagine why someone might prefer not to need a separate wall
wart, whether you prefer one or not!!?  Sounds like a attempt to obscure a
religous statement as a question.  I think the reasons are self-evident, but
I'll list them anyway in case you are really asking:

1 - It's simpler.  You plug in the USB and you're done.

2 - It's cheaper.  A PIC programmer requires several voltages, some of them
adjustable.  It will require some power supplies anyway.  The extra cost of
starting with 5V as apposed to some other voltage from a wall wart is less
than the cost of the wall wart, the connector, and the extra cost of
mounting a thru-hole connector (you want to use thru hole for parts that
take mechanical stress) on the board.

3 - It's less hassle.  You don't have to carry the extra wall wart around
with you, remember which wall wart is for which device, have to make sure
there are two adjacent power outlets available (one for the wall wart and
one more covered up by the wall wart), and remember to plug everything in.

4 - It's universal.  USB is always electrically and mechanically the same.
Wall power varies a lot accross the globe.  There are such things as
"universal" wall warts that take any wall power, but not for $2.  In cany
case, mechanically there is no such thing as a universal wall wart.

5 - It's shipped easier.  A $2 wall wart is just a transformer.  Even a
small one will be heavier than most programmer circuit boards.  The wall
wart I use for my EasyProg and ProProg programmers is one of the smallest
and cheapest around.  I just checked, and it's still over 3 times heavier
than the USBProg circuit board.

> I personally don't give a rip about the supposed inconvenience of a
> wall wart.

That's your call, but I'm surprised you can't see how others would find it
inconvenient.  By the way, the USBProg can be separately powered from a 5V
supply and used from a serial line via a RS-232 to TTL converter.  I don't
populate the connectors for this, but you can either add them yourself or
solder wires to it.  I guess the USBProg is everything you asked for ;-)


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\02@144920 by Herbert Graf

flavicon
face
On Sat, 2006-12-02 at 13:57 -0500, spam_OUTtachyon_1TakeThisOuTspamemail.com wrote:
> Of all the microcontroller programmers out there, I see a preoccupation
> with making a big deal out of the fact that programmer X needs no
> external power supply (wall wart).
> Instead a great deal is made about complicated DC-DC converters and
> charge pumps, and USB power, and serial port power, and even application
> circuit power.
>
> This all makes no sense to me whatsoever. Why bother with all the extra
> complexity, expense, and possible power problems when you can simply hook
> up a $2 wall wart?
> I personally don't give a rip about the supposed inconvenience of a wall
> wart. It's not like I'm programming while riding a motorcycle. I'm quite
> content to use my desk at home or work.
>
> Can someone please explain to me what, if any, advantage there is to not
> using and external power supply? I can't think of any.

Well, you've obviously never misplaced a wall wart.

This is a COMMON issue, since there is no "standard" walwart. It is very
likely that most devices that use a wall wart can only use THAT walwart,
all the other walwarts in the home likely have a different voltage,
capacity, polarity or plug type.

Sure, the "answer" is "don't misplace the walwart", but that doesn't
change the fact of how FRUSTRATING it is to want to use a device but
can't because you can't find the bloody walwart.

After my move about a year ago there is STILL one device I have which I
can't find the walwart for. Fortunately it's a router, which I've got
several, but it's still annoying.

A second reason, one much less important to most, but important to me,
is working in the field. I have some PICs in the field that are FAR away
from the nearest outlet. When I need to update their firmware I can
leave them in place, plug my ICD2 into my laptop (USB power) and
reprogram them on the fly. I don't do this much, but when I can it's
MUCH quicker then removing the device, bringing it to a power outlet and
reprogramming that way. I've even debugged with my ICD2 in the field
this way, a few iterations, problem fixed, and I'm done. If I needed a
walwart I'd have to lug the device back and forth multiple times, and
wouldn't be able to in circuit debug with the real world stimuli.

TTYL

2006\12\02@190417 by Piclist

flavicon
face
As I type this, I have 3 power strips under my desk, with a bunch of
wallwarts connected to it.  If I get anything else that uses a walwart
that I'll have to get a 4th strip.  There is space on the 3 strips right
now for about 3 or 4 powercords, but a walwart will not fit in any of
those free spots.  Might have something to do with that?


-Mario

{Original Message removed}

2006\12\02@203002 by Tony Smith

picon face
> Of all the microcontroller programmers out there, I see a
> preoccupation with making a big deal out of the fact that
> programmer X needs no external power supply (wall wart).
> Instead a great deal is made about complicated DC-DC
> converters and charge pumps, and USB power, and serial port
> power, and even application circuit power.
>
> This all makes no sense to me whatsoever. Why bother with all
> the extra complexity, expense, and possible power problems
> when you can simply hook up a $2 wall wart?
> I personally don't give a rip about the supposed
> inconvenience of a wall wart. It's not like I'm programming
> while riding a motorcycle. I'm quite content to use my desk
> at home or work.
>
> Can someone please explain to me what, if any, advantage
> there is to not using and external power supply? I can't think of any.


They're heavy, so they fall out of the socket.
Heavy means expensive shipping.
Big, so they take up 2 sockets.
Always on, adds to your power bill.
They get hot and melt occasionally.
Different plugs.
Country specific.  (ok, laptops aren't)
Unregulated, so 12v is anywhere from 12v to 24v.
Sometimes AC, sometimes DC.
You can't use the 12v walwart from another device, even if the plugs match.
Lighter load means higher voltage means smoke emmissions.
Can't get replacements when you lose them.
If you can, they are $$$.
Device doesn't say what is needed when you lose the wallwart.
Oh, you lose them.
Thin cables get tangled, knotted and broken.
Cases are welded shut, can't repair them, even for a broken wire.
...and so on.

Yean, not a whole lotta love for the wallwart out there.

I've got a large box full of them, everything from 2v DC to 15v AC (NetComm
modems).  For stuff like ADSL routers & external USB hubs, I run them off
the PC, and the wallwart goes in the box.

On a related note, I've decided I hate permanently attached power leads (TV,
DVD etc) so I've been replacing them with IEC ones.  Now I can pull the DVD
out of the cabinet by unplgging it.

I also encountered a flaw in the 'turn off your devices to save power' meme.
A Sony CD player I have has the transformer inside (quite normal) but the
power switch is attached to the low voltage side.  So even when off it draws
power, staying nice & warm for the cockroaches.  Thanks Sony!

Tony

2006\12\02@204531 by peter green

flavicon
face

> They're heavy, so they fall out of the socket.
ahh tiny flimsy american sockets
> Heavy means expensive shipping.
yep
> Big, so they take up 2 sockets.
see above
> Always on, adds to your power bill.
yep
> They get hot and melt occasionally.
never seen one actually melt though i do recall cooling a laptop power brick by rubbing the plastic with damp clothing (on the side that had no openings in it) and allowing the moisture on the plastic to boil off.
> Different plugs.
> Country specific.  (ok, laptops aren't)
> Unregulated, so 12v is anywhere from 12v to 24v.
> Sometimes AC, sometimes DC.
yep theese are all issues
> You can't use the 12v walwart from another device, even if the
> plugs match.
if the plugs match, the voltages match and the power ratings match then why not.
> Lighter load means higher voltage means smoke emmissions.
> Can't get replacements when you lose them.
> If you can, they are $$$.
only if you buy the manufacturers branded ones.

> Device doesn't say what is needed when you lose the wallwart.
in my experiance they usually do, generally in small embossed print by the input socket

> Oh, you lose them.
yep
> Thin cables get tangled, knotted and broken.
true, otoh seas of full thickness mains cable aren't exactly nice to handle either.
> Cases are welded shut, can't repair them, even for a broken wire.
yeah this is an issue though some i have seen are screwed together instead.



2006\12\02@215232 by Tony Smith

picon face
> > They're heavy, so they fall out of the socket.
> ahh tiny flimsy american sockets

> > Big, so they take up 2 sockets.
> see above


I'm in Australia.  Our sockets are a bit better, but the plugpacks still
fall out or make a general nuisance of themselves.

In front on me at the moment is a six outlet powerboard, it has 3 plugpacks
and 3 mains cables plugged into it.  Since you can't put 2 plugpacks side by
side, they are alternated, so plugpack / cable / plugpack etc.

However, now I can't get the mains cables out since they are blocked by the
plugpacks.

I've seen powerboards with sockets spaced further apart.

Hooray for USB/Fireware. That's knocked the number of plugpacks down a bit.
Old IBM PCs used to have a 12v out socket for running speakers.  Shame it
didn't become standard.

Tony

2006\12\02@224829 by tachyon 1

picon face
OK gang, I get it.
Good points on the anti-wall-wart front.
I have to admit that I was thinking about this from a whole different
angle.
I was thinking more along the lines of systems you can build at home from
the parts box.
I'd still really like to find a good, home build-able, JDM like
programmer. Something that will take a 40 pin ZIF socket, program nearly
all PIC's, something that I can build without having to drive 3 hours to
"Bob's rare European electronics emporium" to get parts for.
Something that has good software or better, MPLab support, and to really
add to my wish-list, has software support on Linux.

Any suggestions?

BTW, anyone know of any Linux apps that work with El Cheapo?

Later, and thanks for setting me straight on why wall-warts suck.

--

Search for products and services at:
http://search.mail.com

2006\12\02@225504 by Breesy

flavicon
face
On the flip side, what is the advantage of using a wallwart? I cannot
think of any..

Tony Smith wrote:
{Quote hidden}

2006\12\02@225622 by tachyon 1

picon face
Yeah,
Given all the spare standardized molex power connectors in a PC, I never
got why PC's don't also have a standard external power plug too.
Maybe a nice, daisy chainable DIN plug with +12, +5, and maybe +3.3 volt
pins.

Oh well.

 {Original Message removed}

2006\12\02@234623 by Tony Smith

picon face
> -----Original Message-----
> From: .....piclist-bouncesKILLspamspam@spam@mit.edu
> [piclist-bouncesspamKILLspammit.edu] On Behalf Of Breesy
> Sent: Sunday, 3 December 2006 2:55 PM
> To: Microcontroller discussion list - Public.
> Subject: Re: [PIC] Why the preoccupation with bus powered programmers?
>
> On the flip side, what is the advantage of using a wallwart?
> I cannot think of any..
>
> Tony Smith wrote:
> >>> They're heavy, so they fall out of the socket.
> >>>      
> >> ahh tiny flimsy american sockets
> >>    


Cheap.

Compliance is almost a non-issue.  Since your device isn't mains powered
(the wall wart has it's own compliance tag), you avoid all that 'will your
thing electrocute people?' and related stuff.  Not that a lot of
manufacturers care anyway...

Easier to manufacture & design your device as well, since you don't have to
deal with mains power.  Ok, it's not that hard, but people just don't want
to cram a  transformer into a router or USB hub.

Tony


2006\12\03@000012 by Tony Smith

picon face
>   Hooray for USB/Fireware. That's knocked the number of plugpacks down
>   a bit.
>   Old IBM PCs used to have a 12v out socket for running
> speakers. Shame
>   it
>   didn't become standard.
>
>   Tony
>
>
>  Yeah,
> Given all the spare standardized molex power connectors in a
> PC, I never got why PC's don't also have a standard external
> power plug too.
> Maybe a nice, daisy chainable DIN plug with +12, +5, and
> maybe +3.3 volt pins.
>
> Oh well.


IBM Aptivas, IIRC had the 12v socket.  From '98 or so.

5v, 9v & 12v pretty much covers it.  5v for USB hubs, wireless mice etc, 12v
for speakers and external HDDs, 9v for modems & routers (what's wrong with
12v?).  Low power 5v can run off USB anyway.

I take a backplane, drill a hole and put a DC socket in it.  5v & 12v you
can wire straight up, for 9v I put a regulator in.  It would be nice if each
voltage was a different size plug (so you can plug a 12v cable in 5v but not
vice versa) but life doesn't work that way.

Tony

2006\12\03@001422 by Phillip Coiner

picon face

Hi Tony

The only good thing I can think of is that you can have a smaller
housing/enclosure because the transformer and rectifier are over on the wall
in its wart.

Now theoretically speaking you can design you product to use one low voltage
and just throw in a wall wart with your product for the country you are
shipping to and bingo you are world wide but trust me this not nearly as
simple as it sounds.
There is a dog's breakfast of plug styles and compliance requirements.
Many of the plug style look close but won't work...e.g. prongs too short to
skinny too fat ad nauseam

I have been shipping the same 220V wall wart to my Korean customers/dealers
for five years.
Last week I get a message from Hyundai with an attached picture that says
I'm shipping the wrong plug style.
The picture of the plug they have is definitely different but I'm thinking
what in the heck did they do with the hundreds of transformers I sent for
the last five years.
In Singapore they have the new UK style and older buildings have the old
German/Euro two pronged 220 looking thing without an earth connection.

In Russia they have the same prongs as the old German ones but they have a
round thing with an earth connection...as do other countrys tin the former
CIS
If you take the cover off the socket you can plug in a Euro 220 provided you
are brave enough to live with out the earth wire connected or you can trim
the corners off with a pocket knife of an old Euro syle ...that's what I did
with my Dell wall wart for my laptop when I was in Moscow and Kazan.

Japan has 110V just like the US.

We have a large shipment to Brazil that is being held up for you guessed it
wall transfomers.
Everyday finished products sit in my factory my profit margin decreases.

Now when you buy/desing in these transformers/wall warts they have to be
RoHs (reduction of hazardus subtances and WEE  compliant (a recycling
directive).
They also have to have UL CE certification and a dozen other certifications.

I spend probably six hours a week dorking with wall transformers evey week
of the year.
So I have a hate hate relationship with the things.
The things allow me to make my product compatable world wide but provide me
with daily hassles and my customers love them as just as much as the as
comments here suggest/imply.



Phillip
Things should be as simple as possible but no simpler














Phillip Coiner
CTO, GPS Source, Inc.


Your source for quality GNSS Networking Solutions and Design Services, Now!
{Original Message removed}

2006\12\03@010254 by Xiaofan Chen

face picon face
On 12/3/06, .....tachyon_1KILLspamspam.....email.com <EraseMEtachyon_1spam_OUTspamTakeThisOuTemail.com> wrote:

> I'd still really like to find a good, home build-able, JDM like
> programmer.
JDM like --> not a good PIC programmer.

> Something that will take a 40 pin ZIF socket, program nearly
> all PIC's, something that I can build without having to drive 3 hours to
> "Bob's rare European electronics emporium" to get parts for.

To program nearly all PICs? You have to buy commercial programmers
like ICD2. PICkit 2 is pretty good. But you'd better buy them ready-made.

> Something that has good software or better, MPLab support, and to really
> add to my wish-list, has software support on Linux.

MPLAB support integrated --> only programmers from Microchip. Right now
the MPLAB support for PICKit 2 is not that good but they are working on it.

> Any suggestions?
It seems you will need PICkit 2 or ICD2 based on your requirement (but
they can not be easily built by your own).

If you want a decent programmer you can easily built, maybe you can
try EPIC+ (http://pikdev.free.fr/IMAGES/EPIC+-modified.png) provided
your computer has a parallel port.

Other choice can be EasyProg/WISP628/Kit150 but I think PICkit 2 is
a better value than all of them.

> BTW, anyone know of any Linux apps that work with El Cheapo?

Try Piklab or Pikdev. Pikdev supports many cheap programmers.
PIklab adds PICkit 1/2/ICD2 support and integrated gputils/sdcc/JAL
along with MPLAB C18/C30 support under Linux.

Pikdev: http://pikdev.free.fr/
Piklab: http://piklab.sourceforge.net/
GNUPIC: http://www.gnupic.org/
gputils: http://gputils.sourceforge.net/
gpsim: http://www.dattalo.com/gnupic/gpsim.html
sdcc: http://sdcc.sourceforge.net/

Regards,
Xiaofan

2006\12\03@012546 by Xiaofan Chen

face picon face
On 12/3/06, Xiaofan Chen <xiaofancspamspam_OUTgmail.com> wrote:
>
> Other choice can be EasyProg/WISP628/Kit150 but I think PICkit 2 is
> a better value than all of them.
>
> > BTW, anyone know of any Linux apps that work with El Cheapo?
>
> Try Piklab or Pikdev. Pikdev supports many cheap programmers.
> PIklab adds PICkit 1/2/ICD2 support and integrated gputils/sdcc/JAL
> along with MPLAB C18/C30 support under Linux.
>

By the way, PICkit 2 is well supported under Linux with pk2
along with PS+ with picp.
http://home.pacbell.net/theposts/picmicro/
http://groups.google.com/group/pickit-devel

Wisp628 is also well supported under Linux.
http://www.robh.nl/picsoft.php

For parallel-port based programmer, you can try either Pikdev/Piklab
or odyssey (http://vasco.gforge.enseeiht.fr/). I personally do not have
any experiences with them but my colleague told me that EPIC+
from MeLabs is quite okay.
http://www.melabs.com/products/epic.htm

Regards,
Xiaofan

2006\12\03@013424 by Xiaofan Chen

face picon face
On 12/3/06, Breesy <@spam@webmasterKILLspamspambreesy.com> wrote:
> On the flip side, what is the advantage of using a wallwart? I cannot
> think of any..
>

Many advantages. They are cheap and readily available. They make
your product design much easier.

Even USB powered programmer might need external supply. For
example, PICkit 2 does not work well with 16F7x7 since often the
USB supply will drop below 4.5V to proper erase 16F7x7s.

Actually you will need a wallwart anyway for your other things. Most
of the programmers will need one. ICD2 has one. Promate III has a
big one. Many demo boards need 9V supply or 5V.

USB power is in general not well protected, not well defined, not
so reliable. It is good to be used to power the programmer. But it
will have problems to power the target. Many PICkit 2 users
managed to blow their PICkit 2 with short-circuit target.
http://forum.microchip.com/tm.aspx?m=192175

Regards,
Xiaofan

2006\12\03@063031 by Eamon Skelton

flavicon
face
tachyon_1@email.com wrote:

> I'd still really like to find a good, home build-able, JDM like
> programmer. Something that will take a 40 pin ZIF socket, program nearly
> all PIC's, something that I can build without having to drive 3 hours to
> "Bob's rare European electronics emporium" to get parts for.

I use a JDM type programmer to program 40 pin PICs It has an
18 pin ZIF socket for smaller PICs and a 40 pin DIL socket
for those rare occasions when I need to zap a PIC16F877A.



> Something that has good software or better, MPLab support, and to really
> add to my wish-list, has software support on Linux.
>
> Any suggestions?

The JDM works great with picprog and Pikdev.
http://hyvatti.iki.fi/~jaakko/pic/picprog.html
http://pikdev.free.fr/

You will need a PC with a 'proper' serial port. Laptops
with +/- 5V serial ports and USB to serial adapters probably
won't work. It works fine on all five of my desktop PCs,
ranging from an old Pentium 133 to a nice new Athlon-64.

> BTW, anyone know of any Linux apps that work with El Cheapo?
>
This guy seems to have had some success:
http://uvasux.googlepages.com/elcheapo

Good Luck.

E.S.


--
linux 2.6.17

2006\12\03@095159 by olin piclist

face picon face
> I'd still really like to find a good, home build-able, JDM like
> programmer.

Why does it have to be JDM-like?  Your requirements for "good" and "JDM"
conflict.

> Something that will take a 40 pin ZIF socket, program
> nearly all PIC's, something that I can build without having to drive
> 3 hours to "Bob's rare European electronics emporium" to get parts
> for.

My EasyProg (http://www.embedinc.com/easyprog) satisfies these requirements
except maybe the "nearly all PICs" part.  It does support all the reasonable
hobby PICs.

Microchip has a very broad product line, which is important for finding a
tight fit for high volume designs where a extra $.05/unit makes a
difference.  Only a small set of PICs make sense for hobby use.  Except for
special peripherals like CAN or USB, you should just get the biggest bestest
available in the footprint.  Unless space is really really a concern,
hobbyists are probably best off considering the 28 pin the "standard"
package.  Go to 40 pins only when you need the extra pins, and maybe down to
18 or 8 pins on occasion you really need the small size.  Buy enough of the
main part you use to get the first price break, then use that for all
projects.  You probably won't use the chip's full capability for most
projects, but now you have a bunch of chips you can re-use and interchange.

Given the view above, here is the list of reasonable hobby PICs:

 28 pin: 18F2620  (get 10 of these and use for everything)
 40 pin: 18F4620  (only when you need more pins}
 18 pin: 18F1330  (do you really really need the smaller package?)

If you've gotten good with PICs and you know what you're getting into, you
could possibly try a 12F675 for 8 pins, but then you have to get into the
quirks of the PIC 16 family.

> Something that has good software or better, MPLab support, and to
> really add to my wish-list, has software support on Linux.

