Really? What do you expect? 25W amplifier can be simple, 2000W version not.
I do think it would be fun to see who can comes up with the most simple 300 to 400 watt amp design.
Since 2000 watts
Would be a interesting approach. Availability of 300 volt transistor for audio , has options getting slim these days.
As mentioned highly unlikely 2000 watts is actually needed.
Even 400 watts, needs all the same protection circuits and would be a challenge with transistor availability
and far far more realistic.
A 4000 volt amp transformer isn't very small or easy to obtain for 2000 watts. likely be many many voltage taps
to make your life easy. 30 volt opamp , 12 volt relays, 24 volt fans the list goes on and on. lol
So playing around with a 800 VA to 1000 VA transformer and getting a working amplifier is still a complicated circuit.
to Barely touch 200 to 400 watts
I would keep your goals to around 300 to 400 watts.
And again depending on the actual application and speaker load.
Being all the same protection is needed. Soft start, cooling, dc protection, short protection, thermal protection etc etc.
Cant just " skip" those
Looking at the top top of the line speakers in live sound doing 1500 watts thermal rating.
The actual 10% distortion voltage is more like 500 to 600 watts.
A amplifier fault, will vaporize those speakers instantly.
So a complicated circuit should be expected, unless destroying expensive speaker loads is a hobby.
The speaker load will be 10x to 20x the cost of a single amp.
And very very old school Crow Bar protection had complete intentions of absolutely saving the speaker load.
The tradeoff would completely destroy the amplifier to make sure that happened.
With big capacitor banks and turn on surges, you dont want a "false trigger" kicking a crowbar off, or on in this case lol. Good luck
Since 2000 watts
Would be a interesting approach. Availability of 300 volt transistor for audio , has options getting slim these days.
As mentioned highly unlikely 2000 watts is actually needed.
Even 400 watts, needs all the same protection circuits and would be a challenge with transistor availability
and far far more realistic.
A 4000 volt amp transformer isn't very small or easy to obtain for 2000 watts. likely be many many voltage taps
to make your life easy. 30 volt opamp , 12 volt relays, 24 volt fans the list goes on and on. lol
So playing around with a 800 VA to 1000 VA transformer and getting a working amplifier is still a complicated circuit.
to Barely touch 200 to 400 watts
I would keep your goals to around 300 to 400 watts.
And again depending on the actual application and speaker load.
Being all the same protection is needed. Soft start, cooling, dc protection, short protection, thermal protection etc etc.
Cant just " skip" those
Looking at the top top of the line speakers in live sound doing 1500 watts thermal rating.
The actual 10% distortion voltage is more like 500 to 600 watts.
A amplifier fault, will vaporize those speakers instantly.
So a complicated circuit should be expected, unless destroying expensive speaker loads is a hobby.
The speaker load will be 10x to 20x the cost of a single amp.
And very very old school Crow Bar protection had complete intentions of absolutely saving the speaker load.
The tradeoff would completely destroy the amplifier to make sure that happened.
With big capacitor banks and turn on surges, you dont want a "false trigger" kicking a crowbar off, or on in this case lol. Good luck
Last edited:
If you bother to sift through it all the actual amplifier is quite simple. So then you add rail switches, which are not that complicated. And an array of protection systems, each of which is not that hard, but in aggregate makes pages and pages of schematic. Even the power supply is simple as SMPS goes, but unfortunately is NOT easy to design. A regular supply would be, just heavy heavy heavy.
Like I said before, you just need to already know how to build amplifiers before tackling one of these. Even true of 400 watts, but at least at that level you’re not dealing with bolting down two 30 pound transformers and doing 4-8 rail voltages (plus auxiliaries).
I have done making a 1000w rms amp which is giving 1060w rms at 2 ohm load before clipping . I am planning a startup and for that I need high quality and strong amplifiers , as I have made a 1000w rms amp with >1% distortion and soft start - cooling-dc protection - thermal protection . Now I want to make a 2kw at 4ohm ,,,, then I will go for 4kw or 5kw at 4ohm load (obvoiusly class H or Class D).
Big amp (200-400 watt ish like CS800) making, or chip amps and single 40V supplies with one pair of 2N3055’s and an output capacitor?
Why don’t you show us that design so the smart people can make suggestions on how to squeeze out more power? What class was that amplifier BTW?
Looks like a classic blameless with more power transistors. What supply voltages did you run it on?
One way to to it.
For more power, more voltage would be used.
As noted, the challenge is moving from to 180 volt transistors to 300 volt transistors.
So what is available with decent specs changes. Is the challenge
It is not being sold or marketed as a commercial unit. Just personal use.
So for personal use it is not being used or abused in un known manners.
As a commercial unit there would be numerous issue. No need to nit pick.
For personal use the lack of many protection circuits and clipping recovery is less a issue.
Just cranking up voltage by itself gets a little ridiculous. The current draw gets astronomical. And you have to do SOMETHING about second breakdown. Cascoding works but it has stability issues to work through and does nothing about heat and current draw. But with emitter follower outputs, it really IS just a matter of adding rail commutators to the existing circuit to go class H. That is, assuming a full complementary output which is no big deal these days. I do favor using op amp input stages as it eliminates the need for a 300 volt input stage, but a Blameless works just fine. Cascoding the input stage is another option, but may not offer the user as much protection as the op amp option if the input stage gets shorted. Eventually you need to go to EF3 outputs, but hell I do THAT starting at 40 watts because I think it sounds better, especially if under biased. You just run out of Hfe and want MORE with 40+ amp output currents. I prefer to run my VAS’s at 10 mA or less. Second breakdown starts kicking in at 20-30 mA at 300V on the types that get used there. The idle HAS to be lower to allow current swing.
There are 10,000 things to nit pick over, but you can start doing that at 10 watts. No need to go to high power for all that to kick in. As powers go up, the need for output and internal protection goes up because the result of accident or internal fault can be more deadly. Even a momentary VAS overload has to be addressed - not just to make to clipping waveform look pretty but to prevent blowing the MJE15034/5 pair and everything that follows it. On a 15 watt amp, you don’t even need DC protection. Sending full 20V rail voltage to a 1.5” VC on an 6 or 8 inch woofer won’t burn it up before you can get to the off switch. Even a 18SW115 won’t take 130VDC very long, and in a large system you may not even notice one woofer out in the array right away.
There are 10,000 things to nit pick over, but you can start doing that at 10 watts. No need to go to high power for all that to kick in. As powers go up, the need for output and internal protection goes up because the result of accident or internal fault can be more deadly. Even a momentary VAS overload has to be addressed - not just to make to clipping waveform look pretty but to prevent blowing the MJE15034/5 pair and everything that follows it. On a 15 watt amp, you don’t even need DC protection. Sending full 20V rail voltage to a 1.5” VC on an 6 or 8 inch woofer won’t burn it up before you can get to the off switch. Even a 18SW115 won’t take 130VDC very long, and in a large system you may not even notice one woofer out in the array right away.
Yeah, exactly! Just increasing the voltage isn't practical because the current draw becomes massive and second breakdown becomes a real issue
Would you happen to recall the manufacturer and part number for that heat sink? It looks like it could be useful in a cooling tunnel.
Tom