Unfortunately my stereo build of the Aleph J hasn’t reached a happy conclusion yet. I initially tested the boards using a Plitron transformer I had bought a while ago for my F5 build. Before I had the chance to buy the additional transformers needed for the Aleph J and F4 builds, Antek went up in flames. I decided to use the Avel Lindberg from Parts Express, but they only have the 18V+18V @ 250VA. So I bought two of them because I really wanted to finish the Aleph J. The DIYAudio store PSU boards are out of stock, so I used a pair of BrianGT boards I had stashed away. I ended up with the noisiest amplifier I have ever built.
As I started to debug the assembly, I found one more issue with the Aleph J PCB. The board layout reflects the original schematic Nelson posted here:
http://www.diyaudio.com/forums/attachments/pass-labs/95797d1194037871-aleph-j-schematic-aleph_j_sch.pdf
But in post #34, Nelson confirms that there is an error in the schematic and Papa’s Little Helper (Zen Mod) provided a corrected diagram in post #37:
http://www.diyaudio.com/forums/pass-labs/111197-aleph-j-schematic-2.html#post1341880
The amplifier noise in my build had three components:
1) Ground loop – greatly reduced with a modified wiring layout.
2) Transformer induced noise - Eliminated by rotating the transformers. I had never stacked toroids before but ZM did it and says it works fine so, why not?
3) Transformer mechanical noise (60Hz buzzing sound) – This is the puzzling one. I have used Avel Lindberg transformers before and know they are well wound. My preliminary conclusion is that the Ultrafast recovery rectifiers (MUR1520) are the culprit. I am working on an experiment this weekend to confirm it.
It is just a minor inconvenience having to tear down part of the build but definitely an opportunity to learn new things and better understand the design.
As I started to debug the assembly, I found one more issue with the Aleph J PCB. The board layout reflects the original schematic Nelson posted here:
http://www.diyaudio.com/forums/attachments/pass-labs/95797d1194037871-aleph-j-schematic-aleph_j_sch.pdf
But in post #34, Nelson confirms that there is an error in the schematic and Papa’s Little Helper (Zen Mod) provided a corrected diagram in post #37:
http://www.diyaudio.com/forums/pass-labs/111197-aleph-j-schematic-2.html#post1341880
The amplifier noise in my build had three components:
1) Ground loop – greatly reduced with a modified wiring layout.
2) Transformer induced noise - Eliminated by rotating the transformers. I had never stacked toroids before but ZM did it and says it works fine so, why not?
3) Transformer mechanical noise (60Hz buzzing sound) – This is the puzzling one. I have used Avel Lindberg transformers before and know they are well wound. My preliminary conclusion is that the Ultrafast recovery rectifiers (MUR1520) are the culprit. I am working on an experiment this weekend to confirm it.
It is just a minor inconvenience having to tear down part of the build but definitely an opportunity to learn new things and better understand the design.
Thanks guys for the follow up.
6L6, buzzforb
You guys are overworked me seems. Too many amps, too little time.
grimberg
I was planning to use the Fairchild Stealth diodes for rectifying (ISL9R3060P2). I wonder if this would be better maybe because of it's soft recovery not requiring snubbers. I know the BrainGT PSU board don't provide for snubbers. I also have the BrianGT PSU boards, on which I am using 6 X 33000uF T-HA Panasonics per board.
6L6, buzzforb
You guys are overworked me seems. Too many amps, too little time.
grimberg
I was planning to use the Fairchild Stealth diodes for rectifying (ISL9R3060P2). I wonder if this would be better maybe because of it's soft recovery not requiring snubbers. I know the BrainGT PSU board don't provide for snubbers. I also have the BrianGT PSU boards, on which I am using 6 X 33000uF T-HA Panasonics per board.
HEre ya go gentlemen. Biased perfectly. My offset is little high, as I need a little more Vgs on the gates of lower fets, due to higher Vga at current source bias. I remember this isuse before, but lickily the boards have so many places to trim all necessary biases. What can I say. I think the Aleph is one of defining sounds of Nelson.
Attachments
Henry,
I replaced the MUR1520 with MSR1560 which is a soft recovery diode, and that lowered the transformer noise. I then tried two external diode bridges connected to the CRC bank on the board and the mechanical noise is now just residual. There is no practical way for me to easily assemble a second bridge using MSR1560 devices and determine whether the nature of the devices alone, or that combined with the dual bridge configuration resolved the issue. However, I just saw this very timely exchange in another thread:
http://www.diyaudio.com/forums/pass-labs/97540-f4-power-amplifier-143.html#post3462472
GKU,
You are right and I stand corrected.
I replaced the MUR1520 with MSR1560 which is a soft recovery diode, and that lowered the transformer noise. I then tried two external diode bridges connected to the CRC bank on the board and the mechanical noise is now just residual. There is no practical way for me to easily assemble a second bridge using MSR1560 devices and determine whether the nature of the devices alone, or that combined with the dual bridge configuration resolved the issue. However, I just saw this very timely exchange in another thread:
http://www.diyaudio.com/forums/pass-labs/97540-f4-power-amplifier-143.html#post3462472
GKU,
You are right and I stand corrected.
Yes Grimberg, and that leaves me with a conundrum. I have an Amplimo 625VA trafo for the AJ with dual 18V secondaries, but I also have the black BrianGT PSU boards that only provides for a single bridge configuration since it seems to have been designed for a trafo with center-tapped secondaries. Which shall I use? On evidence using dual bridges seems to be what is required.
I used dual bridges in my KSA50 and that amp is super quiet. In fact I have 8 bridges in my KSA50 since I have a separate feed from the trafos for the Front-End.
I used dual bridges in my KSA50 and that amp is super quiet. In fact I have 8 bridges in my KSA50 since I have a separate feed from the trafos for the Front-End.
Henry,
That is exactly what I did. After replacing the MUR for MSR devices I ended up removing them from the board and using two bridges attached to the bottom of the chassis. I am sure the Ultrafast Recovery (MUR) and Soft Recovery (MSR) devices have their place, but I don't believe either one is a good option for the PSU of a Class A power amplifier.
I agree. Read here - https://www.passdiy.com/project/articles/power-supplies
(The article formatting is messed up due to the frame on the webpage... but it's still a great read.)
GBPC3502 Fairchild Semiconductor | GBPC3502-ND | DigiKey
I used them in my KSA 50 monoblocks and Aleph 5 monoblocks. If one ever gets destroyed by the charging ripple of the 68 mF capacitor, I have a tray full of replacements ready to go into action.
I used them in my KSA 50 monoblocks and Aleph 5 monoblocks. If one ever gets destroyed by the charging ripple of the 68 mF capacitor, I have a tray full of replacements ready to go into action.
Ja right. Next they will tell us amps sound better when constructed with gold plated caps. I remember way back that Jimmy Hughes claimed his amps sounded better when he removed the screws from the covers. He's the same guy that wanted to blow his ears clean before doing listening sessions, until some docter explained to him how stupid that was.
6L6, OK, I won't remove the caps, but I will highlight all the safe options.
6L6, OK, I won't remove the caps, but I will highlight all the safe options.
Last edited:
Henry,
This is the input capacitor I used. It is available at both Mouser and Digikey.
BFC241671005 Vishay/BC Components | Mouser
BFC241671005 Vishay BC Components | BC2076-ND | DigiKey
This is the input capacitor I used. It is available at both Mouser and Digikey.
BFC241671005 Vishay/BC Components | Mouser
BFC241671005 Vishay BC Components | BC2076-ND | DigiKey