Bleeding caps safely

[EFK]

Can someone offer a definitive yes or no answer on the following methods of bleeding voltage off the amp to ensure safety when working inside? I've read of many methods but tough to find definitive, 'yes this is safe, or 'no you're going to get zapped.'

(1) Amp powered on and standby off - ready to play in other words - pull power cord out of wall.

(2) Jumper from pin 1 of preamp tube (lets say V1 unless it doesn't matter which) to ground. Any sparks or fireworks here?

[Diablo]

Originally Posted by EFK

Can someone offer a definitive yes or no answer on the following methods of bleeding voltage off the amp to ensure safety when working inside? I've read of many methods but tough to find definitive, 'yes this is safe, or 'no you're going to get zapped.'

(1) Amp powered on and standby off - ready to play in other words - pull power cord out of wall.

(2) Jumper from pin 1 of preamp tube (lets say V1 unless it doesn't matter which) to ground. Any sparks or fireworks here?

Method 1 does not necessarily work on all amps because the large filter capacitors can still hold a large charge (high voltage). If there are bleeder resistors built into the amp, then this method can work if you allow enough time for the charge on the capacitors to bleed off through the resistors to ground.

Method 2 will work, but I would add a large capacity ( high wattage and low resistance) resistor in the middle of your jumper cable so you don't get any big sparks. I use a 50W-300 ohm resistor to dissipate the energy. Make sure your jumper cable is well insulated so you don't accidently become part of the grounding path. I like to ground out pin 3 on my power tubes because I can hear when the cap charge is gone by listening through the speakers. You will be surprised how long the charge takes to discharge completely. Then wait a minute and go back and discharge again through pin 3, and you will find out there is still some charge.

[Enzo]

I'll agree and offer a different outlook.

In many amps, the standby switch does nothing to the discharge process. MAny Marshalls for example switch off the high voltage AC behind the rectifier for a standby. So the bridge rectifier diodes will block any discharging there. You are at the mercy of any bleeder resistors. You can just pull the plug from the wall, and it won;t matter which way teh standby is set.

On other amps, the standby is between the first filter stage and the second. SO lacking bleeders, the amp can discharge all except the main filter stage. You don;t want that, so there you SHOULD leave the standby switch in the run position.

The message so far then is to always leave the standby in RUN whether it matters or not.

If the amp has bleeder resistors then, the amp will discharge itself. If it doesn;t then it sits there charged. I don't consider the above to be a reliable method of discharing the filter caps. I always check with a meter. If they are discharged, great. If they still hold a charge, then some sort of active discharging is required, such as below.

CLipping a ground wire to pin 1 of a 12AX7 will slowly discharge the power supply. SInce the typical 12AX7 has a 100k or larger plate resistor, I don;t see the need for an additional resistaor for discharge. If you had 400v on the tube's B+, then that 100k resistor would limit discharge current to 4ma. I would not concern myself with any tiny spark that 4ma might generate.

That works reliably, but slowly. Watch with your meter and see.

Diablo's special jumper is a good way to discharge faster. Pin 3 of a power tube - assuming 6L6 type socket wiring - is convenient. I usually try to find the main filter itself. Ground a clip wire witha resistor in the middle of it, and use it to ground off B+ at pin 3 or wherever. He uses 50w 300 ohms, I usually grab a 1000 ohm 10 watt. It isn't a critical thing. The resistor doesn;t have to be as heavy as it would if the circuit were remaining powered. It might start at 500v, but it drops a ton almost instantly, so the resistor doesn;t have a chance to heat up much.

[Albert Kreuzer]

Originally Posted by Enzo

The resistor doesn;t have to be as heavy as it would if the circuit were remaining powered. It might start at 500v, but it drops a ton almost instantly, so the resistor doesn;t have a chance to heat up much.

Except when you forget to take it out before powering up (don't ask...).

Cheers,
Albert

[Enzo]

Oh, believe me ALbert, I have done my share of bone headed things.

[EFK]

Thanks very much for the clarifying responses guys. I appreciate it.

I was checking B+ and had a meter clipped to pin 3 when I pulled the plug from the wall - I watched the voltage drop down over the course of @ 5 seconds to 2 volts (from 481). Is this indicating the presence of bleeder resistors?

[Diablo]

Originally Posted by EFK

Thanks very much for the clarifying responses guys. I appreciate it.

I was checking B+ and had a meter clipped to pin 3 when I pulled the plug from the wall - I watched the voltage drop down over the course of @ 5 seconds to 2 volts (from 481). Is this indicating the presence of bleeder resistors?

Yes, or a short in the wiring....
Do you have a schematic for this amp?
Verify on the schematic that there are bleeder resistors.
Also, while you're in the amp, you can measure for voltage on all the electrolytic caps just to make sure it's safe inside.

[Chuck H]

I don't think that indicates the presence of bleeder resistors. I've noticed that most new electrolytics simply don't hold a charge very long at all. Older amps could store a charge. But with new electrolytics even without bleeders I've measured only a few volts on the main cap in the time it takes to pull a chassis. And this is with new Sprague Atoms. I have an amp opened up right now with Atoms and no bleeders. I'll monitor voltage from the time of shutoff and report back.

As a peripheral discussion (not hijacking) what difference is there between the old caps that do hold a charge and newer caps that don't? And how does this effect tone and amplifier performance?

Chuck

[Chuck H]

Ok. I just checked my findings. I turned the amp on and even played it for a bit to get the caps "working". Then with my meter in place and reading 356 volts on top of the main filter I switched the amp off. In about five seconds the voltage dropped to 10 and hovered there. This amp has Sprague Atoms and no bleeders. It's about a year old and exhibits no ghosting hum or any other problems. I've seen similar behavior on other amps with new electrolytics.

Chuck

[Diablo]

Boy that's wierd Chuck. Caps are supposed to store charge, that's their only purpose. I have no idea if electrolytic caps are of different composition today and yesteryear.

[martin manning]

I'm thinking that since the cathodes are at working temperature, the tubes will continue to conduct idle current for a few seconds (until they cool down) after you cut the power. That might be long enough to drain most of the charge.

MPM

[bnwitt]

method 1 (leaving power and standby switches on and unplugging the amp from the wall while strumming your guitar) will take the amp down to about 20 volts.

Method 2 will drain the entire power supply through the 100k plate resistor safely and take it down to 0 volts if you keep the jumper in place for about 20seconds.

Method two is my method of choice.

[Chuck H]

Originally Posted by Diablo

Boy that's wierd Chuck. Caps are supposed to store charge, that's their only purpose. I have no idea if electrolytic caps are of different composition today and yesteryear.

I guess it's not that wierd. I've seen this behavior repetedly with new caps. I suppose they do store a charge long enough to be effective as a reserve for audio power amps but don't hold a charge long term like a battery...Like the older caps did. In fact, if you research it you'll find that you have to get a specialty cap now if you intend to use it as a battery or power fault backup.

Has anyone noticed a tonal difference between current production caps that are prone to this behavior and older caps that hold a charge longer than a half a second?

Chuck

[Chris / CMW amps]

Originally Posted by bnwitt

method 1 (leaving power and standby switches on and unplugging the amp from the wall while strumming your guitar) will take the amp down to about 20 volts.

I know certain amps that will still have enough juice to shock you with this method.

Once a year I do forget and it reminds me to always use this method:

find main filter cap or a B+ point (standby on, power OFF, mainscable not connected to mains) and then putting a 1k/>5W from that to ground.

[Diablo]

Originally Posted by Chuck H

I guess it's not that wierd. I've seen this behavior repetedly with new caps. I suppose they do store a charge long enough to be effective as a reserve for audio power amps but don't hold a charge long term like a battery...Like the older caps did. In fact, if you research it you'll find that you have to get a specialty cap now if you intend to use it as a battery or power fault backup.

Has anyone noticed a tonal difference between current production caps that are prone to this behavior and older caps that hold a charge longer than a half a second?

Chuck

A little googling showed that some capacitors are built with an internal bleed resistor so that the cap discharges within a few seconds of powering off. It's a safety feature. Not sure if this is universal in modern electrolytic cap production.

[bnwitt]

Originally Posted by Chris / CMW amps

I know certain amps that will still have enough juice to shock you with this method.

Depending on how well you're grounded, 20 volts is enough to shock you too.
I've been nailed pretty good off of a 24 volt power supply before.

[Enzo]

And as someone mentioned, the tubes continue to conduct. Try the experiment with the tubes removed and see if it drops as fast.

[Diablo]

If I get energized, I may just charge up a new Illinois cap with my amp and some alligator leads, then disconnect the leads and wait 5 minutes to measure the voltage stored in the cap.

[Diablo]

Originally Posted by Diablo

If I get energized, I may just charge up a new Illinois cap with my amp and some alligator leads, then disconnect the leads and wait 5 minutes to measure the voltage stored in the cap.

OK, did the experiment and this is what I measured.
22 mfd/500V Illinois Capacitor charged to 417 volts.
Removed voltmeter, then removed alligator clips connecting cap to power supply. Waited 5 minutes and measured 280 volts on cap.

Then I repeated the experiment with a 40 mfd/500V cap (twist lock triple), 36 years old removed from my Ampeg V4. I only charged the cap up to a little over 400 volts because it took quite a bit longer to charge and the charge rate slowed down a lot once it reached 400 V. Removed voltmeter, then alligator clips, waited 5 minutes and measured 164 volts.

I conclude that new electrolytic caps store charge at least as good as old electrolytic caps. Both can have dangerous amounts of voltage stored after 5 minutes. Also, the voltmeter drains current and voltage much faster than the internal leakage of the cap.

[Chuck H]

Originally Posted by Diablo

Also, the voltmeter drains current and voltage much faster than the internal leakage of the cap.

This, along with the tubes current draw continuing when the amp is shut down must be responsible for my test results. I left my meter connected for the entire test and I only shut the power switch, not the standby.

Chuck

[EFK]

So the meter itself can act as a power drain? That's extremely interesting and would explain to some extent why I got the readings I did; I too left it hooked up the entire time.

[defaced]

A volt meter should have a very large internal resistance, on the order of megaohms. This is to not load down a circuit while taking readings. However for this kind of study, it may be enough to effect the results.

[Enzo]

10 Meg is a very high resistance, but it is a lot lower than an open. SO it will discharge a cap. Also, when you try to measure something like signal level in a circuit where the circuit elements are 1 meg pots and resistors and 470k resistors and the like, then it becomes a significant circuit load and skews readings. From time to time I have to suggest to someone to conect their meter in parallel with some high value resistor to measure the voltage across it instead of readign to ground from each end and subtracting. Meter impedance is why.

[Robmosis]

Could someone please explain what the B+ point is? I have heard this a lot but am not sure what it is.
Also what does a bleed resister on an amps schematic usually look like? I have a schematic for my amp but am in the precess of learning how to read and understand it. I will post it for you guys/gals.

[Diablo]

B+ refers to the high DC voltage supplied to the power tube plates. Typically 430-570 volts, and it's positive.

Bleed resistors are R133, R134, R135, 136, 138, 139, 141, 142.
They are in parallel with the big value filter caps.

That sure is one complicated schematic, not what I would want to tackle for a first try at building/modding/repairing.

[Robmosis]

Thanks very much Diablo. Yea my noob head spun when I found the schematic for my amp. But I am learning off of other older schematics, most from Dave Hunters book, not the one for my amp. Just like to try understand different areas of schematics.
Thanks again Rob

[Steve A.]

Originally Posted by Diablo

OK, did the experiment and this is what I measured.
22 mfd/500V Illinois Capacitor charged to 417 volts.
Removed voltmeter, then removed alligator clips connecting cap to power supply. Waited 5 minutes and measured 280 volts on cap.

Then I repeated the experiment with a 40 mfd/500V cap (twist lock triple), 36 years old removed from my Ampeg V4. I only charged the cap up to a little over 400 volts because it took quite a bit longer to charge and the charge rate slowed down a lot once it reached 400 V. Removed voltmeter, then alligator clips, waited 5 minutes and measured 164 volts.

I conclude that new electrolytic caps store charge at least as good as old electrolytic caps. Both can have dangerous amounts of voltage stored after 5 minutes. Also, the voltmeter drains current and voltage much faster than the internal leakage of the cap.

When using a DMM to measure voltages like that you might also measure the currents involved- or measure the voltage drop across a resistive load.

When using a DMM in HVAC repair work I often got what I called "phantom voltages" when measured without a load, but would disappear once a load was hooked up. I never got phantom voltage readings with an analog meter since the meter itself was a more significant load. BTW the phantom voltages were often a parasitic effect like with a deenergized wire that was run in a conduit with an energized wire. Or a power transformer with an open internal thermal fuse on the secondary.

However, I am a firm believer in the power of electrolytics! My question was just on the measurements you posted (you would always measure the voltage of a battery under a significant load).

Another test that you can make on new and old caps- first charge them up as you did and then drain the charge with a bleeder jumper. Remove the jumper and then check for voltage after 5 minutes. (When I've done this I've gotten small sparks indicating the presence of some voltage.)

FWIW I've trashed a few digital meters measuring electrolytics- these were not Flukes but they never did give me accurate dc voltage readings after that...

Steve Ahola

P.S. So nobody here suggested the tried'n'true method of old school electronics techs- BAS: Big Assed Screwdriver... Place the base of the screwdriver shaft firmly against the chassis and just aim the tip at the ungrounded lead on the cap. Keee-r-r-r-r-ackkkkk!