Built the best I could so far with the schematic. Waiting for laterals from profusion, ordered May 1st, still havent shipped. I have a matched pair of jfets at 9ma idss. I know the damping will suffer. Should I get a set of jfets matched at 4.5-6 instead off ebay? This is a 3.2 build with FQA drivers.
Attachments
Good progress gsrchrisu
For the higher jfet IDSS value the effect on lowering the damping factor is already reduced by using the FQA drivers so you might get still the same or even higher DF than the original version 3.0. You might try it anyway like that and see how it goes. EBay is quite risky anyway to get genuine parts like that…. You ate better off with sellers on this audio forum or you can get the LSK/LSJ series which have also low IDSS value.
Fab
For the higher jfet IDSS value the effect on lowering the damping factor is already reduced by using the FQA drivers so you might get still the same or even higher DF than the original version 3.0. You might try it anyway like that and see how it goes. EBay is quite risky anyway to get genuine parts like that…. You ate better off with sellers on this audio forum or you can get the LSK/LSJ series which have also low IDSS value.
Fab
What is the lowest IDSS that I should consider? I have seen some at 4, is that too low. The sellers on ebay that I would use are punkydawgs or Truna?(he is on these boards, from Japan) I am going to try my 9's and see how it sounds.
Is R13 is missing ?
When ever bending metal , a form should always be used . So rather bending the leads on the MOSFET's freehand ,
recommend bending them around a 1/4 " drill bit .
I used a 6-32 machine screw to fasten the FQP3 MOSFET's to the heat sink .
The hole in the Exicon MOSFET's is large enough to accommodate a 6-32 , but there is no margin for error .
So consider using a 4-40 machine screws for the Exicons .
If you are waiting for the Exicon MOSFET's to arrive , you can still test the USSA 3.2 without them , as shown on post #92 .
Highly recommend using a dual bench supply for this , and start off with a low voltage supply rails .
Maybe +/- 10 Vdc . Keep an eye on the current when you switch on the bench supply , and slowly bring the supply voltage up to +/- 24 Vdc .
This allows you to get familiar with how the trimmers work .
Actually , I taped up the jeweler's screw driver , so if it was dropped , it wouldn't short anything .
I measured the voltage drop across the source resistors to obtain the bias current of the FQP3 MOSFET's .
.
Yes in the schematics r13 is crossed out and just RV1 is listed, but I see I would then possibly need to jump r13 to complete the circuit. I am also missing r19/20 since the values are not clear in the schematic. I planned on asking once I get back to working on the Pcb. I may PM you to ask some of your values and I am still not totally clear on setting the current for the FQA mosfets. Will def be trying first with no laterals and just a power supply. Profusion actually finally shipped today so hurray.
" I am still not totally clear on setting the current for the FQA mosfets. "
Gsrchrisu , I am absolutely no expert on this . However , as FAB coached me , the key is to look at the Yfs ( transconductance ) curve
of a MOSFET , which is not linear .
recall Yfs = delta Id / delta Vgs
https://electronics.stackexchange.com/questions/316397/change-of-transconductance-for-mosfet
The key is to set Id ( bias current ) high enough so that the MOSFET is operating on a section of the transconductance curve that is approx linear .
Unfortunately , the datasheet for the FQP3n30 doesn't show this curve .
https://www.onsemi.com/pdf/datasheet/fqp3n30-d.pdf
Have a look at my post #88 and #89 . I tested the USSA 3.2 without the Exicon MOSFET's , and had a look at the output on a scope .
When the bias was too low , there was a compression effect .
Note my DIY generator creates a waveform rises and falls based on the number e , its not a triangle wave .
Probably absolute minimum bias for the FQP's is somewhere around 55 mA .
So recommend testing the USSA 3.2 without the Exicon's , play around with the bias current and have a look on the output using scope .
Increase the bias current to when the waveform is no longer compressed and that's the minimum bias current .
.
Gsrchrisu , I am absolutely no expert on this . However , as FAB coached me , the key is to look at the Yfs ( transconductance ) curve
of a MOSFET , which is not linear .
recall Yfs = delta Id / delta Vgs
https://electronics.stackexchange.com/questions/316397/change-of-transconductance-for-mosfet
The key is to set Id ( bias current ) high enough so that the MOSFET is operating on a section of the transconductance curve that is approx linear .
Unfortunately , the datasheet for the FQP3n30 doesn't show this curve .
https://www.onsemi.com/pdf/datasheet/fqp3n30-d.pdf
Have a look at my post #88 and #89 . I tested the USSA 3.2 without the Exicon MOSFET's , and had a look at the output on a scope .
When the bias was too low , there was a compression effect .
Note my DIY generator creates a waveform rises and falls based on the number e , its not a triangle wave .
Probably absolute minimum bias for the FQP's is somewhere around 55 mA .
So recommend testing the USSA 3.2 without the Exicon's , play around with the bias current and have a look on the output using scope .
Increase the bias current to when the waveform is no longer compressed and that's the minimum bias current .
.