I am asking @zintolo if he can explain for us how adding that extra resistor works.
In the meantime, see post #38. It would be very easy to add that extra resistor to to my "bias module."
I hope I got those values correct. Did I?
I also stumbled across this tidbit from a different person while looking for the Allen Wright approach:
You need a LARGE cathode cap if you use one per tube. When there are 2 or 4 tubes per cathode resistor, a smaller value will do, since the two currents change in opposite directions.
https://audiokarma.org/forums/index.php?threads/cathode-bias-schemes.248933/
It seems that the value of 220uF as specified for two tubes would better be increased to 470uF (or more depending on who you ask) when using one per tube. Perhaps this is why I can't hear much if any difference having gone from 220uF to 470uF in my existing amp, which shares one cap and one resistor per pair of tubes. It may be a little "cleaner" sounding now with 470uF, but in an A-B test I expect that nobody would be sure.
So much learning and so little time.
In the meantime, see post #38. It would be very easy to add that extra resistor to to my "bias module."
I hope I got those values correct. Did I?
I also stumbled across this tidbit from a different person while looking for the Allen Wright approach:
You need a LARGE cathode cap if you use one per tube. When there are 2 or 4 tubes per cathode resistor, a smaller value will do, since the two currents change in opposite directions.
https://audiokarma.org/forums/index.php?threads/cathode-bias-schemes.248933/
It seems that the value of 220uF as specified for two tubes would better be increased to 470uF (or more depending on who you ask) when using one per tube. Perhaps this is why I can't hear much if any difference having gone from 220uF to 470uF in my existing amp, which shares one cap and one resistor per pair of tubes. It may be a little "cleaner" sounding now with 470uF, but in an A-B test I expect that nobody would be sure.
So much learning and so little time.
Last edited:
I forgot that I asked the question earlier this year.
It seems my "bias module" in post #38 is a compact, viable solution for biasing each tube individually, but 470uF instead of 220uF seems prudent.
Last edited:
Hi, I wrote it here: https://www.diyaudio.com/community/threads/easy-diy-tube-amp.384618/post-6995533
but it was written by Allen himself that the value must have been in the 14% range.
Apparently not on this forum, as AW has only 16 posts on this forum.
This are the results I've found on my database (not performed by me).
Allen trick NOT applied (see 3rd harmonic and side bands):
Allen trick applied:
but it was written by Allen himself that the value must have been in the 14% range.
Apparently not on this forum, as AW has only 16 posts on this forum.
This are the results I've found on my database (not performed by me).
Allen trick NOT applied (see 3rd harmonic and side bands):
Allen trick applied:
That's very interesting. Nice to see actual test results. He said it also reduces IMD. I wonder how that resistor accomplishes this and how he calculated it? I have searched online, and I can't find anything about it. It's a shame he is gone. I accidentally stumbled across your post while looking for something else.
Also, I assume a 1/2 watt resistor would be adequate for the task in my little EL84 PP amp?
Also, I assume a 1/2 watt resistor would be adequate for the task in my little EL84 PP amp?
IIRC he suggested to have 2W resistors there, and the purpose was to enhance interaction between cathodes.
That's odd to me, I've never seen a push pull with such a strong even harmonics, unless it is measured at very low power.
Maybe someone has pointed it out already but It seems to me that 68 ohm maybe nothing more than a way to moderate the charge of 2x 8mF caps and avoid shorting cathodes to ground at power up.
Do the Belton sockets have a "death grip" on the tubes? Some sort of "Micalex" type sockets came with my ST-35 kit, and they are horrible. I'm not sure if they are Belton or not. The insertion and removal force is so severe that one day they may break the circuit boards.
I'm looking for good PCB-mount sockets that do not have a death grip on the tubes. I don't mind the porcelain/ceramic Chinese porcelain sockets that come with the kit, but after a dozen tube changes they no longer have an even grip. It varies from socket to socket.
Is there any such thing as a "low insertion force" tube socket? That would be a real benefit for PCB-based amplifiers.
The octal chassis-mount ones I used were a bit tight. I used an old tube to work them out a bit.
Belton 9 pin sockets are very tight at first. Use an old tube in and out a few times and they become just tight. My preferred socket over the generic white ceramic sockets.
S.
S.
Chemi-Con, Nichicon, Rubycon, and VishayBC - for the large electrolytic power supply capacitors (7 of them) does brand name make any difference at all as far as the reliability of the product? I want a reliable brand with good quality control. Mouser sells all 4 of these brands.
To me, a capacitor rated for long-life and 105C is useless if it actually lives a short life due to poor quality control or manufacturing defects or exaggerated specifications or if it degrades way below specification over time due to poor quality. Using 450V or 500V caps instead of 400V will help I am sure, but that's only true if the capacitor is good quality and the manufacturer is truthful and making a quality product to begin with.
