Finding a short on a prototype board

[Charlie E.]

Yep! have to lift the pins (Veerrry Carefully...) and then air-wire
them to the correct places. Well, I have ten boards to try and get it
right on... ;-)

Charlie

Did the first board, but it won't program. I think in all that high
current testing I probably baked the processor. On to board 2...

Charlie

 

[Don Lancaster]

If you're a hobbyist, and want to get away from the teevee set, it
might make sense to build your own power supply. But for $70 or some
such, a metered, finished bench supply doesn't make sense to build.

Ebay has lots of power supplies too...

http://www.ebay.com/itm/Precision-V...986?pt=LH_DefaultDomain_0&hash=item256f8b216a

I've been real impressed with this...

http://www.mpja.com/0-30V-0-3A-Mini-Bench-Supply/productinfo/9615+PS/


So much so that I threw away a bunch of Heathkits of comparable rating.

But you can still get them via < James Knight <[email protected]> >

--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email: [email protected]

Please visit my GURU's LAIR web site at http://www.tinaja.com

 

[Jamie]

How on earth do you manage without one?

I have four on my bench.

Never heard of wireless power?

Jamie

 

[tm]

Never heard of wireless power?

Jamie

Yep, AA, AAA, C, D,...

 

[Spehro Pefhany]

On Thu, 7 Jun 2012 18:01:36 -0700 (PDT), the renowned

For a few shekels more you can cut the ripple voltage by a factor of
200...
http://www.mpja.com/0-30V-0-3A-Variable-Benchtop-Power-Supply/productinfo/9601+PS/

That 200mV spec (RMS!) on the cheapies is a bit much (they're
switchers), but for $50 apiece they'd be fine to have a few around. My
Rigol is rated at even better, but it's 20dB more expensive again.

It's nice to have some of the old linear ones around that don't have
HF noise on the outputs, but eventually there's usually a switcher in
there (often followed by a linear for sensitive stuff).


Best regards,
Spehro Pefhany

 

[Charlie E.]

Get most of the solder off and then stick a loop of very fine wired behind
the pin, heat & lift!


You mentioned $1200 - that sounds expensive for proto boards are they huge
or special in some way?

No, just ordered 10 boards. They are abut 2"x2", and have about 60
components. Ordered them with all SMT components on them, since
several of the ICs are fine pitched, and my hands are not as steady as
they once were...

I got bids from half a dozen vendors, and this one was best by a few
hundred dollars!

Charlie

 

[josephkk]

That 200mV spec (RMS!) on the cheapies is a bit much (they're
switchers), but for $50 apiece they'd be fine to have a few around. My
Rigol is rated at even better, but it's 20dB more expensive again.

20 dB? Really? That is Agilent and such range and then some. I can get
a nice DSO for that.

 

[John S]

If you're a hobbyist, and want to get away from the teevee set, it
might make sense to build your own power supply. But for $70 or some
such, a metered, finished bench supply doesn't make sense to build.

Ebay has lots of power supplies too...

http://www.ebay.com/itm/Precision-V...986?pt=LH_DefaultDomain_0&hash=item256f8b216a

I bought one. They gave me a tracking number. Then they said they are
out of stock (regardless of what E_ay says) and will refund my money. I
said I would rather have the PS, but they said no more will be available.

 

[Uwe Hercksen]

Just received a batch of prototype boards. Added the through hole
parts to two, and tried to power up. No power!

Hello,

it test the new boards first for the isolation resistance between power
and ground. If there is a short, adding parts is useless.
I suggest you test all other boards first, maybe all of them got the
short. If they all got the short, you can find the location of the short
using high current, of course with a board without parts. When you found
the short, maybe you can repair all other boards with a sharp knife.

Bye

 

[Charlie E.]

Hello,

it test the new boards first for the isolation resistance between power
and ground. If there is a short, adding parts is useless.
I suggest you test all other boards first, maybe all of them got the
short. If they all got the short, you can find the location of the short
using high current, of course with a board without parts. When you found
the short, maybe you can repair all other boards with a sharp knife.

Bye

So, just a note to everyone. Looks like all these processors are
toast. Evidently, just scoping them with my DMM to see the resistance
was enought to blow the PIC processor. Going to see if my board house
can do rework, and replace the chips with new ones with the pins
already lifted...

