What's the best way to drop B+ voltage by 100v or more?

[loudthud]

Was the Heathkit supplying the bias voltage? It looks like you are using fixed bias derived from the HV winding of Big Blue through a 470K resistor.

[stokes]

Like Enzo says,shorten those red wires feeding the power tube grids.By pulling them out from under the board and arching them over you can cut 2 or so inches off.If that doesnt cure it,I would recommend changing those filter caps.I've seen oscillation problems in too many newer Fenders from those cheap asian caps.I've even seen it in a couple amps less than a year old.Always the preamp cap failing to decouple the stage properly.Check to see if you have ac voltage on that last preamp filter cap.That 470 ohm between the main and screen tap should be 2 watts or more,that 1/2 watter is too small.

[mort]

Was the Heathkit supplying the bias voltag

e? It looks like you are using fixed bias derived from the HV winding of Big Blue through a 470K resistor.

Ah you are correct on how it's wired. And I didn't supply anything there when trying to completely power the amp with the Hsathkit. Good catch. Maybe that's why the Heathkit was showing weird current draw?

[R.G.]

On lowering B+:
Any scheme that does not involve inductors, capacitors, or transformers as the main voltage drop elements will dissipate the difference in voltages times the whole current as heat. A zener or amplified zener will dissipate the same heat as a resistor that drops the same voltage at the given current. The zener/amplified zener has the advantage of having a fairly constant output voltage with different currents, while a resistor dropper will vary with current by ohm's law.

A transformer bucking setup could drop the voltage fairly non-dissipatively, as could the very sane idea of inserting an inductor to make an inductor input filter. That is probably the most conceptually simple idea here, and wastes the least power as heat.

Other possibilities include using a high frequency chopper to step down the voltage; this is the same process as the inductor input filter, but lets you diddle with the pulse width and make the output be anything you like instead of the fixed 0.639 times the capacitor input voltage. If you're getting 450 with capacitor input, you'll get (about) 450*0.64 = 288.

If, that is, you always have a minimum load of some amount. The relative values of the inductor and the cap will set the minimum current load to hold it *down* to 288-ish. If the current drops below the "critical current", the voltage will rise, and at zero load hit 450, the cap-input filter value again, as there's no current in the inductor to cause voltage to drop.

[Alan0354]

I've got a little 5F10 Harvard circuit that was built up from spare parts a while back. The PT puts out a pretty high HT and I'm getting ~450vdc off the rectifier. If I wanted to knock that sucker down to 300-350v, what's the best way to go about it?

Fwiw, the problem I'm having is when the volume is up around 75% or higher, the amp squeals/oscillates/whatever you wanna call it. Sounds fine otherwise. In searching this topic I ran across the advice to open the nfb loop but that didn't resolve it. I'm thinking maybe the voltage is just too much.

I drop 40 to 50V on my PT that was for 6L6 to run 6V6. I use a chain of 5W zener diodes 1N53xx type. Of cause, the right way is to use MOSFET and all, BUT you need heatsink. So using a chain of say 8.2V zeners might not be that bad. If you use 10 of the 5W zeners, you don't need heat sink unless you are pulling over 150mA. If you pull higher than that, using lower voltage zener and use more!!!! Believe me, it's still a lot easier than putting in a heat sink for a MOSFET.

I ended up using MOSFET to do power scaling, it was really tempting to even use a chain of zener to do power scaling.

The key is you need to have a large value (100uF)electrolytic across the zener chain. Preferably having a 0.1uF small cap in parallel. Zener makes a lot of popping sound if not bypassed.

[Mick Bailey]

Believe me, it's still a lot easier than putting in a heat sink for a MOSFET.

The chassis is the best heatsink - no separate unit required. One 3.5mm hole is all that's needed. Components mount on the legs of the MOSFET. Neat, compact, reliable and 100% effective.

[Alan0354]

The chassis is the best heatsink - no separate unit required. One 3.5mm hole is all that's needed. Components mount on the legs of the MOSFET. Neat, compact, reliable and 100% effective.

My first Amp with 6V6 got pretty warm, I have the transistor mounted onto the chassis inside and have a 12W heat sink on the opposite side of the chassis. If you draw 150mA and drop 100V, that is 15W. In my 6L6, I use two MOSFET in parallel inside the chassis and heat sink on the other side.

It is good you mount components on the legs. Some people hang long wires. You need to be careful with MOSFET as they can oscillate. When MOSFET oscillates due to parasitic, it can blow because the gate and drain forms a tank circuit and easily over voltage the gate. It die a quiet death, no smoke, stone cold and drain shorted to the source. This is quite common. You have to put a resistor directly on the gate lead instead of wire to the control circuit.