Actual power output vs idle bias

[EETStudent]

How do you determine the actual power output versus what the tubes are idling at?

And I've been overthinking and confusing myself to no end, but is it not correct that in order to acheive maximum power output available the preamp/PI must supply enough voltage swing to fully drive the tubes?

[MWJB]

"How do you determine the actual power output versus what the tubes are idling at?" These are 2 largely separate criteria, there is no direct corellation from one to the other. Output is W AC RMS signal measured at the OT secondary. Idle dissipation is W dc & relates to the amount of heat that the tube can take before burning up. Idle current is set for best tone, ensuring that nothing is going to burn up before its time, there is no one specific set formula/calculation for determining what this point will be, it depends on the amp & design in question and the application...looking at the extremes, ideal idle dissipation might be anywhere from 15% to 150% (SF Champ) plate dissipation. Most Push pull fixed bias amp tend to idle between 50% and 70%, cathode biased amps from 70% to 100%. There are always exceptions to this however and this is where practical experience is key.

As to driving the power tubes, work back from the speaker - let's take the 57 Deluxe/5E3...

To make 12.5W (10VAC RMS) at the speaker, we need 190VAC at each end of the primary (380VAC from end to end with a turns ratio of around 38:1). The power tubes amplify by a factor of around 11 in this amp so we need 17VAC coming off the PI & subesequently 848mVAC pk-pk, or 300mVAC RMS on the input to the PI's previous gain stage. The input/preamp stage has a voltage gain of around x40 so if we see 3.8VAC after the input stage coupling cap & 100mVAC at the input with all controls at 50% rotation (thus with plenty of potential to exceed that clean W RMS rating). In short, making enough voltage at the PI to drive the tubes is rarely an issue because the potential voltage fed into the PI is usually much more than is required for maximum clean power.

[EETStudent]

Great response MWJB. Let me follow up.

Im working on a 2x6AQ5, cathode biased 5E3/5F11 hybrid amp. Idle dissipation is about 8W per tube. But as you said, this is not related to output. Im trying to determine the max clean power available in this amp. Do I need a scope and signal generater to do this?

In my experience with a 5E3 and 5F11, the 5E3 seems to "accelerate" faster than the 5F11. Could it be that the 5F11 doesnt provide that voltage swing as quickly as the 5E3? Thanks for your insight as Im trying to wrap my brain around this stuff

[MWJB]

"Im trying to determine the max clean power available in this amp. Do I need a scope and signal generater to do this?" Well, you need to measure the the max clean power and yes you ideally need a scope & signal generator to do this. However, exact power output will vary somewhat depending on B+ voltage, rectifier used etc. and in reality a lot of guys will play these amps in excess of the clean W RMS rating. Measuring the average OT secondary voltage accross the speaker at a volume & tone that you find pleasing, whether clean or not, should be enough for you to determine whether the amp is at least functioning in ball park terms.

"In my experience with a 5E3 and 5F11, the 5E3 seems to "accelerate" faster than the 5F11. Could it be that the 5F11 doesnt provide that voltage swing as quickly as the 5E3?" There are notable differences between 5E3 and 5F11. 5F11 volume control is wired as a typical voltage divider, 5E3 is not, 5F11 has a slower ramp up. 5F11 has a NFB loop, reducing gain & extending clean headroom. 5F11 is fixed bias and idle currents may be more moderate than the 5E3's cathode biased set up (possibly higher idle currents, but would depend on the amp in question).

5F11 is also missing a 2nd Stage cathode bypass cap, which reduces gain.

[overtone]

"How do you determine the actual power output versus what the tubes are idling at?" These are 2 largely separate criteria, there is no direct correlation from one to the other...

Thanks for this, I think I was under the false impression that biasing colder would also lower the "output power" of the amp.
So to sum up, can I say that the bias point has no relation to my ability to blow speakers, only to killing the tubes, right?

[MWJB]

In some cases lowering idle current may reduce output power, in others it will increase output power...but maximum clean W RMS output, within reason, isn't most players primary concern (take those expensive hand made repro amps, they often sacrifice absolute power for a preferred tonal warmth), tone usually is...and this is what you pay most attention to when biasing, whilst making sure that components run at safe parameters.

Ideal idle current can also be affected by how hard you run the amp, the same amp played at home & in quiet coffee bar gigs may last on the same pair of power tubes for several years, even if biased hot. The same amp used wide open on outdoor stages probably won't. However, it's typically not necessary to run power tubes at their idle limit in the vast majority of cases. If you are running your amp hard enough to blow the speakers, it's because there is excess Wattage hitting them. That wattage on the OT secondary is also present on the OT primary, so unless your speakers are not man enough for the job, you are likely to be damaging both. Note that this may sound like I am contradicting myself, but that's because we are talking about the amp's max rated W in terms of "clean W RMS". Power outputs that are likely to kill speakers & tubes in quick time are typically in excess of this clean W RMS rating (a 50W RMS rated amp, @ say 5% total harmonic distortion, may well make 70W under heavy distortion)...the more time you spend above the clean W rating the harder things are being worked. The hotter you are biased, then the higher the average current and the more heat is produced...potentially killing tubes, OT primaries & possibly speakers.

Very high idle currents may also pull down B+ considerably, reducing power, as well as reducing power tube cathode/negative grid voltages, limiting headroom.

[overtone]

Rest assured MWJB, I'm not blowing speakers, it was just the guitarist inside me trying to paint things black and white, but ha, it is not so clear cut as I hoped!

Your answer has got me ruminating over a related issue of how not to blow a Palmer PDI-03 so I will start another thread, thanks,
best, tony
Edit: new thread http://music-electronics-forum.com/t25631/

[Bruce / Mission Amps]

With regards to "I'm not blowing speakers".... and just to add to this thread...
when the power tubes are operated at high output clipping, the square wave generated is hard on the tubes but also much harder on the speaker! That is, harder then the same "peak power" with a clean sine wave.... the reason is that the square wave has more average "DC" power component (time) and the speaker is thrown out to it's max excursion travel with the DC component ... held there for mili-seconds (sq wave portion is like pulsing straight DC)... then, nearly instantaneously accelerated backwards and held at the reverse max excursion and held for mili-seconds... over and over and over and over.... that will really smack a speaker hard.
It will at least stress the speaker and heat up the voice coil.
Eventually the voice coil or former can become deformed and the speaker will start to buzz or the voice coil can just simply blow open.

[Wilder Amplification]

Idle bias does however have a direct correlation to the input sensitivity of the power amp. This means that it affects how much input signal will be required to drive the output to full output swing (i.e. the "gain" of the circuit). For instance, if you have -35V at the grid relative to the cathode, it will take a signal amplitude of 35Vpk to swing the valve to full saturation.

Output power is dynamic and never constant. It all depends on how sensitive the input is vs how high the input signal amplitude is vs the load impedance at the signal frequency. Given the fact that input signal amplitude is not constant and neither is the speaker impedance (impedance varies with frequency...the "rated impedance" is just a nominal value) and it should be pretty black and white" as to why power output is never constant.