# Basic question...Capacitors and being an idiot

#### Phsion

Joined Mar 10, 2009
6
Hi All, I was hoping to find some help with Capacitors here...The short question is; If I have a Capacitor(the Aluminum Can SMD style...47uF, 35v) feeding +18v to a circuit with 6 Resistors, and I solder the Capacitor on BACKWARDS, can i blow all 6 Resistors? The long questions is; so i have this device...and i broke it. It has card A, and Card B. Card A gets power from the main power source, and then feeds Card B it's power. On Card A, there are four 47uF Capacitors (Paralleled in groups of 2...one set for +18v, the other for -18v), then the ribbon that feeds Card B. Then on Card B, from what i can tell, the +18v first hits all of the Resistors (22.1Ohm SMD, Thick Film?). Here's what happened...somehow the Capacitors got shorted, and the resistors got hit so bad they blew in half. So I went through the whole thing, replaced all the Capacitors, and replaced all the resistors, and then turned the thing on again...And each resistors blew right back up. So i took a closer look, and realized I had soldered one of the capcitors on backwards, so negative was touching positive and vise versa. What i'm trying to figure out is, if i swap that Capacitor around, whats the chance that i fixed it? I really don't want to replace those resistors if i'm just gonna blow them right back up again. Thanks for the help.

Oh, and by the way...those resistors arn't in series...there are 6 similar but slightly different circuits, all next to each other, being fed from the same power. The resistors are just the first thing ( i think) in the circuit.

#### thatoneguy

Joined Feb 19, 2009
6,349
Usually the capacitor will explode.

Check the power supply, and the circuit for a short to ground, to find the actual cause of the symptoms you are seeing.

#### Phsion

Joined Mar 10, 2009
6

#### SgtWookie

Joined Jul 17, 2007
22,201
If the caps are shorted, and they are being charged via the resistors, sure.

A DMM might tell you if a cap has a dead short, but won't tell you the leakage at the rated voltage. You need a power supply capable of sufficient voltage and a series resistor to do that.

You need to find out where the short(s) are. Trace out the circuit. Then divide and conquer. Use Ohm's Law to determine if the resistance readings you're getting are reasonable.

#### Screamtruth

Joined Apr 17, 2006
10
If the caps are shorted, and they are being charged via the resistors, sure.

A DMM might tell you if a cap has a dead short, but won't tell you the leakage at the rated voltage. You need a power supply capable of sufficient voltage and a series resistor to do that.

You need to find out where the short(s) are. Trace out the circuit. Then divide and conquer. Use Ohm's Law to determine if the resistance readings you're getting are reasonable.
Too add to the "stupid questions" list, why do you need a capacitor with leakage current? Or are you just finding what the acceptable rate should be?
(FYI, this is new to me as well!)

Semper Fi Marine.

#### thatoneguy

Joined Feb 19, 2009
6,349
Check resistance (both ohm and on diode check) from both sides of resistors to ground, then power, while the unit is off, and you know the caps are good. If the value is far below any of the resistors, the problem is elsewhere.

To test leakage without a bench supply, some multimeters have an "insulation resistance" range. On a Fluke, switch to resistance manual range, then nanoSiemens (nS, 1nS = 1GigΩ, 1000nS=1MΩ). This range is useful for measuring resistances > 40MΩ, reverse diode leakage, as well as capacitor leakage. If measuring a capacitor and get a high, steady value instead of overload, you are seeing leakage. Compare with a known good cap of exact same type/value to be sure.

On the back of mine is a chart for uF/second on different resistance ranges:
400Ω = 2,600 μF/sec
4kΩ = 300 μF/sec
40kΩ = 30 μF/sec
400kΩ = 3 μF/sec
4MΩ = 0.3 μF/sec

A similar chart can be made for yours with a stopwatch and spare time. The values above are actually about 29 rather than 30 on my meter, but close enough for a fast check without having to move the dial. If you have a "conductance" range, you can roughly test for leakage, it's more appararent when the rate the value changes is different for a similar sized cap, even on resistance ranges.

#### SgtWookie

Joined Jul 17, 2007
22,201
Too add to the "stupid questions" list, why do you need a capacitor with leakage current? Or are you just finding what the acceptable rate should be?
It's not a stupid question at all.

ALL capacitors have "leakage" through their dielectric. Electrically, it acts as a high-value resistor in parallel with the capacitance value, and slowly drains the charge within the capacitor. Problems occur when this resistance value decreases over time; more and more power is dissipated within the capacitor instead of performing useful work. This heats the capacitor; in particular the liquid electrolyte. This causes the electrolyte to boil, forcefully rupturing the capacitor's case. Modern aluminum electrolytics have their cases designed with weak areas, so that the pressure won't build too high before rupturing. The old caps could go out with a real "BANG!". Tantalum caps are also vulnerable to leakage current; they can go bad just by sitting around for half a year.

The formula I posted above was derived from tables that were generated some 50+ years ago. The formula is much easier to use than having to deal with tables.

The original 1957 document that I found the tables in is named ReformerNo1.pdf - you might find it with a Google search. I won't post a direct link here.

Failing that, search for "Electrolytic capacitor reforming"; you'll get plenty of hits.
S/F

#### Screamtruth

Joined Apr 17, 2006
10
It's not a stupid question at all.

ALL capacitors have "leakage" through their dielectric. Electrically, it acts as a high-value resistor in parallel with the capacitance value, and slowly drains the charge within the capacitor. Problems occur when this resistance value decreases over time; more and more power is dissipated within the capacitor instead of performing useful work. This heats the capacitor; in particular the liquid electrolyte. This causes the electrolyte to boil, forcefully rupturing the capacitor's case. Modern aluminum electrolytics have their cases designed with weak areas, so that the pressure won't build too high before rupturing. The old caps could go out with a real "BANG!". Tantalum caps are also vulnerable to leakage current; they can go bad just by sitting around for half a year.

The formula I posted above was derived from tables that were generated some 50+ years ago. The formula is much easier to use than having to deal with tables.

The original 1957 document that I found the tables in is named ReformerNo1.pdf - you might find it with a Google search. I won't post a direct link here.

Failing that, search for "Electrolytic capacitor reforming"; you'll get plenty of hits.
S/F
Bringing back all that old MCCES stuff that I forgot, years ago.

Ended up being a TACP chief anyways..........so I lost all of that useful info.
Semper gumby and thanks!