Amp NOT retaining any voltage on the caps.

[Boxnix]

Is the danger from amps holding onto their voltage something that always happens or something that happens sometimes? Mine goes to zero then slowly builds up to around 16VDC on the positive side of the caps. I replaced all the caps and it still behaves the same. Are some amps just different or is it possible I have a problem?

[Jazz P Bass]

uhmm. Which amp are we dealing with.
The power caps holding voltage depends on the circuit.
When power is turned "off" some circuits are actually still operating, therefore they draw down the caps.

[Boxnix]

It's a 5E3. No mods that I'm aware of.

[Steve Conner]

That's dielectric absorption, or what used to be called "soakage". Perfectly normal behaviour for electrolytic caps.
Understand Capacitor Soakage to Optimize Analog Systems

Having bleeder resistors in the amp will minimize it, good idea for safety's sake. But 16V isn't enough to jolt you.

[R.G.]

The safety regulation people, who will tell you what is NOT safe, but refuse to tell you what is, say that voltages over 42V DC or peak AC is hazardous and should be protected against. They also say that lower voltages which can produce a significant amount of energy - think something like a 1F capacitor loaded up to 10Vdc - are a danger and must be guarded against human contact. Imagine dropping a wrench which landed firmly on the terminals of a car battery. The energy has to be limited to X, which I don't recall. But 16Vdc in a 20-50uF capacitor is under the voltage and energy limits which they say are definitely unsafe.

It's a really, really, really, really, really, really, really, really, really, good idea to put bleeder resistors on any amplifier filter caps, just to ensure that the main charge does finally come down to safe, and that dielectric soakage gets eaten up out of the caps eventually. Size the resistors by wattage, putting a dissipation of 1/4W, 1/2W, 1W, what ever you can spare from the power supply at normal operating voltage. Then put a resistor in there that's rated for twice the actual dissipation for long life and lack of overheating. This generally winds up being 100K to 470K at tube amp capacitor voltages.

By the way, the safety wizards are looking at audio speaker outputs and finding that some of THEM hit more than 42V peak AC, and scratching their chins... hmmm - that's kinda unsafe, even if it's not AC power, isn't it... hmmmm....

[Boxnix]

Sorry I did not phrase my question well. I'm not worried about the 16V, I just wondered if it was a problem that a normal 5E3 circuit would instantly discharge to zero. Maybe that circuit already has a bleeder circuit. I was more concerned I was bleeding to ground somewhere I should not be. It sounds and runs great, I'm sure everything is fine, just wondering.

[Bruce / Mission Amps]

The two power tube's plates (idling at pretty high current) are drawing all that current from the first filter cap when the amp is shut off.
It discharges that cap VERY quickly.
The second cap is connected to the power tube's screens and it will drain a bit too but, discharge backwards into the first filter cap since that is the section connected to the current hog... then the final filter cap is connected to the two preamp tubes, which are idling at around 4ma, so it will drain off too, but because the load is light, it will try to back fill the two previous caps.
This all takes place in a very short amount of time if you shut the amp off when the tubes are hot.
Try this with the amp cold, a NOS 5Y3GT rectifier and the 6V6s, 12AX7 and 12AY7 not even close to lukewarm (like 5-10 seconds of on time) and you'll see the filter caps stay charged up for a much longer time. Then you will get poked by high voltage....!!!

[Boxnix]

Thanks Bruce!

[Jazz P Bass]

Imagine dropping a wrench which landed firmly on the terminals of a car battery.

So do you think we should cut the wrench in half & add a current limiting resistor?
Or maybe outlaw all metal wrenches & make them out of plastic?

[kg]

It's a really, really, really, really, really, really, really, really, really, good idea to put bleeder resistors on any amplifier filter caps, just to ensure that the main charge does finally come down to safe, and that dielectric soakage gets eaten up out of the caps eventually.

the one and only time i really got socked with a lot of joules was because i started digging into an amp under construction that i had left overnight.

didn't have bleeders on it.

AFTER i got hit there was still a couple hundred volts left @ 150uF.

i got lucky.