Yes, that looks good.
I use a 10 R resistor in the cathode to give easier reading of the cathode current as most of my meters are not real accurate in the low mv range. The 10R resistor does not give much feedback, but gives 10X the voltage for setting the cathode current.
Apples and oranges.
I use a 10 R resistor in the cathode to give easier reading of the cathode current as most of my meters are not real accurate in the low mv range. The 10R resistor does not give much feedback, but gives 10X the voltage for setting the cathode current.
Apples and oranges.
Will the Universal Power Supply board that is available at the DIYAudio Store work as the power supply for this amp? If I am way off base please excuse my ignorance. I'm a complete idiot when it comes to electronics.
George, you may have gotten the notice but in case you didn't:
"Mouser Electronics received a lifecycle change notification from the manufacturer regarding a product you had interest in or purchased within the last two years. Please click on the part number to view product details and similar products for replacements.
Manufacturer Part # Mouser Part # PO# Part Status
Bourns 3309P-1-102 652-3309P-1-102 13901034 End of Life
Bourns 3309P-1-104 652-3309P-1-104 13901034 End of Life
"
"Mouser Electronics received a lifecycle change notification from the manufacturer regarding a product you had interest in or purchased within the last two years. Please click on the part number to view product details and similar products for replacements.
Manufacturer Part # Mouser Part # PO# Part Status
Bourns 3309P-1-102 652-3309P-1-102 13901034 End of Life
Bourns 3309P-1-104 652-3309P-1-104 13901034 End of Life
"
Simulation of Feedback options
I’ve been simulating various feedback options for the UDB with 6550 outputs.
The open loop schematic is shown below.
I’ve run simulations of the following feedback options:
· 470K output anode to driver anode (1dB),
· output anode to driver anode plus second LTP anode to first LTP anode 540K (lowest recommended), additional 1.5dB,
· output anode to driver anode plus second LTP anode to first LTP anode plus global (100R/2K7, additional 9.3dB)
· cathode ( 0.5H, 4dB) and
· cathode plus global (100R/2K7, additional 10dB).
The Universal Driver Board is easily capable of swinging the voltage required for cathode feedback.
Output power for these options into 8Ω is shown below:
and for less than 10W
Observations:
The lowest recommended resistor (470K) from the output anode to driver anode only provides 1dB of feedback. This provides some improvement in THD up to about 45W but provides no improvement over open loop above that.
540K between the LTP anodes only adds a further 1.5dB but no improvement in THD.
The addition of 9dB global negative feedback is worse than open loop below 6W but provides significant improvement above that.
Just adding 4dB of cathode feedback provides comparable performance to the output anode to driver anode plus LTP anode and global feedback below 40W and significantly better performance above that.
Adding 10dB of global feedback to the cathode feedback reduces THD to less than 0.1% below 10W and below 0.3% all the way to 60W.
Conclusion:
I'll try, measure and listen to all the options, but cathode feedback looks to be the way to go.
There are at least 2 vendors of the specialist toroidal transformers required. I’m planning on adding bias servos to ensure the output tube quiescent currents are as close as possible.
I’ve been simulating various feedback options for the UDB with 6550 outputs.
The open loop schematic is shown below.
I’ve run simulations of the following feedback options:
· 470K output anode to driver anode (1dB),
· output anode to driver anode plus second LTP anode to first LTP anode 540K (lowest recommended), additional 1.5dB,
· output anode to driver anode plus second LTP anode to first LTP anode plus global (100R/2K7, additional 9.3dB)
· cathode ( 0.5H, 4dB) and
· cathode plus global (100R/2K7, additional 10dB).
The Universal Driver Board is easily capable of swinging the voltage required for cathode feedback.
Output power for these options into 8Ω is shown below:
and for less than 10W
Observations:
The lowest recommended resistor (470K) from the output anode to driver anode only provides 1dB of feedback. This provides some improvement in THD up to about 45W but provides no improvement over open loop above that.
540K between the LTP anodes only adds a further 1.5dB but no improvement in THD.
The addition of 9dB global negative feedback is worse than open loop below 6W but provides significant improvement above that.
Just adding 4dB of cathode feedback provides comparable performance to the output anode to driver anode plus LTP anode and global feedback below 40W and significantly better performance above that.
Adding 10dB of global feedback to the cathode feedback reduces THD to less than 0.1% below 10W and below 0.3% all the way to 60W.
Conclusion:
I'll try, measure and listen to all the options, but cathode feedback looks to be the way to go.
There are at least 2 vendors of the specialist toroidal transformers required. I’m planning on adding bias servos to ensure the output tube quiescent currents are as close as possible.
Bench Test
Above is the start up and adjustment procedure that you documented in post 178 for EvanC:
It’s been very helpful to have this guide to get my board up and running on the bench.
Do any of the values need to change when using EL34’s for output? I’m using 6CG7’s for the driver board. My output transformer is a Hammond 1650N – 4300 primary.
I’m at the point where I need to make the adjustments. I’m getting sound, but if I set R13 to 450 volts the sound cuts out as I raise the bias much above 15ma. I don’t have an oscilloscope and am using cheap multi-meters. Should R13 be set lower? Are my meters too cheap to get a good reading? Can I set up this amp for optimum sound with just multi-meters or do I need a scope?
