Pretty sure I will not get the full wattage with the available B+, but more than what I measured will be fine, however the B+ will increase by about 10 volts by going to fixed bias.
This should hopefully be the final schematic until, I get the 8k transformer.
This should hopefully be the final schematic until, I get the 8k transformer.
An 8k OT should give somewhat less output than a 6.6k OT.
What are your plate and screen voltages measured at full clean output?
What are your plate and screen voltages measured at full clean output?
Here's the measurements. The output measurement is at 400Hz right before the point the sinewave starts to distort.
Now given this is a push pull amp is the choke absolutely necessary to eliminate hum in the speaker, because if not that will eliminate one source of a voltage drop and get more power.
Another option I have is to wire the amp up as ultra linear.
Now given this is a push pull amp is the choke absolutely necessary to eliminate hum in the speaker, because if not that will eliminate one source of a voltage drop and get more power.
Another option I have is to wire the amp up as ultra linear.
UL will not increase output.
Using your voltages, my favorite online calculator (https://www.vtadiy.com/loadline-calculators/power-stage-calculator/) gives me an output around 17.5 W before the OT with either 6BQ5s or EL84s and a 6.6k or a 8k OT.
Even with an OT having a low efficiency of only 85%, the output should be around 15W.
Do you have a chance to try a new pair of EL84s or 6BQ5s?
For full output the clean grid peak-to-peak signal should be more than twice the bias voltage.
Using your voltages, my favorite online calculator (https://www.vtadiy.com/loadline-calculators/power-stage-calculator/) gives me an output around 17.5 W before the OT with either 6BQ5s or EL84s and a 6.6k or a 8k OT.
Even with an OT having a low efficiency of only 85%, the output should be around 15W.
Do you have a chance to try a new pair of EL84s or 6BQ5s?
For full output the clean grid peak-to-peak signal should be more than twice the bias voltage.
I use the Russian equivalent of the 7189.
Given the B+ drops 18V at the full output I suspect that's reducing my available power output.
Given the B+ drops 18V at the full output I suspect that's reducing my available power output.
Last edited:
Yes, but I used your measured voltages at full output with the calculator.
I think you mean 7189 not 7199.
It could be that your tubes produce less power in an EL84/6BQ5 circuit.
But first check the grid signals.
Yes 7189.
I may try a different audio generator and see if there's any difference.
The maximum capacitance that can be used as the first filter cap is 60uF.
So for now I'll just short across the choke and see what I get.
Besides at full output i am exceeding the choke's current rating by 15.6mA.
I may try a different audio generator and see if there's any difference.
The maximum capacitance that can be used as the first filter cap is 60uF.
So for now I'll just short across the choke and see what I get.
Besides at full output i am exceeding the choke's current rating by 15.6mA.
Last edited:
The reason I wanted to try ultra linear is that I don't have to worry about the screen voltage.
Here's the measurements with the choke bypassed. The big drop seems to be the screen voltage.
So I'll try ultra linear real quick just to see what it does.
Here's the measurements wired as ultra linear with the choke installed.
One more set of measurements with the choke bypassed.
Here's the measurements with the choke bypassed. The big drop seems to be the screen voltage.
So I'll try ultra linear real quick just to see what it does.
Here's the measurements wired as ultra linear with the choke installed.
One more set of measurements with the choke bypassed.
Last edited:
I chose not to do ultra linear with the choke bypassed as there was a decent amount of ripple on the output waveform, however the output would be 17.8 watts.
Now how I measure watts is I run the output right up to the point the sinewave just does start to distort then I back it off slightly so my power measurements might actually be lower than other methods.
That said it looks like ultra linear gives me the most output power.
I will also need a higher current choke.
The Hammond 159P should work as it's rated 10H and 125mA.
https://www.mouser.com/ProductDetail/Hammond-Manufacturing/159P?qs=MDqvSWII5j%2BXT3duCqDxIg==
Also the DC resistance is 150 ohms so that will get me slightly more output power as the choke won't drop quite as much voltage.
https://www.hammfg.com/files/parts/pdf/159P.pdf?v=1697661944
Well I ran into a slight problem.
