Searching \ for '[PIC]: ICD Schottkey diode' 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: 'ICD Schottkey diode'.

Exact match. Not showing close matches.
PICList Thread
'[PIC]: ICD Schottkey diode'
2002\03\24@142908 by Bob Barr

flavicon
face
Has anyone who has built their own ICD had problems finding the 1N5817
diode used in the Vpp circuit? I can't find a 1N5817 locally and I'm
just a little too impatient to wait for a Digikey shipment. :=)

I'm thinking of substituting a 1N5819 in place of the 1N5817. The max
Vf spec of the 1N5819 (600 mV @ 1A) is a little higher than the 1N5817
(450 mV @ 1A) and I'm not sure if I can get away with this
substitution. I don't know how critical this difference might be.

Any other substitutions anyone can recommend?


Thanks, Bob

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\24@150709 by Dave Dilatush

picon face
Bob Barr wrote...

>Has anyone who has built their own ICD had problems finding the 1N5817
>diode used in the Vpp circuit? I can't find a 1N5817 locally and I'm
>just a little too impatient to wait for a Digikey shipment. :=)
>
>I'm thinking of substituting a 1N5819 in place of the 1N5817. The max
>Vf spec of the 1N5819 (600 mV @ 1A) is a little higher than the 1N5817
>(450 mV @ 1A) and I'm not sure if I can get away with this
>substitution. I don't know how critical this difference might be.

>Any other substitutions anyone can recommend?

I haven't built my own ICD; but in the Microchip docs for my ICD
unit there's only one 1N5817 shown, labeled CR2, and it's
evidently operating at a very low forward current because it has
a 4.7K resistor in series with it.  If this is the diode you're
talking about, I'd say go ahead and substitute the 1N5819; I
don't see how it could make any difference at this low current.

Dave D.

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\24@161035 by Bob Barr

flavicon
face
On Sun, 24 Mar 2002 20:04:45 +0000, Dave Dilatush wrote:

>
>I haven't built my own ICD; but in the Microchip docs for my ICD
>unit there's only one 1N5817 shown, labeled CR2, and it's
>evidently operating at a very low forward current because it has
>a 4.7K resistor in series with it.  If this is the diode you're
>talking about, I'd say go ahead and substitute the 1N5819; I
>don't see how it could make any difference at this low current.
>

Thanks, Dave. That's the one.  I was fairly sure I could do it but I
was a bit concerned that this might be something really dumb to do and
I wouldn't know it. :=)


Regards, Bob

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\24@170658 by Dave Dilatush

picon face
Bob Barr wrote...

>On Sun, 24 Mar 2002 20:04:45 +0000, Dave Dilatush wrote:
>
>>I haven't built my own ICD; but in the Microchip docs for my ICD
>>unit there's only one 1N5817 shown, labeled CR2, and it's
>>evidently operating at a very low forward current because it has
>>a 4.7K resistor in series with it.  If this is the diode you're
>>talking about, I'd say go ahead and substitute the 1N5819; I
>>don't see how it could make any difference at this low current.
>
>Thanks, Dave. That's the one.  I was fairly sure I could do it but I
>was a bit concerned that this might be something really dumb to do and
>I wouldn't know it. :=)

FWIW, I just measured the forward voltage drop across a 1N5819
with 1 mA through it, which is what CR2 would be passing if the
MCLR/VPP line were pulled down to ground; it was 220 mV.

I don't think it even matters what the forward voltage drop is,
come to think of it: when MCLR is low, CR2's forward drop doesn't
make any difference; and when MCLR is allowed to float to Vcc the
diode will have hardly any current going through it so it's not
going to be dropping much.

Looking at that circuit, it seems to me CR2 has only one mission
in life and that is to block any current from Vpp from flowing
back into the Vcc line during programming.  So I suspect you'll
see no difference at all with the 1N5819.

Dave

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\25@010137 by Bob Barr

flavicon
face
On Sun, 24 Mar 2002 22:04:09 +0000, Dave Dilatush wrote:

>
>FWIW, I just measured the forward voltage drop across a 1N5819
>with 1 mA through it, which is what CR2 would be passing if the
>MCLR/VPP line were pulled down to ground; it was 220 mV.
>
>I don't think it even matters what the forward voltage drop is,
>come to think of it: when MCLR is low, CR2's forward drop doesn't
>make any difference; and when MCLR is allowed to float to Vcc the
>diode will have hardly any current going through it so it's not
>going to be dropping much.
>
>Looking at that circuit, it seems to me CR2 has only one mission
>in life and that is to block any current from Vpp from flowing
>back into the Vcc line during programming.  So I suspect you'll
>see no difference at all with the 1N5819.
>

Thanks for the help, Dave. I've now got a working ICD. All of the
voltages appear to be well within the ranges that I would expect.
For some reason, the ICD self-test fails indicating a bad MCLR
voltage. It's indicating that my Vdd is 6 volts and Vpp is 17 volts
even though they measure at 5.02 and 12.98 volts respectively. As long
as I know that it's only the software's readings that are off (and as
long as the ICD's working), I'm not going to concern myself about
this.

Thanks again, Bob

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\03\25@052541 by Alan B. Pearce

face picon face
>For some reason, the ICD self-test fails indicating a bad MCLR
>voltage. It's indicating that my Vdd is 6 volts and Vpp is 17 volts
>even though they measure at 5.02 and 12.98 volts respectively. As long
>as I know that it's only the software's readings that are off (and as
>long as the ICD's working), I'm not going to concern myself about
>this.

This sounds like the LED has the wrong forward voltage, which someone
mentioned in the past, as this is used as the A/D reference voltage, and
hence if the LED voltage is too low, then the readings will be high.

Use a superbright LED instead of a normal one. IIRC the previous post about
the LED mentioned needing a 2 Volt drop across the LED, which is probably
what a superbright one would do, but if you are using an ordinary one then
you will have only 1.6 to 1.7V across it.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\03\25@061815 by Dave Dilatush

picon face
Bob Barr wrote...

>For some reason, the ICD self-test fails indicating a bad MCLR
>voltage. It's indicating that my Vdd is 6 volts and Vpp is 17 volts
>even though they measure at 5.02 and 12.98 volts respectively. As long
>as I know that it's only the software's readings that are off (and as
>long as the ICD's working), I'm not going to concern myself about
>this.

You should be able to bring these readings back into line by
adjusting the ratio of the R1/R2 voltage divider if getting an
LED with the right forward voltage isn't convenient.  Just
substitute a pot for either one of these resistors and adjust it
until ICD reports values that are close to what you're seeing on
the DMM; then measure the pot and plunk in a fixed resistor of
the same value.

I haven't done this, but looking at the circuit it appears you
should be able to do that.

Dave

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\03\27@005219 by Bob Barr

flavicon
face
On Mon, 25 Mar 2002 11:15:30 +0000, Dave Dilatush wrote:

{Quote hidden}

Dave (and Allen too),

Thanks for the suggestions. I'll look into them. It sounds like it
will be easy enough to get the software readings into line with
reality.

It surprises me that a designer would use an LED's forward voltage as
an ADC reference voltage.
I work a bit on simple hardware designs (more at a 'hack-at-it' tech
level, rather than a professional level). Fortunately, I have access
to a real design engineer to keep me out of any serious trouble. :=)
I'm certainly open to correction but I really don't like this
implementation.

Since the designer specified the LED to be used, he obviously knew its
Vf spec but using it in this way just doesn't sound like a good idea
to me.
The only specs I've seen for an LED's Vf are the maximum and 'typical'
values. Is the Vf among the same type of LEDs from different batches
really consistent enough to trust as a reference voltage? How about
differences among different LED manufacturers for the same type of
LED?

Please feel free to flame away if I'm missing something brain-dead
obvious.


Regards, Bob

--
http://www.piclist.com hint: To leave the PICList
spam_OUTpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu


2002\03\27@060506 by Dave Dilatush

picon face
{Quote hidden}

No flame from me, that's for sure.  Far from missing anything,
you've nailed it right on.

Even a crappy voltage reference like an LM385Z-2.5 (+/- 3%, 150
ppm/C tempco) is ten times better in both respects than the Vf of
any diode or LED; and they're only 20 cents apiece in 10000 lots.

Using forward biased diodes as voltage references is shoddy
design.  It isn't just cheap, it's skanky.

Dave

--
http://www.piclist.com hint: To leave the PICList
.....piclist-unsubscribe-requestKILLspamspam@spam@mitvma.mit.edu


2002\03\27@134206 by M. Adam Davis

flavicon
face
>
>
>Using forward biased diodes as voltage references is shoddy
>design.  It isn't just cheap, it's skanky.
>
It's not just cheap, it's /very/ cheap.  You can get a dozen red leds
for the cost of your inexpensive precision reference, and if you're
building the thing and sourcing the parts then it's a completely valid
design.  When you make a million of them, the 19 cents per LED you save
ends up being nearly $200,000.  You can get red leds from hundreds of
suppliers, and lead time is zero.

So from a design perspective I can see it isn't ideal, but from a
manufacturing, profit perspective it is /much/ better than a voltage
reference.  Remember that this is a value added troubleshooting feature,
and (IIRC when I built mine) it only gave the chip the ability to
monitor the voltages used - it isn't part of any feedback loop.  And it
gives you a power on indicator that you might've added anyway, so you
are killing a few birds with a 1 cent part.

-Adam

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestspamKILLspammitvma.mit.edu


2002\03\27@151959 by Dwayne Reid

flavicon
face
At 01:36 PM 3/27/02 -0500, M. Adam Davis wrote:


>>Using forward biased diodes as voltage references is shoddy
>>design.  It isn't just cheap, it's skanky.
>It's not just cheap, it's /very/ cheap.  You can get a dozen red leds
>for the cost of your inexpensive precision reference, and if you're
>building the thing and sourcing the parts then it's a completely valid
>design.  When you make a million of them, the 19 cents per LED you save
>ends up being nearly $200,000.  You can get red leds from hundreds of
>suppliers, and lead time is zero.
>
>So from a design perspective I can see it isn't ideal, but from a
>manufacturing, profit perspective it is /much/ better than a voltage
>reference.  Remember that this is a value added troubleshooting feature,
>and (IIRC when I built mine) it only gave the chip the ability to
>monitor the voltages used - it isn't part of any feedback loop.  And it
>gives you a power on indicator that you might've added anyway, so you
>are killing a few birds with a 1 cent part.

One other thing to add fuel to the fire:  most *red* LEDs have a pretty
darn reasonable tempco - better than some cheap zener diodes!

Sorry to disagree with Dave but an experienced designer working with a
parts purchaser who understands and obeys a 'Do Not Substitute' label can
effectively use LEDs as voltage references.  But I agree on one thing: its
not a good hobby technique just because of the variety of LEDs available.

dwayne


Dwayne Reid   <.....dwaynerKILLspamspam.....planet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 18 years of Engineering Innovation (1984 - 2002)
 .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
    `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspam_OUTspamTakeThisOuTmitvma.mit.edu


2002\03\27@181322 by Dave Dilatush

picon face
Dwayne Reid wrote...

>One other thing to add fuel to the fire:  most *red* LEDs have a pretty
>darn reasonable tempco - better than some cheap zener diodes!

Indeed they do.  On the other hand, I don't use cheap zener
diodes as voltage references either, for a very good reason: they
aren't accurate enough for my purposes.

I understand full well that there are many applications where
high accuracy is not needed; and I would acknowledge that there
are probably some applications where even something as sloppy as
the forward voltage of a diode or LED might do the job
adequately.

>Sorry to disagree with Dave but an experienced designer working with a
>parts purchaser who understands and obeys a 'Do Not Substitute' label can
>effectively use LEDs as voltage references.

