Can some body please tell me the ideal IDSS range for the differential pair (2SJ109, or 2 x 2SJ74)
Many thanks
Gerd
Many thanks
Gerd
The JFETs are biased by the CCS, 8.3mA
It's a differential, so theoretically one JFET should be able to eat the entire current from the constant current source.
Anything higher than an Idss of 8.6mA (2 times the 4.3mA from the input JFET) never hurts, but in practice any BLue will do. (Violet JFETs are a waste for this, imso)
It's a differential, so theoretically one JFET should be able to eat the entire current from the constant current source.
Anything higher than an Idss of 8.6mA (2 times the 4.3mA from the input JFET) never hurts, but in practice any BLue will do. (Violet JFETs are a waste for this, imso)
Thank you Jacco,
the reason why I asked, was that one of my two AJ build with the PCB and active components from the German diy AAA GB produced distortions at higher frequencies.
After checking the whole PCB and all measuring points I couldn't find any significant differences.
As I have the possibility to measure IDSS with a curve tracer - which I didn't at the begin - I have found that one J109 of the amp with distortions had an IDSS of only 7.00 / 7.15mA.
I still have some J74BL with IDSS > 8.6mA. I will give them a try to see if I can hear an improvement.
Gerd
the reason why I asked, was that one of my two AJ build with the PCB and active components from the German diy AAA GB produced distortions at higher frequencies.
After checking the whole PCB and all measuring points I couldn't find any significant differences.
As I have the possibility to measure IDSS with a curve tracer - which I didn't at the begin - I have found that one J109 of the amp with distortions had an IDSS of only 7.00 / 7.15mA.
I still have some J74BL with IDSS > 8.6mA. I will give them a try to see if I can hear an improvement.
Gerd
Hy Jacco,
I have measured all my 2SJ74BL stock and have selected two matched pairs with an IDSS of 10,55mA each, which I have exchanged against all 2SJ109BL.
The distortions are gone and the Aleph J sounds wonderfull with a high degree of musicallity.
In addition I have measured all the 2SJ109BL stock on IDSS which I purchased from different sources.
Several of them show values on the lower end of specification, means around 6 - 7 mA, which means they are not usable for Pass power amps.
Hence I would recommend to all AJ builders to measure IDSS of 2SJ109BL or 2SJ74BL before soldering them on the board.
Rgds
Gerd
I have measured all my 2SJ74BL stock and have selected two matched pairs with an IDSS of 10,55mA each, which I have exchanged against all 2SJ109BL.
The distortions are gone and the Aleph J sounds wonderfull with a high degree of musicallity.
In addition I have measured all the 2SJ109BL stock on IDSS which I purchased from different sources.
Several of them show values on the lower end of specification, means around 6 - 7 mA, which means they are not usable for Pass power amps.
Hence I would recommend to all AJ builders to measure IDSS of 2SJ109BL or 2SJ74BL before soldering them on the board.
Rgds
Gerd
Bill, I ordered 5% for R16-R23. They all match at .61, but not even close to
.47. These Rs are for current and signal source Fets, correct? Will this be a problem?
Thanks,
Greg
I would suspect that your meter does not zero out. Short the test leads and then subtract that reading from your resistor readings. If it doesn't reasonably add up then I would be looking for a new meter.
And check the batteries.
And check the batteries.
Last edited:
Bill,
Yep, that yielded .13ohms. Never thought of that. Something had to be wrong because 5% of .47 would not be +-.13!
It's a Fluke 8050 True Rms, and oldy but a goodie. I should probably have it calibrated with leads-
Yep, that yielded .13ohms. Never thought of that. Something had to be wrong because 5% of .47 would not be +-.13!
It's a Fluke 8050 True Rms, and oldy but a goodie. I should probably have it calibrated with leads-
I just caught an error on the schematic I have, but I'm not positive it is the latest official one.
R13 is 221 ohms and R14 is 392 ohms, that would have Q3 partially conducting at idle so it is already current limiting. I would imagine that R13 was supposed to be 1K.
If I missed a post about this, my apologies. It did not affect my first build because I was using boards originally set up for the A5.
R13 is 221 ohms and R14 is 392 ohms, that would have Q3 partially conducting at idle so it is already current limiting. I would imagine that R13 was supposed to be 1K.
If I missed a post about this, my apologies. It did not affect my first build because I was using boards originally set up for the A5.
Not really. It will have 0.35-0.4V between base and emitter of Q3 at idle which is not enough to open it.
I don't agree with you, but even if I did, when the swing goes more negative from idle it will definitely be current limiting.
Changed R13 to 1K, a profound difference. The magic is back.
For any skeptics out there, do the math. 2 X bias puts Q3 well into current limiting.
For any skeptics out there, do the math. 2 X bias puts Q3 well into current limiting.
Bill,
The values are the same as NP used in the Aleph 30 for those positions. I agree that it seems like overly aggressive protection. Grey used 221 and 100R respectively in those positions on his version of the AX.
As far as I can extrapolate from the ZTX450 data sheet, it won't conduct 0.1mA until Vbc>.5V. At .6V it still only would draw 1 mA so you'd think it is not a factor in normal operation. I still think I'll use a higher value for R13 since it really is only a protection circuit.
The values are the same as NP used in the Aleph 30 for those positions. I agree that it seems like overly aggressive protection. Grey used 221 and 100R respectively in those positions on his version of the AX.
As far as I can extrapolate from the ZTX450 data sheet, it won't conduct 0.1mA until Vbc>.5V. At .6V it still only would draw 1 mA so you'd think it is not a factor in normal operation. I still think I'll use a higher value for R13 since it really is only a protection circuit.