I like to think my software is pretty good.  You're not going to find a
hobby programmer with MPLAB support.  It's too much work to hook into MPLAB,
and the slim margins on hobby products don't support that.  As for support
on minority operating systems, all I can offer is the protocol
specification.  Wouter's software uses Python, so it should work accross
systems.  If this is important you could hook up a ZIF socket to a Wisp and
you will most of what you want.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\03@105514 by DAX2

picon face
Cyberguys (http://www.cyberguys.com) sells a short (1 ft) extension called a Power
Strip Liberator. It is used in conjunction with a wall wart so that all the
sockets on a power strip may be used.
If this is an echo of a previous post, I apologize. I didn't read them all.

----- Original Message -----
From: <KILLspampiclistKILLspamspammmendes.com>
To: "'Microcontroller discussion list - Public.'" <RemoveMEpiclistTakeThisOuTspammit.edu>
Sent: Saturday, December 02, 2006 7:04 PM
Subject: RE: [PIC] Why the preoccupation with bus powered programmers?

> As I type this, I have 3 power strips under my desk, with a bunch of
> wallwarts connected to it.  If I get anything else that uses a walwart
> that I'll have to get a 4th strip.  There is space on the 3 strips right
> now for about 3 or 4 powercords, but a walwart will not fit in any of
> those free spots.  Might have something to do with that?
>
>
> -Mario
>
> {Original Message removed}

2006\12\03@111746 by Phillip Coiner

picon face
Hi Mario
Radio Shack makes an extension cord power strip thingy called an octopus.
You can plug in a flock of wall warts but it is about as ascetically
pleasing as a garbage dump......not to say that power strips are works of
art or anything.
The whole wall wart is thing is so inelegant that it cries out for a
solution but obviously I'm too stooopid to figure just what that is exactly.
I can just about guarantee it will be one of those duh moments when we see
the solution and whoever does will make a load of cash if/when they get it
right.
Phillip
Things should be as simple as possible but no simpler



Phillip Coiner
CTO, GPS Source, Inc.


Your source for quality GNSS Networking Solutions and Design Services, Now!

{Original Message removed}

2006\12\03@140742 by Wouter van Ooijen

face picon face
> Can someone please explain to me what, if any, advantage
> there is to not
> using and external power supply? I can't think of any.

For an almost 0-components programmer it makes perfect sense to
eliminate one more (very bulky) component. And a wall-wart alone won't
do the job, unless it is a quality regulated one you will have to add a
connector (which one?), an 7805, capacitors, and probably a series
diode.

For an USB-based programmer it makes sense to avoid that extra cable.
Last months I did a PIC assembler class for Informatics students, using
an all-in-one board with programmer, PIC and peripheral. USB-power means
one thing less to worry about when the students are using the board for
the first time. The drawback is that it does not work with some laptops
and probably not with unpowered hubs.

For an ICSP programmer it makes of course sense to draw the power from
the target itself.

Wouter van Ooijen

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



2006\12\03@140742 by Wouter van Ooijen

face picon face
> I'd still really like to find a good, home build-able, JDM like
> programmer.

Which 2 of the 3 do you want?

Wouter van Ooijen

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


2006\12\03@145810 by David VanHorn

picon face
We once shipped a bunch of transaction terminals to Israel, with the
specific wall-wart mains connector asked for by the customer.
Israeli customs seized and destroyed the shipment because of the wall-warts.


Ouch.

2006\12\03@150203 by olin piclist

face picon face
Wouter van Ooijen wrote:
> For an ICSP programmer it makes of course sense to draw the power from
> the target itself.

But then the programmer can't control Vdd, which is necessary for some
chips.  In many instances a hobby programmer can get away with using target
Vdd, but serious professional programmers can't afford to do that.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\03@152143 by tachyon 1

picon face
Very helpful, thank you. I'll check into those recommendations.

Also, I wasn't really clear enough when I said 'JDM Like'.
What I really meant by that was form factor. Small, simple components,
and a single 40 pin ZIF socket (and an ICSP connector).
USB is fine, though I'd just as soon have Parallel (or Serial).

 {Original Message removed}

2006\12\03@154548 by Wouter van Ooijen

face picon face
> > For an ICSP programmer it makes of course sense to draw the
> power from
> > the target itself.
>
> But then the programmer can't control Vdd, which is necessary for some
> chips.  In many instances a hobby programmer can get away
> with using target
> Vdd, but serious professional programmers can't afford to do that.

ICSP and variable-voltage verification is a shaky combination anyway....

BTW: which chip require variable Vdd for programming (!= for
verification)?

Wouter van Ooijen

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


2006\12\03@155131 by Wouter van Ooijen

face picon face
> Also, I wasn't really clear enough when I said 'JDM Like'.
> What I really meant by that was form factor. Small, simple components,
> and a single 40 pin ZIF socket (and an ICSP connector).

Pickit2 and a piggyback ZIF socket with a personality module inbewteen?

DIY150?

If you expect to use the same 40-pin ZIF for various sized PICs you
require the programmer circuit to do quite some pin shuffeling.
Definitely doable, but be aware that your seemingly simple '40 pin ZIF'
requirement requires some extra circuitry.

Wouter van Ooijen

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


2006\12\03@174720 by Bob Axtell

face picon face
tachyon_1@email.com wrote:
>  Very helpful, thank you. I'll check into those recommendations.
>
> Also, I wasn't really clear enough when I said 'JDM Like'.
> What I really meant by that was form factor. Small, simple components,
> and a single 40 pin ZIF socket (and an ICSP connector).
> USB is fine, though I'd just as soon have Parallel (or Serial).
>  
Both Serial and Parallel Ports are vanishing, and it will be very
unusual to see either in a
new PC.  Time marches on.

Personally, I will miss them both, because Microsoft STILL doesn't have
enough reliability
in their USB stuff. But that's the way it is going, and railing about it
will do NO good.

I am hoping that when MS releases XP sp2046 the USB drivers might
finally work.

I just recently bought two identical WinXP PCs that had a DB9 serial
port in them. In
one PC, the DB9 serial port won't work right out of the box. But the USB
section works,
so I use a USB<>serial cable and all is well. Go figure...<G>

--Bob


2006\12\03@175157 by Bob Axtell

face picon face
Wouter van Ooijen wrote:
>>> For an ICSP programmer it makes of course sense to draw the
>>>      
>> power from
>>    
>>> the target itself.
>>>      
>> But then the programmer can't control Vdd, which is necessary for some
>> chips.  In many instances a hobby programmer can get away
>> with using target
>> Vdd, but serious professional programmers can't afford to do that.
>>    
>
> ICSP and variable-voltage verification is a shaky combination anyway....
>
> BTW: which chip require variable Vdd for programming (!= for
> verification)?
>  
NOBODY requires that for hobbyist use. The main problem is that USB
ports are allowed to drop their
voltage to 4.3V legally, and while most PICs will program at that
voltage, few will execute a bulk-erase, which is
the main problem with most USB programmers.

--Bob

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

2006\12\03@180819 by Jan-Erik Söderholm

face picon face
I read Olin's post as that he was talking about Vpp-before-Vdd
(which is harder to implement on self-powered targets). Not the
level as such of Vdd (that very well might be another problem,
of course).

Jan-Erik.


Bob Axtell skrev:
{Quote hidden}

2006\12\03@182056 by Genome

picon face
I thinks most buspowerd programmers cant be used with laptops... they are
more of a convenience, i mean you just plug and play... and if you prefer
them thats -1 cable hanging above your desk... they are cheaper and lots of
info outher if you like to DIY... personaly i would not like to use a
wallwart because they heatup and wasting power... you have to remove them
from the socket everytime you want to turn them off... And most people are
lazy... so they get left on the wall hanging all the time...

<spamBeGonetachyon_1spamBeGonespamemail.com> wrote in message
news:TakeThisOuT20061202185745.16A9886AE0EraseMEspamspam_OUTcal1-1.us4.outblaze.com...> {Quote hidden}

> --

2006\12\03@182348 by peter green

flavicon
face
> I read Olin's post as that he was talking about Vpp-before-Vdd
> (which is harder to implement on self-powered targets). Not the
> level as such of Vdd (that very well might be another problem,
> of course).
yeah, that can be an issue if you disable MCLR.



2006\12\03@184057 by peter green

flavicon
face
note, some quoted stuff rearranged to make responding to it easier.

> I'd still really like to find a good, home build-able, JDM like
> programmer.
JDM like != good, such programmers are totally incompatible with USB-Serial adaptors and put the pics ground at negative voltage making them largely unsuitable for ICSP (you can do ICSP with them but you have to be extremely carefull).

> Something that will take a 40 pin ZIF socket,
there are few programmers that have a big zif socket and support lots of pics in it for the simple reason that it's hard to do, some lines need to be very flexible (anything from an input to a power supply) which means either lots of discrete transistors or special purpose chips.

> something that I can build without having to drive 3 hours to
> "Bob's rare European electronics emporium" to get parts for.
i'm assuming you have found yourself a source of pics, so your best bet is probablly to first build a cheap'n'nasty programmer and then build one of the PIC based intelligent programmer designs (using your cheap'n'nasty programmer to program the pic for it).


>program nearly
> all PIC's,

> Something that has good software or better, MPLab support, and to really
> add to my wish-list, has software support on Linux.
for programming nearly all pics, your best bet is probablly to look for a design for a clone of one of microchips programmers. The number of pics is insane.


2006\12\03@185139 by olin piclist

face picon face
Wouter van Ooijen wrote:
> ICSP and variable-voltage verification is a shaky combination
> anyway....

There are some issues, but there are generally not hard to deal with if
considered in the design from the start.  I have a write that discusses
these issues at http://www.embedinc.com/picprg/icsp.htm.

> BTW: which chip require variable Vdd for programming (!= for
> verification)?

I didn't say "variable", only that the programmer needs to control Vdd.  By
that I meant being able to turn it on and off, and sometimes with specific
timing related to Vpp.  Some PICs require Vpp before Vdd, some the other way
around, and some have specific timing requirements between the two.  As far
as I know, all the PIC 18 can be programmed with Vdd always on, but that's
not the case for all PIC 16, at least not for high voltage programming.

And yes, Microchip does recommend variable Vdd for verification of some
parts.  They make a point of distinguishing between "prototype" and
"production" programmers, with the latter verifying at the Vdd limits.

Then there's the issue of the target voltage being fine for running the PIC,
but not fine for all programming operations, like bulk erase.  Bulk erase is
usually limited to 4.5-5.5V, whereas many PICs can run down to 2V.

Simple programmers that rely on target power can be appropriate for hobby
use where cost is a major concern, but they are not appropriate for
professional or production use.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\03@221121 by John Chung

picon face
I don get it. Why destroy wall-warts and the shipment?
Do look harmless...

John


--- David VanHorn <RemoveMEdvanhornspamTakeThisOuTmicrobrix.com> wrote:

> We once shipped a bunch of transaction terminals to
> Israel, with the
> specific wall-wart mains connector asked for by the
> customer.
> Israeli customs seized and destroyed the shipment
> because of the wall-warts.
>
>
> Ouch.
> --

2006\12\04@001112 by tachyon 1

picon face
I don't blame the Israelies. After all I've heard about wal-warts
lately, it wouldn't surprise me a bit to learn that wal-warts are also
anti-semetic.


 {Original Message removed}

2006\12\04@003231 by tachyon 1

picon face
Well, I find a lot of Windows hardware drivers to be 'lacking'. In fact
XP still has issues with USB multi-card readers.
I don't know how many people I've talked to with brand new PC's with
built in floppy bay multi-card readers that crash
or lock-up when a card is inserted. Also the way WIndows handles
resources management on machines with a lot of
IRQ's seems broken or at least stupid. It seems especially stupid about
dealing with APIC's, though it does do slightly
better on machines with real SMP.
Anyway, my primary personal systems are Linux based and I've had no real
issues with core HW driver stability or
performance. USB seems especially solid.

As for PAR/SER dissappearing, I highly doubt it. It might be going away
on the kind of crap I won't buy anyway
(ie DELL, eMachines, etc. consumer desktops) but there's no way in heck
they'll get rid of it anytime soon.
There's just way too much demand from business customers with legacy
connectivity needs. Especially in industrial,
manufacturing, and production companies. In fact a friend of mine makes a
nice living selling add on PCI PAR/SER
boards to make up the demand for customers who require them. (see
http://www.softio.com/index.htm ) Nice boards BTW.
We used to use a lot of them for our industrial/building controls
customers.


 {Original Message removed}

2006\12\04@005012 by tachyon 1

picon face
 ----- Original Message -----
 From: "peter green"
 To: "Microcontroller discussion list - Public."
 Subject: RE: [PIC] Why the preoccupation with bus powered
 programmers?
 Date: Sun, 3 Dec 2006 23:39:11 -0000


 note, some quoted stuff rearranged to make responding to it easier.

 > I'd still really like to find a good, home build-able, JDM like
 > programmer.
 JDM like != good, such programmers are totally incompatible with
 USB-Serial adaptors and put the pics ground at negative voltage
 making them largely unsuitable for ICSP (you can do ICSP with them
 but you have to be extremely carefull).

 NOTE: See my later post for a clarification of "JDM like"

 > Something that will take a 40 pin ZIF socket,
 there are few programmers that have a big zif socket and support
 lots of pics in it for the simple reason that it's hard to do, some
 lines need to be very flexible (anything from an input to a power
 supply) which means either lots of discrete transistors or special
 purpose chips.

 Or I assume, a PIC and a few less components?

 > something that I can build without having to drive 3 hours to
 > "Bob's rare European electronics emporium" to get parts for.
 i'm assuming you have found yourself a source of pics, so your best
 bet is probablly to first build a cheap'n'nasty programmer and then
 build one of the PIC based intelligent programmer designs (using
 your cheap'n'nasty programmer to program the pic for it).


 > program nearly
 > all PIC's,

 > Something that has good software or better, MPLab support, and to
 really
 > add to my wish-list, has software support on Linux.
 for programming nearly all pics, your best bet is probablly to look
 for a design for a clone of one of microchips programmers. The
 number of pics is insane.

 At this point, I also have to expand my definition of "most PICs"
 I meant most hobbyist PICs, that I'm interested in. Which presently
 means mostly older low end and mid range lines.
 12x and 16x, For example, 12F675, 12F629, 16F628(A), 16F648(A),
 16F87X(A) and a few common 18x you see in projects.

 I apologize that I keep thinking of this list as having mostly a
 hobbyist audience. Which I'm seeing not to be true.

 I built my own PIC programmer for the same reason I built my own
 welder. I'm cheap, I like to do it myself, and I only need to
 use it to solve the hobbyist problems I have of my own, or maybe to
 help friends out.

 -

2006\12\04@015002 by Denny Esterline

picon face
> And yes, Microchip does recommend variable Vdd for verification of some
> parts.  They make a point of distinguishing between "prototype" and
> "production" programmers, with the latter verifying at the Vdd limits.

Does anyone have any actual data about how many parts verify at nominal
voltage, but fail verification at the limits?

It seems to me that the value of verification at Vdd extremes would be
directly related to such statistical information.

Then it begs to ask the question, what do you do if it fails verification at one
of the Vdd extremes? Reprogram? Discard?

-Denny


2006\12\04@031608 by Wouter van Ooijen

face picon face
>   I apologize that I keep thinking of this list as having mostly a
>   hobbyist audience. Which I'm seeing not to be true.

Only half true.

But with those 'refined' requirements you could check Wisp628. No ZIF
and no pin switching, but good for ICSP of 7805-powered targets. And do
consider a pickit2, even if you have to buy it. And check a few ICD2
clones (there are a lot of simplified ICD2 designs around - but be aware
that such simplifications also limit the features).

Wouter van Ooijen

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


2006\12\04@031609 by Wouter van Ooijen

face picon face
> There are some issues, but there are generally not hard to
> deal with if considered in the design from the start.

My market is hobbyist and 'small scale professionals'. For that segment
those issues can be very hard to deal with. But I totally agree that
they can be dealt with by a competent designer when he is present early
enough in the design process.

> I didn't say "variable", only that the programmer needs to
> control Vdd.  By that I meant being able to turn it on and off

OK, agree.

Wouter van Ooijen

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


2006\12\04@031609 by Wouter van Ooijen

face picon face
> I read Olin's post as that he was talking about Vpp-before-Vdd
> (which is harder to implement on self-powered targets). Not the
> level as such of Vdd (that very well might be another problem,
> of course).

OK, that can be solved in a brute force way (literally) :)

Wouter van Ooijen

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


2006\12\04@031609 by Wouter van Ooijen

face picon face
> > ICSP and variable-voltage verification is a shaky
> combination anyway....
> >
> > BTW: which chip require variable Vdd for programming (!= for
> > verification)?
> >  
> NOBODY requires that for hobbyist use. (snip)

Wrong branch of the discussion, this was about Olins remark that
target-powered programmers can not vary the Vdd which he said is
required for some PICs.

Wouter van Ooijen

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


2006\12\04@031610 by Wouter van Ooijen

face picon face
> I just recently bought two identical WinXP PCs that had a DB9 serial
> port in them. In
> one PC, the DB9 serial port won't work right out of the box.
> But the USB
> section works,
> so I use a USB<>serial cable and all is well. Go figure...<G>

The mainboard serial port of my main PC (bought 5 y ago) never worked
OK. That's probably more a driver issue than hardware. I always use my
(Wisp628) programmers with an usb-serial converter.

Wouter van Ooijen

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


2006\12\04@040914 by Alan B. Pearce

face picon face
>In front on me at the moment is a six outlet powerboard, it has
>3 plugpacks and 3 mains cables plugged into it.  Since you can't
>put 2 plugpacks side by side, they are alternated, so
>plugpack / cable / plugpack etc.

At least in Oz & NZ you can piggyback on top of the plugs ... or have they
outlawed those now.

2006\12\04@042454 by Alan B. Pearce

face picon face
>The whole wall wart is thing is so inelegant that it cries out for
>a solution but obviously I'm too stooopid to figure just what that
>is exactly. I can just about guarantee it will be one of those duh
>moments when we see the solution and whoever does will make a load
>of cash if/when they get it right.

The "most right" one I have seen so far is the wart supplied with my Palm.
It is a switcher that is small and light, and has an American 2 parallel pin
connector that folds 90 degrees to tuck it away for packing. In the folded
position you can slide on adapter plugs for other countries, making it as
universal as it could possibly be - assuming you can get the adapter for the
country you are going to next.

And at 5V at about an amp it would make a pretty neat test bench supply.

2006\12\04@043529 by Alan B. Pearce

face picon face
>Then it begs to ask the question, what do you do if it fails
>verification at one of the Vdd extremes? Reprogram? Discard?

I could envisage a production item where the power supply is close to one
extreme of the tolerance, and a chip that failed verification at the extreme
being fitted and getting intermittent errors in operation as a result. Hence
a verification failure caught at programming time, at supply extremes has a
good chance of keeping production out of the quagmire.

2006\12\04@044432 by peter green

flavicon
face
>   > Something that will take a 40 pin ZIF socket,
>   there are few programmers that have a big zif socket and support
>   lots of pics in it for the simple reason that it's hard to do, some
>   lines need to be very flexible (anything from an input to a power
>   supply) which means either lots of discrete transistors or special
>   purpose chips.
>
>   Or I assume, a PIC and a few less components?
Can PIC IO lines source/sink enough current at voltages sufficiantly close to the power rails to power another pic reliablly? i haven't tried it myself but i somewhat doubt it.



2006\12\04@045708 by Richard Prosser

picon face
Nope - at least they're still legal in NZ.

RP

On 04/12/06, Alan B. Pearce <A.B.PearceEraseMEspam.....rl.ac.uk> wrote:
> >In front on me at the moment is a six outlet powerboard, it has
> >3 plugpacks and 3 mains cables plugged into it.  Since you can't
> >put 2 plugpacks side by side, they are alternated, so
> >plugpack / cable / plugpack etc.
>
> At least in Oz & NZ you can piggyback on top of the plugs ... or have they
> outlawed those now.
>
> -

2006\12\04@051443 by Wouter van Ooijen

face picon face
> >   > Something that will take a 40 pin ZIF socket,
> >   there are few programmers that have a big zif socket and support
> >   lots of pics in it for the simple reason that it's hard
> to do, some
> >   lines need to be very flexible (anything from an input to a power
> >   supply) which means either lots of discrete transistors or special
> >   purpose chips.
> >
> >   Or I assume, a PIC and a few less components?
>
> Can PIC IO lines source/sink enough current at voltages
> sufficiantly close to the power rails to power another pic
> reliablly? i haven't tried it myself but i somewhat doubt it.

No, but I am afraid that it will work on your bench but fail (sometimes)
in the field. Same problem as with serial-port-powered programmers.

Wouter van Ooijen

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


2006\12\04@052454 by Gerhard Fiedler

picon face
tachyon_1@email.com wrote:
> As for PAR/SER dissappearing, I highly doubt it. It might be going away
> on the kind of crap I won't buy anyway (ie DELL, eMachines, etc.
> consumer desktops) but there's no way in heck they'll get rid of it
> anytime soon. There's just way too much demand from business customers
> with legacy connectivity needs. Especially in industrial, manufacturing,
> and production companies. In fact a friend of mine makes a nice living
> selling add on PCI PAR/SER boards to make up the demand for customers
> who require them.

