Low output on JTM45 build

[pdf64]

Yes, that's fair. Into 8ohms, using my Fluke, the GTs get about 14.5Vrms with a 3k load (26.3W), about 16Vrms with a 6k load (32W).

[tubby]

Yes, that's fair. Into 8ohms, using my Fluke, the GTs get about 14.5Vrms with a 3k load (26.3W), about 16Vrms with a 6k load (32W).

Very interesting! The above voltages were measured at the onset of clipping? What's your plate voltage?

[Helmholtz]

With EL34s in my original JTM50 the results are different.
I measure 46W into 3.4k and 37W into 6.6k using the original 3.4k Drake 784-128 OT.
B+ at full output is 400V @ 46W and 420V @ 37W.
(Chinese EL34s, biased cold <50%. 70% dissipation increases power output by around 3W.)

Clearly shows the impact of the different shapes of the plate curves.
Also my OT likely has less losses.

[Helmholtz]

I did some tests, comparing GrooveTube KT66 to TubeTown KT66, using a 100:1 probe on one of the anodes, on my JTM45 type build, with a 6k OT.
It looks like the Vsat is about 50V for the GT ().

Great!
Nicely confirms the theory regarding the influence of different Vsat on power output.

(bottom of red trace wave, scale on the left)

Shouldn't it be the red scale on the right?

Does your scope provide cursors for measurements?

[pdf64]

Very interesting! The above voltages were measured at the onset of clipping? What's your plate voltage?

Yes, Vrms at the 8ohm output at the onset of clipping. The THD on the scope analyser is fairly low still at that point, maybe 2%. The Vrms would increase (due to more area under the curve) if I raised the the gain to push it to 5%THD, but it looks heavily clipped by then, so dunno if I trust it.
The voltage at the OT CT is about 450VDC at idle, about 415VDC with the amp putting out an 'onset of clipping' signal.

Shouldn't it be the red scale on the right?

Doh, yes.

Does your scope provide cursors for measurements?

No, not that I've found anyway.

[Helmholtz]

Yes, Vrms at the 8ohm output at the onset of clipping. The THD on the scope analyser is fairly low still at that point, maybe 2%. The Vrms would increase (due to more area under the curve) if I raised the the gain to push it to 5%THD, but it looks heavily clipped by then, so dunno if I trust it.

Took me a while, but meanwhile I tend to agree that in such cases (soft clipping) calculating equivalent sine power from Vpp can be useful .

[pdf64]

Took me a while, but meanwhile I tend to agree that in such cases (soft clipping) calculating equivalent sine power from Vpp can be useful .

Yey, I’m relieved you’ve seen my point!
I saw the Uncle Doug video linked in post #40 a while ago, and it annoyed me, as in the context of guitar amps, I think his conclusion that bias affects output power is incorrect. But it got me thinking how artefacts of non linearity could be removed when assessing output power.
What’s most annoying about the video is that he’s set up his scope such that it’s not feasible to make an appraisal of the waveforms he’s measuring.

[Helmholtz]

I think his conclusion that bias affects output power is incorrect.

But it does increase real sine power.
The Vpp method just gives a potential sine power.
The signal with cold bias and some crossover distortion "contains" less clean power than a full sine having the same Vpp.

[pdf64]

But it does increase real sine power.
The Vpp method just gives a potential sine power.
The signal with cold bias and some crossover distortion "contains" less power than a full sine having the same Vpp.

Agreed, but in the context of guitar amps, I think it's the potential sine wave capability that is most relevent. As it allows direct comparison between amps that may be very linear right up to clipping, eg my JTM45, and amps that are less linear at high power outputs, eg AC30, cold biased Deluxe Reverb. Because the non linearities can add or subtract from the Vrms area under the curve. And we don't use the amps for absolutely linear reproduction, rather the application typically includes at least some degree of overdrive and significant manipulation of the frequency response.
Rather what matters is the max sine wave that it could hypothetically achieve if only it had perfect linearity, not the actual Vrms at clipping.
Bear in mind that 'inside' that non linear waveform, there's the sine fundemental, whose Vpeak is equal to that of the Vpeak of the waveform you're seeing, it's just that there's some harmonics mixed in with it

[Helmholtz]

.
there's the sine fundamental, whose Vpeak is equal to that of the Vpeak of the waveform you're seeing, it's just that there's some harmonics mixed in with it

That's simply not true for peaky signals where uneven harmonics increase the peak value while the hidden sine actually has a lower peak.
That's what happens with cold bias.

[pdf64]

That's simply not true for peaky signals where uneven harmonics increase the peak value while the hidden sine actually has a lower peak.
That's what happens with cold bias.

That's interesting, thanks, I hadn't considered that was possible. Does Zollner etc go into it somewhere?

[Helmholtz]

That's interesting, thanks, I hadn't considered that was possible. Does Zollner etc go into it somewhere?

Look at the picture with the third harmonic here and consider reversed phase of the third harmonic:https://www.electronics-tutorials.ws...harmonics.html

[g1]

I saw the Uncle Doug video linked in post #40 a while ago, and it annoyed me,

Sorry to go off on a tangent, but sometimes he exhibits such a misunderstanding of the most basic fundamentals. He is becoming more and more widely cited by novices, and he has so much stuff out there it would be tough for anyone to try and fact check.
One guy I was helping referred to his video on 'how to bias a quad or quartet'. At around 14m15s., he points at 2 directly connected plates and talks about them having different voltages. And that is how you can tell if they are mismatched? How could you know what you are doing and yet make such a fundamental error? I tried to think of what he may have meant to say, but just could not come up with anything at all. But other times he seems to know exactly what he is doing. The rest of the stuff about bias checking by OT voltage/resistance is pretty good. So I really don't know what to say to people when they cite him as an info source. When it's an hour video, I'm not going to wade through it to find out.