Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: _Steve on May 13, 2022, 05:58:22 pm
-
Hi!
While playing with a newly built JTM45 clone I inadvertently switched to a higher B+ tap without re-biasing the tubes. It sounded incredible, but only once I finished playing I checked my bias probe/meter and they were conducting 60mA(!) @ 434VDC - which is 104% of KT66's max dissipation using the basic P=VA=26W method.
But it sounded so good I wanted to dig into the loadlines etc a bit to try to understand whats going on a little more. When I plugged the numbers into VTADIY however I see a graph that looks pretty good to me(?). And the calculated power output is 40W. I understand that number is at the speaker, but surely it shouldn't be that different to the cathode current measurement?
Can someone help me understand why the difference in the two power numbers and whether that graph is indeed "healthy"?
I find it hard to search for good information about power stage design and theory. Values and graph in the attachments.
-
How exactly are you measuring the current mentioned?
Note that the power dissipated by the output valves at idle is different to the power delivered to the load (obviously not at idle).
JTM45 get the feed for the bias supply from the full HT winding leg. So switching to a higher voltage winding tap should also increase the bias voltage, and so keep things in line, ie both HT and bias supply voltage increase.
That’s how it works on my amps. Idle dissipation increases a little, but provided the bias isn’t set super hot in the lower voltage mode, it’s fine.
It’s a different scenario to many fixed bias Fenders, which take the feed for the bias supply from a dedicated tap on the HT winding.
-
How exactly are you measuring the current mentioned?
Note that the power dissipated by the output valves at idle is different to the power delivered to the load (obviously not at idle).
Hi! I'm measuring the cathode current using a bias probe (an adaptor between the tube and socket). The measurement is at idle.
Sorry if my question was confusing. Im not really asking about the bias voltage changing with the B+, I understand that OK. My question is why/how VTADIY calculates 40W dissipation and why thats so different to my own figure using the cathode current * plate voltage method?? According to VTADIY im only running slightly hot while the cathode current method has me at 104%. I feel like im misunderstanding something???
-
install a 1 ohm 1% from cathode to ground, ohms law won't lie, but software will :icon_biggrin:
-
…My question is why/how VTADIY calculates 40W dissipation and why thats so different to my own figure using the cathode current * plate voltage method?? According to VTADIY im only running slightly hot while the cathode current method has me at 104%. I feel like im misunderstanding something???
The 40W is audio power output, to a load, speaker. Useful work type watts.
Whereas your calculations are in regard of the valve’s dissipation at idle. Wasted heat type watts.
-
OK thanks that makes sense
-
A couple of points -
1 dunno why that calculator uses the idle voltage. Idle voltages are only relevant when there’s no signal. The point being that HT voltages will sag, probably significantly, at higher current draw, ie when the amp is putting out high power. It’s that voltage which matters in regard of power output.
Marshall HT supplies tend to be more saggy than Fender.
2 KT66 anode limit is shown on the chart as 30W (dashed red line).
Being an absolute max limit, that’s something of a con, or at least isn’t directly comparable to limits of other valve types. KT66 design max limit is 25W, design centre limit would probably be about 22W. ie the KT66 anode is not as beefy as that of the EL34, which has a 25W limit under the design centre system.
-
Thats awesome info thanks! It actually answers my follow up questions about the graph :)
-
... dunno why that calculator uses the idle voltage. Idle voltages are only relevant when there’s no signal. ...
Because accounting for Supply Impedance is additional information the user probably doesn't know. I know in the days when such a calculator would be attractive to me, I had no concept of how to figure out Supply Impedance.
And the max-power RMS current to figure B+ drop is bound up in the answer you get from the calculator. So it's kind of an iterative calculation, where you Assume an input ("B+ doesn't change; Impedance is 0Ω"), find an answer, then plug your answer back into a different assumption ("Impedance is 60Ω") and re-evaluate.
But mostly, the needed data is an Unknown.