Chassis voltage question

[Greggy]

Got a question:

I recently completed a Class A single ended amp kit from a reputable dealer. The first time I fired it up the tone was geat but I could hear an intermittent crackling noise from the speaker. Knowing of the dangerous voltages and that I probably had a bad solder joint, I emailed the dealer and he recommended running the amp until the tubes were hot, then unplugging the amp from the wall outlet while leaving the standby and on/off swithches in their on positions. Said the hot tubes would act as a load on the filter capacitors and would drain the voltages.

Just wondering what y'all have to add to this, and more importantly, does this technique work with all tube amps or just single ended Class A types?

 

[aeolian]

I've seen this and shorting the power plug with the switch on. Making up a bleeder jumper is so easy that it's not worth poking in the amp without the caps shunted to ground.

I took a test cable with the insulated hooks and spliced a 10K 2W resistor in the middle and covered it with shrink sleeving. Takes only a minute to drain off the cabs and I leave it on while I'm doing anything in the amp. Caps have a memory thing and can bouce back to a charge even after they're been "drained".

 

[Greggy]

Originally posted by aeolian



............

Caps have a memory thing and can bouce back to a charge even after they're been "drained".

This caught my attention. You mean they can recharge even after pulling the wall plug with the switches on as I did?

 

[loverocker]

No they can't. If there's no source of charge (and the only sources of electrons are the power socket and the other caps), there's no such memory effect.

Both methods will work fine with all types of amp although with an amp I've never seen before (particularly high-tech amps which might use relays in a clever way) I reckon there's always a possibility that a cap remains charged so I sometimes use both.

 

[Greggy]

Originally posted by loverocker
No they can't. If there's no source of charge (and the only sources of electrons are the power socket and the other caps), there's no such memory effect.

Both methods will work fine with all types of amp although with an amp I've never seen before (particularly high-tech amps which might use relays in a clever way) I reckon there's always a possibility that a cap remains charged so I sometimes use both.

Thanks. I fixed the crackling noise issue (had a wire whose insulation burnt through when soldering another wire, and was touching the chassis so probably shorting out). So hopefully I'll never have to open the chassis again.

 

[loverocker]

Originally posted by Greggy
hopefully I'll never have to open the chassis again.

Uhuh... that's how it always starts. In a coupla months you'll find yourself wondering about changing some cathode bypass caps, or tweaking for a bit more gain... then it's only a matter of time before homebrewing an 18W or Tweed Deluxe seems a fine idea! Welcome to the club

 

[Greggy]

I hear ya. The Allen Accomplice kit looks tempting. A deluxe reverb clone. Been wanting one of those for awhile.

 

[John Phillips]

Originally posted by loverocker
No they can't. If there's no source of charge (and the only sources of electrons are the power socket and the other caps), there's no such memory effect.

The memory effect is real and is caused by the electrolyte in the cap acting like a battery. The charge comes from the chemicals in the electrolyte as it slowly returns to its unpolarized state. Leave a high-resistance voltmeter connected across a filter cap (one without a bleed resistor of course) and see what happens, if you don't believe this.

The voltage developed is not usually very large though, and it isn't enough to worry about as a hazard I don't think. But it is enough to confuse resistance readings sometimes if you're trying to measure components that are still in the circuit. It's occasionally enough to cause a little spark if you accidentally short out a cap too, which will make you jump if you think you've drained them...

 

[loverocker]

I sit corrected. But just to clarify here, we're talking of voltages of what order of magnitude?

Having been bitten once (by a mere 240V AC!) I became ultra-cautious, and have often remeasured across caps that I *knew* were discharged (by both of the two methods mentioned above) hours, days, weeks before. And never have I seen anything more than a (totally harmless) volt or so - about what I expect to be left on a cap given that discharging a cap *completely* takes forever, thanks to that pesky time constant and the way that the maths works.

 

[Wakarusa]

Actually, there's a formula you can use to determine how long it takes to discharge a cap through a given resistance

The "recharge" voltage -- no idea as to its magnitude, but, through empirical testing , there isn't much stored energy behind it. Like John said -- just enough to screw with your DMM.

For shorting probes around here I found a batch of old (and fat) test probes. Just big enough that you could solder in a 1W carbon film and still fit the barrel on. 'Gator clip goes on the other end. If memory serves I think I put 1Meg resistors in 'em.

 

[Greggy]

Thanks for the responses.

I'm interested in making one of these "bleeder jumpers" or "shorting probes." I assume you would have a clip on one end attached to a leg of the cap and the other end connected to ground. Does anybody have a picture or drawing of one of these. I'm not 100% for sure how to make one from the descriptions in this thread. Thanks.

 

[loverocker]

Greggy - buy a 10K 2W resisitor, a couple of short lengths of 1A/600V wire, two insulated mini-crocodile clips and some heatshrink tubing. Then it's as simple as:
clip - wire - resistor - wire - clip
while using using the heatshrink to make sure that no metal part except the tips of the crop clip jaws is touchable.

To use it you clip one end to the ground tab on the chassis and then clip the other clip to the + side of the cap. For most amps it takes just a few seconds to get the voltage down to around 10V. Here's a great page for typing in typical values and seeing how the time constant works with discharge: http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capdis.html#c2

I know that some guys like to leave it connected while they work on the amp. Which is OK as long as you remember to remove it before switching back on

Wakarusa Amp - 1Mohm - that makes it take a long time to discharge!

 

[John Phillips]

Originally posted by loverocker
I sit corrected. But just to clarify here, we're talking of voltages of what order of magnitude?

I've never seen anything higher than 50V on a 500V cap (not even sure that high, maybe 20-30V), and usually much less than that - ie a few volts.

It's definitely been enough on a few occasions to cause sparks when working on caps though, and accidentally shorting them.

I think (but I'm not sure) that it's worse the older and more degraded the caps are, too.

Of course, it will only occur on amps like Marshall 50s, Deluxe Reverbs etc where there are no 'stacked' first-stage filter caps with resistors across them to act as built-in bleeders.

I learned the hard way about residual charge (not memory, just from not draining them) in amps like these, when I got a really nasty shock from a 50W Marshall that had been off for at least a couple of days - I checked the cap and found over 200V on it, after some had already discharged through me . Be sure you know what you're doing!

Nowadays I tend to leave an old analog voltmeter (fairly low resistance) clipped across the main cap while I'm working. This not only acts as a good draining method - you can 'track' the voltage as it falls by switching to progressively lower ranges - but it gives a good visual indication that everything is as it should be while you're working.

Like most techs, I've had a fair number of shocks, of varying severity... I'm still here, but it is very highly not recommended.

 

[Greggy]

Originally posted by loverocker
Greggy - buy a 10K 2W resisitor, a couple of short lengths of 1A/600V wire, two insulated mini-crocodile clips and some heatshrink tubing. Then it's as simple as:
clip - wire - resistor - wire - clip
while using using the heatshrink to make sure that no metal part except the tips of the crop clip jaws is touchable.

To use it you clip one end to the ground tab on the chassis and then clip the other clip to the + side of the cap. For most amps it takes just a few seconds to get the voltage down to around 10V. Here's a great page for typing in typical values and seeing how the time constant works with discharge: http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capdis.html#c2

I know that some guys like to leave it connected while they work on the amp. Which is OK as long as you remember to remove it before switching back on

Wakarusa Amp - 1Mohm - that makes it take a long time to discharge!

Just what I was looking for. Mucho thanks!