The amp was fine with the non-inductive load, but soon as I connected the speaker I got a nice oscillation.
I hope I didn't damage the tubes as the screen grids glowed some.
So I'll try 100 ohm resistors in series with the screen grids as I had forgotten grid stoppers are usually used to stop the chance of oscillation.
Now how I measure watts is I run the output right up to the point the sinewave just does start to distort then I back it off slightly so my power measurements might actually be lower than other methods.
That said it looks like ultra linear gives me the most output power.
I will also need a higher current choke.
The Hammond 159P should work as it's rated 10H and 125mA.
https://www.mouser.com/ProductDetail/Hammond-Manufacturing/159P?qs=MDqvSWII5j%2BXT3duCqDxIg==
Also the DC resistance is 150 ohms so that will get me slightly more output power as the choke won't drop quite as much voltage.
https://www.hammfg.com/files/parts/pdf/159P.pdf?v=1697661944
Well I ran into a slight problem.
The amp was fine with the non-inductive load, but soon as I connected the speaker I got a nice oscillation.
I hope I didn't damage the tubes as the screen grids glowed some.
So I'll try 100 ohm resistors in series with the screen grids as I had forgotten grid stoppers are usually used to stop the chance of oscillation.
Last edited:
Ok well boy do I feel stupid.
I had wired the ultra linear taps backwards which turned the audio amp into a square wave oscillator.
At 50mA cathode current at idle I get 6.43Vrms out which gives me 10.33 watts.
At 70mA cathode current at idle I get 6.23Vrms out which gives me 9.70 watts.
So the B+ voltage dropping is causing a drop in output power.
What I may do is leave the amp ultra linear and get the higher current choke as that will drop less B+ voltage under load.
At idle current of 70mA I calculated a drop of 13.65V across the 195 ohm choke.
At idle current of 70mA I calculated a drop of 10.50V across the 150 ohm choke.
That slight voltage difference will likely not make much of a power difference so I may just leave well enough alone for now.
I could go with a lower idle current of 50mA provided that has no negative effects. One bonus is the tubes will then last longer.
At 50mA cathode current at idle I now get 6.51Vrms out which gives me 10.59 watts. So I believe that I will leave it there for now.
Now if I could find the original Hammond power transformer I used which required a resistor to drop the B+ some, I could likely do a choke input filter or even the same capacitor input filter and get more out of the amp.
Here's what's hopefully the final schematic.
I had wired the ultra linear taps backwards which turned the audio amp into a square wave oscillator.
At 50mA cathode current at idle I get 6.43Vrms out which gives me 10.33 watts.
At 70mA cathode current at idle I get 6.23Vrms out which gives me 9.70 watts.
So the B+ voltage dropping is causing a drop in output power.
What I may do is leave the amp ultra linear and get the higher current choke as that will drop less B+ voltage under load.
At idle current of 70mA I calculated a drop of 13.65V across the 195 ohm choke.
At idle current of 70mA I calculated a drop of 10.50V across the 150 ohm choke.
That slight voltage difference will likely not make much of a power difference so I may just leave well enough alone for now.
I could go with a lower idle current of 50mA provided that has no negative effects. One bonus is the tubes will then last longer.
At 50mA cathode current at idle I now get 6.51Vrms out which gives me 10.59 watts. So I believe that I will leave it there for now.
Now if I could find the original Hammond power transformer I used which required a resistor to drop the B+ some, I could likely do a choke input filter or even the same capacitor input filter and get more out of the amp.
Here's what's hopefully the final schematic.
Last edited:
7189/6P14P has crazy transconductance, so it is difficult to find precisely matched pairs for fixed bias operation. You need individual bias adjustment for output tubes for DC balancing. That's what they did in commercial amps, eg Scott.
Yes I am.
So the ideal thing would be to use a second pot and second 1 ohm resistor so I can measure and set each cathode current individually, right?
So the ideal thing would be to use a second pot and second 1 ohm resistor so I can measure and set each cathode current individually, right?
