PSU question

[iknowjohnny]

How do you all determine what voltages you use for the preamp and PI tubes when you build an amp thats not a clone? I've experimented a lot and pretty much know about what i like, but it's 100% and ear thing and i'm just wondering how others who actually know what they're doing (:D) determine what voltages to use.

[Merlinb]

I don't much care what the preamp voltages are- they will simply come out "a bit lower than the PA voltages". So once I have designed the PA using a readily available power transformer, my preamp voltages are pretty much chosen for me. Since the preamp supply voltages don't really matter much to the design (within normal limits) I can then set about biasing and tweaking the preamp to get whatever tone I'm after. Only the relative voltages matter, not the raw supply voltages.

[stoo]

Every tube has a set of characteristics that have been graphed and charted ie. plate current vs bias vs plate voltage ect. I'm not really up on reading these charts but you can learn to use them. Just try looking up ..say....12ax7 specs.
Stew

[iknowjohnny]

Merlin, when you say "chosen for me" do you mean that you just use a particular value for the dropping resistors that you just consider typical? Do you just throw all 10k's in there or is there a thought behind the choice?

[Merlinb]

Merlin, when you say "chosen for me" do you mean that you just use a particular value for the dropping resistors that you just consider typical? Do you just throw all 10k's in there?

Pretty much, yes. A 12AX7 requires so little current that the difference between using a 1k or 10k dropping resistor is pretty immaterial. Obviously I will calculate what the node voltages will turn out to be, to check they won't be anything stupid, but provided they're in the 250 to 450V range, I'm usually happy. I'm not aiming for a "magic voltage" or anything.

Although I will admit that, in general, the higher the voltage the better. So if my raw supply is 400V say, then I might aim to keep the last preamp node up as high as 350V say. Sometimes very large dropping resistors are necessary, though, if you have rather small smoothing caps, but I generally use 47uF for everything, so I can use dropping resistors as small as 2k before I start to worry.

[iknowjohnny]

once you get well under 250 things seem to change a lot more tho don't you think? I've had voltages as low as 100 on V1 and it was nice in some ways, not as dynamic tho which i missed. The bad cat shows 100 on v1 so i figured it was worth trying. Let me ask you this....aside from very low voltages in the say 150-175 range and less, do you feel there is generally more high end as you go up in voltage?

[Structo]

Boy, I don't know about that.

On my 6V6 Rocket I just completed I had all the preamp voltages about 50v too high and it oscillated, motorboated and sounded like ass.

Once I dropped them down to within 20v or so of the target it all came together and sounded great.
But it does have a weird power supply due to it's Hammond roots.

I didn't think you were supposed to exceed 300v on a 12ax7 and as a cathode follower there is that cathode to heater spec of around 180v.

[FYL]

On my 6V6 Rocket I just completed I had all the preamp voltages about 50v too high and it oscillated, motorboated and sounded like ass.

Once I dropped them down to within 20v or so of the target it all came together and sounded great.

I'm pretty sure that voltages didn't play a major role here, time constants and lead dress did. Too little R with small C = motorboating. Higher R = no more motorboating.

R * C > 0.1, with R in Mohms and C in µF

For 22µF and 10K => 22 * 0.01 = 0.22 - good

Cascaded filtering stages using the same TC = possible motorboating. The Rocket uses 9K1 + 20µF for the B4 and B5 nodes.

Add a cramped chassis and Q&D wiring, and you're in for a chopstick session.

[Snapcase]

Too little R with small C = motorboating. Higher R = no more motorboating.

Great info, FYL. I didn't knew that. Thanks for sharing.

[Merlinb]

Let me ask you this....aside from very low voltages in the say 150-175 range and less, do you feel there is generally more high end as you go up in voltage?

Nope, frequency response remains more-or-less the same regardless of applied voltage, but I know what you mean. A kind of "clarity" rather than just more treble.

Higher anode voltages means reduced grid current, and smoother or less fuzzy quality of distortion, less blocking, and a bit more more headroom of course, which gives you a little more control over the dynamics. I these are the things that add up to what people often associate with more high end.

[Structo]

Thanks FYL and Merlin.

As I mentioned this started life as a Hammond AO-29 organ amp, so the power supply and rectifier are a little different from the Rocket PS.

That makes more sense about the TC.

It had very high voltage (380-0-380 and 5U4 rect.) that I dropped with a 25w Zener.
I just looked at it and I guess I ended up with a 10K on B3+ and B4+ so in that regard it is similar.
It uses a JJ can with 40-20-20-20 sections.
For B5+ I had to add a 22uF and 18K resistor on the preamp board since I was out of room in the tiny chassis.
It was difficult getting the voltages where I wanted them because the PT seems rather soft and the tube load brought it down quite a ways.

What is strange is that in it's original design it powered 11 tubes!
But many were the weird 6v tubes such as 6AU6 and 6C4.
But the output is 6V6GT.