I am totally rebuilding my Audio note oto with the addition of seperate power transformers, rectification, chokes for filament and HT, black gates silver cabling etc etc.
F5 F3 Aleph 30 J Jx and oto would be a very interesting comparison especially regarding the 'magical midrange'.
PS the oto after the first round of small initial upgrades was close to AUdio Note P4 silver Signature 300b triode monoblocks with Nickle output cores!!! So would be a good benchmark.
I think might be time to move this conversation to a new thread?????
F5 F3 Aleph 30 J Jx and oto would be a very interesting comparison especially regarding the 'magical midrange'.
PS the oto after the first round of small initial upgrades was close to AUdio Note P4 silver Signature 300b triode monoblocks with Nickle output cores!!! So would be a good benchmark.
I think might be time to move this conversation to a new thread?????
Attached is schematic of front end board for Aleph JX circuit.
This is the 'bridging' board between two output stages that are created from standard AJ boards by not instaling the input circuitry.
Can someone please verify that I have it right, after all the last AX I did was 7 years ago: http://www.diyaudio.com/forums/showthread.php?s=&threadid=5980&highlight=
Also, what is suggested power rating for output to sources resistors (RS) and output to ground resistors (RG) required for X circuit?
This is the 'bridging' board between two output stages that are created from standard AJ boards by not instaling the input circuitry.
Can someone please verify that I have it right, after all the last AX I did was 7 years ago: http://www.diyaudio.com/forums/showthread.php?s=&threadid=5980&highlight=
Also, what is suggested power rating for output to sources resistors (RS) and output to ground resistors (RG) required for X circuit?
Attachments
With a 2-minute look, I cannot see anything wrong. But why not using a single 2SJ74 as current source ? All you need is a single trimpot btw +ve rail and the J74 source, is quiet as hell, saves a lot of complexity.
I use 10k for Rs, someone else uses 22k, I think.
Rg depends on the output power. If you are using 4 MOSFETs for an X, then 33R 3W is absolute minimum, and I would not hesistate to use 5W. For high power versions, I have been using 100R, 100W Caddock TO247s.
So I suggest leaving Rg out of the board, and connect them at the speaker terminals.
But it's a matter of taste.
Patrick
I use 10k for Rs, someone else uses 22k, I think.
Rg depends on the output power. If you are using 4 MOSFETs for an X, then 33R 3W is absolute minimum, and I would not hesistate to use 5W. For high power versions, I have been using 100R, 100W Caddock TO247s.
So I suggest leaving Rg out of the board, and connect them at the speaker terminals.
But it's a matter of taste.
Patrick
> did you have a chance to compare it to stock current source for sonic differences?
Actually ever since my first AX, I have always used cascoded JFETs as current source. So I can't really answer that one. To cascode 2SJ74, I use J174, in the same way as Borbely does with 2SJ108. And J111 to cascode 2SK170. Both will get about 5V Vds on the Toshibas. But if you are using low rail voltages, then a single 2SJ174 without cascode is just fine.
People who have matched 2SJ74s would probably have unmatched ones to spare, so why not use them instead of a BJT + R +ZD + C ....
> With respect to RGs, is it better to reference them to PS ground (main filter caps) or ground at front end board?
You can argue that it is in theory better to have a single star point at the front end. My reasoning is that the current through the Rgs are not small, and I want them a bit away from the frontend circuit. Also, they are only there to offer a reference to ground, to lower absolute DC. Even if you connect them to the power ground instead of the signal ground, all you get is some small amount of common mode. And I declare that unimportant for AX-J. For A-J, you don't need the Rg.
But maybe Nelson would care to comment on this one.
Patrick
Actually ever since my first AX, I have always used cascoded JFETs as current source. So I can't really answer that one. To cascode 2SJ74, I use J174, in the same way as Borbely does with 2SJ108. And J111 to cascode 2SK170. Both will get about 5V Vds on the Toshibas. But if you are using low rail voltages, then a single 2SJ174 without cascode is just fine.
People who have matched 2SJ74s would probably have unmatched ones to spare, so why not use them instead of a BJT + R +ZD + C ....
> With respect to RGs, is it better to reference them to PS ground (main filter caps) or ground at front end board?
You can argue that it is in theory better to have a single star point at the front end. My reasoning is that the current through the Rgs are not small, and I want them a bit away from the frontend circuit. Also, they are only there to offer a reference to ground, to lower absolute DC. Even if you connect them to the power ground instead of the signal ground, all you get is some small amount of common mode. And I declare that unimportant for AX-J. For A-J, you don't need the Rg.
But maybe Nelson would care to comment on this one.
Patrick
