Safety Bleed Resisor?

[elipsey]

How about about a 500k 3W metal oxide power resistor from 1st power supply node on cap can, to cap can ground? I figure this dissipates a little less then 1/2 Watt with 425V plate. Sound reasonable?

[nickb]

How about about a 500k 3W metal oxide power resistor from 1st power supply node on cap can, to cap can ground? I figure this dissipates a little less then 1/2 Watt with 425V plate. Sound reasonable?

How long do you want to wait? Time to discharge = 2.2 * R * C = about 2 minutes for a 500K resistor and 100uf total capacitance. Trade power dissipated versus your patience

I would say a 3W resistor dissipating under 1/2W is very conservative, and that is a good thing.

[ThermionicScott]

Sounds good to me. I'm guessing that if you powered up the amp without signal tubes, the B+ would be more like 450V, and it's still got plenty of power handling capability for that. Since I like to live on the edge, I'd probably find myself using one of my uglier 470k/2W's out of the junk drawer.

[potatofarmer]

If you have a standby switch between the first filter cap and the second, it might be worthwhile to put a second bleeder across the second cap. Easier than always remembering to leave the standby "on" when you shut it off.

[Gingertube]

Consider putting a Red LED in series with the resistor (at the 0V end). LED goes out, its bled. LED still RED - wait some more.
Cheers,
Ian

[nickb]

Consider putting a Red LED in series with the resistor (at the 0V end). LED goes out, its bled. LED still RED - wait some more.
Cheers,
Ian

or LED is RED means YOU'RE DEAD!

[ThermionicScott]

The sky is really the limit with how complicated and clever your bleeder circuit can be... remember that the only "safety" benefit you get is that the power supply caps will be completely drained by the time the next guy reaches in there with the amp turned off. One should NEVER take for granted that high voltages aren't present, though, and should NOT be poking around in tube circuits if they are an idiot or clumsy. (It goes without saying that bleeders convey no safety benefits whatsoever when examining an amp while it's turned on.)

Early amp circuits didn't need bleeder resistors at all, because the folks working on them knew what they were doing (or learned quickly!) That, and hot tubes do a pretty good job of draining power supply caps to a much safer level when the amp is switched off -- that's why amps rarely have bleeder resistors on any power supply caps after the standby switch.

I say this because I'm concerned about people getting a false sense of security and letting their good habits atrophy if they have extensive bleeder resistors and indicators built into the amp. Safety always starts with knowing what you're doing, and being careful.

[Chuck H]

I want to back this strongly... Bleeder resistors not withstanding, practice safe probing procedures as if the amp were live until you have actually tested to see with your own eyes that there is no significant voltage in the amp. To do otherwise is taking an unnecessary risk. Problems with potentially horrifying outcome that can be avoided with seven seconds of effort.

As such, bleeder resistors are a matter of convenience only. Also, since the old amps were made so damn well there are a bunch around that don't have them. I have a clip lead with a 220k/1W resistor in line. It gets attached to the top of the main filter and ground and the standby switch gets flipped to "play" mode before I even reach for my meter. Then I test for live volts. And I poke hard too. Old amps often have an oxide or flux layer on the joint that can interfere with getting an accurate measurement. Please, everyone, make every effort to be safe and never get impatient in these matters.

[Old Tele man]

Back in the "military vacuum tube" days, the "sizing" rule for a bleeder resistor was that it draw/consume 10% of the TOTAL circuit current draw. Why? Because it's presence actually performs a modicum of *voltage & current" regulation.

[Leo_Gnardo]

How about about a 500k 3W metal oxide power resistor from 1st power supply node on cap can, to cap can ground? I figure this dissipates a little less then 1/2 Watt with 425V plate. Sound reasonable?

Sounds good to me.

I've been experimenting with replacing "dynamite stick" 30 uF caps in some Ampegs (B15, B25) also series-connected 100 uF caps (V4), with 30 uF film caps rated 800V (Panasonic) and 900V (Cornell-Dubilier). Parallel pair 60 uF for replacing the series 100's.

First time I tried this on a V4, I found the voltage starting to run away past 900V before the standby switch was closed. "Bleeder resistor" I thought, and stuck on an IRC 1000V rated 3W 330K, problem solved. While in standby mode the cap charges to 550V, well within the safe range. Anybody who tries the film cap solution, take notice, and DO apply a bleed R for peace of mind.

The film-filtered Ampegs work & sound fine. So far so good.

[Old Tele man]

See page 344 in "Principles of Electrical Engineering Series (MIT): APPLIED ELECTRONICS," 2nd Ed, by Truman S. Gray:

http://www.tubebooks.org/Books/gray_...tronics_2e.pdf

[R.G.]

I used the LED in series with the bleeder resistor in the Workhorse amps. I hadn't see that used til that time. After that got shipped, I got to thinking: what you really want is both a bleeder and an indicator. The LED series with the bleeder is OK, but the LED can get very dim when the voltage is hazardous because you have to size the resistor to not over current the LED at full voltage, so the current and light output drop a lot as the voltage goes down. It gets really dim.

A resistor in series with an NE-2 is better - the NE-2 goes completely dark at about 60-70V.

The LED with the bleeder is a good trick to remember. I use more than one in the Workhorse amps. Using one at the first filter cap, after the inductor, and at the end of the preamp B+ string lets you tell at a glance whether plate DC is making it down to the preamps and where it stops. NE-2s would do the same, probably at lower currents.

[loudthud]

Neon bulbs really like to oscillate, and they can do it at ultrasonic and low RF frequencies. They tried using them as panel indicators at Tektronix on the 5000 series lab scopes. They quickly went back to incandescent bulbs and LEDs.