Searching \ for '[PIC] EasyProg with dsPICDEM 28 pin board to 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/microchip/devprogs.htm?key=icd
Search entire site for: 'EasyProg with dsPICDEM 28 pin board to progr'.

Exact match. Not showing close matches.
PICList Thread
'[PIC] EasyProg with dsPICDEM 28 pin board to progr'
2005\12\02@140334 by John Nall

picon face
I'm trying to use Olin's EasyProg programmer to program a ds30F3013
chip.  The EasyProg has PRG16.HEX firmware downloaded into the 16F648A
chip.  The 3013 is plugged into a socket on a dsPICDEM 28-pin board,
which I previously used with the ICD2 as the programmer.  The ICD2
attached to the demo board with an RJ12 cable.  So I plugged  the same
RJ12 cable into the EasyProg instead.  The two boards (EasyProg and demo
board) are powered separately.  When I type "pic_prog" the EasyProg
reports back "Firmware is type EasyProg version 16 by Embed Inc.  Unable
to  read the device ID from the target chip."

Perhaps I just can't do this, but since the EasyProg RJ12 socket is
supposed to  have the same wiring configuration as the ICD2, I  thought
it would work.  Any thoughts will be appreciated.

John

2005\12\02@150529 by olin piclist

face picon face
John Nall wrote:
> I'm trying to use Olin's EasyProg programmer to program a ds30F3013
> chip.  The EasyProg has PRG16.HEX firmware downloaded into the 16F648A
> chip.  The 3013 is plugged into a socket on a dsPICDEM 28-pin board,
> which I previously used with the ICD2 as the programmer.  The ICD2
> attached to the demo board with an RJ12 cable.  So I plugged  the same
> RJ12 cable into the EasyProg instead.  The two boards (EasyProg and demo
> board) are powered separately.

I hope the demo board is not powered at all when you try that, else
"interesting" things could happen.

> Perhaps I just can't do this, but since the EasyProg RJ12 socket is
> supposed to  have the same wiring configuration as the ICD2, I  thought
> it would work.

It does in lots of cases, but there is no way to account for all the things
a target circuit can do to the programming lines.  My ICSP writeup at
http://www.embedinc.com/picprg/icsp.htm might give you some insight.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\02@172247 by John Nall

picon face
Olin Lathrop wrote:
> > I hope the demo board is not powered at all when you try that, else
> "interesting" things could happen.
Well, it *was* powered when I did that, since it is powered when I use
the ICD2 as a programmer rather than the EasyProg.  Since there is no
connection between  the two boards other than through  the RJ12 cable,
why would it be bad?  (I'm not arguing the point, since you generally
know what you are talking about -- just wondering why it is bad.)
> > It does in lots of cases, but there is no way to account for all the
> things
> a target circuit can do to the programming lines.  My ICSP writeup at
> http://www.embedinc.com/picprg/icsp.htm might give you some insight.
OK, I'll take a look at it.  PDQ, as a matter of fact.  The dsPICDEMO is
pretty innocuous.  If I had made a target board rather than just using
what I already had available, I'd probably have wired it pretty much the
same way.  There is nothing there other than a 28-pin socket and support
(power, crystal, etc.) for the socket, plus a couple of LED's.  One to
show that it is powered, and the other that is attached to an I/O pin
for demonstration (flash an LED).  However, I might have a different
opinion after reading your ICSP writeup.

(Of course, I am assuming that I do not have to specifically tell the
EasyProg that it should program via the RF12 line rather than going to
the ZIF socket -- I just figured they  ran in parallel).

John

2005\12\02@181005 by olin piclist

face picon face
John Nall wrote:
> Well, it *was* powered when I did that,
> since it is powered when I use
> the ICD2 as a programmer rather than the EasyProg.

The ICD2 is designed to run with "target power" in some cases.  The EasyProg
is not.

> Since there is no
> connection between  the two boards other than through  the RJ12 cable,
> why would it be bad?

