Electronics > Projects, Designs, and Technical Stuff

Voltage sag problems in tube amp (SS rectifier)

<< < (6/9) > >>

floobydust:
OP you may not have a power supply problem, you might just be wasting the available current.

Example is if you push low bass through a tube amplifier, the output transformer goes very low impedance at low frequencies and you'll draw lots of B+ current and roast output tubes- but the speaker doesn't see it. 47nF/240k is around 14Hz? which I think is way low. You don' t want too much chug.

Another example is if AC balance is terrible, it's not operating efficiently in push-pull and any common-mode going to the output transformer is also wasted as heat in the output tubes.
Your circuit seems imbalanced from the unequal plate resistors 150k/160k, 91k/160k and the entire 6N2 stage seems just wrong to me. You'd be better off adding a tube and using a Fender or Marshall phase-inverter. All that trouble to set DC balance but you want AC imbalance for distortion? It's not done in the power amplifier section, instead always upstream of it.

To prove what I'm saying, need to scope both O/P tube cathodes (resistors). I think 1R is very low, I would use at least 10R which keeps distortion low in the output stage and affects sound character. If you scope the sum between the cathode waveforms, or make a summer with two 1k resistors coming off each cathode to a common point, then to your scope, you can see when symmetry falls apart. Normally (at the sum) there is DC bias with no AC as each O/P tube is anti-phase. You can consider adding an AC balance trimmer to hear what that sounds like.

ELS122:

--- Quote from: floobydust on June 04, 2020, 05:28:55 pm ---OP you may not have a power supply problem, you might just be wasting the available current.

Example is if you push low bass through a tube amplifier, the output transformer goes very low impedance at low frequencies and you'll draw lots of B+ current and roast output tubes- but the speaker doesn't see it. 47nF/240k is around 14Hz? which I think is way low. You don' t want too much chug.

Another example is if AC balance is terrible, it's not operating efficiently in push-pull and any common-mode going to the output transformer is also wasted as heat in the output tubes.
Your circuit seems imbalanced from the unequal plate resistors 150k/160k, 91k/160k and the entire 6N2 stage seems just wrong to me. You'd be better off adding a tube and using a Fender or Marshall phase-inverter. All that trouble to set DC balance but you want AC imbalance for distortion? It's not done in the power amplifier section, instead always upstream of it.

To prove what I'm saying, need to scope both O/P tube cathodes (resistors). I think 1R is very low, I would use at least 10R which keeps distortion low in the output stage and affects sound character. If you scope the sum between the cathode waveforms, or make a summer with two 1k resistors coming off each cathode to a common point, then to your scope, you can see when symmetry falls apart. Normally (at the sum) there is DC bias with no AC as each O/P tube is anti-phase. You can consider adding an AC balance trimmer to hear what that sounds like.

--- End quote ---

well I tried putting in the values for other 'known good' amplifiers into PSUD2 and immediately saw the problems
I am fighting sag because I have a really high B+ voltage and because of that need big dropping resistors for the first stages, because of that and the rather large 30uF capacitors, the voltage goes back up over a long time, ideally, I would want it to go back in about a second, BUT I am pretty sure that the 2.2uF capacitor on the last node would still pass some low frequencies from the gain stage down the power supply chain. and reduce bass. or is that not a problem really? or would it mess with the low frequencies in other ways?
the power transformer and choke aren't really the problem by comparing it to another 'good' amp. well the high B+ is but I want the output plates at a really high voltage.



I1 is the plate supply for the output
I2 is the screen supply for the Output and the cathode follower supply
I3 is the phase inverter supply
I4 is the supply for the 3rd stage (the one driving the AC cathode follower)
R5 is the supply for the 1st and 2nd stages which at idle should draw about 1.5mA at 205V so 136k should be about the same.

the load of I1 and I2 goes from max current draw and to idle current.

here's the simulation:




also, I just love the tone of the paraphase phase inverter. I tried a cathodyne phase inverter but it just was too flubby whatever I tried. and LTP phase inverters don't have enough gain for it.
and yeah I was thinking about putting a balance pot in. I'll probably do it now

ELS122:
maybe I should have independent supply chains for the preamp stages, but now I need multiple PSUD2 files for each chain urgh  :-[

oh and I just found out that you can change the resistive load to be a constant current load, but that wouldn't change it much since it's close to the voltage it should be.

ELS122:
or maybe I'll just use a regulated supply and that will fix most of the problems. yeah I'll just do that.

and while I find a regulator circuit that would work with my cirucit. I'll disconnect the 4uF cap before the choke since that reduces the sag from ~100V to about 40V. but I don't want that low of a B+ supply. with the high ~480V supply which I have now it outputs 54 watts (on paper) (into 16Ohms) and 48 watts (into 12Ohms) by measuring the output voltage and calculating the power.
then I'll just have to get a 4x12 and I can make anyone deaf with just one blast of a power chord  ;D

TimFox:
A regulated power supply for a class-AB or class-B amplifier feeds a pulsating load, and the stabilizer circuit needs to be at least as good as the amplifier.  Why not try a conventional choke-input rectifier circuit?  The reduced output voltage is not due to bad efficiency.  Remember my earlier example calculation of the ripple voltage across the first capacitor:  it charges up to the peak rectifier output (1.4 x rms input), then droops down by a large pk-pk voltage.  The mean voltage at that point has been reduced from the peak voltage by one-half the ripple waveform, whose magnitude is directly proportional to the load current.  Further filtration will reduce the AC magnitude of the ripple, but even without loss the output of the filter will be the mean voltage across the input to the further filtration.  In an ideal choke-input circuit, the input to the choke is the rectified sine wave, whose average value is approximately 0.9 x (rms input).  With no loss in.   the choke (and sufficient current through the choke, see "critical inductance"), that mean voltage is conserved.  Designed properly, the result has good "load regulation" (although none of these circuits has "line" regulation by itself).

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod