Hi everyone. As a first post, I would like to salute everyone and say thanks to all those that post around here. After all, the Solid State section of this forum has been the inspiration for my three most recent projects.
With that out of the way, let's get to the meat.
I've been thinking about building a new amplifier for the last two years or so. I just got to it about 3 weeks ago. My knowledge of electronics is not really good, but I like to experiment.
I've read Randy Slone's "High-power Audio Amplifier Construction Manual" and "The Audiophile's Project Sourcebook". And for now, I think that the most appropriate amplifier for my needs would be the Opti-mos. It's a 200W into 8ohm mosfet amplifier. It requires about 67.5VDC to output this power.
The first part I designed is of course, the power supply. My goal in terms of power is low noise and two channel connected to the same power supply. And according to Mr. Slone, 40000uF per rail should be enough.
The reason of my post is that I would like your suggestions and possible improvement for the following schematic (It is big, but made to be printed on a 8x11 sheet of paper):
Power Supply Schematic
The parts that I already have are the transformer (Plitron 50V+50V, 750VA) and the reservoir capacitors (Nippon Chemi-Con 68000uF, 50V)
Any comments will be appreciated.
Thanks,
Sébastien
With that out of the way, let's get to the meat.
I've been thinking about building a new amplifier for the last two years or so. I just got to it about 3 weeks ago. My knowledge of electronics is not really good, but I like to experiment.
I've read Randy Slone's "High-power Audio Amplifier Construction Manual" and "The Audiophile's Project Sourcebook". And for now, I think that the most appropriate amplifier for my needs would be the Opti-mos. It's a 200W into 8ohm mosfet amplifier. It requires about 67.5VDC to output this power.
The first part I designed is of course, the power supply. My goal in terms of power is low noise and two channel connected to the same power supply. And according to Mr. Slone, 40000uF per rail should be enough.
The reason of my post is that I would like your suggestions and possible improvement for the following schematic (It is big, but made to be printed on a 8x11 sheet of paper):
Power Supply Schematic
The parts that I already have are the transformer (Plitron 50V+50V, 750VA) and the reservoir capacitors (Nippon Chemi-Con 68000uF, 50V)
Any comments will be appreciated.
Thanks,
Sébastien
Hi,
welcome to the forum ( though you've been here for some time by the looks of it .)
A few comments on your power supply:
Inrush current to a 750VA toroid will be substantial. Use a very high current capable on/off switch with a suppressor ( L/C/R or transient VDR types ) or be prepared to replace it every few months. An inrush current limiter may be useful and suitable circuits have been discussed in a number of threads.
If your transformer has two secondary windings ( 4 leads ) consider using two bridges to generate the +/- rails. This lowers the peak currents seen by the diodes and the transformer, decreasing temperature rise and helping reliability. In addition the ripple will be at a higher frequency which may help.
Encapsulated bridges may not be the best solution - but are convenient. The use of high speed diodes with soft recovery characteristics seem to be the general concensus - see the high speed diodes thread. This discussion hasn't reached any firm conclusions yet but a number of devices have been suggested.
James
welcome to the forum ( though you've been here for some time by the looks of it .)
A few comments on your power supply:
Inrush current to a 750VA toroid will be substantial. Use a very high current capable on/off switch with a suppressor ( L/C/R or transient VDR types ) or be prepared to replace it every few months. An inrush current limiter may be useful and suitable circuits have been discussed in a number of threads.
If your transformer has two secondary windings ( 4 leads ) consider using two bridges to generate the +/- rails. This lowers the peak currents seen by the diodes and the transformer, decreasing temperature rise and helping reliability. In addition the ripple will be at a higher frequency which may help.
Encapsulated bridges may not be the best solution - but are convenient. The use of high speed diodes with soft recovery characteristics seem to be the general concensus - see the high speed diodes thread. This discussion hasn't reached any firm conclusions yet but a number of devices have been suggested.
James
Update!
Thanks to nemestra for the different suggestions. Here is an updated version of the schematic . The modifications are as follow:
- Added a soft start circuit (upper part)
- Added a second bridge in the main power supply
- Added a DC protection for my loudspeaker output (lower part)
Altough i'm rather sure of 90% of the wiring, I'd like your opinion on the wiring of the two bridges in the main power supply. And I'd also like to have a validation that the circuit in the red zone will truly output +15V. (Is it right or did I mess up (it is 2AM after all)
As for the use of high speed diodes, I like the simplicity of an integrated bridge and don't believe the sound would benefit enough to consider adding four of these.
Simply because they are rated for 50V and that the power supply outputs 70V. So by putting them up in series with resistors, I can be sure that each capacitor won't be exposed to more than 70V/2 = 35V and still get a decent 1/((1/68000)+(1/68000)) = 34000uF from those caps.
Any comment will (as always) be appreciated.
Thanks,
Sébastien
Thanks to nemestra for the different suggestions. Here is an updated version of the schematic . The modifications are as follow:
- Added a soft start circuit (upper part)
- Added a second bridge in the main power supply
- Added a DC protection for my loudspeaker output (lower part)
Altough i'm rather sure of 90% of the wiring, I'd like your opinion on the wiring of the two bridges in the main power supply. And I'd also like to have a validation that the circuit in the red zone will truly output +15V. (Is it right or did I mess up (it is 2AM after all)
As for the use of high speed diodes, I like the simplicity of an integrated bridge and don't believe the sound would benefit enough to consider adding four of these.
Simply because they are rated for 50V and that the power supply outputs 70V. So by putting them up in series with resistors, I can be sure that each capacitor won't be exposed to more than 70V/2 = 35V and still get a decent 1/((1/68000)+(1/68000)) = 34000uF from those caps.
Any comment will (as always) be appreciated.
Thanks,
Sébastien
Good circuit.
Lot of relay contacts to handle
some "what if's".
Question.
RL1 turns on after soft start, then
RS5 goes from NC to NO which
turns on RL2 which opens the
contacts on rails ?
Lot of relay contacts to handle
some "what if's".
Question.
RL1 turns on after soft start, then
RS5 goes from NC to NO which
turns on RL2 which opens the
contacts on rails ?
Hi Sébastien,
my, you have been busy. Main bridges look fine. Haven't checked your loudspeaker protection cct. but looks similiar to others.
In your 15V power supply you don't need R8, remove it. Also R7 at 680R seems to be much too high. The loudspeaker protection cct. won't load it too much until the relay switches on. You haven't indicated the relay coil impedances on the schematic but if it's a relay designed to switch the main d.c. rails then it will be quite large and hence have a low impedance coil ( a few hundred ohms max. ) With a 680R resistor (R7) before the zener this will drop too much voltage under a relay coil load. You may want to recalculate this value.
Switching the main d.c. rails is one plan, however many designers use the relay to switch the output of the amplifier instead. Again there have been several threads debating the merits of introducing a relay contact into the signal path. For audio use a number of relays have been suggested, Jan Didden (I think) recommending Amplino relays for output protection duty.
Since you're collecting audio books may want to get Audio Power Amplifier Design Handbook by Douglas Self. It has much interesting material including a loudspeaker protection circuit.
Your offset detector may have problems with -ve offsets. You need to carefully consider the component values around the bridge and the transistor gains and current paths. I would breadboard or simulate this circuit independently and satisfy yourself that it works as you expect it to. Also be aware that opposite polarity faults on each channel will cancel and the speakers will thus be unprotected for that case (ok, a rare occurance.)
James
my, you have been busy. Main bridges look fine. Haven't checked your loudspeaker protection cct. but looks similiar to others.
In your 15V power supply you don't need R8, remove it. Also R7 at 680R seems to be much too high. The loudspeaker protection cct. won't load it too much until the relay switches on. You haven't indicated the relay coil impedances on the schematic but if it's a relay designed to switch the main d.c. rails then it will be quite large and hence have a low impedance coil ( a few hundred ohms max. ) With a 680R resistor (R7) before the zener this will drop too much voltage under a relay coil load. You may want to recalculate this value.
Switching the main d.c. rails is one plan, however many designers use the relay to switch the output of the amplifier instead. Again there have been several threads debating the merits of introducing a relay contact into the signal path. For audio use a number of relays have been suggested, Jan Didden (I think) recommending Amplino relays for output protection duty.
Since you're collecting audio books may want to get Audio Power Amplifier Design Handbook by Douglas Self. It has much interesting material including a loudspeaker protection circuit.
Your offset detector may have problems with -ve offsets. You need to carefully consider the component values around the bridge and the transistor gains and current paths. I would breadboard or simulate this circuit independently and satisfy yourself that it works as you expect it to. Also be aware that opposite polarity faults on each channel will cancel and the speakers will thus be unprotected for that case (ok, a rare occurance.)
James
Hello again.
Thanks thylantyr for the remark. It was actually a mistake, this switch actually belong to RLY2 instead of RLY1. This has the effect of keeping the amp off after a failure has been detected, forcing you to manually switch it off then back on again (after proper maintenance (or not if your more dary
))
The following modifications have also been implemented to the schematic:
- Moved the thermal switches so they shut down the amp, and the amp stays off until manually reset.
-Lowered the value of R7 to accomodate the 15V (instead of the original 55V)
-Removed R8
-Added the relay specs
The relay should not interfere since my auxiliary transformer is directly wired to the mains. The reason 680R was there, was for the previous transformer which had a 55V output instead of the current 15V. I think that 225r0 should be adequate leaving 1W of available power for the rest of the circuit.
It's in my "to read" pile. I just haven't started it yet. But at first glance, the circuit herein seems to be equivalent to those illustrated in the Self book.
And so far none of the thread has convinced me. I'd rather just shut the amp off and hope that in the time of the shutdown it won't do too much damage. I'm still not in favor of having a relay in the signal path.
I might sound dumb, but I'm not sure what you mean by -ve offsets. The ground should be common to all. The values herein were used in at least 5 loudspeaker protection circuit I was able to find. I'll breadboard it and see what it does.
Thanks again for all those excellent suggestions.
Sébastien
Thanks thylantyr for the remark. It was actually a mistake, this switch actually belong to RLY2 instead of RLY1. This has the effect of keeping the amp off after a failure has been detected, forcing you to manually switch it off then back on again (after proper maintenance (or not if your more dary
)) The following modifications have also been implemented to the schematic:
- Moved the thermal switches so they shut down the amp, and the amp stays off until manually reset.
-Lowered the value of R7 to accomodate the 15V (instead of the original 55V)
-Removed R8
-Added the relay specs
The relay should not interfere since my auxiliary transformer is directly wired to the mains. The reason 680R was there, was for the previous transformer which had a 55V output instead of the current 15V. I think that 225r0 should be adequate leaving 1W of available power for the rest of the circuit.
It's in my "to read" pile. I just haven't started it yet. But at first glance, the circuit herein seems to be equivalent to those illustrated in the Self book.
And so far none of the thread has convinced me. I'd rather just shut the amp off and hope that in the time of the shutdown it won't do too much damage. I'm still not in favor of having a relay in the signal path.
I might sound dumb, but I'm not sure what you mean by -ve offsets. The ground should be common to all. The values herein were used in at least 5 loudspeaker protection circuit I was able to find. I'll breadboard it and see what it does.
Thanks again for all those excellent suggestions.
Sébastien
- Status
- Not open for further replies.