Remember the capacitor plague? https://en.wikipedia.org/wiki/Capacitor_plague
To me, a capacitor rated for long-life and 105C is useless if it actually lives a short life due to poor quality control or manufacturing defects or exaggerated specifications or if it degrades way below specification over time due to poor quality. Using 450V or 500V caps instead of 400V will help I am sure, but that's only true if the capacitor is good quality and the manufacturer is truthful and making a quality product to begin with.
Remember the capacitor plague? https://en.wikipedia.org/wiki/Capacitor_plague
Last edited:
Absent some reason to stay away from Nichicon power supply caps, I guess I'll use those. All seem to be available from DigiKey at 450VDC or 500VDC 105C. Vishay/BC is significantly more expensive.
For C1 and C2, instead of 220uF, a 270uF at 500V is available and fits perfectly. Probably not a significant increase in capacitance I guess.
I was tempted to use the included four caps for the power supply board since it's removable and they are easily changed if they go bad, I just don't want to deal with any issues since I intend to keep this amplifier for a very long time. It's worth $25 to me to put in known reliable replacements not to have to deal with it if anything ever happened. The three 100uF on the amplifier board would be very difficult to replace, so those three definititely will get a known reliable replacement for $16.41.
Total power supply cap cost: $43.52 tax included.
For C1 and C2, instead of 220uF, a 270uF at 500V is available and fits perfectly. Probably not a significant increase in capacitance I guess.
I was tempted to use the included four caps for the power supply board since it's removable and they are easily changed if they go bad, I just don't want to deal with any issues since I intend to keep this amplifier for a very long time. It's worth $25 to me to put in known reliable replacements not to have to deal with it if anything ever happened. The three 100uF on the amplifier board would be very difficult to replace, so those three definititely will get a known reliable replacement for $16.41.
Total power supply cap cost: $43.52 tax included.
Last edited:
Reduced IMD. Those products are IM products between the applied 1 kHz and the power supply ripple.
I still can't figure out how/why the Allen Wright trick with the extra resistor works (the 68 ohm resistor in post #31). Also, 8000 uF is bigger than most suggest, but certainly smaller than the infinite value some have suggested in theory.
After several months of waiting, my new output transformers arrived from Primary Windings in the UK. Musical Power Supplies custom wound a new power transformer for me so all voltages should be exact this time. The new output transformers look a little silly because they are so massive, but they do fit and the price was fair. They ought to get the job done.
Most of the parts in in my shopping carts at Mouser and DigiKey. "All" I have to do now is build it. That requires some drilling to mount the new transformers followed by painting everything black. I have again plugged the holes where the magic eye tubes should be. This time I put a round metal backing plate behind each hole. It looks better than the pop-in plugs. I custom made in my graphics program a top plate without those two holes, but the minimum order to have one made is $29 at https://sendcutsend.com/ I might just have a top plate made to cover the whole chassis, with proper holes for the new transformers. I'll probably go ahead and do that. It would be a more finished look, and it would replace both the front plate where the holes for the tubes are and the metal plate at the rear that I removed to make the larger transformers fit. $29 seems like a small price to pay in the overall cost of this build with the new transformers and premium parts. I'm hoping that this will become a "forever" amplifier for me, so $29 to make it look nicer would be worth it.
Most of the parts in in my shopping carts at Mouser and DigiKey. "All" I have to do now is build it. That requires some drilling to mount the new transformers followed by painting everything black. I have again plugged the holes where the magic eye tubes should be. This time I put a round metal backing plate behind each hole. It looks better than the pop-in plugs. I custom made in my graphics program a top plate without those two holes, but the minimum order to have one made is $29 at https://sendcutsend.com/ I might just have a top plate made to cover the whole chassis, with proper holes for the new transformers. I'll probably go ahead and do that. It would be a more finished look, and it would replace both the front plate where the holes for the tubes are and the metal plate at the rear that I removed to make the larger transformers fit. $29 seems like a small price to pay in the overall cost of this build with the new transformers and premium parts. I'm hoping that this will become a "forever" amplifier for me, so $29 to make it look nicer would be worth it.
Very compact, please double check the power transformer orientation against the output transformers, to avoid hum.
That issue comes up every time someone sees the laminations going in the same direction. The laminations do not control the orientation of the magnetic field. The coil winding does. On the PT the coil is wound horizontally, parallel with the chassis. On the OPTs the coils are wound vertically, perpendicular to the chassis. Therefore, the magnetic fields are at right angles to one another. The coil and lamination orientations are the same as the ones that come with the kit, and there is no hum.
For barely more expensive than the custom plate, wouldn't you like to opt for a slightly larger Hammond chassis ?
The small, narrow chassis is one of the things I like best about this amplifier. 🙂