Charlie

 

[Spehro Pefhany]

So, just a note to everyone. Looks like all these processors are
toast. Evidently, just scoping them with my DMM to see the resistance
was enought to blow the PIC processor. Going to see if my board house
can do rework, and replace the chips with new ones with the pins
already lifted...

Charlie

I suspect you have another problem. Those chips can take more current
than any sensible DMM will put out. A power supply across those
connections might damage them, but not a DMM.


Best regards,
Spehro Pefhany

 

[ehsjr]

Yikes, a 90 watt metered bench supply for $50. That seems impossible.

I've got the same supply. I removed the regular pots and
installed 10 turn pots in their place. Took a little work,
but well worth the effort. (Had to drill the panel holes
a bit larger.) The supply works well - the only problem
I see is the current meter showing 20 mA (no load) as the
minimum. i.e.: If your load draws 10 mA, the meter show 30 mA;
if you don't connect a load it shows 20 mA.

Ed

 

[JW]

[...]

Yikes, a 90 watt metered bench supply for $50. That seems impossible.

I've got the same supply. I removed the regular pots and
installed 10 turn pots in their place. Took a little work,
but well worth the effort. (Had to drill the panel holes
a bit larger.) The supply works well - the only problem
I see is the current meter showing 20 mA (no load) as the
minimum. i.e.: If your load draws 10 mA, the meter show 30 mA;
if you don't connect a load it shows 20 mA.

There's probably a pot that you can adjust for that...

 

[Charlie E.]

Hi All,
Hoping ya'll will have some brilliant suggestion to help me out...
again!

Just received a batch of prototype boards. Added the through hole
parts to two, and tried to power up. No power!

After a bit of testing, determined that my regulated VCC is shorted to
ground! :-( So, I started looking for the obvious - solder bridges,
vias shorted to an adjacent pad, going over my design files in detail
for maybe an error on my part, etc. No joy, can't see where the error
is.

So, my problem, how can I locate this elusive short. Any ideas? It is
a populated board, so I can't just crank up current till it indicates
the problem 'visually' (and aromatically!) ;-)

Charlie

Another update on this problem. As you may recall, I had found the
short, repaired it, but was still unable to program the MCUs. I sent
the boards back, and replaced the chips, and they still wouldn't
program. Finally, I put the scope on the programming connections, and
could see the clock and data lines were changing, but the programmer
wasn't seeing it. I finally decided it must be my programming cable,
so I ordered an 'official' Microchip adapter for my programming
header.

It arrived this afternoon, and I am now able to program my boards,
EVEN the one that I didn't replace the chip on!

As far as I can tell, on the programming cable that I had built, the
program data line, even though it passed every continuity check I
could think of, was not connecting to the programmer. I can only
assume that the programmer didn't like the RJ11 connector that I was
using, and didn't make good contact. ARRGGH!

Now, I have a lot of debugging and troubleshooting ahead of me!

Thank you all for your help!

Charlie

 

[Spehro Pefhany]

Another update on this problem. As you may recall, I had found the
short, repaired it, but was still unable to program the MCUs. I sent
the boards back, and replaced the chips, and they still wouldn't
program. Finally, I put the scope on the programming connections, and
could see the clock and data lines were changing, but the programmer
wasn't seeing it. I finally decided it must be my programming cable,
so I ordered an 'official' Microchip adapter for my programming
header.

It arrived this afternoon, and I am now able to program my boards,
EVEN the one that I didn't replace the chip on!

As far as I can tell, on the programming cable that I had built, the
program data line, even though it passed every continuity check I
could think of, was not connecting to the programmer. I can only
assume that the programmer didn't like the RJ11 connector that I was
using, and didn't make good contact. ARRGGH!

Now, I have a lot of debugging and troubleshooting ahead of me!

Thank you all for your help!

Charlie

They can get kind of fussy about long cables. I don't like to go
over about 6" (150mm).


Best regards,
Spehro Pefhany

 

[Charlie E.]

They can get kind of fussy about long cables. I don't like to go
over about 6" (150mm).


Best regards,
Spehro Pefhany

For this one, I took another cable, and just put a new RJ11 end on it.
I think that the RJ11 socket on the ICD3 is a little tempermental. I
have had problems with them before not making good contact on some
connectors.

Charlie