Because the EasyProg tries to control the Vdd line.  If your board is also
trying to control it, something might go poof or just not work right.

> (Of course, I am assuming that I do not have to specifically tell the
> EasyProg that it should program via the RF12 line rather than going to
> the ZIF socket -- I just figured they  ran in parallel).

Yes.  Since they are in parallel, you have to make sure nothing is in the
ZIF socket while using the ICSP port.  If not, other "interesting" things
could happen, some of which would be bad.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\02@182332 by John Nall

picon face
Olin Lathrop wrote:
> > The ICD2 is designed to run with "target power" in some cases.  The EasyProg
> is not.
>  
OK.  Good enough.   I have kind of gotten in the mindset lately of
giving every board a separate power supply. :-)

> > Because the EasyProg tries to control the Vdd line.  If your board is also
> trying to control it, something might go poof or just not work right.
>  
I don't think the board is trying to control that line.  (You as a
piclist member probably do not have a particular interest, but if you
put on one of your other hats -- that of  the EasyProg developer -- you
might want to take a look at the schematic of the board.  I'm attaching
it as an attachment, just in case).
> > Yes.  Since they are in parallel, you have to make sure nothing is in the
> ZIF socket while using the ICSP port.  
Yeah, I knew that.  Nothing was in the ZIF socket.  Note that this is
not a big deal -- I'll just continue to use the ICD2 to program 30F
chips.  But I'd really like to see an lower-priced alternative to the
ICD2, and I think the EasyProg fits that bill.  I have recommended it to
a couple of people who know that I have been working with the 30F
family, and I pretty much assumed that everything worked as advertised
without ever actually using it myself (on a 30F chip -- I've used it on
other chips, of course, and it works fine.  But that is with the ZIF
socket).

John

2005\12\02@183946 by olin piclist

face picon face
John Nall wrote:
> I don't think the board is trying to control that line.

But that's exactly what the job of the power supply is.  The EasyProg is
trying to make Vdd one voltage and your power supply another voltage.  At
best nothing blows up.  In any case programming is unlikely to work.

> I'll just continue to use the ICD2 to program 30F chips.
> But I'd really like to see an lower-priced alternative to the
> ICD2, and I think the EasyProg fits that bill.

So have you tried it without the target board powered and it still didn't
work?  Keep in mind that the EasyProg can only source a few 10s of mA on
Vdd.  If the target works with the ICD2 and Vdd doesn't require too much
current and doesn't have too much capacitance, I'd expect the EasyProg to
work with the same target circuit.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\02@191601 by Xiaofan Chen

face picon face
On 12/3/05, Olin Lathrop <spam_OUTolin_piclistTakeThisOuTspamembedinc.com> wrote:
> John Nall wrote:
> > I don't think the board is trying to control that line.
>
> But that's exactly what the job of the power supply is.  The EasyProg is
> trying to make Vdd one voltage and your power supply another voltage.  At
> best nothing blows up.  In any case programming is unlikely to work.
>

Is it possible to add a command line switch to tell that EasyProg not to
vary voltage and/or not to control the voltage? I know this may require
a firmware update as well.

I think it is good to have a power supply and variable Vdd. This is one
aspect where EasyProg better than Wisp628 (the other is the dsPIC/PIC12
base line support). But it may be good to have the option to switch this off.

Regards,
Xiaofan

2005\12\02@200625 by olin piclist

face picon face
Xiaofan Chen wrote:
> Is it possible to add a command line switch to tell that EasyProg not to
> vary voltage and/or not to control the voltage? I know this may require
> a firmware update as well.
>
> I think it is good to have a power supply and variable Vdd. This is one
> aspect where EasyProg better than Wisp628 (the other is the dsPIC/PIC12
> base line support). But it may be good to have the option to switch
> this off.

