here is a schematic i found on the inetrnet ...the author claims 400W with 80+80 rails but there is a couple of things i dont get ....
1) what determine the bias of this amplifier ???
2) does the D4 1N4148 need to be in thermal contact on heatsink?
3) in this amplifier there is no VI limmiter as is in bipolar amps there is a protection circuit that protects mosfets from overdrive ... is this enough to protect also from the output section ??? in example shortcircuit conditions
4) presuming that its done by mistake and the load for the amplifier is 2 ohms will this circuit be able to protect at this condition also ??
finaly in the complete schematic ( available if needed ) there is an op amp that monitors curent flow on R35 220R and then conected to the output relay
what can it be protected if we monitor the curent flow in this resistor ???
for reference outputs are BUZ901-906
regards sakis
1) what determine the bias of this amplifier ???
2) does the D4 1N4148 need to be in thermal contact on heatsink?
3) in this amplifier there is no VI limmiter as is in bipolar amps there is a protection circuit that protects mosfets from overdrive ... is this enough to protect also from the output section ??? in example shortcircuit conditions
4) presuming that its done by mistake and the load for the amplifier is 2 ohms will this circuit be able to protect at this condition also ??
finaly in the complete schematic ( available if needed ) there is an op amp that monitors curent flow on R35 220R and then conected to the output relay
what can it be protected if we monitor the curent flow in this resistor ???
for reference outputs are BUZ901-906
regards sakis
Attachments
Looks similar to an early Hitachi implementation 
1. The 1N4148, not sure this would be very precise tbh. lateral mosfets only need a low gate voltage to turn on & flow some current. 0.7V might be enough, this would be 0.35V per gate. I'm not sure tbh. A variable resitor would have been better imo.
2. No, they are lateral mosfets & thus have a negative temperature coefficient, if they warm up then current through them will fall.
3. Maybe. The mosfets are limited to a 4V gate drive into a short circuit when idling, im not sure at all how much current this would allow a mosfet to pass. When a music signal is applied the gates can go to a higher voltage. Well assume you'll only short things out when connecting them up - not in operation
4. The thing would have a hard time driving a 2ohm load with rail voltages of +/- 80V with only 6 lateral mosfets
I should imagine that monitoring R35 (220R) will show when the VAS enters class B operation. This will only happen if the rather odd protection circuit is forced to operate or gate drive is trying to exceed 12.7V. Either condition is indicative of some kind of short on the output or very heavy load.
Not sure i like the idea of even using laterals with no source resistors but i didn't design the thing
1. The 1N4148, not sure this would be very precise tbh. lateral mosfets only need a low gate voltage to turn on & flow some current. 0.7V might be enough, this would be 0.35V per gate. I'm not sure tbh. A variable resitor would have been better imo.
2. No, they are lateral mosfets & thus have a negative temperature coefficient, if they warm up then current through them will fall.
3. Maybe. The mosfets are limited to a 4V gate drive into a short circuit when idling, im not sure at all how much current this would allow a mosfet to pass. When a music signal is applied the gates can go to a higher voltage. Well assume you'll only short things out when connecting them up - not in operation
4. The thing would have a hard time driving a 2ohm load with rail voltages of +/- 80V with only 6 lateral mosfets
I should imagine that monitoring R35 (220R) will show when the VAS enters class B operation. This will only happen if the rather odd protection circuit is forced to operate or gate drive is trying to exceed 12.7V. Either condition is indicative of some kind of short on the output or very heavy load.
Not sure i like the idea of even using laterals with no source resistors but i didn't design the thing
Last edited:
"The thing would have a hard time driving a 2ohm load with rail voltages of +/- 80V with only 6 lateral mosfets "
The Hafler DH500 did 400W/4R, 600W/3R, and 500W/2R with 3 pair, but it ran on ±93V at idle. The XL600 increased it to 4 pair and 750W/2R.
Original Hafler outputs were about 1A/1V, so clamping the gates to 4V would be about 4A per device.
The Hafler DH500 did 400W/4R, 600W/3R, and 500W/2R with 3 pair, but it ran on ±93V at idle. The XL600 increased it to 4 pair and 750W/2R.
Original Hafler outputs were about 1A/1V, so clamping the gates to 4V would be about 4A per device.
As event horizon mentions, it's "copy" of the old Hitachi application.
Here's the Maplin amp that's similar, you will see that Lateral FET's don't need temperature compensation like bjt's and HEXFET's. Also those 1k gate resistors seem far to high.
That IN4148 to provide bias is useless IMO.
Here's the Maplin amp that's similar, you will see that Lateral FET's don't need temperature compensation like bjt's and HEXFET's. Also those 1k gate resistors seem far to high.
That IN4148 to provide bias is useless IMO.
Attachments
Hi Sakis, if it is 2SK1058 and 2SJ162 used, then you can throw away that crazy protection circuit utilising Q4, Q5, ZD2, ZD6, D3, D7, ZD3, ZD5, R10, R12, R16, C13 and C10.
The 2SK1058 and 2SJ162 has gate protection zeners built in.
