Re: MFA Magus preamp hum/buzzing sound

Jon:

Please note the interpolations:

On Nov 5, 11:26 pm, Gray Glass <glassg...@xxxxxxxxx> wrote:
Peter (& West),

I have seen rectifiers fail shorted. In my experience, that is how they
most commonly fail in lower voltage DC applications. I don't see too many
in mains rectification circuits, but it DOES happen from time-to-time.

OK... you have seen something that I have not seen in 35 years - a
mains rectifier failed-short. Assume that is the case, however. In a
bridge situation, that would require at least two of the rectifiers to
fail in this way to give a path for more than a very few volts of AC
to reach the secondary to be measured. And, of course, that would
immediately cut the B+ in half. The unit would go silent under those
conditions.

As far as electrolytics go, here is my rule of thumb:

1. An electrolytic that tests bad is bad
2. An electrolytic that tests good may be bad

ESR meters are pretty useful, but they usually tests caps at a fraction of
their working voltage, and that is a practical limitation in our hobby.

If the cap has begun to fail, the ESR meter will show this as compared
to a properly functioning cap, but I agree that the only legitimate
way to test a cap is under full operating voltage out-of-circuit.

When testing capacitors out of circuit, I use an adjustable high voltage DC
power supply. I connect a 1000 ohm, 1 or 2W resistor between the minus lead
of the cap and ground. I then measure the voltage across the resistor with
a VOM set to the voltage scale, i.e. voltage measured across the resistor
corresponds to current drawn through the capacitor.

Poor Man's Ammeter. A similar device using a 1ohm, 5 watt resistor in
series with one mains-lead may be used to test current draw with a
VOM. Measure across the resistor to read current draw of the device
under test.

I gradually increase the voltage and keep an eye on the meter. This is a
useful way to spot defective non-elecrolytics pretty fast. Electrolytics
are a little harder because they will draw current with voltage changes
until they stabilize. It is also a good way to reform electrolytics, as you
can keep an eye on the current as you change slowly increase voltage.

Everyone knows that an old electrolytic that hasn't been used in awhile can
experience catastrophic failure with immediate application of high voltage,
so before I condemn an electro, I reform it first.

A really useful bench tool is a variable A/C supply with built-in ammeter
and voltmeter. You can ramp up voltages and look for excess current draw,
usually before toasting the defective part.

I keep (two actually I feel a bit piggy over it) Heath IP-5220s for
this purpose.

http://www.oaktreevintage.com/web_photos/test_gear/Heathkit_IP-5220_web.jpg

Voltage ramping (or reforming) is best done with any tube rectifiers removed
and replaced with SS equivalents, such as those sold under the brand name
"Copper Cap." Otherwise, no B+ flows until you reach the turn-on voltage of
the rectifier. Common sense, right?

Jon: You know better! Common Sense isn't. But unless there are
prevailing reasons to attempt a reform, I tend not to trust this
expedient as a permanent fix. Yes, it is possible, but the result is
still an old, questionable cap that is now pretending to be good for
an unknown period. On the other hand, I very nearly always "form" (&
test) new electrolytics before installing them. And equipment that has
sat about for 2-20 years between uses does get the approach you
describe. Another hint: A 40 year old electrolytic NOS is just as
suspect as a 40 year old electrolytic of unknown condition installed.


A scope is very helpful to show amount of A/C ripple.

It certainly is. Most AC ripple will come from Tube Rectifiers,
however, not from silicon bridges.

Electros that leak current sometimes leak electrolyte, and over time almost
all get warm or hot. So one quick (and low-tech) way to gauge a failure is
to check the can temperature. B+ dropping resistors located under some caps
will warm them up, too, so you have to watch for that too.

And remember - discharge electros before & after testing them.

Why? You don't like char-marks and divots on tools? The feel of bits
of molten metal on your skin? Finding yourself across the room in a
heap with no memory of how you got there?

Actually, a fully-charged large-value electro can be useful in
"healing" open field-coil and transformer windings. Not something to
do often, but if the alternative is scrap, worth a try.

Peter Wieck
Wyncote, PA

.

Relevant Pages

  • Re: MFA Magus preamp hum/buzzing sound
    ... I have seen rectifiers fail shorted. ... As far as electrolytics go, here is my rule of thumb: ... their working voltage, and that is a practical limitation in our hobby. ... You can ramp up voltages and look for excess current draw, ...
    (rec.audio.tubes)
  • Re: X7R ceramics for DC blocking ?
    ... says "If you keep the DC Bias high wrt the AC component there isn't ... applying DC-bias would avoid having the cap ... Particularly with electrolytics as I remember. ... Electrolytics don't become non-linear until the reverse voltage exceeds ~ 100mV. ...
    (rec.audio.tech)
  • Re: Audio Op Amp power
    ... making the cap a a larger value, you reduce the signal voltage across the cap in ... discussion is about the alleged failings of zero bias electrolytics as coupling caps, ... I'm aware of the voltage coefficient of certain ceramics especaially Hi-K ... If you have evidence of a voltage coefficient with electrolytics I'd be interested to see it. ...
    (sci.electronics.design)
  • Re: Audio Op Amp power
    ... making the cap a a larger value, you reduce the signal voltage across the cap in ... Hence a 100uF coupling cap ... Getting a 100uF cap is hardly a problem with electrolytics. ... I'm aware of the voltage coefficient of certain ceramics especaially Hi-K ...
    (sci.electronics.design)
  • Re: Audio Op Amp power
    ... making the cap a a larger value, you reduce the signal voltage across the cap in ... I fail to see why that's a reason to suggest ceramics. ... Getting a 100uF cap is hardly a problem with electrolytics. ... in another thread that you were aware of the voltage coefficient ...
    (sci.electronics.design)
Loading