Need help identifying location for standby switch

[thehoj]

safer still use a 330k resistor, the time to discharge (~5RC) isn't too much slower.

220k x 40uF x 5 = 44s
330k x 40uF x 5 = 66s

You have to be real fast with the screwdriver to get the case off and hit the caps in these times...

If I did use 330K resistors I could probably get away with 1/2w resistors..?
The reason I ask is because I don't have 1w or higher resistors above 47k.

[defaced]

The resistors are just a path to ground and a voltage reference, so the size isn't really that important (within reason). 330k with 300 volts across it gives you a dissipation of about 0.27 watts, so you're safe with a 1/2 watt resistor. Since in reality the voltage across the caps (and resistors) will be lower, more like 240v, they'll typically be operating with less dissipation than the 0.27 watts from above, actually more like 0.17 watts.

[Old Tele man]

...the "rule-of-thumb" for "sizing" the bleeder resistor is that its value should draw about 1/10th of the 'average' current...give or take.

...so, if the average idle current (all tubes) is 80mA, then:

R = (450V)/(0.080A/10) = 56.3K

...330K would draw less than 682 µA...sorta low.

...another "benefit" of the bleeder resistor is that it acts as a "constant" load on the capacitors which help smooth power supply ripple, especially downstream at the preamp tubes.

[thehoj]

So I'll end up with this

[Old Tele man]

...not quite, you only need the first two caps, not one across each following down-stream capacitor.

...they're "needed" across the "stacked" capacitors to "equalize" the voltage drops/loads across the two capacitors in series...without them, one capacitor or the other will "hog" the voltage (over-voltage) leading to eventual (soon) failure. If you're familiar with how transistors in parallel will have one transistor "hogging" more current than the other...here, with capacitors, it's voltage that gets "hogged."

[Enzo]

I would nit pick that the purpose of the standby switch is to silence the amp, not to kill B+ to the power tubes. Killing B+ will silence the amp of course, but so will grounding the grids, killing the screen supply, lifting the cathodes, etc. ALl of which and more are used to make standby switch functions in amps.

Leaving B+ on power tube plates and killing B+ to screens and other downstream B+ nodes is how Peavey has wired standby switches on their amps for years.

[thehoj]

...not quite, you only need the first two caps, not one across each following down-stream capacitor.

...they're "needed" across the "stacked" capacitors to "equalize" the voltage drops/loads across the two capacitors in series...without them, one capacitor or the other will "hog" the voltage (over-voltage) leading to eventual (soon) failure. If you're familiar with how transistors in parallel will have one transistor "hogging" more current than the other...here, with capacitors, it's voltage that gets "hogged."

Is your use of the words "eventual (soon) failure" exaggerated at all? These resistors were not in the original design of the amp I'm working on, and the "stacked" caps without bleed resistors did have a decently long life.

Here is a revised schematic:

[Enzo]

The reason the caps lasted without the resistors is that they had reasonably close characteristics on their own.

Getting away with it doesn;t make it a good idea. For the cheap price of the two resistors, you get insurance against that potential failure mode.

[R.G.]

If I did use 330K resistors I could probably get away with 1/2w resistors..?
The reason I ask is because I don't have 1w or higher resistors above 47k.

So wire four 47K resistors in series. That gives you 184K ohms, but four times the dissipation ability.

...the "rule-of-thumb" for "sizing" the bleeder resistor is that its value should draw about 1/10th of the 'average' current...give or take.

I'm not trying to be argumentative, but if you're putting in bleeders for safety, you design them to pull the power supply below 42V before you can get the covers off. We were forced by corporate edict to get capacitors below 10J of charge in 10S after turnoff, but these aren't mainframe computers. Under 42V is what most safety agencies think is a safe(ish) voltage.

I believe the 10% rule was for inductor-input filters.

...another "benefit" of the bleeder resistor is that it acts as a "constant" load on the capacitors which help smooth power supply ripple, especially downstream at the preamp tubes.

Again, I think this was for inductor input filters. With capacitor input filters, any current increase increases ripple voltage.

Whenever you see caps in series, you should feel this itch to put equalizing resistors in there. If you don't any difference between the two capacitors' leakage or loading will shift the DC voltage balance, and may cause one to be overvolted just because its partner was leaky. However, that can't happen here, because the rectifier setup is such that the caps are both charged to the transformer voltage each half-cycle of AC, but discharged in series, so the transformer forces them to share voltage equally. It's putting equal voltages into them. But that's a special case of how this power supply is wired. If you drove the same two caps from a rectifier bridge, you'd need the resistors to equalize DC.

In this case, they're just to keep the caps pulled down to safe voltages when you service it.

[Steve Conner]

You have to be real fast with the screwdriver to get the case off and hit the caps in these times...

What if you have the amp already out of the case and are tinkering with it on the bench?

As others have pointed out, the balancing resistors are a good idea for series-connected caps. I think in this particular case, the voltage doubler circuit forces equal voltage across the two series capacitors, because it charges them individually: one on the positive half-cycle, the other on the negative. So the resistors are probably not strictly needed for balancing. But you need them as bleeders anyway.

[soundmasterg]

I've got an amp project that started life as a Conn organ pwoer amp. It had a common power transformer using a half wave voltage doubler circuit, with three output transformers, each with it's own pair of power tubes and a phase inverter tube. It was all cathode biased and I added a seperate preamp chassis, (one preamp for each power amp, so total of three amps in one head, sharing a power transformer....I know I'm crazy but its been a godo learning project) made everything fixed bias, and added a full wave doubler. I had to use screen standbys to be able to operate each power section seperately from each other and it works great. I also had to add some bias tracking circuitry.

I'm also modding a Sovtek MIG 100U into a Vox AC100 clone, and I used their power supply. They had a series connection for the main caps on the B+ but no resistors, and they were using a bridge rectifier. They did have the power trans winding center tapped, with the CT going to the junction of the two caps so they see equal voltage, and the bleeder resistors aren't required. Moreover, the amp discharged itself as quick as any typical amp with the bleeders installed.

Greg

[thehoj]

Interesting discussion. To clarify for my particular situation though, the two 330k resistors over those first two series caps are all I need to discharge all of the filter caps in that circuit?
I'm not concerned about the length of time. I will make sure that I count a minute before going inside of the amp.

[Old Tele man]

Interesting discussion.

...there are many "facets" to knowledge and experience...six, half-dozen, square-root of 36.

To clarify for my particular situation though, the two 330k resistors over those first two series caps are all I need to discharge all of the filter caps in that circuit?

...yes.

[tubeswell]

With bleeder/equalising resistors, don't you also have to take account of the start-up surge when calculating power rating? (I was always under the impression you should use 2W resistors around the 220k mark for each cap?)

[thehoj]

I ended up getting some 220k 1w resistors from a friend, hopefully that'll be fine.. Better than 1/2w anyways.