Six years ago I built my F5 with extremely well matched Toshiba 2sk170BL and 2sj74BL (P-channel with Idss about 1.1mA higher than N-channel to create perfect complementary pair - P-channel scaled down to match the curves). IRFP240 and IRFP9240 Vgs matched at 1.3A by Alex (he always matches MOSFETs perfectly!).
I am very happy with it, it sounds great with most speakers.
Recently I found 80pcs of used 2SJ554 and 2SK2955 that I purchased years ago and I matched 2 pairs. Photos attached. I noticed that they have a much higher gain than Vishay MOSFETs, also lower Vgs, so I thought it might be a good idea to create another F5 with better complementary pair, NEC 2SK163M and 2SJ44M with Idss at about 11mA, they have lower gain than Toshiba JFETs. I have done a simulation (ASC attached), but don't have models for HITACHI MOSFETs so I used TOSHIBA which is quite close (thanks to Patrick EUVL). Changed input JFETs resistors to 2.2k as they have lower input capacitance than Toshiba JFETs. Lowered MOSFET gate resistors as HITACHI/RENESAS has higher Cgs capacitance. Also, MOSFETs source resistors reduced to 0R1. Biased at 1.36A. Of course, I would use NTC thermistors that are not on the schematics. The gain is 15.4dB. At 1W the 2nd harmonics at -84.5dBFS.
What do you think about it? Any suggestions, do you think it's worth trying?
I am very happy with it, it sounds great with most speakers.
Recently I found 80pcs of used 2SJ554 and 2SK2955 that I purchased years ago and I matched 2 pairs. Photos attached. I noticed that they have a much higher gain than Vishay MOSFETs, also lower Vgs, so I thought it might be a good idea to create another F5 with better complementary pair, NEC 2SK163M and 2SJ44M with Idss at about 11mA, they have lower gain than Toshiba JFETs. I have done a simulation (ASC attached), but don't have models for HITACHI MOSFETs so I used TOSHIBA which is quite close (thanks to Patrick EUVL). Changed input JFETs resistors to 2.2k as they have lower input capacitance than Toshiba JFETs. Lowered MOSFET gate resistors as HITACHI/RENESAS has higher Cgs capacitance. Also, MOSFETs source resistors reduced to 0R1. Biased at 1.36A. Of course, I would use NTC thermistors that are not on the schematics. The gain is 15.4dB. At 1W the 2nd harmonics at -84.5dBFS.
What do you think about it? Any suggestions, do you think it's worth trying?
Attachments
Last edited:
just go for it
fun is most spoiled with reluctance
do you recommend any changes, do you think it won't
with rising temperature? Thermistor should prevent this
in your shoes, I would start with 0R22 source resistors, then depending how well circuit behaves, would think of decreasing them
according to datasheets I foumd in the net - there is positive TempCo, so having source resistors is clever thing ..... and really entire game development depends of your entire thermal management and interaction with source resistors and NTCs
though , looking at SOA graph - do not use speakers under 8R load ..... at 40Vdc (part seeing near amp clipping) they're puny critters
or use 2 pairs, for calmer sleep
according to datasheets I foumd in the net - there is positive TempCo, so having source resistors is clever thing ..... and really entire game development depends of your entire thermal management and interaction with source resistors and NTCs
though , looking at SOA graph - do not use speakers under 8R load ..... at 40Vdc (part seeing near amp clipping) they're puny critters
or use 2 pairs, for calmer sleep
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.