For some PICs this would be possible, although the EasyProg will always put
a load on Vdd since it uses an active pullup and passive pulldown.  Some
PICs require Vpp before Vdd to enter programming mode, so this won't work
for those.  If you don't want the EasyProg to control Vdd and the PIC can
accept Vdd before Vpp, then you can just not hook up the Vdd line.  However
the EasyProg will still do two verify passes although both will be at the
target Vdd voltage.  I suppose a command line option could support this
better.

I'm still not sure though that John has actually tried programming his dsPIC
with the target circuit power off.  I expect this can be made to work, but
we need more details of what happens when he tries it.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\02@200913 by John Nall

picon face
Olin Lathrop wrote:
> > So have you tried it without the target board powered and it still didn't
> work?
I tried it without the target board powered and it still doesn't work
(exact same error message).  And, of course, you might be wondering if
something got screwed up.  I pulled the 3013 chip out and checked it --
before trying to program it, it had some firmware in it that blinked an
LED.  That program still works.  So it appears that nothing got harmed
with the target board being powered.)

John

2005\12\02@202220 by olin piclist

face picon face
John Nall wrote:
> I tried it without the target board powered and it still doesn't work
> (exact same error message).  And, of course, you might be wondering if
> something got screwed up.  I pulled the 3013 chip out and checked it --
> before trying to program it, it had some firmware in it that blinked an
> LED.  That program still works.  So it appears that nothing got harmed
> with the target board being powered.)

It wasn't the target chip I was worried about.

You can use the PIC_CTRL program to see if the EasyProg can drive the lines
properly.

 pic_ctrl -vdd 5

should make Vdd go to 5V, and -VDD 0 should make it go to 0V.  -VPP works
similarly except that only 13 and 0 are legal values.  You can use the other
options to make sure the clock and data lines can be wiggled too.  You may
want to short the debug jumper else the PIC_CTRL setting may only last 5
seconds before everything goes back to off.

If all the voltages can be wiggled from the programmer and the levels look
right, then try putting 47pF on the PGD and PGC lines as close to the target
PIC as possible.  This often makes a difference, especially with dsPICs.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\02@203926 by John Nall

picon face
Olin Lathrop wrote:
> >You can use the PIC_CTRL program to see if the EasyProg can drive the
> lines
> properly.
>
>  pic_ctrl -vdd 5
>
> should make Vdd go to 5V, and -VDD 0 should make it go to 0V.  -VPP works
> similarly except that only 13 and 0 are legal values.  You can use the
> other
> options to make sure the clock and data lines can be wiggled too.  You
> may
> want to short the debug jumper else the PIC_CTRL setting may only last 5
> seconds before everything goes back to off.
>
> If all the voltages can be wiggled from the programmer and the levels
> look
> right, then try putting 47pF on the PGD and PGC lines as close to the
> target
> PIC as possible.  This often makes a difference, especially with dsPICs.
.
OK.  I'll report back tomorrow on the results  Right now, my child bride
is making demands on my time, because her pottery kiln is not doing what
it is supposed to do.  It is tripping a 60 amp breaker, at 220 volts.  
Takes a lot of power to turn that silicon into glass, I guess!  So I
have to remove my PIC hat and put on my electrical-contractor hat. Talk
about contrasts in thinking . . . (5 VDC and a few milli-amps, and then
suddenly switching to coping with 230 VDC and 60 amps!!)

John


2005\12\03@171300 by John Nall

picon face
Olin Lathrop wrote:
> > You can use the PIC_CTRL program to see if the EasyProg can drive
> the lines
> properly.
>
>  pic_ctrl -vdd 5  [etc]

Well, I have about decided, after doing the tests you suggested and
reading the stuff on ICSP with the EasyProg that there is just a basic
incompatibility between the demo board  and  the EasyProg.  
Specifically, I  think that it is because the demo board requires its
own power supply, and of course the EasyProg won't go along with that,
although the ICD2 will.

OK -- no big problem.  I've got a 28-pin ZIF socket around here
somewhere, and will make up something on a piece of perf board and just
run the RJ12 cable from the EasyProg to that instead of trying to use  
the demo board (for the purpose of using the EasyProg to program a 3013
chip, I mean).

Just running the 5 lines from the RJ12 cable to the appropriate pins on
the 3013 would work without anything further needed, wouldn't it?  
(MCLR, Vdd, Vss, PGC and PGD).  Not really ICSP, I suppose, but just
more of an auxiliary ZIF socket to the one that comes with the
EasyProg.  (I'm not ignoring your suggestion of a couple of 47 pf
capacitors -- I'll stick those on the perf board).

2005\12\03@175138 by olin piclist

face picon face
John Nall wrote:
> Olin Lathrop wrote:
>>> You can use the PIC_CTRL program to see if the EasyProg can drive
>> the lines
>> properly.
>>
>>  pic_ctrl -vdd 5  [etc]
>
> Well, I have about decided, after doing the tests you suggested and
> reading the stuff on ICSP with the EasyProg that there is just a basic
> incompatibility between the demo board  and  the EasyProg.
> Specifically, I  think that it is because the demo board requires its
> own power supply, and of course the EasyProg won't go along with that,

There is no need for that to be a basic incompatibility.  I don't see how
the demo board "requires it's own power supply" when being programmed.  Why
does anything need to be on except the PIC?  How much Vdd current does it
draw when the power supply is off?  You never said what the results of your
PIC_CTRL tests were, so I can't really say much more.

> Just running the 5 lines from the RJ12 cable to the appropriate pins
> on the 3013 would work without anything further needed, wouldn't it?

Most likely not, especially if using the standard Microchip ICD2 cable.  If
this is a stand along ZIF socket, put the 100pF caps on PGC and PGD right
where they enter the board, and 100 ohms in series with PGD from there to
the PIC pin.  Then put a 100nF or 1uF ceramic bypass cap on Vdd to ground as
close as possible to the PIC.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\03@180412 by John Nall

picon face
Olin Lathrop wrote:
> > There is no need for that to be a basic incompatibility.  I don't see how
> the demo board "requires it's own power supply" when being programmed.  Why
> does anything need to be on except the PIC?  How much Vdd current does it
> draw when the power supply is off?  You never said what the results of your
> PIC_CTRL tests were, so I can't really say much more.
>  
I was trying to get you off the hook, in case this was really not of
much interest.  :-)  But if you are determined to get to the root of  
the problem, then that is fine  -- I would love to do that.  I'll get
the numbers and report back.


2005\12\04@130227 by John Nall

picon face
Olin Lathrop wrote:
> > There is no need for that to be a basic incompatibility.  I don't see how
> the demo board "requires it's own power supply" when being programmed.  Why
> does anything need to be on except the PIC?  How much Vdd current does it
> draw when the power supply is off?  You never said what the results of your
> PIC_CTRL tests were, so I can't really say much more.
>  
OK.  I'm sending the schematic of that board as an attachment in a
separate off-list message.  I did all the pic_ctrl tests, and everything
worked fine.  Each pin of the ds30F chip had the correct voltage.  So
(a) the proper pins are connected, and (b) the EasyProg is doing what it
is supposed to do.  However, the original error is still the same.

I then rigged up a breadboard circuit with nothing except the ds 30F
chip, and went directly from the RJ12 plug to the pins.  The EasyProg
read that one OK.  I first did a pic_prog on it, and it identified it as
a 2010 chip.  Then did a pic_read on it, and it read it ok.

So there is just an incompatibility with that dsPICDEMO board.  If you
want to continue to troubleshoot it and figure out what the problem is,
I'll be more than happy to do whatever  tests you suggest.  

John

2005\12\04@150555 by olin piclist

face picon face
John Nall wrote:
> I then rigged up a breadboard circuit with nothing except the ds 30F
> chip, and went directly from the RJ12 plug to the pins.  The EasyProg
> read that one OK.  I first did a pic_prog on it, and it identified it as
> a 2010 chip.  Then did a pic_read on it, and it read it ok.
>
> So there is just an incompatibility with that dsPICDEMO board.  If you
> want to continue to troubleshoot it and figure out what the problem is,
> I'll be more than happy to do whatever  tests you suggest.

I looked at the schematic, and there isn't much there.  I think the EasyProg
should be able to work with that board, although there are a few issues.

First, the jumpers J2 and J3 must be to the RF2/RF3 position.

Also there is no filtering on PGD.  I keep saying this, but I don't get any
feedback from you that you've done anything about it.  Put 47pF caps to
ground on the PGC and PGD lines close to J1.  I would solder them right to
the bottom of the connector.

If that doesn't allow it to work, then it may be that the 5V regulator is
being driven backwards when Vdd is applied to the PIC and the 220uF cap
causes too large a time constant for Vdd to decay when the EasyProg thinks
its set it to zero.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\04@152055 by John Nall

picon face
Olin Lathrop wrote:
> > First, the jumpers J2 and J3 must be to the RF2/RF3 position.
Right.  I had them set that way.
>
> > Also there is no filtering on PGD.  I keep saying this, but I don't
> get any
> feedback from you that you've done anything about it.  Put 47pF caps to
> ground on the PGC and PGD lines close to J1.  I would solder them
> right to
> the bottom of the connector.
You haven't had any feedback from me on that, correct.  That is because
I have not yet done it, being kind of hesitant to solder stuff onto that
board until you looked at the schematic.  After all, even though I saw
no reason why it should not work, there was always the chance that you
would immediately spot something that gave it a thumbs down.  But now
that you have looked at it, and see no reason why it should not work,
then I guess I'll take that next step, of tacking on the caps.

2005\12\04@155621 by olin piclist

face picon face
John Nall wrote:
> But now
> that you have looked at it, and see no reason why it should not work,
> then I guess I'll take that next step, of tacking on the caps.

I find that dsPIC don't program without the caps most of the time.

******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\04@213005 by Chen Xiao Fan

face
flavicon
face

Olin wrote:
> If you don't want the EasyProg to control Vdd and the PIC can
> accept Vdd before Vpp, then you can just not hook up the Vdd line.  
> However the EasyProg will still do two verify passes although both
> will be at the target Vdd voltage.  I suppose a command line option
> could support this better.

That would be a good option to add.
1) good for new flash PICs which do not require variable Vdd verify.
Actually most of them do not require this. Still variable Vdd is
necessary for the 18J device and PIC24/dsPIC33 device running at 3.3V
or less. EasyProg and PICkit 2 are both future-proof.
2) good for EasyISP as well but is has not been updated for long time.

Olin wrote:
>I find that dsPIC don't program without the caps most of the time.

Yes that is at the top of Microchip Forum. Olin is the
one who proposed the solution. I think this may also help
with some other 18F chips besides dsPICs. PGC/PGD crosstalk
seems to be a big problem for ICD2.

Microchip perhaps designed the demo board before they find
out the problem. ;-) The fact John's ICD2 works with the
board does not guarantee other people's ICD2 will work.

Regards,
Xiaofan

2005\12\05@135413 by John Nall

picon face
Olin Lathrop wrote:
> > Also there is no filtering on PGD.  I keep saying this, but I don't
> get any
> feedback from you that you've done anything about it.  Put 47pF caps to
> ground on the PGC and PGD lines close to J1.  I would solder them
> right to
> the bottom of the connector.

It is going to be a while before I try the 47 pF caps, because I don't
have any and the local Radio Shack doesn't have any, either.  I'll have
to stick some in the next order to Mouser, which is just about due to go
in anyway.  But it will take a few days for stuff to get delivered.  So
don't think that I've forgotten to try this -- you WILL get feedback!!  
:-)   (In the meantime I'm going to try a couple of 100 pF caps, which I
do have on hand, and see what that does).

John

2005\12\05@140312 by Robert Young

picon face
>
> It is going to be a while before I try the 47 pF caps,
> because I don't
> have any and the local Radio Shack doesn't have any, either.  
> I'll have
> to stick some in the next order to Mouser, which is just
> about due to go
> in anyway.  But it will take a few days for stuff to get
> delivered.  So
> don't think that I've forgotten to try this -- you WILL get
> feedback!!  
> :-)   (In the meantime I'm going to try a couple of 100 pF
> caps, which I
> do have on hand, and see what that does).
>
> John
>

Use 4 caps, 2 sets of 2 paralleled to get close to the 47pF

Rob

2005\12\05@144931 by Dave Wheeler

flavicon
face
Robert Young wrote:

{Quote hidden}

I think Rob meant 2 sets of 2 in series. Caps in parallel increases the
overall value (or it used too !!)

Dave


2005\12\05@152819 by Robert Young

picon face

> >
> >Use 4 caps, 2 sets of 2 paralleled to get close to the 47pF
> >
> >Rob
> >  
> >
> I think Rob meant 2 sets of 2 in series. Caps in parallel
> increases the
> overall value (or it used too !!)
>
> Dave

Duh, yes that is what Rob meant!  Time for my nap.

Rob

2005\12\05@163135 by John Nall

picon face
Robert Young wrote:
> > Use 4 caps, 2 sets of 2 paralleled to get close to the 47pF
>  

Right...now if only I had four of them, instead of just two.  :-(


2005\12\05@165844 by Wouter van Ooijen

face picon face
> Right...now if only I had four of them, instead of just two.  :-(

just use the two. determining the optimal value for such caps is a black
art, I would be surprised if the value that was quoted to you was within
a factor 4 of the actual optimal value. Luckily the working range will
liklely be very wide: the caps must be high enough to kill small
cross-talk signals and damp reflections, yet not so high that the timing
of the programming signals is affected.

Wouter van Ooijen

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


2005\12\05@180651 by olin piclist

face picon face
John Nall wrote:
> (In the meantime I'm going to try a couple
> of 100 pF caps, which I do have on hand, and see what that does).

I didn't mean 47pF to be an exact value.  100pF is fine, just don't go any
bigger.  Adding 100pF caps will definitely tell you whether the problem is
PGC/PGD crosstalk or something else.  With 100pF on each line, I think
crosstalk can be eliminated as a cause of failure.

Unfortunately if the caps don't fix the problem, the remaining culprit I can
think of is the 220uF cap C9 which might act like it's on Vdd depending on
how the regulator VR1 works.  The only way to get around that without
changing the hardware is to change the Vdd on/off wait times in the EasyProg
firmware.  They are currently both set to 25mS.  The source code is
available, so you could make your own version with longer wait times.

Unfortunately ICSP is not something every circuit it suited for with every
programmer.  These things are best considered up front when the circuit is
designed.  Of course in this case you didn't have that option since the
circuit came from elsewhere.  If I had designed this, I would probably have
stuck a diode between C8 and C9 and not put a full wave bridge on the front,
just a blocking diode.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\05@182510 by olin piclist

face picon face
Robert Young wrote:
> Use 4 caps, 2 sets of 2 paralleled to get close to the 47pF

It's not that critical.  I originally used 22pF for that, but since then
I've found cases where 47pF was marginal.  On my ICSP web page
(http://www.embedinc.com/picprg/icsp.htm) I no recommend 100pF.  I wouldn't
go any bigger though since then you've got too large a time constant
compared to the 2Kohm PGC and PGD impedence of the EasyProg.  100pF sounds
rather heavyweight, so I said 47pF because he was adding this to an existing
board for which I didn't have schematics at the time.  In other words, I
somewhat picked 47pF out of the air.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\06@085117 by John Nall

picon face
Wouter van Ooijen wrote:
> > . . . just use the two. determining the optimal value for such caps is a black
> art, I would be surprised if the value that was quoted to you was within
> a factor 4 of the actual optimal value. Luckily the working range will
> liklely be very wide: the caps must be high enough to kill small
> cross-talk signals and damp reflections, yet not so high that the timing
> of the programming signals is affected.
>  
.
If it were anyone but Olin giving the advice,  then I would just have
assumed that 100 pf would probably be OK.  But since he was so precise
(47 pf -- not 50, not 45, but 47!) I figured perhaps he had run involved
calculations on a Cray and determined the specific value.  :-)   Of
course, since then he has confirmed that the 100 pf can be used as well,
so long as I do not go higher than that.

2005\12\06@105242 by John Nall

picon face
Olin Lathrop wrote:

>> Unfortunately if the caps don't fix the problem, the remaining culprit I can
>think of is the 220uF cap C9 which might act like it's on Vdd depending on
>how the regulator VR1 works.  The only way to get around that without
>changing the hardware is to change the Vdd on/off wait times in the EasyProg
>firmware.  They are currently both set to 25mS.  The source code is
>available, so you could make your own version with longer wait times.
>  
>
Well,  the caps didn't fix the problem.  No change whatsoever.  So much
for  that.  I'm pulling the plug on the project.  However, I really do
not need to get around it, since just running the EasyProg to a
breadboarded socket works fine.  For that matter, you might consider
adding an "accessory board" to your product line, for people who want to
use the EasyProg with a dsPIC (and presumably also for the 24F line),
since I don't know of another competitive board right now  given that
the 28-pin dsPICDEMO board won't work.  It wouldn't  take much -- just a
circuit board, a 28-pin ZIF socket, an RJ12 socket, a diode, couple of
capacitors and perhaps a couple of bells and whistles (some option
jumpers and a couple of LED's).  You could probably sell it for $9.95
and still make a good profit.  :-)  The dsPICDEMO board is $79.

Thanks for your help and suggestions, as always.

John


2005\12\06@105937 by olin piclist

face picon face
John Nall wrote:
> But since he was so precise (47 pf -- not 50, not 45, but 47!)

In the world of capacitors, 47pF is a round number.  Go try and find a 50pF
cap and you'll see what I mean.  I was really thinking "about half the
maximum of 100pF" and rounded to the nearest common value.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\06@112152 by olin piclist

face picon face
John Nall wrote:
> For that matter, you might consider
> adding an "accessory board" to your product line, for people who want to
> use the EasyProg with a dsPIC (and presumably also for the 24F line),

I have considered doing a board with a RJ-12 socket and PIC sockets for
various pinouts.  Every time I thought about it though, the numbers didn't
work out.  To do it right with ZIF sockets would make it too expensive, and
I suspect volumes would be very low to begin with.

> It wouldn't  take much -- just a
> circuit board, a 28-pin ZIF socket, an RJ12 socket, a diode, couple of
> capacitors and perhaps a couple of bells and whistles (some option
> jumpers and a couple of LED's).  You could probably sell it for $9.95
> and still make a good profit.  :-)

A ZIF socket alone costs about that.  Even without any sockets (so someone
could add their own whatever flavor they liked), it would probably take at
least $5 to produce in quantity of 100 to 200 at a time.  With the handling
for each one and getting the design time paid for would make it more than
most people are willing to pay for it.  That's one of the problems selling
to hobbyists.  They're cheap.  Because of that many potential products don't
exist because there's no room to make money.

> The dsPICDEMO board is $79.

I am somewhat seriously considering a QuickProto for the 28 pin dsPICs which
would sell for considerably less than $79.  However volumes for the existing
QuickProto-01 are lower than what I expected, so this isn't a high priority
for me now.

I am working on a different new product though.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\06@114245 by John Nall

picon face
Olin Lathrop wrote:

> >  That's one of the problems selling
> to hobbyists.  They're cheap.