The second part of this paragraph seems to contradict the first part. If
people make a nice living selling parallel/serial add-on boards, that's
probably at least partly because these ports are going away in the normal
computer systems.

When there's only 10% of customers left that wants/needs them, you're
starting to find more and more motherboards (including good ones) that
don't have them anymore. And more and more of the people who need them will
use add-ons (PCI, USB). Which makes sense to me, even though that makes
these ports more expensive.

Gerhard

2006\12\04@055151 by Jan-Erik Söderholm

face picon face
I thought he said *control* Vdd, not ?

Jan-Erik.



Wouter van Ooijen wrote :

> ...this was about Olins remark that
> target-powered programmers can not vary the Vdd which he said is
> required for some PICs.
>

2006\12\04@061020 by Tony Smith

picon face

{Quote hidden}

Industrial users buy a lot of those.  Still lots of ISA backpanes in the
world.

In the real world, serial is almost dead (the occasional 56k modem), and I'm
surprised that parallel has hung on for so long.  Must be a lot of old
printers out there.

A Dell laptop I had recently had both a serial & a parallel port, but no
PS2.  I thought that was odd.  Had to buy a USB/PS2 converter.  Since
laptops have USB, NIC & modems ports, what are people plugging into them?

I still use serial & parallel for CNC & control stuff.  I've got lots of
cards!

Tony

2006\12\04@061022 by Tony Smith

picon face
> >The whole wall wart is thing is so inelegant that it cries out for a
> >solution but obviously I'm too stooopid to figure just what that is
> >exactly. I can just about guarantee it will be one of those
> duh moments
> >when we see the solution and whoever does will make a load of cash
> >if/when they get it right.
>
> The "most right" one I have seen so far is the wart supplied
> with my Palm.
> It is a switcher that is small and light, and has an American
> 2 parallel pin connector that folds 90 degrees to tuck it
> away for packing. In the folded position you can slide on
> adapter plugs for other countries, making it as universal as
> it could possibly be - assuming you can get the adapter for
> the country you are going to next.
>
> And at 5V at about an amp it would make a pretty neat test
> bench supply.


I've got a couple of those from Palm as well, but mine came with a bunch of
connectors for the various countries.  Apple do them as well.

Switchmode plugpacks don't cause too many problems since they're lighter &
thinner, but I've only ever seen 5v versions.  (Nokia are 6v?)  Laptops
don't count as they have a cable.  You'd think they'd be cheaper than the
old ones.  Especially with copper at $US7k/ton, down from $US9K!  Just goes
to show what economies of scale and existing capital can do for you.

I doubt the plugpack is ever going to go away.  It would be nice to be able
to have a standard where stuff can be powered off the PC.  There's a
high-power variant of USB, essentially a second port that just supplies
power.  Shows up on Dell laptops.  Possibly 'not-a-standard' thing.

The USB Hub power supplies are nice too, 4-port ones come with a 2 amp
supply, and I've seen 4 amp (!) ones as well.

Tony

2006\12\04@070036 by olin piclist

face picon face
Denny Esterline wrote:
> Does anyone have any actual data about how many parts verify at nominal
> voltage, but fail verification at the limits?

Probably not outside of microchip.  I can't imagine a programmer verifying
at nominal voltage and the Vdd limits.  That would require 3 verification
passes.  Most programmers out there do only one.  A few professional
programmers like the PM3 and mine do two.  I don't know of a single
programmer that does three.

> Then it begs to ask the question, what do you do if it fails
> verification at one of the Vdd extremes?

Consider the PIC not programmed of course.  You might want to try again, but
if it persists you have a bad PIC.  Certainly the PIC needs to be tossed.
Whether to toss or repair the product depends on the economic particulars.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\04@081631 by Xiaofan Chen

face picon face
On 12/4/06, Tony Smith <RemoveMEajsmithEraseMEspamEraseMErivernet.com.au> wrote:
> A Dell laptop I had recently had both a serial & a parallel port, but no
> PS2.  I thought that was odd.  Had to buy a USB/PS2 converter.  Since
> laptops have USB, NIC & modems ports, what are people plugging into them?

My old Dell Inspiron 600M at home and pretty-new Dell Latitude D610 at work
both have serial/parallel port.

Why do you need PS/2? There are plenty of USB mouse/keyboard.

Interestingly, the IT guys just installed a port replicator for the
D610 and it only
adds PS/2 ports and S/PDIF to the existing ports.

Regards,
Xiaofan

2006\12\04@084958 by Tony Smith

picon face
> > A Dell laptop I had recently had both a serial & a parallel
> port, but
> > no PS2.  I thought that was odd.  Had to buy a USB/PS2 converter.  
> > Since laptops have USB, NIC & modems ports, what are people
> plugging into them?
>
> My old Dell Inspiron 600M at home and pretty-new Dell
> Latitude D610 at work both have serial/parallel port.
>
> Why do you need PS/2? There are plenty of USB mouse/keyboard.
>
> Interestingly, the IT guys just installed a port replicator
> for the D610 and it only adds PS/2 ports and S/PDIF to the
> existing ports.
>
> Regards,
> Xiaofan


The one I had was a Latitude as well.

I wanted to use a 'real' keyboard, and we only had PS2 ones floating about.
At this company, every had a laptop.  (I liked the little Toshiba Presario I
had for a while.  Cute, + SD card reader.)

I can't recall the last time I plugged something into a parallel port
(non-nerd stuff).  For business use, all you need is the network port.  Even
most 'home' printers are USB now.  Can you even buy parallel printers?  Why
do a lot of laptops still have them?

Tony

2006\12\04@090827 by Herbert Graf

flavicon
face
On Mon, 2006-12-04 at 22:09 +1100, Tony Smith wrote:
> In the real world, serial is almost dead (the occasional 56k modem), and I'm
> surprised that parallel has hung on for so long.  Must be a lot of old
> printers out there.
>
> A Dell laptop I had recently had both a serial & a parallel port, but no
> PS2.  I thought that was odd.  

Yes, and it bit me pretty bad once. When I was buying a laptop a serial
port was very important since the GPS units I had were both serial. It
wasn't until I bought the laptop and tried using the SECOND GPS that I
discovered it didn't have a PS2 port. The PS2 port of course power many
GPS serial devices. Had to actually go out and buy a USB GPS device on
the road, very annoying, but all my fault.

TTYL

2006\12\04@090946 by Herbert Graf

flavicon
face
On Mon, 2006-12-04 at 21:16 +0800, Xiaofan Chen wrote:
> On 12/4/06, Tony Smith <RemoveMEajsmithspam_OUTspamKILLspamrivernet.com.au> wrote:
> > A Dell laptop I had recently had both a serial & a parallel port, but no
> > PS2.  I thought that was odd.  Had to buy a USB/PS2 converter.  Since
> > laptops have USB, NIC & modems ports, what are people plugging into them?
>
> My old Dell Inspiron 600M at home and pretty-new Dell Latitude D610 at work
> both have serial/parallel port.
>
> Why do you need PS/2? There are plenty of USB mouse/keyboard.

Many "portable" devices use the PS2 port for power (when USB isn't
enough). I have a serial GPS unit and a USB CD writer that both rely on
the PS2 port for power. There are many other examples. TTYL

2006\12\04@091613 by Michael Rigby-Jones

picon face


{Quote hidden}

I think I'd have kludged a USB/PS2 power converter rather than paying for a new GPS!

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2006\12\04@102048 by Herbert Graf

flavicon
face
On Mon, 2006-12-04 at 14:15 +0000, Michael Rigby-Jones wrote:
{Quote hidden}

On the road? Didn't have any tools, any equipment, and I NEEDED the GPS
working for that trip, didn't have much choice.

The funny thing is the USB GPS I bought is a SERIAL unit with PS2 for
power! It just comes with a USB dongle with a USB-RS232 converter in it
and a PS2 recepticle! Grr...

The benefit however is it's a more modern GPS chipset, to it locks on
cold start much faster then my "few year old" GPS, and also can recover
from dropouts MUCH faster.

GPS technology is one of those things where the receivers have improved
many magnitudes over the years. The first GPS receiver I purchased
sometimes took up to 10 minutes to lock on cold start, took up to a
minute to recover on dropout, was a power HOG, and often didn't lock on
more then 4 birds (I think it was a 6 receiver design). And don't even
think about not giving it a 100% unobstructed view of the sky.

My latest GPS receiver cold starts < 1 minute, recovers from drop out in
seconds and takes so little power you barely notice.

TTYL

2006\12\04@110055 by Mark Rages

face picon face
On 12/3/06, Olin Lathrop <spamBeGoneolin_piclistSTOPspamspamEraseMEembedinc.com> wrote:
>
> And yes, Microchip does recommend variable Vdd for verification of some
> parts.  They make a point of distinguishing between "prototype" and
> "production" programmers, with the latter verifying at the Vdd limits.
>

Can you give an example of a flash PIC where M'chip recommends this?
I looked through the programming specs once, and only found this
distinction for PROM based chips.  Admittedly, my search wasn't
exhaustive.


Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2006\12\04@113549 by Wouter van Ooijen

face picon face
> Can you give an example of a flash PIC where M'chip recommends this?
> I looked through the programming specs once, and only found this
> distinction for PROM based chips.  Admittedly, my search wasn't
> exhaustive.

IIRC it is in (almost?) all programming specifications. In which one did
you not find it?

Wouter van Ooijen

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


2006\12\04@113935 by Bob Axtell

face picon face
Mark Rages wrote:
> On 12/3/06, Olin Lathrop <KILLspamolin_piclistspamBeGonespamembedinc.com> wrote:
>  
>> And yes, Microchip does recommend variable Vdd for verification of some
>> parts.  They make a point of distinguishing between "prototype" and
>> "production" programmers, with the latter verifying at the Vdd limits.
>>
>>    
>
> Can you give an example of a flash PIC where M'chip recommends this?
> I looked through the programming specs once, and only found this
> distinction for PROM based chips.  Admittedly, my search wasn't
> exhaustive.
>
>
> Regards,
> Mark
> markrages@gmail
>  
Mark, Microchip identifies programmers that verify at low and high VDD
ranges as "production"
programmers, and those that only operate at normal VDD as "hobbyist".  
The ICD2 is therefore
a "hobbyist" level programmer. But I can't remember where I saw the spec.

--Bob




2006\12\04@114703 by peter green

flavicon
face

> A Dell laptop I had recently had both a serial & a parallel port, but no
> PS2.  I thought that was odd.  Had to buy a USB/PS2 converter.
PS/2 ports on laptops were always a bit of a mess. Firstly there was a couple of different pinout standards for the combination. Secondly there was the need for extra hardware to handle combining of external and internal signals and possiblly hotplugging of the external stuff (though in my experiance said hotplugging was never very reliable).

far far simpler to use the PS/2 lines on the chipset for the internal stuff only and let people use USB for external stuff.


