Power amp input sensitivity

[naciketas]

Hi, I was trying to verify the theory learned and when it comes to the power amp I was disappointed because the numbers aren't as expected: it's a class A PP KT88 6k6 OT, 200Va, cathode biased (460R and 2200uF each tube) at about -15v @ 55mA, so I was expecting a 30Vpp signal at the grid would make the anode clip, but to my surprise clipping (almost perfectly simmetrical) occurs at an input signal of about 16Vpp, that gives me half the headroom expected and burns all my dreams of glory!
And the worst thing is that I can't explain why!

So I must surrender and ask for help to this knowledgeable forum, thanks.

[Stevem]

I have two questions first, what are you running such a powerful tube ( 42 watts) at such a low plate voltage even in terms of running in class A which you are 50 volts low of anyway, and why do you have that huge 2200uf bypass cap?
The tube spec in PP with cathode bias with 400 volts on the plate calls for a common 560 ohm cathode resistor, yet your are running a 460 ohm resistor on each with even less plate voltage!

[naciketas]

Yes i can understand the points, but I was interested in Having a low power amp, class a in order to have the possibility to switch from se to pp, big bottles mainly because I thought I could have more headroom and of course for their tone. But it seems there's not all that gain in terms of headroom.

[Stevem]

I should have posted before that when you have that low a plate voltage it's the same thing as having a vari B+ power supply to lesson a Amps output and part of that mod to a amp is installing a master volume control before the output tubes due to the same excess signal drive issues you have!

[naciketas]

I've scaled the whole amp and, to give a complete view of the project, it has an express pre and PI.
The idea was to have more clean range using a different tube, but what puzzels me is that early clipping: why does it clip with a 15Vpp sinewave when it is biased @ -15v? In my ignorance I was expecting that when the signal goes up to 30Vpp. What am I missing?

[Jana]

First, are you sure it is the power tubes that are clipping and not your phase inverter?

Your numbers don't make sense. With 15 volts of bias across the 460 ohm, I get 32 ma. not 55 ma. Then, when calculating the wattage, (200 - 15) * .032 = about 6 watts. These are a 42 watt tube and you only have them biased to 6 watts. That is really cold. Can someone else check my ciphering?

[naciketas]

Put one probe on the anode (pin 3) and the other on the control grid (pin 5), so it is only the power tube.
My mistake, Rk is actually 264R.
Actual measurements:
Va=212 Va-k=197
Vg2=205 (500R Rs - Ig2 2.5 mA)
Vk=15.18
Rk =264
15.18/264=0.0575A
minus screen current (2.5 mA) gives about 0.055 mA and 10W.

[Firestorm]

Well that's still not Class A. It's incredibly hard to have push-pull Class A stay there without falling into Class B. I bet your tubes are cutting off.

Or is this SE mode. I can't quite follow what you're doing.

[naciketas]

Basically I want to play using one tube doing so by grounding the grid (se mode) of the other and whenever i need a boost, e.g. during solos, switch in pp mode. The choice of operating the amp in class a is because in se mode the signal is simmetrically amplified (and clipped), while in class ab this is only possible in pp when the signals of the tubes are combined in the ot. I was trying to verify what i learned, maybe i need to go back to basic again...

[Stevem]

I do not follow your thinking on this at all!
If you are say playing with a certain level dirt on a rhythm line, and this level of dirt was due to clipping from both the preamp and output stage and you go to solo and cut I another output tube your clipping of the output stage will change to be on the cleaner side of things!

[naciketas]

Normal playing: one tube, clean to mean with guitar pot
Solo playing: two tubes, double power, enough boost for my needs.
Problem is that I don't get enough clean range despite biasing @ -15.

[Jana]

How did you come up with this configuration?
Why 200 volts?
Why 6k6 OT?
Why biasing at less than 25% plate dissipation?
Why KT88 tubes?

None of your choices make any sense to me and I would be interested in why you made these choices.

If you were going to use a KT88 with 200 volts on the plates and expect it to be in class A, you would need to bias to around 200 ma. Your OT would probably need to be about 1/4 the impedance you currently have. But, I'm not sure that running a KT88 at that low of a voltage is even practical.

Even if you do get twice the power with two tubes, which I don't think you will, that is only 3 db.

[naciketas]

As mentioned earlier, 3 (or 6) db more is just suiting my need to boost the solos.
I used the genalex datasheet, where you could find curves for even g2=100V. The amp is where it it supposed to be except for the lack of clean range and let me insist on my primary question: why does it clip when the signal is 15Vpp being the pa biased @ -15V? As I believed to undestand, clipping should occur @ 30Vpp.

[Jana]

Because you are biased very cold--meaning close to class B if not already in class B. The tubes are only a volt or two away from cutoff. You don't have a 15 volt range available on both sides of the bias voltage.

[martin manning]

What is the load impedance, and what are you doing about it when you bring on the second tube? You should be reducing the load by half when you do that.

A shematic is worth a thousand words, and we are all native speakers in that language.