Actually, we kind of prefer the term "thrifty."  :-)  But I hear what
you are saying.  Oh well.  As I said earlier, running the RJ12 lines
from the EasyProg to a basic breadboard setup is very easy and works
fine.  I'm going to go ahead and set it up as a permanent accessory,
using some strip board.

> > I am working on a different new product though.


Send me one, and I'll debug it for you.  Murphy's Law seems to apply
more strictly to me than to other people.  :-(


2005\12\06@115409 by Peter van Hoof

face picon face
--- Olin Lathrop <.....olin_piclistKILLspamspam@spam@embedinc.com> wrote:

> John Nall wrote:
> > But since he was so precise (47 pf -- not 50, not 45, but 47!)
>
> In the world of capacitors, 47pF is a round number.  Go try and find
> a 50pF
> cap and you'll see what I mean.  I was really thinking "about half
> the
> maximum of 100pF" and rounded to the nearest common value.


Capacitors follow E6, E12, E24 ... etc series just like resistors
see dirs.org/wiki-article-tab.cfm/preferred_values
or www.logwell.com/tech/components/resistor_values.html
For an explanation


Peter van Hoof

2005\12\06@121102 by John Nall

picon face
Peter van Hoof wrote:

>> Capacitors follow E6, E12, E24 ... etc series just like resistors
>see dirs.org/wiki-article-tab.cfm/preferred_value
>  
>
>  Thanks for pointing me to the wiki-article -- that provided a  good
explanation.  Now I know.  (I wondered before how they arrived at all
those weird values, but just always kind of assumed that pitchers of
beer and a dart board were involved.  :-)

2005\12\06@121824 by William Chops Westfield

face picon face

On Dec 6, 2005, at 8:23 AM, Olin Lathrop wrote:

>> For that matter, you might consider
>> adding an "accessory board" to your product line, for people who want
>> to
>> use the EasyProg with a dsPIC (and presumably also for the 24F line),
>
> I have considered doing a board with a RJ-12 socket and PIC sockets for
> various pinouts.  Every time I thought about it though, the numbers
> didn't
> work out.

Microchip sells "AC162049 - Universal Programming Module for ICD2",
which would presumably work, being little more than an RJ12, a 40pin
ZIF, and a bunch of jumpers (sorta kludgey, actually.)  They want
$40 for it, apparently (i thought I remembered it being cheaper when
I bought it, but perhaps not.)

Something better ought to be within the realm of DIY PCBs...

BillW

2005\12\06@164845 by Wouter van Ooijen

face picon face
> If it were anyone but Olin giving the advice,  then I would just have
> assumed that 100 pf would probably be OK.  But since he was
> so precise
> (47 pf -- not 50, not 45, but 47!)

47 is one of the magic values that are used for components, like 22, 33,
39, etc. If he had said 50 or 45 I would have suspected a very
deliberate choice.

Wouter van Ooijen

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



2005\12\06@170955 by Peter

picon face

>>> But since he was so precise (47 pf -- not 50, not 45, but 47!)

47 +/- about 5% is somewhere between 45 and 50 (not included), even more
so for +/- 10% ;-)

Peter

2005\12\07@164429 by John Nall

picon face
Olin Lathrop wrote:

> > I have considered doing a board with a RJ-12 socket and PIC sockets for
> various pinouts.  Every time I thought about it though, the numbers
> didn't
> work out.

Even though this particular horse is dead, I'm going to beat him a
little bit more.  :-)  Then consider putting a little blurb with the
EasyProg documentation on how to make an accessory board to program a
30F chip.  It doesn't  take much, but it might not be obvious to someone
who is new at this.  You could sketch it out, the same way you did the
drawing of the RJ12 cable on the last page of your ICSP blurb.  In fact,
that pinout sketch, with the corresponding PIC pins for a 30F chip
added, and a little sketch showing the 47 pf caps (I used 100 pf and
they worked fine, by the way) would probably be sufficient.  A little
time, no expense, and the problem is solved for most people without ever
having to go further.

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