That equation will not work out properly.....since your R is the parallel combination of the 360 Ohm Cathode resistor and the " Looking in" the cathode internal resistance...roughly 1/gm per tube....
One thing to consider about the cap you choose. For cathode bias the capacitor is in the signal path of the output. And the larger the cap the greater its influence on the signal. So, a balance between all the factors yields the best results. Or, go with a fixed bias scheme which will be more easily adjusted for individual tubes and removes the large electrolytic cap from the signal path.
Last edited:
Not just any old book. The RDH is the Bible in this domain. And you should have cited it.
Everyone, Putting all this in context. The OP has a Kreisler 11-102. It is a radiogram and the amplifier circuit is very similar to the Magnavox push pull console amplifiers (paraphase). No disrespect to the OP, but this is hardly super HiFi. I guess the output transformers are none too large and rolling the low frequency off may well be a benefit in this situation.
Not sure how it finally ended up or what value caps you used, but if it sounds good or better to you that is what counts.
And you sparked a very informative 'discussion'
Alan
Not sure how it finally ended up or what value caps you used, but if it sounds good or better to you that is what counts.
And you sparked a very informative 'discussion'
Alan
Gees, you know how to crush a guy, and here I thought it was the best $25.00 that I have ever spent. But reading some of the suggestions made me think that I might be making a silk purse out of a sow’s ear. I really just wanted to safeguard the valves. Maybe I was just worrying to much. The set sounds good at the moment but once I finish the job i’ll report back. It’s all good info that I am sure other people will benefit from as well.
What you cite from RDH pdf does not apply to your situation here...
Mixing grapes and oranges with monkeys and zebras.... LOL
While on the topic of low frequency cutoff and the OP's desire to safeguard the amplifier's output valves, could I suggest replacing the coupling caps from driver stage to output stage grids? *This* is also the most desirable location for a dominant low frequency "pole" (actually zero, but) for several good reasons. So, I'd recommend very large (so large that they're a decade or more from working band) cathode bypass caps. And grid couplings caps chosen to set the dominant rolloff.
This geometry minimizes loop stability issues with the output transformer's varying inductance with signal, minimizes whatever electrolytic capacitor linearity issues remain, and minimizes overload recovery time. (Well designed class A1 and AB1 circuits clip at the output stage's grids. The smallest possible time constant here is the bestest.
All good fortune,
Chris
This geometry minimizes loop stability issues with the output transformer's varying inductance with signal, minimizes whatever electrolytic capacitor linearity issues remain, and minimizes overload recovery time. (Well designed class A1 and AB1 circuits clip at the output stage's grids. The smallest possible time constant here is the bestest.
All good fortune,
Chris
According to schematic on post #37, both output channels are pushpull, most probable in class A. Then if tubes and driving are properly balanced, cathode signal must be very low, ideally nul. Considering a small amount of mix from channel to channel may increase stereo sensation, the common cap bypass must not be so critical., including no cap at all.
Actually, the input coupling cap could be an even better place to limit bandwidth, assuming noise is not a problem. The main point is that a coupling cap is a better way to do it than a cathode bypass.jgf said:
As I said earlier, a common cathode bypass cap may be omitted so its value if present is not too critical. If split into separate bias and bypass for each output valve its value becomes important so it is not just a matter of halving the existing common value.
Loop feedback around a coupling cap's low frequency rolloff, a cathode bypass cap's low frequency rolloff, and an output transformer's (varying with signal level) low frequency rolloff , at a bare minimum, is two rolloffs too many. Each of these contributes phase shift that adds to the total. So the parts of the amplifier enclosed by the feedback loop need one of these to dominate the phase shift slope for stability's sake.
When an optimally designed class A1 or class AB1 amplifier clips, the output stage's grids are driven to more positive than the cathode and conduct, charging the coupling capacitor and changing output valve bias until it has bled back to normal. Bleed-down time varies linearly with the value of the coupling cap, so it wants to be small from this standpoint.
The cathode bypass cap doesn't provide a clean rolloff, more of a shelf, and the output transformer's rolloff varies with signal level, obviously not ideal. That long argument leaves the coupling cap. (Rolloffs outside the feedback loop don't count, so this inside-the-feedback-loop decision must still be made.)
Besides, the original caps are old enough to be grandparents and need replacing for reliability's sake!
All good fortune,
Chris
While reading this I have been wondering if in this case it may not be all that critical to ensure that the correct value caps are arrived at. I play a lot of music from the www and then Bluetooth it to my stereogram. So i’m breaking all of the audio buff rules by using lossy digitised music and then adding Bluetooth compression as well. In the end it would be interesting to work out just how much of the original song is being amplified. In reality i’m missing out on a lot of high and low notes anyway. But in saying that, it must be the combination of valve amp and speakers that makes it worth listening to.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.