2006\12\04@114703 by peter green

flavicon
face

> In the real world, serial is almost dead (the occasional 56k
> modem), and I'm
> surprised that parallel has hung on for so long.  Must be a lot of old
> printers out there.
also there seem to be a lot of dongles for specialist software that rely on having a paralell port.


2006\12\04@114924 by peter green

flavicon
face


> -----Original Message-----
> From: EraseMEpiclist-bouncesspamEraseMEmit.edu [@spam@piclist-bounces@spam@spamspam_OUTmit.edu]On Behalf
> Of Wouter van Ooijen
> Sent: 04 December 2006 16:36
> To: 'Microcontroller discussion list - Public.'
> Subject: RE: [PIC] Why the preoccupation with bus powered programmers?
>
>
> > Can you give an example of a flash PIC where M'chip recommends this?
> > I looked through the programming specs once, and only found this
> > distinction for PROM based chips.  Admittedly, my search wasn't
> > exhaustive.
>
> IIRC it is in (almost?) all programming specifications. In which one did
> you not find it?
iirc the 18Fxx2/xx8 specs and the 18fxxxx specs both say that a programer should have two programable voltage supplies but doesn't actually say that you should use them for anything!



2006\12\04@120545 by Mark Rages

face picon face
On 12/4/06, Bob Axtell <spamBeGoneengineerspamKILLspamneomailbox.com> wrote:
{Quote hidden}

Well, the reference is what I'm asking for :)

I've seen this bit of information passed around the list as folklore,
but I can't find a source for it, except for the old PROM/EPROM parts.
I'll be the first to admit, I haven't read all the documents on
Microchip's website.

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2006\12\04@120853 by Mark Rages

face picon face
On 12/4/06, Wouter van Ooijen <TakeThisOuTwouter.....spamTakeThisOuTvoti.nl> wrote:
> > Can you give an example of a flash PIC where M'chip recommends this?
> > I looked through the programming specs once, and only found this
> > distinction for PROM based chips.  Admittedly, my search wasn't
> > exhaustive.
>
> IIRC it is in (almost?) all programming specifications. In which one did
> you not find it?
>
> Wouter van Ooijen
>

OK, I must be looking in the wrong place.  Let's start with the
12F675, since I've got its programming spec open right now.  Which
page is it on?

(URL: http://ww1.microchip.com/downloads/en/DeviceDoc/41191D.pdf )

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2006\12\04@121134 by Alan B. Pearce

face picon face
>also there seem to be a lot of dongles for specialist software
>that rely on having a paralell port.

A lot of these are moving to USB dongles though.

2006\12\04@123530 by Hazelwood Lyle

flavicon
face
> On 12/4/06, Wouter van Ooijen <TakeThisOuTwouterKILLspamspamspamvoti.nl> wrote:
> > > Can you give an example of a flash PIC where M'chip
> recommends this?
> > > I looked through the programming specs once, and only found this
> > > distinction for PROM based chips.  Admittedly, my search wasn't
> > > exhaustive.
>

Most Interesting. I found the following in AN910
"PICmicro® Device Programming: What You Always
Wanted to Know (But Didn't Know Who to Ask)"
Page 16
Quote Follows...
VERIFICATION
Reading back a program memory location and verifying
its contents is a highly recommended practice; it makes
certain that each memory cell has been properly written
or erased. Regardless of the memory technology, all
programming algorithms include a read-back step
immediately following the write. This is done with a
"Read Data" command for devices using the mid-range
method, or a Table Read for PIC18 devices.
For OTP devices, a major issue in verification has been
establishing the parameters for a properly programmed
location. Minute differences in cells may mean that a
cell properly written to at the operating VDD might not
show as properly erased at VDDMIN, or not be
consistently read as programmed at VDDMAX.
With these possibilities in mind, Microchip has
developed programming algorithms that take these
possibilities in account. When the memory has been
completely programmed, verification is performed at
both the VDDMIN and VDDMAX of the part. This method,
known as intelligent verification, ensures that each cell
has a good "program margin" (it will read correctly
across the entire operating range) as well as a good
"erase margin" (it will consistently read as '1' after
being erased). The need for a range of supply voltages
is one reason why ICSP specifications have called for
a well-regulated supply voltage with a resolution of
±0.25V.
The majority of Flash devices do not require the
intelligent verification algorithm. The same internal
write timing that makes overprogramming unnecessary
also provides an adequate programming margin for
each cell across the entire operating range. All that is
required for verification is a successful read-back at the
device's operating voltage. The reliability of Flash
technology has eliminated the need for intelligent verification
in those algorithms and the ±0.25V resolution
requirement for VDD. (A well regulated power supply, of
course, is still a good idea.)

2006\12\04@150131 by olin piclist

face picon face
>> And yes, Microchip does recommend variable Vdd for verification of some
>> parts.  They make a point of distinguishing between "prototype" and
>> "production" programmers, with the latter verifying at the Vdd limits.
>
> Can you give an example of a flash PIC where M'chip recommends this?
> I looked through the programming specs once, and only found this
> distinction for PROM based chips.  Admittedly, my search wasn't
> exhaustive.

I was recently looking thru a 16F84 programming spec of all things and I saw
that message.  It was in most programming specs when I started designing
programmers, so look for chips like the 16F877.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\04@164112 by Wouter van Ooijen

face picon face

> OK, I must be looking in the wrong place.  Let's start with the
> 12F675, since I've got its programming spec open right now.  Which
> page is it on?
> (URL: http://ww1.microchip.com/downloads/en/DeviceDoc/41191D.pdf )

you got me. 1.1 explicitly states that only two voltages are needed.

I have been looking through a few programming specifications, but I
can't find the note I wanted to refer you to. I get the feeling
Micrichip has removed the note from the current versions of the
programming specifications?

Wouter van Ooijen

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


2006\12\04@164114 by Wouter van Ooijen

face picon face
> Most Interesting. I found the following in AN910
> "PICmicro® Device Programming: What You Always
> Wanted to Know (But Didn't Know Who to Ask)"
> Page 16
> Quote Follows...

two fragments:

> The majority of Flash devices do not require the
> intelligent verification algorithm.

> The reliability of Flash
> technology has eliminated the need for intelligent verification
> in those algorithms and the ±0.25V resolution
> requirement for VDD. (A well regulated power supply, of
> course, is still a good idea.)

Now what is the truth: the *majority* or *all* flash devices don't need
the variable-Vdd verification?

Wouter van Ooijen

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


2006\12\04@164114 by Wouter van Ooijen

face picon face
> Mark, Microchip identifies programmers that verify at low and
> high VDD
> ranges as "production"
> programmers, and those that only operate at normal VDD as
> "hobbyist".  
> The ICD2 is therefore
> a "hobbyist" level programmer. But I can't remember where I
> saw the spec.

the uChip term is "development", not "hobbyist".

Wouter van Ooijen

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


2006\12\04@171838 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Most Interesting. I found the following in AN910

> two fragments:
>
>> The majority of Flash devices do not require the intelligent
>> verification algorithm.
>
>> The reliability of Flash technology has eliminated the need for
>> intelligent verification in those algorithms and the ±0.25V resolution
>> requirement for VDD. (A well regulated power supply, of course, is
>> still a good idea.)
>
> Now what is the truth: the *majority* or *all* flash devices don't need
> the variable-Vdd verification?

Given that Olin found something to the effect that reading back at
different voltages is required for the 16F84, but apparently in many other
programming specs this is missing, couldn't it be that it's in a few Flash
devices necessary (in those where it is stated in the programming spec),
but in many not? And that this is caused by some detail in the Flash
process that they not necessarily are publicly commenting on?

If so, what is the difference between a development programmer and a
production programmer in terms of programming process? (Of course there are
other criteria: command line operation, features like overlaying serial
numbers over the hex file etc.)

Gerhard

2006\12\04@180339 by Wouter van Ooijen

face picon face
> If so, what is the difference between a development programmer and a
> production programmer in terms of programming process?

none, as uChip defines these terms the only difference is in the Vdd
used for verification.

> (Of
> course there are
> other criteria: command line operation, features like
> overlaying serial
> numbers over the hex file etc.)

no, those are not relevant to the prototype-development distinction as
uChip defines these terms.

Wouter van Ooijen

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


2006\12\04@181848 by Xiaofan Chen

face picon face
On 12/5/06, Alan B. Pearce <.....A.B.PearcespamRemoveMErl.ac.uk> wrote:
> >also there seem to be a lot of dongles for specialist software
> >that rely on having a paralell port.
>
> A lot of these are moving to USB dongles though.

Then there are quite some parallel port based JTAG debugger.
In front of me there is the TI MSP430 FET. There is a  USB version
as well but it is double the price of the parallel version.

The ARM Wiggler is based on parallel port. The USB counter part
is much more expensive.

Regards,
Xiaofan

2006\12\04@230618 by tachyon 1

picon face
How hard would it be to convert the Wisp628 to self power (eg. the
infamous wall-wart)?
Also, do you have more detail on the circuit listed on the WIsp28 page
referring to using a game connector to make an
external board?
I'd like to do something similar, make an external board with a ZIF and a
header to connect to the ICSP from the WISP628.
This board could also have the power connector.
I also thought you could easily put a row of pin headers on either side
of the ZIF, you could use jumper wires from the
ICSP header to setup any pin configuration for any PIC in the socket.

Thanks

 {Original Message removed}

2006\12\04@232938 by tachyon 1

picon face
I don't see the contradiction. I said I couldn't see them disappearing.
Not the they would remain prevalent.
I think it will take at least another hw generation for USB to replace
them in industrial and legacy applications.
The software and hardware requirements in a lot of these applications are
pretty static. One major controls
company only recently transitioned off of OS/2 on the PC controllers. And
in that case it wasn't because it
couldn't do the job, but because of IBM's dropping of support. It's like
that, until legacy ports are dropped by
the applications using them in these applications, people will continue
to use them.


 {Original Message removed}

2006\12\05@001504 by tachyon 1

picon face
Yeah, good point. THere's a lot of them too.

 ----- Original Message -----
 From: "peter green"
 To: "Microcontroller discussion list - Public."
 Subject: RE: [PIC] Why the preoccupation with bus powered
 programmers?
 Date: Mon, 4 Dec 2006 16:45:19 -0000



 > In the real world, serial is almost dead (the occasional 56k
 modem), and I'm
 > surprised that parallel has hung on for so long. Must be a lot of
 old
 > printers out there.
 also there seem to be a lot of dongles for specialist software that
 rely on having a paralell port.

