Hi guys,
I bought an EB-199/320 All-Fet phono preamp kit several years ago and was told recently that it would be possible to reduce the input voltage noise by cancelling the J-Fets' source resistors (8 per channel) and even the offset-trimming potentiometer (P1) only if I use extremely tight matched J-Fets.
Borbely provided J74-BL and K170-BL Toshiba J-Fets showing an Idss between 8.8mA and 9.7mA with a 6.8 Ohms value for the source resistors (R4 to R11) and a 10 Ohms value for P1 : in this case, the total drain current is 20mA (5mA per J-Fets' pair) as confirmed by the 3V voltrage measured across the 150 Ohms drain resistor.
- does anybody know the Idss value for the new tightly matched J-Fets octet (4 pairs) if we only cancell all the 8 source resistors ?
- same question if we cancel the source resistors and also P1 ?
Thanks.
I bought an EB-199/320 All-Fet phono preamp kit several years ago and was told recently that it would be possible to reduce the input voltage noise by cancelling the J-Fets' source resistors (8 per channel) and even the offset-trimming potentiometer (P1) only if I use extremely tight matched J-Fets.
Borbely provided J74-BL and K170-BL Toshiba J-Fets showing an Idss between 8.8mA and 9.7mA with a 6.8 Ohms value for the source resistors (R4 to R11) and a 10 Ohms value for P1 : in this case, the total drain current is 20mA (5mA per J-Fets' pair) as confirmed by the 3V voltrage measured across the 150 Ohms drain resistor.
- does anybody know the Idss value for the new tightly matched J-Fets octet (4 pairs) if we only cancell all the 8 source resistors ?
- same question if we cancel the source resistors and also P1 ?
Thanks.
Below 7.0
My friends 320 from Erno has GR devices , no source resistors, just 10r trimpot maybe his around 3.5 idss.
I have my 320 with 4 x 2SK147 and 4 x 2SJ72 idss around 6.5
With only Vishay bulk metal foil 10R trimpot, no source resistor It draws 27 ma in the front end.
The 147/72 has 17ohm internal resistance instead of 34 ohm for 170/74 so quieter .
John curl uses his best matched idss 2SK146 / 2SJ73 sets for Constellation Orion phono and uses 5 ohm trim pot. The Constellation Perseus uses not the best matched devices and 10R trimpot
You can't get rid of both source resistors and trim pot. Too hard to match devices to get no offset and they all drift from thermal effects unless you pot them all in a big block of thermal epoxy and them you can't change one if it goes bad.
So a nice idea, but not practicle. Toshiba jfets are more stable than Linear System for idss drift for intial selection and less drift over time. So if you can find enough Toshiba GR devices you can try to match your sets, but i think really hard to find them now.
If you change the jfet idss current you have to change the the 150 R resistors to get 3.0 volts with the new front end total current. I'm using 110 ohms with 27ma current.
My friends 320 from Erno has GR devices , no source resistors, just 10r trimpot maybe his around 3.5 idss.
I have my 320 with 4 x 2SK147 and 4 x 2SJ72 idss around 6.5
With only Vishay bulk metal foil 10R trimpot, no source resistor It draws 27 ma in the front end.
The 147/72 has 17ohm internal resistance instead of 34 ohm for 170/74 so quieter .
John curl uses his best matched idss 2SK146 / 2SJ73 sets for Constellation Orion phono and uses 5 ohm trim pot. The Constellation Perseus uses not the best matched devices and 10R trimpot
You can't get rid of both source resistors and trim pot. Too hard to match devices to get no offset and they all drift from thermal effects unless you pot them all in a big block of thermal epoxy and them you can't change one if it goes bad.
So a nice idea, but not practicle. Toshiba jfets are more stable than Linear System for idss drift for intial selection and less drift over time. So if you can find enough Toshiba GR devices you can try to match your sets, but i think really hard to find them now.
If you change the jfet idss current you have to change the the 150 R resistors to get 3.0 volts with the new front end total current. I'm using 110 ohms with 27ma current.
Last edited:
I have my first 320 from Erno in 2010. He was supplying K170/J74 in GR grade with IDSS between 3.5mA and 4mA. No source resistors needed and R3,12 were 300R.
