cviller said:Thanks Choky for helping me with this!
I tried to move all ground wires to gnd terminal on my psu board and only running one channel but I didn't see any noticeable improvements, but I discovered something else. There is a very large difference in my CCS transistors source resistor voltages. One reads 450mV whilst the other is 200mV (across a 0.47 resistor).
I did match all my transistors to within 0.01V at 1A and 20sek. Could it be my resistors or should I try with some different transistors? The gate voltages were the same in reference to ground.
recheck resistors , mebbe you have R22 disguised as R47 😉
if they are the same-change mosfet.......and don't scratch a head about this....
regarding hum...........if rearranging gnd connections didn't solve your problem- you must play further with cap bank..........try with PSUd.....and- your Fish deserve chokes instead resistors .............
Shouldn't the circuit be able to reject +/- 100mV ripple on rails (I know it is not optimal, but anyways)?
I realized that I can run psud in linux, so I'll play around with it.
I realized that I can run psud in linux, so I'll play around with it.
cviller said:Shouldn't the circuit be able to reject +/- 100mV ripple on rails (I know it is not optimal, but anyways)?
I realized that I can run psud in linux, so I'll play around with it.



output stage is SE...........think..........
mebbe it's fine with output CCS,but certainly not with lower half
Ripple
Hi Cviller,
the problem is the current source for the input J-Fets. You can not measure the ripple voltage here because it is a ripple current
If you think, you have a time constant 33k x 0,22uF you are wrong. The input impedance of this CCS is very low, so you have no filtering at all.
I would recommend to use 10k + 22k and to tap this arrangement with a 100uF cap.
Ivo
Hi Cviller,
the problem is the current source for the input J-Fets. You can not measure the ripple voltage here because it is a ripple current

If you think, you have a time constant 33k x 0,22uF you are wrong. The input impedance of this CCS is very low, so you have no filtering at all.
I would recommend to use 10k + 22k and to tap this arrangement with a 100uF cap.
Ivo
Re: Ripple
Thanks for your idea. I just did a quick sketch to make sure that I understand you correctly the 20k resistor should be 22k - too quick there.
Wouldn't this 100u give a massive turn on thumb?
Ila said:Hi Cviller,
the problem is the current source for the input J-Fets. You can not measure the ripple voltage here because it is a ripple current![]()
If you think, you have a time constant 33k x 0,22uF you are wrong. The input impedance of this CCS is very low, so you have no filtering at all.
I would recommend to use 10k + 22k and to tap this arrangement with a 100uF cap.
Ivo
Thanks for your idea. I just did a quick sketch to make sure that I understand you correctly the 20k resistor should be 22k - too quick there.
Wouldn't this 100u give a massive turn on thumb?
Attachments
Ripple
Hi once again,
in a different set up (but with identical input stage structure) I had the same problem. The circuit below proved to be the solution. Perhaps you have to experiment with the values because there is a trade of between hum noise and turn on thumb.
Ivo
Hi once again,
in a different set up (but with identical input stage structure) I had the same problem. The circuit below proved to be the solution. Perhaps you have to experiment with the values because there is a trade of between hum noise and turn on thumb.
Ivo
Attachments
Thanks Ivo! It looks very promising. I might just add some relays to the output if the thumb is to insane. 😉
Just tried your trick Ivo, but it didn't fix it. Perhaps my problem is in the output ccs.
But thanks a lot for looking into my schematics.
But thanks a lot for looking into my schematics.
Hi cviller,
,sorry. I do not believe that the output CCS is the reason, it must be something else.
Another tip:
- The PNP-transistors in the input CCS are false connected. You have reversed the emitter/collector. The transistors are still operating but with a very low current gain.
- Try lower feedback resistors (scale down a bit)
Good luck and let me know.
Ivo



