Help me out here guys not much experience with mosfets and i am running in to problems .
I have an amplifier to repair based on 2SJ 118 and 2SK 413 no schematic available though the rest of circuit is pretty much generic MJE340-350 drivers and MJE 340-350 Vas and current source .
Outputs are obsolete so replaced with different pin out 2SJ162 2SK1058 easy to do
Bias though hit the sky not possible to adjust lower than 1A while the original bias was about 400ma per device Class A
Noticed that the bias schema was also generic created with one zener diode and one trimmer ..replaced the zener diode with 100R and out of the blue bias is adjustable in a reasonable level and remains stable
Then the question is about gate resistors and most schematics i ve seen with2SJ162 2SK1058 have gate resistors from 82-470R while the original amp with 2SJ 118 and 2SK 413 had gate resistors of 3K9
any help please ?
Kind regards
sakis
I have an amplifier to repair based on 2SJ 118 and 2SK 413 no schematic available though the rest of circuit is pretty much generic MJE340-350 drivers and MJE 340-350 Vas and current source .
Outputs are obsolete so replaced with different pin out 2SJ162 2SK1058 easy to do
Bias though hit the sky not possible to adjust lower than 1A while the original bias was about 400ma per device Class A
Noticed that the bias schema was also generic created with one zener diode and one trimmer ..replaced the zener diode with 100R and out of the blue bias is adjustable in a reasonable level and remains stable
Then the question is about gate resistors and most schematics i ve seen with2SJ162 2SK1058 have gate resistors from 82-470R while the original amp with 2SJ 118 and 2SK 413 had gate resistors of 3K9
any help please ?
Kind regards
sakis
Last edited:
not sure if this helps but the parts you used have a much lower threshold voltage compared to the originals... so its probably worth changing the zeners over to something like 3v (see the transfer characteristic graphs for 1058 vs 413 to get an idea). gate resistors can't be causing big differences in clipping behavior etc....
transconductance of the 1058 types are half that of the 413 so maybe worth dropping the source resistors also (maybe by half?)
good luck!
cheers
weird that the gate stoppers they used seem so high.
transconductance of the 1058 types are half that of the 413 so maybe worth dropping the source resistors also (maybe by half?)
good luck!
cheers
weird that the gate stoppers they used seem so high.
How many times do we go through this? 
You used the wrong MOSFETs as substitutes. Proper ones are 2SK1529/2SJ200 (a bit of extra current handling at 10A vs 8A of the originals).
K1058/K162 are laterals, low treshold, negative tempco, low Gm.
K413/J118 are DMOS PI-channel devices, higer treshold, POSITIVE tempco (so need a bias servo), high Gm. K1529/J200 and K1530/J201 are direct descendants of the above.
Did you not slaim thousands of amps fixed, you should know the difference?!
You used the wrong MOSFETs as substitutes. Proper ones are 2SK1529/2SJ200 (a bit of extra current handling at 10A vs 8A of the originals).
K1058/K162 are laterals, low treshold, negative tempco, low Gm.
K413/J118 are DMOS PI-channel devices, higer treshold, POSITIVE tempco (so need a bias servo), high Gm. K1529/J200 and K1530/J201 are direct descendants of the above.
Did you not slaim thousands of amps fixed, you should know the difference?!
Window shopping Magnatec. ALFET Lateral MOSFETs or Mosfets/Modules - Lateral MOSFETs - Class D
NOT RELEVANT, please do not confuse the issue.
2SJ118/2SK415 are NOT lateral. They are PI trench MOSFETs, so akin to verticals, low treshold, like other Toshiba audio MOSFETs. And as someone else pointed out, the pinout is different.
The most proper repair is to replace with the same device. If obsolete then there is that Toshiba device that has closer character than the latfet, and adjust to restore similar performance.
Latfet has lower Vgs so most probably you want to see no more than 2V between NMOS gate and PMOS gate. The original mosfet has up to 2V-5V of Vgs threshold so you may find the amp has been designed around 5V between gates. If this high potential is applied to latfet you will find higher bias so it will be very hot. Yes you need to lower the bias setting if you have to use lateral.
Latfet has negative temp coefficient, the original has positive. In the original, most probably the compensation is poor (I believe you can see terrible cap based compensation also in the circuit) so that they need to slow down the output with such a high gate stopper. Usually you will run into problem lowering them so must be adjusted carefully.
The zener diode is the simplest device for bias tracking. When the temperature is hot or there is more current the resistance is lower. So for output stage having positive temperature coefficient you will find that increasing the resistance in place of the diode will reduce bias. That's why when you change positive tempco device into negative one, you may want to replace the diode with resistor, or move the diode to different location where increased resistance will increase the bias (such as between NMOS and PMOS gates)
Latfet has lower Vgs so most probably you want to see no more than 2V between NMOS gate and PMOS gate. The original mosfet has up to 2V-5V of Vgs threshold so you may find the amp has been designed around 5V between gates. If this high potential is applied to latfet you will find higher bias so it will be very hot. Yes you need to lower the bias setting if you have to use lateral.
Latfet has negative temp coefficient, the original has positive. In the original, most probably the compensation is poor (I believe you can see terrible cap based compensation also in the circuit) so that they need to slow down the output with such a high gate stopper. Usually you will run into problem lowering them so must be adjusted carefully.
The zener diode is the simplest device for bias tracking. When the temperature is hot or there is more current the resistance is lower. So for output stage having positive temperature coefficient you will find that increasing the resistance in place of the diode will reduce bias. That's why when you change positive tempco device into negative one, you may want to replace the diode with resistor, or move the diode to different location where increased resistance will increase the bias (such as between NMOS and PMOS gates)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.