--

Search for products and services at:
http://search.mail.com

2006\12\05@022004 by Wouter van Ooijen

face picon face
>  How hard would it be to convert the Wisp628 to self power (eg. the
> infamous wall-wart)?

wall-wart + 7805 + associated stuff

> Also, do you have more detail on the circuit listed on the WIsp28 page
> referring to using a game connector to make an
> external board?

?

> I'd like to do something similar, make an external board with
> a ZIF and a header to connect to the ICSP from the WISP628.

Note that one ZIF, connected in a fixed way, will not do all PICs.

> I also thought you could easily put a row of pin headers on
> either side
> of the ZIF, you could use jumper wires from the
> ICSP header to setup any pin configuration for any PIC in the socket.

human errors will be included for free ....

uchip has such a board for the ICD2.

why are you so interested in ex-circuit programming? in my experience I
use it very little.

Wouter van Ooijen

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


2006\12\05@025802 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> (Of course there are other criteria: command line operation, features
>> like overlaying serial numbers over the hex file etc.)
>
> no, those are not relevant to the prototype-development distinction as
> uChip defines these terms.

Yes, but I used a more "real-life" distinction, as the Microchip
distinction seems to be not very useful anyway in the light of what they
say about their Flash technology.

Gerhard

2006\12\05@030935 by Xiaofan Chen

face picon face
On 12/5/06, Gerhard Fiedler <RemoveMElistsspamspamBeGoneconnectionbrazil.com> wrote:
> Wouter van Ooijen wrote:
>
> >> (Of course there are other criteria: command line operation, features
> >> like overlaying serial numbers over the hex file etc.)
> >
> > no, those are not relevant to the prototype-development distinction as
> > uChip defines these terms.
>
> Yes, but I used a more "real-life" distinction, as the Microchip
> distinction seems to be not very useful anyway in the light of what they
> say about their Flash technology.

In real life, none of the Microchip tools (PICkit 2, ICD2, Promate II,
Promate III) are good for mass production. PICkit 2 and ICD2 are both
very fragile. Promate II+ICSP adapter sucks based on my previous job
experience. Promate III with integrated ICSP adapter is better. It does
have the features like SQTP and command line version. However I
think it is still a bit fragile since I've seen various reports in Microchip
forum that that it has been destroyed. I used it once for rework (only
hundreds of boards) and it worked fine though.

Regards,
Xiaofan

2006\12\05@032221 by Wouter van Ooijen

face picon face
> >> (Of course there are other criteria: command line
> operation, features
> >> like overlaying serial numbers over the hex file etc.)
> >
> > no, those are not relevant to the prototype-development
> distinction as
> > uChip defines these terms.
>
> Yes, but I used a more "real-life" distinction, as the Microchip
> distinction seems to be not very useful anyway in the light
> of what they
> say about their Flash technology.

But when you make your own distinction what's the relevance? (seriously)
Most people seem to prefer a windows interface, but I prefer a command
line. So which of the two is more prototype and which is more
development?

For your own distinctions you should use a different set of words, maybe
indeed the dreaded 'professional' versus 'hobbyist'.

Wouter van Ooijen

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


2006\12\05@050904 by Tony Smith

picon face
> > In the real world, serial is almost dead (the occasional
> 56k modem),
> > and I'm surprised that parallel has hung on for so long.  Must be a
> > lot of old printers out there.
>
> also there seem to be a lot of dongles for specialist
> software that rely on having a paralell port.


That could be it.  Lots of cu$tom $oftware out there, I guess.  I haven't
seen a dongle in use for years.  Hmmm, 1998 on a custom database box.  I
played with some USB ones about 5 years ago, still got them in a box here
somewhere.

Tony

2006\12\05@060244 by Xiaofan Chen

face picon face
On 12/5/06, Tony Smith <spamBeGoneajsmith@spam@spamspam_OUTrivernet.com.au> wrote:
>
> That could be it.  Lots of cu$tom $oftware out there, I guess.  I haven't
> seen a dongle in use for years.  Hmmm, 1998 on a custom database box.  I
> played with some USB ones about 5 years ago, still got them in a box here
> somewhere.
>

I just got one USB dongle. ;-( It is for IAR ARM C compiler base line edition.

2006\12\05@080629 by olin piclist

face picon face
Wouter van Ooijen wrote:
> But when you make your own distinction what's the relevance? (seriously)
> Most people seem to prefer a windows interface, but I prefer a command
> line. So which of the two is more prototype and which is more
> development?
>
> For your own distinctions you should use a different set of words, maybe
> indeed the dreaded 'professional' versus 'hobbyist'.

I think there are three broad catagories of use for PIC programmers:
production, professional development, and hobbyist.

Production:
 - Does the same job repetitively.
 - Probably embedded in a production test/calibration fixture,
   mechanically making this easy is a plus.
 - Must be automatable from command scripts or another program.  Native
   program rarely used directly by the operator.
 - Operator is relatively unskilled, low paid, often not on site.
 - Reliability is critical.
 - Price is not a high priority.

Professional development:
 - Used by engineer or technician as needed.
 - Should be scriptable so that can be automatically run as part of
   firmware build script if desired.
 - Must be a reliable tool that can be counted on.
 - Should just work, lots of fiddling to get working not tolerated.
 - Price is medium priority, must be good tool to be considered.

Hobbyist:
 - Price is top priority.
 - More inconveniences and unreliability tolerated as price get lower.

Note that these catagories are largely mutually exclusive.  A programmer
that meets the needs of one will do a poor job of meeting the needs of the
others.  Of course this doesn't mean that all hobbyists want hobby-level
tools, but the term "hobbyist" for that particular use pattern of a tool is
still a relevant label that I think most people would understand.

When I first developed the EasyProg, I wasn't expecting to make any profit.
It was a first attempt at directly selling a product, setting up a web site
where things can be bought, and a lot of associated logistics with
production, stocking, shipping, etc, that is a lot more envolved than most
people realize who have never done this before.  In the end I think I did
actually make a little money on the EasyProg in materials, but definitely
not if I were to account for my time.  I did this on the side, so in this
instance I considered my time more like a hobbyist.  This was something I
wanted to do, and kept separate from real business activity.  The point is I
don't think there is a way that a for-profit commercial venture can provide
a hobbyist programmer.  This is why you only see hobby programmers out there
produced by other hobbyists.

At the other end of the spectrum, it is possible to make a profit providing
a production programmer.  The ProProg is my best product, partly because the
low price sensitivity allows for a decent profit margin, and partly because
most customers buy multiple units.  I haven't looked, but I guess the
average is about 4/customer.  I know the maximum is 14.  After all, you
don't want a single broken test fixture to hold up all the production of
your product.  Then engineering probably has a fixture too for support.  And
once a product has been used in a test fixture and a comfort level has been
established, you are going to do the same thing in future fixtures until
they carry you off the production floor feet first.

I don't know yet whether a professional programmer can be a viable
commercial venture, but I'm going to find out shortly with the USBProg.  For
those who keep asking what "real soon now" means this week, I got the first
production sample, identified a few issues (it would be good if they would
actually *read* the BOM), and the first lot is being built right now with
the issues fixed (supposedly).  I'm supposed to receive them in a "couple of
weeks", but I'm figuring early January is more realistic.  I will definitely
post a announcement here when they are available.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\05@132454 by Wouter van Ooijen

face picon face
> The point is I don't think there is a way that a for-profit commercial

> venture can provide a hobbyist programmer.  This is why you only see
hobby
> programmers out there produced by other hobbyists.

I am not sure those boundaries are realy sharp, just as the boundaries
between hobbyist and professional are blurring (even when you apply the
(IMHO only correct) definition 'a professional earns his living doing
that work'). How would you classify my Wisp628? I certainly do make a
profit on these things, and I do earn (half of) a living selling stuff.
Also note that at least one customer bought a bunch (~ 32) Wisp628's to
build a production rig. This was some years ago, the last message from
him was that he had succeeded and the rig was shipped to the far east
for production work.

Wouter van Ooijen

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


2006\12\05@141540 by olin piclist

face picon face
Wouter van Ooijen wrote:
> How would you classify my Wisp628? I certainly do make a
> profit on these things,

I think of the Wisp628 as being on the high end of hobby programmers.  As
you said, the lines are blurry.  Mostly I say that because it has it's own
controller, and is therefore much closer to "just works" operation than
kludge programmers.  I am actually surprised you do make a profit from it.
Have you truly accounted for all your time, not just parts?  If so, then I'm
wrong and it is possible to provide a commercial hobby programmer
profitably.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2006\12\05@145116 by Matt Pobursky

flavicon
face
On Tue, 5 Dec 2006 19:24:51 +0100, Wouter van Ooijen wrote:
>> The point is I don't think there is a way that a for-profit
>> commercial venture can provide a hobbyist programmer. This is why
>> you only see hobby programmers out there produced by other
>> hobbyists.
>>
> I am not sure those boundaries are realy sharp, just as the
> boundaries between hobbyist and professional are blurring (even
> when you apply the (IMHO only correct) definition 'a professional
> earns his living doing that work'). How would you classify my
> Wisp628? I certainly do make a profit on these things, and I do
> earn (half of) a living selling stuff. Also note that at least one
> customer bought a bunch (~ 32) Wisp628's to build a production rig.
> This was some years ago, the last message from him was that he had
> succeeded and the rig was shipped to the far east for production
> work.

I think what Olin means is the math just doesn't work out for a
product like a hobbyist programmer as a viable way to make a living.

As an illustration:

1. Add up all the sales you've made of your Wisp628 programmer since
day 1. Subtract from that all your real costs (parts, assembly labor,
shipping, returns/refunds, taxes, etc.). These are the real cash
amounts you make/lose in building the units. It's a rough gross
margin number.

2. Now add up every single hour you've ever spent on the hardware
design, software design, documentation for the end user and
production, answering production questions, technical support (both
online/forums and direct emails/calls from customers).

3. Now take #1 and divide it by #2. That would give you a rough gross
hourly rate. I suspect most people would be fairly shocked at how low
the hourly rate will be.

If you're going to try and live on this income you also have to
subtract you other taxes, office expenses and other overhead. So you
really won't even have #3, when all is said and done.

For a hobbyist, this isn't a big deal. After all, you're doing it
because you like it and you have more time than money (or you are
independently wealthy and don't care about the money) ;-)

For me, as a professional consultant it is a big deal. My time is
always in short supply so I have to choose carefully where to spend
it. If I can't recoup the up-front engineering and production setup
costs, plus sustain an equivalent hourly rate of at least $50/hr. in
the margin of the ongoing sales then the product isn't even worth
considering for me. I simply will not make enough money and I can
make more doing other products or consulting work at a much higher
rate.

