Design Help 2 - Please NO PIC

Indeed.

A client sent me a PCB with the chip I designed loaded in it, for me to generate a test plan.

Powered it up... VDD shorted.

Probably along traces trying to find what was shorted... kablooie!!

Fortunately I was wearing magnifiers... the Tantalum shrapnel scratched the lenses :-(

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
       If Bush was a "MORON", what does that make Obama ?:-)

^^^^^^ |____Probing (spell checkers :-(

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
       If Bush was a "MORON", what does that make Obama ?:-)

o

Solid aluminium electrolytics may well be better, but I've no reason to suppose that the regular parts are much different. Do you?

-- Bill Sloman, Nijmegen

--
Sure. The foil\'s cleaner, the anodizing is more precise, and the
electrical forming is a lot better controlled now than it was 30 years
ago.

Just compare package volume VS capacitance voltage product for what was
available then and what\'s available now and it should be pretty clear
that run-of-the-mill aluminum electrolytics are a _lot_ better now than
they were then.

For low leakage, take a look at Xicon\'s LLRL line:

http://www.tawelectronics.com/xicon/XICON_LLRL.pdf

With about 0.002CWV or 400nA (whichever is worse) of leakage current, 

they\'re, I believe, about the best available.

How do your old caps measure up?
 
JF

:

ck

rs

e no

And why do you think that?

And why is this going to make the leakage behaviour better than that of modern tantalum capacitors?

-- Bill Sloman, Nijmegen

--
Because, unlike you, I stay informed re. matters technical and don\'t go
around spouting irrelevant 30 year old nonsense.

If you think I\'m wrong, cite an example to prove it.

--
Replaced example:

"For low leakage, take a look at Xicon\'s LLRL line:

http://www.tawelectronics.com/xicon/XICON_LLRL.pdf

With about 0.002CWV or 400nA (whichever is worse) of leakage current, 

they\'re, I believe, about the best available.

How do your old caps measure up?"

They can\'t, of course, so you decide to snip what you can\'t respond to
what will put you in a bad light, and then you change the subject.

Typical Sloman chicanery.

:

ck

rs

e no

Thanks for the link John, if I ever need a low leakage aluminum I'll try some.

If anyones interested, the heating did increase the discharge rate of the nichicon cap I was looking at. At 60 degree's C the voltage dropped by 10% in one hour. Then at something near 100C (+/-10C) the voltage dropped by 25% in 30 minutes.

George Herold

Hmm... Why does he need to connect the 9V battery during a few days???

Pere

To "form" the oxide on the electrolytic capacitor. That procedure fills in voids and reduces leakage.

Cheers, James Arthur

I take it, this is the HE series 1000uF/50V with leakage spec'd at

500uA.

Earlier, and I gather it was around 20C, "in 20 minutes the voltage has dropped by less than 10%!" I gather that at 60C it remains less than 10% in 20 minutes... in fact, in 60 minutes!

You also mentioned earlier a 'fast droop in voltage after the initial charge', which is a different effect. Is this initial reflex included when saying 10%?

I take it that this works out to just a few microamps, but not yet quite the 500uA leakage spec?

Anyway, thanks for taking the trouble here and talking about it.

Jon

e:

:

ote:

back

ble

hours

e

are no

s n

Hi Jon, I just set this up quickly on the corner of my lab bench. I stuck the capacitor into a white protoboard, attached a 1 G ohm resistor to one leg and measured the voltage from the other side of the resistor to the opposite side of the capacitor with my 10 Meg ohm input impedance voltmeter. This gave me a 1/100 voltage reading and discharged the cap at only a 10-20 nA rate. I applied 20 (something) volts to the cap for a few seconds and then recorded the voltage at different times

Time Voltage (mV)

2:26 213 2:31 202 2:44 195 3:00 190 4:00 185

8:36 AM 156 (So this was the next morning 16 hours later)

I then read the heat idea and taped a 100 ohm aluminum housed power resistor to the top of the cap, tapped a thermal couple to the side, stuck some melamine foam around the outside and top. I also placed a book on top of the foam to try and hold the contraption together. (I don't want you to think this was some precision measurement.) I applied some power to the resistor,

Time Voltage(mV) Temp

10:25 150 54C 11:18 137 61C 11:29 136 61C (I then cranked up the temperature a bit more) 11:40 128 93C 11:45 120 95C 11:52 110 98C Something is beginning to stink I reduced power 12:00 100 95C Heat off, removed the foam. 12:11 98 25C End of the experiment

There was a bit of noise in the measurements as you might expect with the 1G ohm source driving the voltmeter. As I got near it to make a measurement the voltage would jump and then settle back.

The good news is that at room temperature the caps hold their charge for a long time.

George H.