Calculating power output in the design stage?

[audio_moksha]

Hello everyone

I'm in the process of designing a guitar amplifier but I'm unsure about how to design for a particular power output. So, in order to calculate the total power of an amp do you simply multiply the power value of every stage? For eg. a 0.2W for a Cathode bias 12AX7 times a 15W AB fixed bias 6L6 (taking circuit effeciency into account)? Also how do you calculate the power for Push Pull Class AB fixed bias topologies at the design stage? I know you can find the power output by connecting a 1 ohm 2W resistor between cathode and ground and take into account the grid current but that is only after you've built the circuit.

Any help from you experienced folks out there would be helpful. Also do you know of any links related to push/pull class AB amp theory out there? Thanks to all of you once again.

My end goal is to build a guitar amp with a switching 10W, 18W and 30W power output using a cathode based preamp stage followed by a cathode biased push/pull stage and finally a fixed bias class AB push/pull topology.

[PeteN]

May I direct you to Miles' excellent guide to calculating (among other things) power (and distortion) output from an AB stage:
linkies

To summarise:
Find maximum current in one of the (presumably) two output tubes (i.e. at Vgk = 0), this is I_p. R_L is naturally your load (1/4 output trafo's Anode-Anode impedence, if you have two tubes only in AB). In his example Ip is 262mA, RL is 1100 ohms:

I_RMS I_p / sqrt(2)
I_RMS= 262.2E-3 / sqrt(2)= 0.185A_RMS
Po= R_L * I_RMS^2
Po= (1100)(0.185^2)= 37.65W (Not including the ineviatble core and copper losses in the OPT)

I think unless you're in the marketing dept. for an unscrupulous amp manufacturer, you only consider the output power of the final stage, not the drivers etc. too!

Hope that helps! If the summary isn't clear the main article should be

[ray_moth]

A good site to explain tube-based guitar amp design isAiken Amps.
Basically, as PeteN says, power is not passed on from one stage to the next, it all depends on the OP stage. The way I'd approach it is to look up data sheets for a few OP tube types, This will tell you the available power and suitable conditions for use, including voltage and current for plate and screen, load impedance and bias. For the power levels you're considering, there are a few possibilities, including push-pull EL34 or 6L6. There are plenty of datasheets available for download at Frank Filipse's site.

[audio_moksha]

Peter, thanks a lot; this was EXACTLY what I was looking for. Miles' guide is truly excellent and a nice way around composite curves! Just one thing, what does VT stand for?

ray_moth, the Aiken Amps page was also very helpful and the articles definitely helped fill in the gaps.

As far as my design is concerned, I have a 12AX7 cathode biased preamp feeding an intermediate 12AX7 PP Class A stage and finally a 6L6 PPP Class AB power amp stage using four tubes. This lets me get 10W, 15W and 30W of power.

In order to go from 15W to 30W I have a toggle switch between the cathode and ground on ONE of the two 6L6 PPP pairs. Switched on, the connection to ground is completed and I go from 15W to 30W. When the switch is off, the pair is no more connected to ground and instead the cathodes are directly connected to each other with no resistors etc. Would this work? I'm thinking that when off, the switch effectively turns this 6L6 pair into a buffer stage.

One last question to save my sorry ***: is it possible to have a single ended circuit feed a PP OPT? If so how should I connect my plate to the OPT? You see, I would very much like this amp to work in SE mode when switched to 10W. I would not be too sad if this is not possible.

Once again thank you so much ray_moth and PeteN

[ray_moth]

what does VT stand for?

Vacuum tube (I think).

I have a 12AX7 cathode biased preamp feeding an intermediate 12AX7 PP Class A stage and finally a 6L6 PPP Class AB power amp stage using four tubes. This lets me get 10W, 15W and 30W of power.

The 12AX7 is not a good choice, IMHO, to drive 6L6, especially in PPP. The reason is that the 12AX7 really needs a high impedance load. 6L6 needs quite a low value for its grid 'leak' resistance and is an overload for the 12AX7. 12AT7 would make a better driver.

In order to go from 15W to 30W I have a toggle switch between the cathode and ground on ONE of the two 6L6 PPP pairs. Switched on, the connection to ground is completed and I go from 15W to 30W. When the switch is off, the pair is no more connected to ground and instead the cathodes are directly connected to each other with no resistors etc. Would this work?

One problem I see is that, by eliminating one parallel pair, you require double the load impedance. For your purposes, it would probably be better to leave the PPP stage intact but have switchable attentuation at the input of the amp. The max. power possible doesn't change but the signal voltage required to achieve it does. One advantage would be that you retain all the headroom available from the PPP stage for transients.

is it possible to have a single ended circuit feed a PP OPT?

This is a complex issue. An OPT that is meant for PP use cannot tolerate net direct current through the primary. It will cause the core to saturate. In a balanced PP circuit, the plate currents of the OP tubes cancel, so no problem, but in an SE circuit there is no such balance.

To use a PP OPT for SE duty, you would need to use a constant current sink on the half of the primary that you're not using for signal and adjust it to draw the same current as the OP tube(s), to try to maintain the current balance. SE has to be Class A, so there is no net variation in current draw with signal strength.

The CCS could use the 6L6 you're not using, since pentodes/tetrodes make very good CCSs. Problem is, the switching would be complex. Both the bias requirements and the plate load for 6L6 SE class A would be different than PP Class AB1.

[audio_moksha]

[QUOTE]Originally posted by ray_moth
[B]
Vacuum tube (I think).


The 12AX7 is not a good choice, IMHO, to drive 6L6, especially in PPP. The reason is that the 12AX7 really needs a high impedance load. 6L6 needs quite a low value for its grid 'leak' resistance and is an overload for the 12AX7. 12AT7 would make a better driver.

Thanks for the prompt reply ray_moth. Yes, I think I will take your advice and try the AT7. But I have a feeling that good guitar overdrive tones are created due to this very impedance mismatch between 12AX7 and 6L6 as I see them used in so many classic designs. I guess the best way to do it would be to actually try them both.

One problem I see is that, by eliminating one parallel pair, you require double the load impedance. For your purposes, it would probably be better to leave the PPP stage intact but have switchable attentuation at the input of the amp. The max. power possible doesn't change but the signal voltage required to achieve it does. One advantage would be that you retain all the headroom available from the PPP stage for transients.

The reason I asked this is because I saw this kind of topology on the Mesa Boogie Lonestar amp that uses this topology to switch between 10W SE, 50W AB and 100W AB. Quite interesting considering that did it using just one PP OPT. For the sake of simpliciry I might simply go class A PP for the 10W option. But how would the doubling of load impedance affect my power if it is just joing to be 2 of the 4 6L6's in PPP. I should still get less power? If so why would it not work? I would really prefer to do the power switching this way if I can....

Thanks ray_moth!