Now the amp is even more simple and trust me it works well and sounds pretty great and power is ample. Mosfets will swing about 5 -7V from the rail without clipping.
You can play with the gate resistor values to adjust the speed. I would use 470R.
I think you will enjoy this amp, it is very lively and sounds much more powerful than it is. If your supply is beefy, it will produce astonishingly clean and powerful low bass
Kind regards
Nico
The 2SK1058 and 2SJ162 has gate protection zeners built in.
Now the amp is even more simple and trust me it works well and sounds pretty great and power is ample. Mosfets will swing about 5 -7V from the rail without clipping.
You can play with the gate resistor values to adjust the speed. I would use 470R.
I think you will enjoy this amp, it is very lively and sounds much more powerful than it is. If your supply is beefy, it will produce astonishingly clean and powerful low bass
Kind regards
Nico
well Nico ...first of all it is very nice to talk to you and hope that u and family are doing just fine!!!
now ... power of 250 or 400 w is not a real problem ... get a rowbust suuply , high rails , many mosfets , goood cooling and more or less you are done ...
the thing is that most of my amplifires are supposed to be for light PA use and abuse is a fact ...they are going to be pushed
so i will go for a design that is VI+LDR limited eventhough sonics might not be perfect but it will guarrantee that the machine will conitnue to play even if abused ....
if this mosfet design is good to your opinion , simple to make , produces a lot of power and can be re-designed as idiot proof then it will start make sense ...
otherwise i cannot use it
( it is amazing that the cr@@piest construction i made one year ago is still alive and kicking eventhough abused daily ...reason : too small trafo that actually sugs under load so there is no voltage enough to make damage ....expect sound to be terrible but it seems that the costumer doesnt care (??? )
now ... power of 250 or 400 w is not a real problem ... get a rowbust suuply , high rails , many mosfets , goood cooling and more or less you are done ...
the thing is that most of my amplifires are supposed to be for light PA use and abuse is a fact ...they are going to be pushed
so i will go for a design that is VI+LDR limited eventhough sonics might not be perfect but it will guarrantee that the machine will conitnue to play even if abused ....
if this mosfet design is good to your opinion , simple to make , produces a lot of power and can be re-designed as idiot proof then it will start make sense ...
otherwise i cannot use it
( it is amazing that the cr@@piest construction i made one year ago is still alive and kicking eventhough abused daily ...reason : too small trafo that actually sugs under load so there is no voltage enough to make damage ....expect sound to be terrible but it seems that the costumer doesnt care (??? )
Last edited:
I recently repaired / updated a mono Rauch DVT 250m amplifier.
1000 watts into 2 ohms, 2kva of transformers @ 80 -0 - 80 rails
It has 8 pairs of Hitachi Lateral mosfets, and efficient heatsinking
Practically if you want to run 2 ohms, and the PSU is powerful enough, 8 pairs per channel is about right for 80-0-80 rails.
3 pairs will drive about 450W @ 4 Ohms, but 4 pairs and a decent sized psu is better.
1000 watts into 2 ohms, 2kva of transformers @ 80 -0 - 80 rails
It has 8 pairs of Hitachi Lateral mosfets, and efficient heatsinking
Practically if you want to run 2 ohms, and the PSU is powerful enough, 8 pairs per channel is about right for 80-0-80 rails.
3 pairs will drive about 450W @ 4 Ohms, but 4 pairs and a decent sized psu is better.
Hi Sakis,
Thanks for asking, we are well. Xoc1 is correct, you need a few more output devices in parrallel and this you can by just adding a few more devices including the gate and source resistors. For PA use you may want to use fan cooling, but yest provided the PSU can supply, this amplifier will deliver.
Earlier this year, I did a 350 watt amplifier using 2Sk1058/2SJ162 for Gokhan Nalbant (my Turkish Friends) using 8 pairs per channel, it was according to them quite a nice amp capable of high power and pretty good sound, but these devices unfortunately are very expensive.
I don't want to give Gokhan's e-mail adress since I am not sure if he is a DIY member and I don't know his handle.
Regards
Nico
Thanks for asking, we are well. Xoc1 is correct, you need a few more output devices in parrallel and this you can by just adding a few more devices including the gate and source resistors. For PA use you may want to use fan cooling, but yest provided the PSU can supply, this amplifier will deliver.
Earlier this year, I did a 350 watt amplifier using 2Sk1058/2SJ162 for Gokhan Nalbant (my Turkish Friends) using 8 pairs per channel, it was according to them quite a nice amp capable of high power and pretty good sound, but these devices unfortunately are very expensive.
I don't want to give Gokhan's e-mail adress since I am not sure if he is a DIY member and I don't know his handle.
Regards
Nico
I am currently trying out some Renesas SK1058 /SJ 162 Mosfets I bought here.
1 pcs x 2SJ162 & 1 pcs x 2SK1058 MOSFET RENESAS on eBay (end time 01-Jun-10 15:37:15 BST)
Price $7.20 per pair
I have yet to try a full power test, but VGS seemed to be well matched on the 8 pairs I bought.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.