John Curls JC Parasound I think puts out about 90V peak . . . see his earlier post. If you want high power and not bridged, you need high rail voltages. My original 250W power amp was +-75V (measured in Japan with 100VAC line voltage) but in Taiwan I measure +-85V - the mains regulation is not as tight.
With these big amps, you need many OP devices . . .
With these big amps, you need many OP devices . . .
I am do not assume that high power amp automatically has to give up resolution compared to a low power amplifier, but to achieve same quality in low level is much more difficult. Construction and lay out will be critical to achieve low noise and low distortion.
Usually, high power amp is using for PA when you listening in crowded of people. You can not notice noisy amp 😀. Any way you can not hear low level signal in this condition.
There are two separate issues. Design and layout is critical with respect to noise and unwanted buzz/hum, but there is absolutely no reason why a high power amplifier should sound worse at low level and why it should have higher distortion at low level. In fact, opposite is true in a well designed power amplifier. It is also possible to get very low noise, it depends mostly on design of input stage.
Bimo,
I admit that many high powered amplifiers have been designed for PA applications and the requirements there are different than what we want in a home amplifier. But there have also been other power amps built for home use that have reached into the 300 to 4oo watt range, not often but there are examples out there. Sometimes it has taken running a pair of amplifier in bridged mode, I have done that in the past with some Parasound amplifiers but it is many times a questionable proposition to do that. I have taken out an entire bank of output devices in one of my Parasound amps doing that into a questionable load impedance. Between Bonsai and Ostripper it looks like we could have some very nice amplifiers that could work well without running in bridged mode, not saying that can't be done well, but that is usually a poorly done implementation. A well designed push pull bridged amplifier would be a nice addition to these high powered amplifiers.
I admit that many high powered amplifiers have been designed for PA applications and the requirements there are different than what we want in a home amplifier. But there have also been other power amps built for home use that have reached into the 300 to 4oo watt range, not often but there are examples out there. Sometimes it has taken running a pair of amplifier in bridged mode, I have done that in the past with some Parasound amplifiers but it is many times a questionable proposition to do that. I have taken out an entire bank of output devices in one of my Parasound amps doing that into a questionable load impedance. Between Bonsai and Ostripper it looks like we could have some very nice amplifiers that could work well without running in bridged mode, not saying that can't be done well, but that is usually a poorly done implementation. A well designed push pull bridged amplifier would be a nice addition to these high powered amplifiers.
With big amps, you can run the first 10-20W in class A (you need a very good heatsink of course) and can therefore get a lot of the class A benefits - clean high register sound, plenty of bass power etc. Assume 10-12 pairs of output devices, each running at 130-150mA - that will give ~20W of class A.
10 - 12 pairs/130mA ..... almost 1.5A idle current with 2 x 70V supply rails it would dissipate 210W of heat per channel at idle. This would be very difficult task for any heatsink.
And it would need a good driver, a MOSFET preferably. It would make very nice amplifier, if some of you guys will be able to build this I would applaud him 🙂
And it would need a good driver, a MOSFET preferably. It would make very nice amplifier, if some of you guys will be able to build this I would applaud him 🙂
JC uses mosfet drivers afaik, so did Accuphase (Madrigal) in their high end model, can't recall model number now...
Yes he does, I know which parts. This is not a problem. But the dissipated power IS.
I would decrease the idle current per pair, to get something like 15mV across Re (0.22 ohm). Then we have 68mA per pair and 680 - 816 mA of idle current. Something like 100W dissipated power per channel is still enough but possible to cool with a large conventional fin shaped heatsink.
I would decrease the idle current per pair, to get something like 15mV across Re (0.22 ohm). Then we have 68mA per pair and 680 - 816 mA of idle current. Something like 100W dissipated power per channel is still enough but possible to cool with a large conventional fin shaped heatsink.
I'll stick with the JC-1 heatsinks and almost 30W of Class A power, and 500W of Class A-B power, and 0.15 ohm emitter resistors. Why not? It's doable.
Yes - I agree is doable.
I think your JC amp though has very substantial heatsinks. The big modushop 5U housing has good heatsinks and they run about 10 deg C above ambient with about 75-80W dissipation (I'll need to measure it this week end to be sure).
To PMA's point, the 200-210W I am proposing may be on the high side - only way around it is a bigger set of sinks - I dont like fans, so this is not my preferred approach.
Sims indicate 10ppm at 420W 8 Ohms 20 kHz and with AFEC about 3 ppm (~15ppm without AFEC). SR 200V/us with FE filter - so meeting the 1V us per pk OP voltage is not a problem.
Concept circuit will be posted up tomorrow.
OK, Pie in the sky specification time! How about a all Class A push pull (input to output), balanced input, balanced output, with water cooled or refrigerated heatsinks (a la Seymour Cray). I believe the loudspeaker is the weakest link. We need more efficient loudspeakers, and yes bi and tri amplification greatly helps. GRH
These CFA's (and my beloved VFA's) are already class A from input to
the final driver stage.
I've "plugged in" these IPS's to a class A OPS .... sub-ppm to 2ppm full
output is all thats left.
A/B X-over distortion is the main component of the typical 10-40ppm that is usually encountered.
PS - run the "NX" at first watt class A or into my big genesis OPS , 1ppm
is the rule. (below)
OS
Attachments
😎With Bonsai's history of demonstrated excellence, we are in store for a serious HIGH powered amplifier. Some Dunlap Clarke like transformers will be needed of course. And maybe a computer control of things here and there, like cooling and display.😀 
I understand that cooling fans can be somewhat noisy but aren't there some very quite fans that could be used that would take care of the heat problems, perhaps only going on when a set point temperature is reached? If you are driving a speaker system with hundreds of watts of power who could possibly hear the fans at that level? I guess at this level is where the class D amplifiers will come into play when they are finally optimized and the sonic qualities are brought up to the level of what can be done with a discrete design.
He he. I don't know about the Bonsai excellence bit, but the high power bit we can definitely do.
Peter baxandall devised this test to find out the extend of the voltage and current limits of an amp.
In fact he had a nice piece of equipment that let you show the power envelope on a scope.
He used 50Hz sine wave and 2kHz impulses. The sine wave exercised the output voltage and the impulses exercised the current capability at the specific output voltage.
This was the test which with he discovered a flaw in the 405 protection system, which lead to the 405-2 with a redesigned protection circuit.
I built this thing and it works as advertised.
Jan
I'd like to see a nicely integrated bal/bal circuit as well. Brute force power -- al la scaled up from low power circuits is OK. However, lets also think some about this.... here's what i do with VFA, just to give some ammo to all the CFA idea creators here.
View attachment Bal-Bal.pdf
THx-RNMarsh
Last edited:
