Hum problem with jfet Aleph 3 (babbelfish)

[cviller]

Hi

I thought it would be better to start a new thread with my problem. Perhaps I'm not the only one struggling with hum.
http://www.diyaudio.com/forums/showt...26#post1119626

I have built a different version of Zen Mod's babbelfish, with two pairs of output transistors. It sounds very good (I had it working without noise for some reason), but now I have problems removing the noise and adjusting the bias.

I have tried to connect all ground wires the same place, but that didn't fix my problem. I have also swapped all bjt's and tried removing c102.

My supply is a crc with 2x15mF->0.2R->4x15mF and my ripple is +/- 25mV with one channel running, but I want to use it to power both channels. The output ripple is 10mV

I have attached my schematics with measured values and I have some pictures of my setup here:
http://viller.org/audio/2007jan_jfish/hum/

Please tell me if I need to supply more information. I have a 60MHz scope.

Questions:
How much ripple should it be able to reject?
Is my supply simply too tiny? (I do however have 90mF per rail while aleph 3 has 88mF)
Do I have any fundamental errors in my schematics?
How do I fix it?

[Zen Mod]

Originally posted by cviller
Hi

I thought it would be better to start a new thread with my problem. Perhaps I'm not the only one struggling with hum.
http://www.diyaudio.com/forums/showt...26#post1119626

I have built a different version of Zen Mod's babbelfish, with two pairs of output transistors. It sounds very good (I had it working without noise for some reason), but now I have problems removing the noise and adjusting the bias.

I have tried to connect all ground wires the same place, but that didn't fix my problem. I have also swapped all bjt's and tried removing c102.

My supply is a crc with 2x15mF->0.2R->4x15mF and my ripple is +/- 25mV with one channel running, but I want to use it to power both channels. The output ripple is 10mV

I have attached my schematics with measured values and I have some pictures of my setup here:
http://viller.org/audio/2007jan_jfish/hum/

Please tell me if I need to supply more information. I have a 60MHz scope.

Questions:
How much ripple should it be able to reject?
Is my supply simply too tiny? (I do however have 90mF per rail while aleph 3 has 88mF)
Do I have any fundamental errors in my schematics?
How do I fix it?

adjusting the bias?

if you have problems with that-just decrease slightly R124 in your schematic

I see that you now have similar currents per mosfet ; what was issue? mosfet or resistor?

you must tell us now how you made supply .

mebbe diode bridges are tiny....... or you need to rearrange CRCs

[jeapel]

hi

technical details for my normal A3 with irf9610
-one transfo for both channels + 2 x 35A bridge +2 x 68000uF
-after separate LC sections for right and left
-total cap 408000uF +4 x 1.2mH
-25 v rail and 2A bias
-power supply ripple is 25mV
-hum is very low on output
-oversize heatsink for around 40C to 45C after 3 hrs
-exactly 50% ac gain

-just to check your power supply my supply ripple for both
channels 4A is also 25mv ac and just a small hum on
(92db speaker) no problem so your supply seem ok

-maybe a ground loop so if you remove all input what
happend to hum on speaker and also ground input

-maybe transfo magnetic induction so ?? try
to isolate with metal sheet

-when you said "problem to adjust bias" same problem
for me my aleph ccs was unstable but with 1nF cap
on base and collector q103 (mpsa18 in normal A3)
it s ok now

bye

[cviller]

Originally posted by Zen Mod



adjusting the bias?

Sorry - I meant biaspoint or dc offset.

Originally posted by Zen Mod

I see that you now have similar currents per mosfet ; what was issue? mosfet or resistor?

I changed the mosfets.

Originally posted by Zen Mod

you must tell us now how you made supply .

mebbe diode bridges are tiny....... or you need to rearrange CRCs

I have two toroids with 10x17V at 1.5 each. I use two 25A bridges (one for each rail)
http://pdf1.alldatasheet.com/datashe...TC/FB2506.html

capacitors are in total 12x15000uF

[cviller]

Originally posted by jeapel
hi

technical details for my normal A3 with irf9610
-one transfo for both channels + 2 x 35A bridge +2 x 68000uF
-after separate LC sections for right and left
-total cap 408000uF +4 x 1.2mH
-25 v rail and 2A bias
-power supply ripple is 25mV
-hum is very low on output
-oversize heatsink for around 40C to 45C after 3 hrs
-exactly 50% ac gain

-just to check your power supply my supply ripple for both
channels 4A is also 25mv ac and just a small hum on
(92db speaker) no problem so your supply seem ok

-maybe a ground loop so if you remove all input what
happend to hum on speaker and also ground input

-maybe transfo magnetic induction so ?? try
to isolate with metal sheet

-when you said "problem to adjust bias" same problem
for me my aleph ccs was unstable but with 1nF cap
on base and collector q103 (mpsa18 in normal A3)
it s ok now

bye

Thanks for your data - makes me think I might be able to work this out.

I just tried to see if a metal plate between psu and amp changed anything, but nothing changed on the scope.

[JC Fardo]

Originally posted by cviller

I have two toroids with 10x17V at 1.5 each. I use two 25A bridges (one for each rail)
http://pdf1.alldatasheet.com/datashe...TC/FB2506.html

capacitors are in total 12x15000uF

Hi
Is the rated VA of trafos sufficient? Seems some tinny.
Cheers

[Zen Mod]

Originally posted by cviller


Thanks for your data - makes me think I might be able to work this out.

I just tried to see if a metal plate between psu and amp changed anything, but nothing changed on the scope.

seems irrelevant - if I conclude OK-that you have hum even in breadboard variant,where toroid was miles away from main pcb

if then you haven't hum.......there is a culprit

[cviller]

I think I might have found the error - or at least I have found something which is wrong. The CCS regulation seems to be oscillating insanely. I took a picture (with my left hand - sorry!) of the scope output. I'm looking at the node connected to basis of Q103 and R133,R125 and R123.

The scope is set to 5V and 2us

[jeapel]

did you try 1nF on base and collector

the official aleph 30 design has 1nF

and it s work in my aleph 3

[cviller]

Can I use 10nF?