I have a couple of questions to start, if I could get some help here.
I've been reviewing this thread and the babblefish thread looking for ideas.
I've been slowly assembling these AJ boards from PD.
If I use just one IRF240 per rail, should I drop the source resistor to .22 or so?
I am also considering using a set of SouthSemi SJEP120R100's as well. These have different input and output capacitances, and I believe this influences what source resistance to use. The IRF240 is 1300pf input 400pf output, and the R100 is 640pf input 103pf output.
I may use several of the SemiSouth devices in this circuit, as well, 2 per rail, and not sure if I need to lower/raise the source resistance. I dont beleive the positive rail/negative rail need to match.
I've been reviewing this thread and the babblefish thread looking for ideas.
I've been slowly assembling these AJ boards from PD.
If I use just one IRF240 per rail, should I drop the source resistor to .22 or so?
I am also considering using a set of SouthSemi SJEP120R100's as well. These have different input and output capacitances, and I believe this influences what source resistance to use. The IRF240 is 1300pf input 400pf output, and the R100 is 640pf input 103pf output.
I may use several of the SemiSouth devices in this circuit, as well, 2 per rail, and not sure if I need to lower/raise the source resistance. I dont beleive the positive rail/negative rail need to match.
for one IRF240 - look at Babelfish schematic , and you'll see 0R27 source resistors
Source resistances for SJEP ....... voltage .....
amny things are burried in these lines from Papa's pdf - First Watt Model J2 :
ok - just looking at data for 8 ohms - pretty everything is as with Aleph J
consumption - 100W per channel , same gain , same output power ......
also - looking at input impedance ......
that means several things :
both LTP gate/feedback resistors are 9K1 & 91K pairs
voltage across output Jfets and their source resistors are in same ballpark as on Aleph J output mosfets , or ~ 1V less
EDIT : not in same ballpark ...... more like 0V5 across source resistors and some 1V5 for Ugs - looking at lower Jfet , and data in datasheet
in case of upper jfet - these two source resistors aren't necessary (summed) in value as lower Jfet's source resistor ( which isn't drawn on simplified J schmtc) ....... they're probably each same as lower Jfet's source R
END of EDIT
xformer is - as already seen from pictures- same as in Aleph J ...... that means PSU voltage and Iq are the same as in Aleph J
current through input LTP is twice of current through Aleph J's LTP , from two ..... 3 ......... 4 reasons :
their drain resistor must be in same time as low is possible ( small voltage across ) and adequate load for left LTP part ...... and to preserve sweet spot characteristics , established by Papa .... meaning - enough current , but small enough dissipation .....
and I told hat I'll not chase J schematic ......
if you need exact J2 schematic - send me two pairs of these critters
there are too many "if's" without real thing on bench , and I see few different solutions worth trying for each ( of many
) part of amp , and they all are in spirit of Papa's simplified J .....
or just ask Papa is magic number of sold J's anywhere near ......
Source resistances for SJEP ....... voltage .....
amny things are burried in these lines from Papa's pdf - First Watt Model J2 :
ok - just looking at data for 8 ohms - pretty everything is as with Aleph J
consumption - 100W per channel , same gain , same output power ......
also - looking at input impedance ......
that means several things :
both LTP gate/feedback resistors are 9K1 & 91K pairs
voltage across output Jfets and their source resistors are in same ballpark as on Aleph J output mosfets , or ~ 1V less
EDIT : not in same ballpark ...... more like 0V5 across source resistors and some 1V5 for Ugs - looking at lower Jfet , and data in datasheet
in case of upper jfet - these two source resistors aren't necessary (summed) in value as lower Jfet's source resistor ( which isn't drawn on simplified J schmtc) ....... they're probably each same as lower Jfet's source R
END of EDIT
xformer is - as already seen from pictures- same as in Aleph J ...... that means PSU voltage and Iq are the same as in Aleph J
current through input LTP is twice of current through Aleph J's LTP , from two ..... 3 ......... 4 reasons :
their drain resistor must be in same time as low is possible ( small voltage across ) and adequate load for left LTP part ...... and to preserve sweet spot characteristics , established by Papa .... meaning - enough current , but small enough dissipation .....
and I told hat I'll not chase J schematic ......
if you need exact J2 schematic - send me two pairs of these critters
there are too many "if's" without real thing on bench , and I see few different solutions worth trying for each ( of many
) part of amp , and they all are in spirit of Papa's simplified J .....or just ask Papa is magic number of sold J's anywhere near ......
Last edited:
ZM,
I am not looking yet to fully copy the J2, no. I am looking to optimize the AJ board to handle these parts a bit.
I assume I could change the resistors R1,R2,R4 to match the J2, but I am not sure that lowering R4 would allow less feedback, monkey the gain up etc. I wouldn't mind around 12db of gain anyways.
It would be nice to have an alternative platform to get these P-JFETs working in without getting Papa's copyrights infringed.
I am waiting for a BA-1 board which also looks close, and I might be able to cascode parts here a bit easier to give some more current on the front end.
The questions remains in my mind, is running 2 PowerJets (besides more ecomomical) a smarter idea, and bias them to 1A or so each, or spread over the sink a bit better, and in this circuit makes more sense to use as designed.
Of course I may do both and see if I can get an idea of the sweetspot with my ears and mind.
I am not looking yet to fully copy the J2, no. I am looking to optimize the AJ board to handle these parts a bit.
I assume I could change the resistors R1,R2,R4 to match the J2, but I am not sure that lowering R4 would allow less feedback, monkey the gain up etc. I wouldn't mind around 12db of gain anyways.
It would be nice to have an alternative platform to get these P-JFETs working in without getting Papa's copyrights infringed.
I am waiting for a BA-1 board which also looks close, and I might be able to cascode parts here a bit easier to give some more current on the front end.
The questions remains in my mind, is running 2 PowerJets (besides more ecomomical) a smarter idea, and bias them to 1A or so each, or spread over the sink a bit better, and in this circuit makes more sense to use as designed.
Of course I may do both and see if I can get an idea of the sweetspot with my ears and mind.
You can drop them right in, if you wish. You will probably
find that you no longer need the Gate-Drain compensation
cap.
Some help please.
One of my AJ channel plays at a lower volume (soft) compared to the other channel. I think the input signals are not being amplified. I have changed Q3 (ZTX450), R18 source resistor and Q7 and Q8 (IRFP240) to no effect. The voltage across source resistor R18 is 0.78V compared to 0.46V (across R19, R16, R17). Voltage across R7 is 4.6V. Speaker output voltage is around 22mVDC (I have around -6.1mvDC on the output for the other channel).
Any suggestions are appreciated. Tx.
One of my AJ channel plays at a lower volume (soft) compared to the other channel. I think the input signals are not being amplified. I have changed Q3 (ZTX450), R18 source resistor and Q7 and Q8 (IRFP240) to no effect. The voltage across source resistor R18 is 0.78V compared to 0.46V (across R19, R16, R17). Voltage across R7 is 4.6V. Speaker output voltage is around 22mVDC (I have around -6.1mvDC on the output for the other channel).
Any suggestions are appreciated. Tx.
Just a thought: are the jfets in both amplifiers identical? I noticed in my jfet amp that the two channels sounded different (one played softer with less authority). I changed out the jfets in the weaker amp and solved the problem.
Theoretically, if the offset is close to zero then the voltage drop across R18 plus R19 will equal the drop across R16 plus R17. I suspect a bad reading somewhere to start with. Once you find that I imagine the one fet is either damaged or very badly matched.
Best, Bill
Thank you for all the suggestions.
Bill - you are right. I remeasured - voltage across r18 = 0.622v, r19 = 0.546v, r16 = 0.551v, r17 = 0.606v, Vout = -12.6mv. I am using a 5% source resistor - awaiting for a supply of Mills 1% resistors from partsconnexion.
I am now using a 2SJ109BL for the jfets - same effect of low volume. I previously have a matched 2sj74 pair.
I have also checked the speaker + connections - it is not connected to ground.