Another tip:
- The PNP-transistors in the input CCS are false connected. You have reversed the emitter/collector. The transistors are still operating but with a very low current gain.
- Try lower feedback resistors (scale down a bit)
Good luck and let me know.
Ivo
SL electronics and the 2SJ109
Hello folks,
I am thinking about making a few channels of Babbelfish. Sound and Light Electronics appear to have the 2SJ109. I think they might also have access to the 2SK389.
Does anyone know if they are a reputable group?
Sound & Light Electronics
3043 Washington Ave
El Monte, CA 91731-3357
http://www.slelectronics.net
Regards,
George.
Hello folks,
I am thinking about making a few channels of Babbelfish. Sound and Light Electronics appear to have the 2SJ109. I think they might also have access to the 2SK389.
Does anyone know if they are a reputable group?
Sound & Light Electronics
3043 Washington Ave
El Monte, CA 91731-3357
http://www.slelectronics.net
Regards,
George.
Ila said:Hi cviller,
- The PNP-transistors in the input CCS are false connected. You have reversed the emitter/collector. The transistors are still operating but with a very low current gain.
Do you have any idea how many times I have checked this - datasheet->layout->pcb??? many!
As I wrote in my hum thread - I discovered that one pnp was burned. So even though I couldn't see the error in layout - I tried to flip the pair to the other side and now I have one channel dead silent!!!!!
Thanks for all the help!
cviller said:
Do you have any idea how many times I have checked this - datasheet->layout->pcb??? many!
As I wrote in my hum thread - I discovered that one pnp was burned. So even though I couldn't see the error in layout - I tried to flip the pair to the other side and now I have one channel dead silent!!!!!
Thanks for all the help!
what you mean-"dead silent"
making music without hum,or not even making music?
I hope first......
You're right - could be misundestood. Music was never the problem, it has been good all along, but now the buzz has gone. Only hiss from preamp left (ear <5cm from speaker) and the wonderful jfet mos enhanced music left!
cviller said:You're right - could be misundestood. Music was never the problem, it has been good all along, but now the buzz has gone. Only hiss from preamp left (ear <5cm from speaker) and the wonderful jfet mos enhanced music left!
my respond in "other" thread .......
I know that I triple (at least) checked everything..........
Malotron said:Is an AlephX-Babblefish pcb a glimmer in Oly's eye...maybe?![]()
-Mal
not yet ;
that will wait a little ;
I hope that I'll soon receive pair of Aleph X pcbs ,gift from Demolition Man himself.............Mr. AR2

they will be -fer sure- excellent test bed for Babelfish JX ........ which I plan in near future .
anyway-fresh from Oly's Cousine- pcbs for bal or unbal Babelfish J (pic attached) .
I'll make final negotiation on Monday for first (little) batch production ........
Attachments
Oly's Cousine- pcbs for bal or unbal Babelfish J (pic attached) .
Looking good. I'll take a couple.Messing with a single sided AX-J...what a gaggle!
-Mal
Attachments
Balanced Babbelfish
Hi there ZenChoky,
Which of the many schematics in this thread is the new board based upon? Would it be from Post#92?
Regards,
George.
Hi there ZenChoky,
Which of the many schematics in this thread is the new board based upon? Would it be from Post#92?
Regards,
George.
Re: Balanced Babbelfish
tonight .....please 😉
I hope I hope I hope that I'll manage tomorrow at last that Paypal account I need for small Babelfish J pcb batch........in glory of Papa 😉
I have all opportunities to make pcbs of exceptional quality......but some lousy byrocratic things (as one that Serbia is not in Paypal covering) really bothers me.....
I'll squeeze schmtc pic later tonight
GeorgeBoles said:Hi there ZenChoky,
Which of the many schematics in this thread is the new board based upon? Would it be from Post#92?
Regards,
George.
tonight .....please 😉
I hope I hope I hope that I'll manage tomorrow at last that Paypal account I need for small Babelfish J pcb batch........in glory of Papa 😉
I have all opportunities to make pcbs of exceptional quality......but some lousy byrocratic things (as one that Serbia is not in Paypal covering) really bothers me.....
I'll squeeze schmtc pic later tonight
- Status
- Not open for further replies.
- Home
- Amplifiers
- Pass Labs
- Babbelfish J PCBs