Dated back in 2011, I took the phono amp to my friend's factory to desolder the K170/J74 GR then put them in a curve tracer. The IDSS were not tightly matched, from 3.55 to 3.90 but the main problem of mine was one J74GR had a different curve than the others. I also measured the other BL grade jfets and they were not tightly matched either. The biggest discovery were even the dual jfets did not have tight IDSS within the same jfet too! I asked Erno but he said he only matched them under IDSS and he did not know the effect if one of the jfet has a different curve and he can do nothing abut the dual jfets.
I was curious and purchased four bags of jfets - J74GR, K170GR, J74BL and K170BL from a Yokohama supplier (each bag is 200 pieces). I didn't purchase the dual jfets because I think they were too expensive at that time. Recently, one of the dual jfet was gone and I have to pay $16 plus shipping from a guy in NZ.
My friend asked his workers to measure the IDSS of each jfet (1600 of them) and compared with those supplied from Erno. Over 90% of them had the same curve and out of 200 of each GR bag, there were not more than 15 of them within the range of 3.5 to 4mA. I chose 8 of each GR with tight IDSS (matched within 0.1mA) with same curve pattern putting them back into the 320 phono - the result was, I can't tell the difference exactly but it "feels" better.
Later, a friend of mine purchased another 320 kit from Erno's appointed successor in Taiwan (Mr. Chan) and he supplied the 320 kit with BL grade for the first input stage plus the source resistor (dale). I asked Erno (he was still replying at that time) and he said there should be no difference between the Taiwan kit and his. However, after completion, we can hear more hiss from the speaker. After changing those BL grade into GR grade, the hiss was lowered.
I believe the source resistors make a difference. Perhaps better quality source resistor can also help.
Dated back in 2011, I took the phono amp to my friend's factory to desolder the K170/J74 GR then put them in a curve tracer. The IDSS were not tightly matched, from 3.55 to 3.90 but the main problem of mine was one J74GR had a different curve than the others. I also measured the other BL grade jfets and they were not tightly matched either. The biggest discovery were even the dual jfets did not have tight IDSS within the same jfet too! I asked Erno but he said he only matched them under IDSS and he did not know the effect if one of the jfet has a different curve and he can do nothing abut the dual jfets.
I was curious and purchased four bags of jfets - J74GR, K170GR, J74BL and K170BL from a Yokohama supplier (each bag is 200 pieces). I didn't purchase the dual jfets because I think they were too expensive at that time. Recently, one of the dual jfet was gone and I have to pay $16 plus shipping from a guy in NZ.
My friend asked his workers to measure the IDSS of each jfet (1600 of them) and compared with those supplied from Erno. Over 90% of them had the same curve and out of 200 of each GR bag, there were not more than 15 of them within the range of 3.5 to 4mA. I chose 8 of each GR with tight IDSS (matched within 0.1mA) with same curve pattern putting them back into the 320 phono - the result was, I can't tell the difference exactly but it "feels" better.
Later, a friend of mine purchased another 320 kit from Erno's appointed successor in Taiwan (Mr. Chan) and he supplied the 320 kit with BL grade for the first input stage plus the source resistor (dale). I asked Erno (he was still replying at that time) and he said there should be no difference between the Taiwan kit and his. However, after completion, we can hear more hiss from the speaker. After changing those BL grade into GR grade, the hiss was lowered.
I believe the source resistors make a difference. Perhaps better quality source resistor can also help.
Last edited:
data sheets for2SJ72 and 2SK147.
i have 6.5 -7.0 idss devices ( not much choice about IDSS even when they were available)
these are the input devices for Vendetta Research phono preamp.
original mid 80's Borbely's first Audio Amateur preamp and 1988 MC phono stage used these devices, but by early 90's they were out of production.
2SK369 is same as 2SK147 but less dissipation 400mw
i have 6.5 -7.0 idss devices ( not much choice about IDSS even when they were available)
these are the input devices for Vendetta Research phono preamp.
original mid 80's Borbely's first Audio Amateur preamp and 1988 MC phono stage used these devices, but by early 90's they were out of production.
2SK369 is same as 2SK147 but less dissipation 400mw
Last edited:
But with higher current cascode fets Q9-Q12 ,2sk246 and 2sj103 have lower vgs voltage than minimum 2V for the drains of Q1-Q4 to work properly. I think that you have about 1V at drains of input J-fets.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.