Any other suggestions are appreciated. Tx
Bill - you are right. I remeasured - voltage across r18 = 0.622v, r19 = 0.546v, r16 = 0.551v, r17 = 0.606v, Vout = -12.6mv. I am using a 5% source resistor - awaiting for a supply of Mills 1% resistors from partsconnexion.
I am now using a 2SJ109BL for the jfets - same effect of low volume. I previously have a matched 2sj74 pair.
I have also checked the speaker + connections - it is not connected to ground.
Any other suggestions are appreciated. Tx
You can drop them right in, if you wish. You will probably
find that you no longer need the Gate-Drain compensation
cap.
Ok, sorry for my ignorance, I looked around a bit, my guess this is C2,C3 - 220uf caps. But I am not sure.
I seem to have run into a similar problem as twitchie and mthoi.
With FETs not attached, I am only getting 3.5v across R7.
I wired a 500 ohm pot with a 510ohm resistor (ZM advice), and this did not vary it enough.
So I read a few possibilities here.
1. Remove R27, or go up to 100K
2. Lower R7 (680 ohms?)
3. Remove Q4, or put 777r resistor across it.
With IRF240's connected, R7 stays at 3.5v
With SemiSouth's connected (had to try anyway) I get 2.15v across R7, about 100mv offset, and 600mv on the Power JFETs. It hums, plays music for a second, then my test speaker starts thumping. Good times. I realize the SouthSemi devices conduct at a lower Vgs.
Anyways, I've tried replacing with J109 with two matched J74's at 8.8ma, got the IN- shorted to ground. I think my front end needs modification for sure, to get it running with one pair of 240's, then will dry 'dropping in' The power Jfets.
Any ideas? I figured I would ask before removing or changing more parts and losing my 2Sj's to further expermentation
With FETs not attached, I am only getting 3.5v across R7.
I wired a 500 ohm pot with a 510ohm resistor (ZM advice), and this did not vary it enough.
So I read a few possibilities here.
1. Remove R27, or go up to 100K
2. Lower R7 (680 ohms?)
3. Remove Q4, or put 777r resistor across it.
With IRF240's connected, R7 stays at 3.5v
With SemiSouth's connected (had to try anyway) I get 2.15v across R7, about 100mv offset, and 600mv on the Power JFETs. It hums, plays music for a second, then my test speaker starts thumping. Good times. I realize the SouthSemi devices conduct at a lower Vgs.
Anyways, I've tried replacing with J109 with two matched J74's at 8.8ma, got the IN- shorted to ground. I think my front end needs modification for sure, to get it running with one pair of 240's, then will dry 'dropping in' The power Jfets.
Any ideas? I figured I would ask before removing or changing more parts and losing my 2Sj's to further expermentation
current through LTP is (Uz1-Ube)/R8 = (9V1-0V65)/1k=8,45mA , well - in range of that
so - current through R7 gives you voltage across it 8,45mA/2 x R7
check Z1 ,Q2
so - voltage across R7 is completely dependent on input LTP's CCS (Z1, Q2,R8)
so - speaking of your solutions , I can say :
1. changing value of R27 is just for fine setting of Iq of output stage
2. depends what Papa tell ya what's adequate Ugs for SS output
3 . friendly speaking - rubbish
current through LTP is (Uz1-Ube)/R8 = (9V1-0V65)/1k=8,45mA , well - in range of that
so - current through R7 gives you voltage across it 8,45mA/2 x R7
check Z1 ,Q2
so - voltage across R7 is completely dependent on input LTP's CCS (Z1, Q2,R8)
so - speaking of your solutions , I can say :
1. changing value of R27 is just for fine setting of Iq of output stage
2. depends what Papa tell ya what's adequate Ugs for SS output
3 . friendly speaking - rubbish
Thanks ZM,
Here are voltages loaded with SSR100 devices.
Z1- 9.1v (seems I used right part)
Q2 -
CE- -12.78v
BE - -558v
CB -13v
R8 - 8.48
Can load up readings with IRF's later this week. Got to go to O-HI-O.
