Dropping B+ before filter caps

[EETStudent]

Is it possible to drop the B+(via resistor) before the filter caps. I got caps rated at 450V, but the incoming voltage will be around 480V. Maybe drop a resistor where the choke would be? Meanwhile, 480V would still go to the power tubes, but the voltage drop would go across the caps into preamp

[Bruce / Mission Amps]

Is it possible to drop the B+(via resistor) before the filter caps. I got caps rated at 450V, but the incoming voltage will be around 480V. Maybe drop a resistor where the choke would be? Meanwhile, 480V would still go to the power tubes, but the voltage drop would go across the caps into preamp

If the power tranny has a high voltage center tap, use a few 5 watt, reverse biased (cathodes pointed to ground), zener diodes in the lead to ground.
I'd use no more then three to five 9v-12v diodes in series.

[EETStudent]

Sorry Bruce, but didnt quite get that. Your sayin instead of resistor, go with 3-5 zeners in that place?

[cbarrow7625]

The zeners will provide a constant voltage drop rather than a voltage drop that varies under load.

With a resistor, you may get 450V with no signal but as soon as the power tubes begin to draw current ohm's law will kick in with a vengance and you will drop additional volts across the resistor.

It's the same kind of effect you get out of a tube rectifier...but in a bad way. You will get excessive voltage drop that will make the amp sound mushy & flabby.

The zeners will drop a constant voltage regardless of the current draw.

Chris

[52 Bill]

What I think you are asking, is can you reduce the voltage from your 480 volt power supply down to 450 volts, to work with the filter caps that you already have, but still use the 480 volts for the output stage. In an word, NO.

What Bruce is suggesting, will reduce the overall power supply voltage by adding the zeners to the output of the transformer.

What I would suggest, is to build a totem stack for your input capacitor. Look at the power supply design of most Fender amps, and you'll see two series wired caps with paralleled resistors.

The voltage of the caps in this case will add up, so 2-350 volt caps will handle 700 volts. The 2-220K resistors make sure that the voltage across each of the two caps will be exactly 1/2 the applied voltage. The capacitance divides when wired in series, so 2-70 uf caps will become the equivalent of a 35 uf cap.

Additionally, the two paralleled resistors will also act as bleeders for the power supply.

Hope this helps.

[stokes]

Sorry Bruce, but didnt quite get that. Your sayin instead of resistor, go with 3-5 zeners in that place?

What you are missing here is that Bruce is telling you to put the diodes between the CT and ground,rather than connecting the CT directly to ground.Not "instead of the resistor".You can use a number of smaller ones in series as Bruce has stated,or if you check with Mouser,they have larger ones.I have used a 50v 50 watt reverse diode in the same application with no problem.Just be aware you wont drop the full stated voltage,I think the exact voltage I used was about 56v,but it dropped forty something volts.Also this was in a 2X 6V6 amp,so I wasnt even close to the 50 watt rating of the diode.

[Bruce / Mission Amps]

Sorry Bruce, but didnt quite get that. Your sayin instead of resistor, go with 3-5 zeners in that place?

Hmm... I think 52 Bill's answer is probably more to the point if you are only trying to increase the voltage handling rating the filter caps.
Sorry if I misunderstood what you were asking.
Don't use power resistors for dropping B+ to the PA as it is a waste of current and the regulation of the B+ rail goes way off.

[EETStudent]

Yeh i believe Bill 52's answer is what i was looking for, but good information Bruce. So Bill, your saying that instead of having one 450V filter cap to ground as the input cap, i could use the arrangement that you mentioned to accomodate the extra voltage? And then the next cap may be able to handle the voltage after its dropped by the the resistor ahead of it.

[52 Bill]

Yes, that is exactly what a lot of manufacturers have done for a lot of years.

[EETStudent]

What if you put two caps in parallel, and then put a third one in series with those two. Would that effectively increase and voltage and also capacitance?

[bbartilson]

What if you put two caps in parallel, and then put a third one in series with those two. Would that effectively increase and voltage and also capacitance?

Caps and resistors add and subtract in value just like one another, but in reverse.

2 100 ohm resistors in series = 200 ohms
2 100 ohm resistors in parallel - 50 ohms

2 100 uf caps in parallel = 200 uf
2 100 uf caps in series = 50 uf

The formula is the same for resistors in parallel and for caps in series.

For each cap in the circuit, you get take the value, and make it a fraction - 1 over the value - in our example 1/100. You add each inverted value together, and invert that result 1/sum. That's the effective value of what you created.

Taking your circuit suggestion - two in series and one in parallel with those:

1 / (1/200 + 1/100) = 66.66667 uf


Capacitor life is also a affected by the voltage across them. It's okay to run 400v across 450v caps, but they don't last as long that way. What I'm saying is that it's not a waste to use two 450v caps in a ~500v circuit. Or to stack two in parallel and those resulting two in series to end up with both the capacitance and the voltage rating you desire.

But that all rests upon the design parameters of the circuit in question as to how much of either you need. I just answered the theory question for you. I didn't skim the rest of the thread to see what exactly you're trying to do.

I wouldn't suggest using three in the arrangement you suggested. If you're going to stack them, do it in singles or pairs.

-B

[Howlin' Mad Mac]

Yes you can put a resistor before the reservoir & create a voltage drop across it. I've seen this done & even done it myself however with a tube rectifier it will increase the sag effect. Now when I did this it was in a SE design (generally not known for much sag) & the rectifier was a hybrid 1N4007 full wave/5Y3GT combination (5Y3GT's don't have a lot of sag.) It is my understanding you see this often deployed in SS rectified amplifiers & some SE designs to do just that... create sag, although if this will work with your application I don't know enough to say. Anyway hope maybe this helps.

[Chuck H]

If you want 480Vp the easiest answer is to change the first filter arrangement to accomodate. If, for example, you want 100uf of filtering on the first node that will handle 500 volts you can use two 220uf/350 volt caps in series with a 220k resistor in parallel with each cap (helps equalize voltage across the caps). This would handle 700 volts. 480 volts would be a walk in the woods for this main filter.