cheap caps drain quickly on shutdown : are they bad?

[airfrankenstein]

Hi,

Sorry, might be a beginner's question.
I built an ac15 inspired clone last year which sounds fine overall, but I'm a bit perplexed by how quickly the caps drain themselves when I turn the amp off....they drop right down to 6 volts ! These are cheap caps from the local electronics store.
Should I swap them out for better quality ?
Can something else be the cause (choke, soldering, etc...)?


Thanks

[xtian]

A#1 = does the amp sound good?

I've observed that filter caps drain pretty quickly on shutdown. With all tubes installed and hot, they drain the power reservoir. Conversely, if you remove the tubes, you should observe that your caps stay charged after power down.

[Phil_S]

Do you know if there is a bleeder resistor installed?

[airfrankenstein]

Thanks for the quick replies :

Amp sound good.
I dont think there's a bleeder resistor. i included the ps section i'm using.
I'll try powering down without the tubes.
I began wondering about this because I noticed that my Ampeg Jet J-20's caps drain much slower.

[rdjones]

Newer commercial amp designs are recommended to have bleeders in place so the B+ drains off to less than 10% within 5 minutes.
Not sure when this came about but probably sometime in the '80s or '90s.
The value needed to achieve this is fairly high and shouldn't effect tone.

If the amp has a high idle current (ie "Class A") the supply would drain off quickly if the caps were small.

rd

[Tillydog]

I began wondering about this because I noticed that my Ampeg Jet J-20's caps drain much slower.

Your Ampeg has 2 x 47u caps, the AC15 has 2 x 16u in the same place - 1/3rd of the capacitance will drain 3x quicker under the same load. (I would also hazard a guess that the bias current drawn by the EL84s in your AC15 was higher than that through the 6V6s in your J-20)

[airfrankenstein]

Thanks, seems quite obvious.

[teemuk]

I never knew that the price tag of capacitors could alter the laws of physics.

[tictac]

There was an article in Vintage Guitar magazine a few years back that talked about this subject, even had plans for a cap tester you could make...

Seems that an electrolytic cap that drains quickly if it's "disconnected from the circuit" could be bad.

A down and dirty way to check a filter cap already installed in an amp would be to...
1) disconnect the ungrounded end of the cap.
2) clip a voltmeter accross the cap.
3) Using a pair of test leads with aligator clips, connect a 9 volt battery across the cap. (the + terminal of the battery to the disconnected lead of the cap, - terminal to the grounded lead)

4) You should now be reading 9VDC on your voltmeter
5) Disconnect the + terminal of the battery and watch the voltmeter

If the cap can't hold 70 to 80% of the charge for 10 seconds consider it bad.
This was from an article written by Gerald Weber of Kendrick Amplifiers and
I don't know where he got his numbers from but common sense alone should tell you if a filter cap can't hold a charge of 9V you surely don't want it in your amp. I've ferreted out a few bad caps this way and the bad ones usually go to zero volts pretty fast. Compare your results with some known good caps or other caps within the amp for a reality check.

I know some people don't care for Gerald but he does offer a few decent tips once in awhile, even if only by trial and error...

Hope this helps...

TT

[vibratoking]

If the cap can't hold 70 to 80% of the charge for 10 seconds consider it bad.

Sorry, but that is incorrect! I do not agree with that statement and would never use that method to check a cap. Use of that method will cause you to throw away every good cap in your amp, unless you screw up the measurement in some other way.

How long the cap should hold a charge is related to the size of the cap and the amount of resistance in the circuit. The time constant is the product RC and the value on the cap should decay to 37% of it's initial value in one time constant. Take a 10uF cap, a 1Meg resistor, and 9V initial value. The time constant is 10u x 1Meg = 10s and the voltage on the cap should decay to about 3.3V after 10 seconds. Increased voltage also can have an effect on how the dielectric might breakdown and leak.

[martin manning]

In the procedure described above step 1) is disconnect the ungrounded end of the cap. This means the R in the RC circuit is the DMM input impedance, which is around 10M for a good meter. So in the case of a 10uF cap the voltage should (in theory) be at 90% of it's initial value after 10 seconds. Of course the voltage decay depends on the size of the cap- a 22uF should be at better than 95% after 10 seconds.

[vibratoking]

My point is the following: DO NOT blindly apply Weber's, or anyone elses, assumptions and expect that the result will be correct. Do not assume anything unless you like making mistakes and being mislead. Learn to turtle instead of just eating...or just keep begging for more. Martin is correct, many meters have an input resistance of 10M Ohms, but not all. I have seen plenty of 1M Ohm meters. You also may have other impedances in the circuit that you have not realized or accounted for which will make the blind application incorrect. Turtle on brother! mmmmm turtle.

Make sure you understand your meter, what a time constant is, and what exponential decay is before you attempt to measure the voltage decay of a cap to determine if it is good or not. There are plenty of ways to arrive at an improper result. A begger might decide to check a 100pF cap, then what? No turtle soup, that's what. Here is a good link that does the math for you and also graphically displays HOW an RC circuit behaves. This is BASIC theory that anyone who uses a cap should know.

http://hyperphysics.phy-astr.gsu.edu/hb ... apdis.html

[tictac]

About my last post...

1) The info in my last post was from a Vintage Guitar magazine article written by Gerald Weber not me...

2) The context was electrolytic filter caps not other types...

3) I've tried this method a few times and have never seen a brand new cap go to zero instantaniously but I have seen some old bulging caps do so... I wouldn't want a filter cap new or other wise that couldn't hold a measly 9V charge in my amp...

The afor mentioned article is no replacement for test equipment like a capacitance meter that can test ESR and the like but it's the only article I've ever seen written on the subject so I thought I'd share it...

TT