I'd do the following: Increase the 1st filter capacitor to 60 µF, the 5AR4 maximum rating, and connect the output transformer to this one, hence circumventing the choke. A PP amplifier allows for this. Get, as yet suggested, individual bias adj potentiometers for each power tube. Get individual cathode resistors of 10 Ω for each power tube to ease cathode current measurement.
Best regards!
Best regards!
Leave it, but swap the OT CT connection (nothing else!) from after to before the choke. I suspect the hum resulting from the sub par PI.
Best regards!
Best regards!
I suspect the hum is coming from the plate circuit of the 6J5 due to inadequate filtering so I could increase the value of the 3.2k resistor to provide better filtering.
Only reason I used the original value of 1.2k was due to a receiver getting its B+ from that point and I just left the resistor when I made a separate B+ supply for the receiver given the phase splitter isn't very picky about the B+ voltage it is fed. I added a 2k in series when I needed to drop extra voltage to get the screens closer to 300Vdc.
Only reason I used the original value of 1.2k was due to a receiver getting its B+ from that point and I just left the resistor when I made a separate B+ supply for the receiver given the phase splitter isn't very picky about the B+ voltage it is fed. I added a 2k in series when I needed to drop extra voltage to get the screens closer to 300Vdc.
Yes, that's the point. A Concertina phase inverter has a very poor CMRR, making it susceptive for supply induced hum. Paraphase and LTP inverters (especially LTP's with a CCS tail) are much (!) better here and in other aspects.
Best regards!
Best regards!
Oh ok I didn't know that, thanks.
I'll ditch the choke and try a 10k resistor in place of the 3.2k resistor. That stage draws little current. Plus it was designed to operate at around 250Vdc B+ anyways. The first filter cap will then be 62uF given the 22uF and 40uF caps will be in parallel.
If I do decide to add a second bias pot to adjust each individual tube, I'll add a SPDT switch so that the banana jack can be selected between either cathode as I don't have enough room on the chassis for another banana jack.
Also given I'm likely going to be above 100mA current, I wouldn't be able to use that choke anyways nor would I be able to use the 100mA meter.
If I wanted even more output power I could ditch the 5AR4 and go with diodes to get a higher B+, however I prefer the rectifier tube given it brings the B+ up as the tubes warm up instead of instantly like diodes would.
That said I may buy this 200mA meter given its low cost and figure out a place to mount it.
https://www.amazon.com/YXQ-0-200mA-Current-Accuracy-Measurement/dp/B07JLRQ419
I'll ditch the choke and try a 10k resistor in place of the 3.2k resistor. That stage draws little current. Plus it was designed to operate at around 250Vdc B+ anyways. The first filter cap will then be 62uF given the 22uF and 40uF caps will be in parallel.
If I do decide to add a second bias pot to adjust each individual tube, I'll add a SPDT switch so that the banana jack can be selected between either cathode as I don't have enough room on the chassis for another banana jack.
Also given I'm likely going to be above 100mA current, I wouldn't be able to use that choke anyways nor would I be able to use the 100mA meter.
If I wanted even more output power I could ditch the 5AR4 and go with diodes to get a higher B+, however I prefer the rectifier tube given it brings the B+ up as the tubes warm up instead of instantly like diodes would.
That said I may buy this 200mA meter given its low cost and figure out a place to mount it.
https://www.amazon.com/YXQ-0-200mA-Current-Accuracy-Measurement/dp/B07JLRQ419
Last edited:
Tried that and still got AC ripple, but not as much.
So that tells me the AC ripple is coming from the 7189 push pull stage, although I do not know why unless the tubes are so mismatched that it's causing the AC ripple to make it through to the speaker. I do know the + alternation of a sinewave distorts right before the - alternation does.
For now I will keep the amp like this as I do not necessarily need every bit of power these tubes can do, however I do want to fix the problem so that I can remove the band aid disguised as a choke.
That said I do have a mutual conductance tube tester and I found a third 7189 so maybe I can compare all three on that tester and select the two that read the closest to each other.