Interpretation of capacitor ohms reading

[DuderMcMerican]

I recently replaced a few capacitors in my PCB, and I jumpered a burnt trace.

The jumper works nicely, but one concern I have is that one of the 47uf caps is showing 0 resistance on analog multimeter. My understanding is that on-board cap tests with multimeter are all but useless EXCEPT when cap reads 0 ohms (needle pegs and doesn't regress back) or infinite resistance (needle does nuttin). In my case--0 ohms--the online says it's a shorted (bad) cap. Well, i pulled one leg out of the board and tested resistance again, and the meter read exactly how it should (needle pops to no resistance but quickly regresses back to greater resistance)..

I went on google to ask Jeeves the meaning of these results (thinkin maybe there's a short elsewhere on the board??), but all i found is that ppl says it's a bad cap.. But it's a brand new Japanese cap from Digikey and it tests fine when disconnected from board, so I'm thinkin it's prob not the cap, but idk.. I'm hoping I can just chalk it up to "cap tests with ohm meters aren't reliable when cap is in circuit..." and not worry about it, but other info suggests 0 ohms--even when in circuit on board--is trouble. Anyone have thoughts on this?

Btw, the analog multimeter i use has max ohm setting of only x1k, in case that matters, but i did also try my cheapo digi meter with max 2000k ohms and it still read <1ohms with no regression.

[stj]

Re: Interpretation of capacitor ohms reading

what is the board, got foto's?

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

It's from the EEC IV on a 1990 Mustang. No, I don't have pics right now, but I can take some. What would you want the pics to show?

[stj]

Re: Interpretation of capacitor ohms reading

what it connects with.
because it does sound like something is shorted.

[stj]

Re: Interpretation of capacitor ohms reading

i just found this:
http://www.kvitek.com/ford/

is this the pcb?

?

[keeney123]

Re: Interpretation of capacitor ohms reading

So an analog meter's supply voltage from it's battery will be a much greater voltage than a digital meter. The fact the board is reading 0 ohms indicates that the board has a short. If you ohm the board where the cap goes without the cap in the board are you reading 0 ohms?

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

i just found this:
http://www.kvitek.com/ford/

is this the pcb?

?

It's similar but not my exact one. I'll take pics tonight and post

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

So an analog meter's supply voltage from it's battery will be a much greater voltage than a digital meter. The fact the board is reading 0 ohms indicates that the board has a short. If you ohm the board where the cap goes without the cap in the board are you reading 0 ohms?

Hmm.. Haven't tried but good idea. Just to be clear (I'm a newb) do you mean ohm across the pads where the cap legs go? If I recall, there are no visible traces connected to those pads so I don't think following traces and ohming them would be an option..

[stj]

Re: Interpretation of capacitor ohms reading

check all the diodes and transisters for shorts.

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

check all the diodes and transisters for shorts.

Okay. Will do. Here are pics of the board and cap in question. The one solder joint is still a little assed up from when i pulled the leg out.

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

So i ohm'd my diodes and got mixed readings. Some read 0 ohm one direction and max ohms the opposite direction, some read 0 one direction and 1-2ohms the opposite, and some read 0 ohms both directions (and i think one read 1 ohm one direction and 2 ohms the other. I'm guessing that the "no deflection" diodes--0 ohms in both directions--indicate they're in the shorted circuit? Those were the yellowish ones in the vicinity of the cap in question.

In case it matters, I used the x1k setting on my analog meter. The only other ohm setting is x10. I also have a digi meter with more range selections..

[keeney123]

Re: Interpretation of capacitor ohms reading

So i ohm'd my diodes and got mixed readings. Some read 0 ohm one direction and max ohms the opposite direction, some read 0 one direction and 1-2ohms the opposite, and some read 0 ohms both directions (and i think one read 1 ohm one direction and 2 ohms the other. I'm guessing that the "no deflection" diodes--0 ohms in both directions--indicate they're in the shorted circuit? Those were the yellowish ones in the vicinity of the cap in question.

In case it matters, I used the x1k setting on my analog meter. The only other ohm setting is x10. I also have a digi meter with more range selections..

So yes you would ohm across were the pads of the cap was. If it reads 0 ohms the short is on the board not the cap. When one ohms for a short you use the smallest scale. So the times oneK scale means every digit is time 1000. The times 10 scale is every digit is times 10. The times 10 scale is the smallest scale. The digital meter would work better on this surface mount board. If you find on the times 10 scale it is still shorted then it may be possible to find the short with a good digital meter. Certain Fluke meters can find shorts with-in 3 inches.

[stj]

Re: Interpretation of capacitor ohms reading

dont let this transistor lean back and short against the other part!

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

So yes you would ohm across were the pads of the cap was. If it reads 0 ohms the short is on the board not the cap.

Ohh okay. I'll do that, but i don't quite understand the logic.. The short was there with cap in place, the cap tested fine with one leg pulled, and the short was still there when i reinstalled cap leg.. Wouldn't that already indicate the short isn't in the cap? I'm just tryin to understand why pullin the cap and ohming the pads would provide new info.

When one ohms for a short you use the smallest scale. So the times oneK scale means every digit is time 1000. The times 10 scale is every digit is times 10. The times 10 scale is the smallest scale. The digital meter would work better on this surface mount board. If you find on the times 10 scale it is still shorted then it may be possible to find the short with a good digital meter. Certain Fluke meters can find shorts with-in 3 inches.

Oh yeah.. I knew the x1k meant x1000. My mistake reporting the 1and 2 ohms..

I'm having trouble understanding the logic behind reducing the ohm setting.. If I'm gettin 0 ohms in both directions at x1k then wouldn't i surely get the same reading at x10? After all, 0x1000 is zero just as 0 x10 is 0. Seems that x1000 would be a more rigorous test as anything >0 would register as resistance before it would for x10, no?

Please know that I'm not questioning your advice. I'm just trying to understand the logic as I'm very new to all of this.



***EDIT: Upon further reflection, I think I now understand the reason to use the lowest ohm setting possible. That makes sense. However, I'm still confused about the testing across cap pads with the cap removed... how does that give more information on identifying whether the short is in the cap or the board after the one-leg-pulled test already indicated that the cap is okay?

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

dont let this transistor lean back and short against the other part!

Oops.. Thanks for noticing that. I'll take care of that and retest when I get home today.

[keeney123]

Re: Interpretation of capacitor ohms reading

Ohh okay. I'll do that, but i don't quite understand the logic.. The short was there with cap in place, the cap tested fine with one leg pulled, and the short was still there when i reinstalled cap leg.. Wouldn't that already indicate the short isn't in the cap? I'm just tryin to understand why pullin the cap and ohming the pads would provide new info.



Oh yeah.. I knew the x1k meant x1000. My mistake reporting the 1and 2 ohms..

I'm having trouble understanding the logic behind reducing the ohm setting.. If I'm gettin 0 ohms in both directions at x1k then wouldn't i surely get the same reading at x10? After all, 0x1000 is zero just as 0 x10 is 0. Seems that x1000 would be a more rigorous test as anything >0 would register as resistance before it would for x10, no?

Please know that I'm not questioning your advice. I'm just trying to understand the logic as I'm very new to all of this.



***EDIT: Upon further reflection, I think I now understand the reason to use the lowest ohm setting possible. That makes sense. However, I'm still confused about the testing across cap pads with the cap removed... how does that give more information on identifying whether the short is in the cap or the board after the one-leg-pulled test already indicated that the cap is okay?

I always question a persons advice if I do not know something, that is not a bad thing.
I told you about ohming the pads because you where wondering about shorted caps in the beginning of the article. I wanted to show you how to tell if the cap was shorted. I figured by you doing this you would learn the above statement. So yes you are correct if you take the cap out of the board and the short goes away then that would indicate that the board was the problem and if you have any doubt then ohm the board.
So if you are looking for a short then you are looking for a reading under 200 ohms. The times 10 scale is more accurate in the lower scale of a ohm reading. So if you have a percentage of accuracy lets say as an example 1%. On the X10 scale if the reading is say 20 ohms your accuracy would be 0.2, 200 ohms would be 2 ohms. Now if we look at The X1000 scale the 1% accuracy decreases by 100 from o ohms to 1000 ohms. So a reading of 20 ohms would have and accuracy of 2ohms and 200 ohms would be 20 ohms and then 1000 ohms would have 10 ohms. This is a simple percentage difference for a simple definition. It really comes down to how many digits the meter is capable of and how accurate those digits are that are displayed.

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

I always question a persons advice if I do not know something, that is not a bad thing.
I told you about ohming the pads because you where wondering about shorted caps in the beginning of the article. I wanted to show you how to tell if the cap was shorted. I figured by you doing this you would learn the above statement. So yes you are correct if you take the cap out of the board and the short goes away then that would indicate that the board was the problem and if you have any doubt then ohm the board.
So if you are looking for a short then you are looking for a reading under 200 ohms. The times 10 scale is more accurate in the lower scale of a ohm reading. So if you have a percentage of accuracy lets say as an example 1%. On the X10 scale if the reading is say 20 ohms your accuracy would be 0.2, 200 ohms would be 2 ohms. Now if we look at The X1000 scale the 1% accuracy decreases by 100 from o ohms to 1000 ohms. So a reading of 20 ohms would have and accuracy of 2ohms and 200 ohms would be 20 ohms and then 1000 ohms would have 10 ohms. This is a simple percentage difference for a simple definition. It really comes down to how many digits the meter is capable of and how accurate those digits are that are displayed.

Okay. Thanks for all the info! I'll double check with the lower ohm settings.

Should i systematically go around board and pull a leg from diodes/transistors to test for shorts? Kinda like what i did for the cap?

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

dont let this transistor lean back and short against the other part!

I checked the transistors and they aren't actually leaning against other components. I think they just look that way because of the picture angle.

What's the best way to systematically test diodes/tramsistors for shorts?

[stj]

Re: Interpretation of capacitor ohms reading

meter them, any that show a short (0 ohms or near it) between any 2 pins.
if you find one - lift it and re-test.

[DuderMcMerican]

Re: Interpretation of capacitor ohms reading

meter them, any that show a short (0 ohms or near it) between any 2 pins.
if you find one - lift it and re-test.

Okay, thanks!

UPDATE: I metered the cap in question using the x10 setting and got different (better?) redults. With leads matched to cap polarity, i got 9 ohms (0.9 x 10) followed by a deflection to 11 ohms (1.1 x 10). Reversing polarity yielded a steady 2 ohms (0.2 x 10). Is this good? I will still test diodes and transistors now that i know how to do it.