I would like some hints for a design that I have to start in the near future.
I am not asking that somebody else is designing for me, just to get some open minded ideas from competent people.
So the design constraints are : load is 2 Ohm at best, maybe al little worse.
Power supply is already there, 65-0-65 V. This should account for ab 1 kW r.m.s. on 2 Ohm. THD I am tolerant, 0.1% would suffice. Band pass is important from the extreme low (ideally DC) to 10kHz .
But I would like to concentrate on the output stage, at the moment.
So would MOSFETs bring any advantage? I know by sure that with 10 pairs BJT (TO3s) it can be done , but also I know that thermal runaways are a problem.
Would it be good to go for class H with a such low rails?
The input and VAS I would prefer IC, also do you have hints for a really rugged one, that can take care of some kind of SOA protection as well?.
Any constructive suggestion will be very appreciate.
Thanks in advance to all
effebi
I am not asking that somebody else is designing for me, just to get some open minded ideas from competent people.
So the design constraints are : load is 2 Ohm at best, maybe al little worse.
Power supply is already there, 65-0-65 V. This should account for ab 1 kW r.m.s. on 2 Ohm. THD I am tolerant, 0.1% would suffice. Band pass is important from the extreme low (ideally DC) to 10kHz .
But I would like to concentrate on the output stage, at the moment.
So would MOSFETs bring any advantage? I know by sure that with 10 pairs BJT (TO3s) it can be done , but also I know that thermal runaways are a problem.
Would it be good to go for class H with a such low rails?
The input and VAS I would prefer IC, also do you have hints for a really rugged one, that can take care of some kind of SOA protection as well?.
Any constructive suggestion will be very appreciate.
Thanks in advance to all
effebi
Last edited:
The view on BJT or MOSFET is split from what I have seen on here.
Personally I get on well with MOSFETs. So long as you use a Vbe multiplier they are usually fine. If your arent sure go for lateral MOSFETs as they arent so fussy about thermal runaway.
I dont usually bother with protection for a home amp but I do have a DC offset trap using a PIC micro on my disco amp. I suppose it depends what the chances are of shorting the output during normal use.
Personally I get on well with MOSFETs. So long as you use a Vbe multiplier they are usually fine. If your arent sure go for lateral MOSFETs as they arent so fussy about thermal runaway.
I dont usually bother with protection for a home amp but I do have a DC offset trap using a PIC micro on my disco amp. I suppose it depends what the chances are of shorting the output during normal use.
Superb ampli, I had seen already the thread , thank you.
But I really doubt that it can do 1000 W on 2R with only 5 pairs
, PA service.BTW do MOS show any real advantage on very low output impedences?
effebi
I have problems to understood your extra - short comments, sorry.
If you mean as a simple amp, you are calculating more than double the power you can obtain with such rails. If you mean bridged, you must consider that the load seen by a bridged ampli is half than the actual load. I do not belive that is good to go so low as under 1 Ohm, unless you have tried so already with your beautiful amp.
Anyway I am bound to stick with 65-0-65 rails.
thanks
effebi
This makes perfectly sense.
Did you show the schematic , BTW?
Is the driver circuit capable to handle the extra 5 pairs?
effebi
Last edited:
With that kind of power and current consumption, perhaps a class D amp working in class G operation to form an envelope to a linear amp would be the most efficient way to produce a low distortion arc welder, er, rather 1KW @ 2R.
It just seems like a lot of extra heat to waste in using a totally linear amp for 1KW output power.
It just seems like a lot of extra heat to waste in using a totally linear amp for 1KW output power.
I agree completely. Unfortunately I never did a class D stuff. I will study that, I hope it should not be too different from some PWM drive.With that kind of power and current consumption, perhaps a class D amp working in class G operation to form an envelope to a linear amp would be the most efficient way to produce a low distortion arc welder, er, rather 1KW @ 2R.
Do you really belive that it is possible to reach 0.1 % THD that way? At a first look a class H seems to me more reassuring.
Thank you.
effebi
CBS240 offering classG with classD as envelope and the amplifier is linear.
Like this:
http://www.diyaudio.com/forums/solid-state/179475-its-only-need-50v-45v-outputs.html
http://www.diyaudio.com/forums/soli...-hybrid-linear-switching-power-amplifier.html
Isn't it better to use 2x500W? or at least start with lower power first.
Like this:
http://www.diyaudio.com/forums/solid-state/179475-its-only-need-50v-45v-outputs.html
http://www.diyaudio.com/forums/soli...-hybrid-linear-switching-power-amplifier.html
Isn't it better to use 2x500W? or at least start with lower power first.
The links you posted are very interesting.CBS240 offering classG with classD as envelope and the amplifier is linear.
Like this:
http://www.diyaudio.com/forums/solid-state/179475-its-only-need-50v-45v-outputs.html
http://www.diyaudio.com/forums/soli...-hybrid-linear-switching-power-amplifier.html
Isn't it better to use 2x500W? or at least start with lower power first.
What you mean 2 x 500 W? bridge? Parallel?
A i said, I do need 1kW on 2R, 65-0-65 rails.
I am not afraid about hi-power, in other fields i was designing multi kW devices.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.