If you were to add onto the end of that last sentence the phrase,
"for some purposes, provided he is fully aware of what
performance he can realistically expect from them" then I totally
agree.

But if what you meant by the above is that forward-biased LEDs
are absolutely, positively every bit as good as voltage reference
ICs in all applications, no matter what accuracy, stability or
repeatability you're trying to achieve, I would disagree
vehemently.

>But I agree on one thing: its not a good hobby technique just because
>of the variety of LEDs available.

Agreed.

Dave D.

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestspamspam_OUTmitvma.mit.edu


2002\03\28@031232 by Peter L. Peres

picon face
Using LEDs a voltage reference is pretty common in minimal cost designs in
my experience. There is a temperature coefficient problem but when it does
not matter it's ok. The LEDs from the same batch (as they tend to be used
for this) are pretty close to each other in parameters (better than 5% -
more like 2% usually - note that 2% zeners usually carry a premium price
tag). The Vf must be determined experimentally at a certain current.

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\28@111440 by Dwayne Reid

flavicon
face
At 11:09 PM 3/27/02 +0000, Dave Dilatush wrote:
>Dwayne Reid wrote...
>
> >Sorry to disagree with Dave but an experienced designer working with a
> >parts purchaser who understands and obeys a 'Do Not Substitute' label can
> >effectively use LEDs as voltage references.
>
>If you were to add onto the end of that last sentence the phrase,
>"for some purposes, provided he is fully aware of what
>performance he can realistically expect from them" then I totally
>agree.
>
>But if what you meant by the above is that forward-biased LEDs
>are absolutely, positively every bit as good as voltage reference
>ICs in all applications, no matter what accuracy, stability or
>repeatability you're trying to achieve, I would disagree
>vehemently.

Sometimes I *really* hate the lack of face-to-face nature of mailing lists
- subtle nuances that are ASSUMED by the sender are completely invisible to
the reader.

Dave - you are right.  I did not mean to imply that a LED is a suitable
reference for ALL applications.  I didn't say that!

This thread started out by discussing the use of a LED as the reference for
the a/d convertor in the Microchip ICD - great accuracy is NOT
required!  The designer of the ICD needed a low cost, repeatable reference
- and could specify the what parts the purchaser gets.  I don't think that
they considered that hobbyists would be copying their design - it would not
have been part of the design spec.  But several list members have decried
the design as sloppy and unprofessional and I disagree.  The ICD works as
it is supposed to.  The fact that LEDs with different part numbers have
different forward voltages is just not an issue so long as LEDs with the
SAME part number have the same Vf.

The problem is that someone copying the ICD sees a LED and just assumes
that all LEDs behave the same.  Then, when the ICD doesn't work because
they grabbed any old LED from their junk box, they accuse the designer of
incompetence.  I'm trying to set the record straight: LEDs make acceptable
references for some applications.

dwayne


Dwayne Reid   <@spam@dwaynerKILLspamspamplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 18 years of Engineering Innovation (1984 - 2002)
 .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
    `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\28@170724 by Dave Dilatush

picon face
Dwayne Reid wrote...

>Sometimes I *really* hate the lack of face-to-face nature of mailing lists
>- subtle nuances that are ASSUMED by the sender are completely invisible to
>the reader.

Amen to that; see below.

>The problem is that someone copying the ICD sees a LED and just assumes
>that all LEDs behave the same.  Then, when the ICD doesn't work because
>they grabbed any old LED from their junk box, they accuse the designer of
>incompetence.

I wasn't saying that the ICD's designer was being incompetent; I
was saying that Microchip was being cheap.

That ICD, in my opinion, is one VERY clever device and has been
worth every one of the 15,900 pennies I paid for it--except, that
is, for the 19 pennies that Microchip could have spent giving me
a decent voltage reference.  :)

>I'm trying to set the record straight: LEDs make acceptable
>references for some applications.

Agreed.

Dave

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...