Thanks, Jacques
Hi George,
Above is the start up and adjustment procedure that you documented in post 178 for EvanC:
It’s been very helpful to have this guide to get my board up and running on the bench.
Do any of the values need to change when using EL34’s for output? I’m using 6CG7’s for the driver board. My output transformer is a Hammond 1650N – 4300 primary.
I’m at the point where I need to make the adjustments. I’m getting sound, but if I set R13 to 450 volts the sound cuts out as I raise the bias much above 15ma. I don’t have an oscilloscope and am using cheap multi-meters. Should R13 be set lower? Are my meters too cheap to get a good reading? Can I set up this amp for optimum sound with just multi-meters or do I need a scope?
Thanks, Jacques
I finished two KT-88/Hammond 1650-RA mono-blocks. They’ve been up and running, generally without incident, for several weeks. However, I did lose one 5AR4 and had a close call with a second. In reading about possible causes for rectifier tube failure, it was mentioned that too much high voltage filter capacitance could be a cause.
The power supply, high voltage circuit finishes with a “100uf Temco, Aux cap.” I looked around for this component and the closest I could find was a very large automotive capacitor, too big to be useful in a finished amp; so, instead, I installed a Vishay-BC 100uf 500v electrolytic cap. Wonder if that was mistake.
The power supply, high voltage circuit finishes with a “100uf Temco, Aux cap.” I looked around for this component and the closest I could find was a very large automotive capacitor, too big to be useful in a finished amp; so, instead, I installed a Vishay-BC 100uf 500v electrolytic cap. Wonder if that was mistake.
The first capacitor right after the rectifier tube can't be too high if that's what you mean? It should be about 30uf at the most and there is a very popular SS diode mod that can help the tube rectifier last much longer in big amps.
You simply add UF4007 (best ) or IF4007 Diodes in series with the PS trans secondary leads to pins 4 & 6 of the GZ-34(5AR4) rect. tube. You can use unused pins on the tube socket to mount the diodes if you wish, otherwise you can use heat shrink tubing on the diodes to protect from shorts.
Sonics are unaffected!
tube rectifier diode mod
You simply add UF4007 (best ) or IF4007 Diodes in series with the PS trans secondary leads to pins 4 & 6 of the GZ-34(5AR4) rect. tube. You can use unused pins on the tube socket to mount the diodes if you wish, otherwise you can use heat shrink tubing on the diodes to protect from shorts.
Sonics are unaffected!
tube rectifier diode mod
Last edited:
The diode mod helps with the PIV rating of the tube, However it does not address inrush current.
The maximum capacitor value for the first capacitor after the rectifier is 60uF. Many current production tubes don't like that value and are better off at 40-47uF (which normally run on the high side and may actually be 60uF).
https://drtube.com/datasheets/5ar4-amperex1958.pdf
The second filter cap can be much larger as the choke or resistor acts as a current limiter so the tube does not see the full capacitance.
The maximum capacitor value for the first capacitor after the rectifier is 60uF. Many current production tubes don't like that value and are better off at 40-47uF (which normally run on the high side and may actually be 60uF).
https://drtube.com/datasheets/5ar4-amperex1958.pdf
The second filter cap can be much larger as the choke or resistor acts as a current limiter so the tube does not see the full capacitance.
Thanks for the comments. My original question was whether the 100uf 500v cap I installed in the final position could be responsible for my rectifier tube issues. No one seemed to think that was likely. Instead there is concern that the first cap, 47uf 500v, might be too much. Of course that was the spec in the designer's p/s schematic (post 202). I measured the cap in my power supply and it read 44.9 uf; so it is not out of spec. Apparently, the question is whether it would be worthwhile to try a 30uf cap in that spot. I'm not sure what the overall impact would be, however.
I do believe George has added that Diode "Mod" to the PS on his PCB's some years ago and it should help greatly for the tube failure. If George spec'd that value for the first cap it should be fine and it measures under 60 uF. Many new tubes aren't as rugged as the ones made back in the day.
Last edited:
The power supply for the monoblock version of the UDB is based on the SSE schematic. While the schematic doesn't show the diodes going to pins 4 and 6 per the suggested mod above, the later SSE boards have it.
So it looks like a good idea to me and one I am assuming that George has blessed.
So it looks like a good idea to me and one I am assuming that George has blessed.
V2 setup and PP signal swing
I'm puzzled by the 450V recommendation for the plates of V2. Even assuming a 500V supply this leaves very little room for signal swing.
Worst case you'd want the MOSFET sources to swing 150Vpp. This would mean the same for the V2 anodes.
It seems to me that 350V on the V2 anodes would be more appropriate.
Am I missing something here?
I'm puzzled by the 450V recommendation for the plates of V2. Even assuming a 500V supply this leaves very little room for signal swing.
Worst case you'd want the MOSFET sources to swing 150Vpp. This would mean the same for the V2 anodes.
It seems to me that 350V on the V2 anodes would be more appropriate.
Am I missing something here?
Last edited: