Audio coupling/decoupling caps with higher capacitance than their rated values

[lti]

I have four electrolytic caps that were used as coupling or decoupling caps and had no DC voltage across them. All four have a measured capacitance that is higher than the rated value.

Two of these are 0.22uF 50V caps. One measures 0.38uF and the other measures 0.4uF. The other two are 1uF 50V caps and measure 650uF and 900uF! These values do not change when the test leads are reversed. There seems to be excessive leakage.

Can I bring the capacitance down to the rated value and reuse these caps? I don't plan on using these for any critical applications.

[PCBONEZ]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

Uf shouldn't change when you reverse leads.
.
If you use the resistance function vice capacitance you should see them charge/discharge from the meter when you reverse leads.
The range you use and the uF will affect how fast they charge/discharge.
If it's too fast you won't see it, so experiment with the range.
.
.
The 1uF definitely look bad.
The others might just be meter error.
-
Check a new good cap with the meter to make sure the meter isn't off.
.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

I have tested new caps with the meter and they all give their expected values.

The 1uF caps are the input coupling caps in my cheap computer speakers.

[PCBONEZ]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

If the meter reads uF accurately at 0.22uF then those are probably bad too.
- That assumes they are the typical +/-20% caps.
Some caps have goofy tolerances like +100%/-20% and things like that.
If there is an 'M' that is the mark[code] for +/-20% uF.
.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

I measured .22uF and 1uF caps with the same meter. The 0.22uF caps don't have a tolerance listed. I don't remember the tolerance rating in the 1uF caps, but it probably isn't +90000%.

After running DC voltage through the 0.22uF caps, their capacitance values have decreased.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

When I checked the capacitance of the 1uF caps, I did not remove them from the board. When they were tested out of circuit, they both tested good, even though there is nothing in parallel with them.

[PCBONEZ]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

Would normally expect lytics to have uF go down as they age.
Are these some other kind of cap?

Would also expect uF to go down with high leakage.

If they look to be weird just replace them.
Not like those would cost a bunch.
.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

Those are electrolytic caps and are really crappy brands. The 1uF caps are "D S" CD110 series and the 0.22uF caps are Jackcons.

[Logistics]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

Are you going to upgrade the input and output caps to films? You can get 1% 1uF metallized polypropylenes from Parts Express. If space is concern then I would at least update to bipolar/non-polarized caps in place of all lytics in the intput and output signal paths.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

The 1uF caps were not actually needed. Look at the "$10 computer speakers" thread. The 0.22uF caps have already been replaced by film caps.

[Longbow]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

I have four electrolytic caps that were used as coupling or decoupling caps and had no DC voltage across them. All four have a measured capacitance that is higher than the rated value.

Signal coupling caps sometimes don't have a lot of dc across them. Decoupling caps are used to filter the supplies and are generally connected from a supply line to ground. If there is no voltage across them, your supply is bad, not the capacitor. Or if the cap is shorted, something else on the board should be smoking.

Two of these are 0.22uF 50V caps. One measures 0.38uF and the other measures 0.4uF.

Why bother measuring them? If they pass a signal, then forget it.

The other two are 1uF 50V caps and measure 650uF and 900uF!

I believe you misread your beloved capacitance meter. You were probably seeing 650pf and 900pf.

There seems to be excessive leakage.

If a cap is leaking, you will not get huge increases in capacitance readings.

Can I bring the capacitance down to the rated value and reuse these caps?

How do propose to go about changing the capacitance values?

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

Signal coupling caps sometimes don't have a lot of dc across them. Decoupling caps are used to filter the supplies and are generally connected from a supply line to ground. If there is no voltage across them, your supply is bad, not the capacitor. Or if the cap is shorted, something else on the board should be smoking.

The 1uF caps were installed on the input of a TDA2822M amplifier IC with the positive side connected to the IC. Coupling caps are not needed on the input of that IC.

Why bother measuring them? If they pass a signal, then forget it.

Those were signal decoupling caps. They were not doing their job.

[PCBONEZ]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

The 1uF caps were installed on the input of a TDA2822M amplifier IC with the positive side connected to the IC. Coupling caps are not needed on the input of that IC.

Unless where ever the signal is coming from has a DC bias voltage on it.
.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

Since the positive side of the cap was connected to the IC and the negative side was connected to the input signal, the DC bias voltage would have to be negative for the caps to actually be useful.

[Th3_uN1Qu3]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

No. The caps would have blocked the DC voltage regardless of their polarity. Sure, they would have blown up given a positive DC bias of really high amplitude. But they would have done their job.

The TDA2822 also uses output coupling caps unless it's in bridge mode (which it isn't) so passing DC to the speakers would have been impossible anyway. But if the input signal had a large DC bias, without the input coupling caps it would distort the hell out of the sound.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

I knew that the caps would have blocked DC voltage but blown up.

If I was concerned about the DC bias, I would have installed nonpolarized caps in those positions. I have plenty of dead electronics with surface mounted ceramic caps that I could have soldered onto the back of the board. I did that to my Sony boombox.

I did get two good 1uF caps out of this. I keep running out of those.

[PCBONEZ]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

A signal caps purpose is to -block DC- while letting the signal through.

I doubt they would have used parts they didn't need.
That goes against their whole 'as cheap as possible' theme.
.

[Longbow]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

The 1uF caps were installed on the input of a TDA2822M amplifier IC with the positive side connected to the IC. Coupling caps are not needed on the input of that IC.

I beg to differ. Keep in mind that the IC itself has an internal biasing system which can be altered by hooking up stray d.c. to the input. Low uF value capacitors are usually sufficient because the input Z of the IC is typically high. If you still believe that input coupling caps are not needed, then remove them and have a look at the d.c. on the output pin of the IC. Report back.

Those were signal decoupling caps. They were not doing their job.

Let's be clear about the term decoupling. Power supply lines are filtered, or "decoupled" in order to prevent the supply lines from affecting the operation of the stage in question, and also from affecting other stages connected to the same supply lines. Good engineering practice includes a resistor or choke in combination with the filter capacitor (decoupling cap), which effectively isolates each stage from the others. These days, you see circuitry that omits the resistor or choke completely, but that is beside the point. Supply lines are meant to be very low impedance, just another way of saying they should essentially be AC grounds.

Small value capacitors within the signal processing path are almost always intended for waveshaping purposes, or to prevent oscillation by rolling off upper and lower frequency limits. So it would be perfectly normal to see signal on the ungrounded end of caps in the range of .22 uF.

[Th3_uN1Qu3]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

If you still believe that input coupling caps are not needed, then remove them and have a look at the d.c. on the output pin of the IC. Report back.

In his case, since it is a single supply IC with capacitor output, he should have exactly half the supply voltage at its output. I know the TDA2822 inside out, and usually those caps have accompanying resistors which together form a highpass filter, because it's pointless to stress those little speakers with subbass frequencies that they will never be able to reproduce. Heck, most of those barely make it down to 80Hz.

[lti]

Re: Audio coupling/decoupling caps with higher capacitance than their rated values

If those caps were required, they would have been in the application circuit in the datasheet for that IC.

The 0.22uF caps were connected between the audio signal and ground in a tone control circuit.