Matt Pobursky
Maximum Performance Systems

2006\12\05@151608 by William Chops Westfield

face picon face

On Dec 5, 2006, at 11:51 AM, Matt Pobursky wrote:
>
> 3. Now take [gross margin] and divide it by [total hours spent.]
> That would give you a rough gross hourly rate. I suspect most
> people would be fairly shocked at how low the hourly rate will be.
>
Yes, but that's typical of MANY small business endeavors, and quite
a few not-so-small businesses that compute their salaries based on
8-hour days even though few of their employees work so little.

BillW

2006\12\05@182128 by Wouter van Ooijen

face picon face
> I think of the Wisp628 as being on the high end of hobby
> programmers.  As
> you said, the lines are blurry.  Mostly I say that because it
> has it's own
> controller, and is therefore much closer to "just works"
> operation than
> kludge programmers.  

I think even the hobby world is shifting away from serial-port-abusing
programmers, both because they prove not-always-reliable and because the
ports are slowly disapperaing. To I think the Wisp628 and the likes will
in time be the low-end (hobby?) programmers.

> I am actually surprised you do make a
> profit from it.
> Have you truly accounted for all your time, not just parts?  

The design has not changed for years, and was almost complete before I
started as a business, so maybe I did not calculate enough time into the
equation.

Wouter van Ooijen

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


2006\12\05@182129 by Wouter van Ooijen

face picon face
> 1. Add up all the sales you've made of your Wisp628 programmer since
> day 1. Subtract from that all your real costs (parts, assembly labor,
> shipping, returns/refunds, taxes, etc.). These are the real cash
> amounts you make/lose in building the units. It's a rough gross
> margin number.

Note: I sell mainly kits, so the per-Wisp628 labour is very low. The
assembled Wisp628's I sell are assembled by Olimex, not by me (OK, I put
in the chips and I solder a crystal).

> 2. Now add up every single hour you've ever spent on the hardware
> design, software design, documentation for the end user and
> production, answering production questions, technical support (both
> online/forums and direct emails/calls from customers).

Yes, there is quite some cost in that. But I sell other things beside
the Wisp628, so a good reputation is worth some time spent.

As I said in another reply: the Wisp628 design is old, based on the yet
older WISP, which I made when it was all 'only' a hobby. So maybe the
equation would be different if I had to do a totally new design now.

Wouter van Ooijen

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

2006\12\05@210552 by Xiaofan Chen

face picon face
On 12/6/06, Wouter van Ooijen <TakeThisOuTwouterspamspamvoti.nl> wrote:
> > 2. Now add up every single hour you've ever spent on the hardware
> > design, software design, documentation for the end user and
> > production, answering production questions, technical support (both
> > online/forums and direct emails/calls from customers).
>
> Yes, there is quite some cost in that. But I sell other things beside
> the Wisp628, so a good reputation is worth some time spent.
>
> As I said in another reply: the Wisp628 design is old, based on the yet
> older WISP, which I made when it was all 'only' a hobby. So maybe the
> equation would be different if I had to do a totally new design now.

I think Rob Hamerling (author of xwisp2, http://www.robh.nl/picsoft.php)
contributes a lot to the software for Wisp628.

Since Microchip has so many PICs and so many programming
specifications, I think software development is kind of the difficult
point for hobbyist programmer. Therefore it might make more
business sense to sell clone programmers. There are plenty
of PS+ and ICD2 clones. I think we will start to see PICkit 2 clones
as well. In fact there are already some kits sold in various Chinese
forums. The low price of the original PICkit 2 might make the profit
margin a bit low for people in the developed world...

Regards,
Xiaofan

2006\12\05@214010 by John Chung

picon face

--- Wouter van Ooijen <wouterEraseMEspamvoti.nl> wrote:

{Quote hidden}

 Wouter and Olin has build a good reputation for PIC
by providing programmers that work. That is a fact.
Even they did not make a fortune with the programmer's
they do have a very good reference for future work on
PIC and such. Future employers may also look at these
reference before awarding them with the contract.


John



____________________________________________________________________________________
Want to start your own business?
Learn how on Yahoo! Small Business.
http://smallbusiness.yahoo.com/r-index



____________________________________________________________________________________
Do you Yahoo!?
Everyone is raving about the all-new Yahoo! Mail beta.
http://new.mail.yahoo.com

2006\12\06@075001 by Gerhard Fiedler

picon face
Olin Lathrop wrote:

{Quote hidden}

IMO that pretty much sums it up. I'd probably add that a professional
development programmer should provide possibilities to integrate into the
most commonly used development environments used by professionals in the
targeted area. That may be covered by "scriptable" or not, depending on the
environments.

An example is MPLAB... it doesn't lend itself well for integrating
scriptable tools. I once tried to make my make scripts MPLAB-compatible (so
that I can run them both from other, "script-friendly", environments and
from MPLAB), but that seemed to take a lot of work, so I abandoned it. Who
wanted to work with MPLAB had to maintain their own MPLAB projects and keep
them in sync with the make files :)

I don't know how common use of MPLAB is among professional PIC developers,
but it's probably significant.

FWIW, it was easier to integrate ICD2 programming into my make scripts
using AutoIt3. (ICD2 is almost not a professional development programmer...
:)

Gerhard

2006\12\09@210957 by Josh Koffman

face picon face
On 12/4/06, Xiaofan Chen <RemoveMExiaofancEraseMEspamspam_OUTgmail.com> wrote:
> My old Dell Inspiron 600M at home and pretty-new Dell Latitude D610 at work
> both have serial/parallel port.

The Latitude D620 comes with only a serial port...no parallel. The IBM
T series come with a parallel port but no serial.

So if I could get work to buy me one of each...

Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
       -Douglas Adams

2006\12\09@220714 by Xiaofan Chen

face picon face
On 12/10/06, Josh Koffman <@spam@joshybearRemoveMEspamEraseMEgmail.com> wrote:
> On 12/4/06, Xiaofan Chen <EraseMExiaofancspam@spam@gmail.com> wrote:
> > My old Dell Inspiron 600M at home and pretty-new Dell Latitude D610 at work
> > both have serial/parallel port.
>
> The Latitude D620 comes with only a serial port...no parallel. The IBM
> T series come with a parallel port but no serial.
>

You can buy the port replicator for D620.

2006\12\10@050205 by Peter Bindels

picon face
On 04/12/06, @spam@tachyon_1spam_OUTspam.....email.com <spamBeGonetachyon_1EraseMEspamemail.com> wrote:
>  Well, I find a lot of Windows hardware drivers to be 'lacking'. In fact
> XP still has issues with USB multi-card readers.
> I don't know how many people I've talked to with brand new PC's with
> built in floppy bay multi-card readers that crash
> or lock-up when a card is inserted.

That's not just Windows to blame, it's mainly bad architecture for
them. The devices themselves are crap too.

I've written a part of a driver for a multi-card reader that plain
crashed on a certain command that was required (by the standard it
claimed to support) to be present and working. The device just
wouldn't give any response anymore. The command? "Get Supported
Commands".

It also crashed on a few other types of commands. I would expect
Windows to do the same, especially since the devices themselves are
that bad.

> Anyway, my primary personal systems are Linux based and I've had no real
> issues with core HW driver stability or
> performance. USB seems especially solid.

It's called "hacking around the known / reported bugs". Microsoft
doesn't nearly get as many "hey Linus, I got this and this device and
it's doing this wrong, here's the diff" messages as Linux does.

2006\12\15@052309 by tachyon 1

picon face
Or for the ThinkPad.
Plus, ThinkPads aren't made by Dell, nor do they tend to explode. ;')

 ----- Original Message -----
 From: "Xiaofan Chen"
 To: "Microcontroller discussion list - Public."
 Subject: Re: [PIC] Why the preoccupation with bus powered
 programmers?
 Date: Sun, 10 Dec 2006 11:07:13 +0800


 On 12/10/06, Josh Koffman wrote:
 > On 12/4/06, Xiaofan Chen wrote:
 > > My old Dell Inspiron 600M at home and pretty-new Dell Latitude
 D610 at work
 > > both have serial/parallel port.
 >
 > The Latitude D620 comes with only a serial port...no parallel. The
 IBM
 > T series come with a parallel port but no serial.
 >

 You can buy the port replicator for D620.

--

Search for products and services at:
http://search.mail.com

2006\12\15@054724 by Xiaofan Chen

face picon face
On 12/15/06, tachyon_1spamBeGonespamemail.com <RemoveMEtachyon_1@spam@spamspamBeGoneemail.com> wrote:
>  Or for the ThinkPad.
> Plus, ThinkPads aren't made by Dell, nor do they tend to explode. ;')
>

Are you sure?

www-307.ibm.com/pc/support/site.wss/document.do?sitestyle=lenovo&lndocid=BATT-LENOVO

2006\12\16@022734 by tachyon 1

picon face
Yeah, I know, but AFAIK only Dell's have actually experienced explosions
and fires.
The other reacalls were pre-emptive by manufacturers using the same
battery.
IMO, it's probably Dell's charging circuit and thermal design. or rather
lack of.


 {Original Message removed}

2006\12\16@024110 by Mark Rages

face picon face
On 12/16/06, .....tachyon_1@spam@spamEraseMEemail.com <.....tachyon_1RemoveMEspamemail.com> wrote:
>  Yeah, I know, but AFAIK only Dell's have actually experienced explosions
> and fires.
> The other reacalls were pre-emptive by manufacturers using the same
> battery.
> IMO, it's probably Dell's charging circuit and thermal design. or rather
> lack of.
>

No, Thinkpads and Macs have also exploded.

Read this account of an exploding Thinkpad:
http://zeniv.linux.org.uk/~telsa/boom/

(Alan Cox is one of the top Linux developers.)

Or see pictures of a burned Powerbook:
http://www.engadget.com/2006/08/06/another-powerbook-violently-explodes/

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2006\12\16@212728 by tachyon 1

picon face
The ThinkPad 600 is more than 5 years old. It is not part of the recall,
and it stated in that page that that was an aftermarket battery.
Basically it is not relevant to the current Sony battery recall issue.

I am still enclined to think that the problem is due to poorly engineered
charging and battery management circuits and software on the part of the
companies that have had this problem. Time will tell I guess. Not that
the Sony product is perfect, it may well be marginal or even outright at
fault.

 {Original Message removed}

More... (looser matching)
- Last day of these posts
- In 2006 , 2007 only
- Today
- New search...