Thanks but that thread is somewhat...very long. Could you tell me which has actually been built and tested with a complex load? Thanks. And i don't need more than 50 watts. You know the hardest thing is to fit the diy amp in a proper enclosure. So i don't do that but prefer active speakers. There are restrictions to meet such as heatsink size transformer size etc.
No... I don't want to fight for that - I'm a very friendly guy
But this is only "true" for some simplified calculations. If you look at the accurate BJT model - like Ebers-Moll equation - BJT is a transconductance device, just like a JFET. The independent variable in the Ebers-Moll equation is Vbe. Not base current!There's simply no little demon in the base, counting electrons moving through the base, and dispensing β times that many electrons into the collector
No it is semiconductor physics. But that is truly academic. There is a standard textbook about this. I don't even know where to find it in my library. But a long time ago i designed very fast logic with bjt, ECL and such. I mean the chips. That makes direct use of the physics of semiconductors.
Hi Michael,
The principle is very simple yet efficient.
The first Vbe spreader, on the left (Q3, Q4) sets the quiescent current at no signal conditions - as usual.
The second Vbe spreader (Q5, Q6) - sets the minimal bias, making sure the collector current of the inactive shoulder comes close to zero, but never touches it, preventing the output transistors from closing completely.
Then, there is a diode clamp (D1-D4) - that's the mechanism, making this bias limiting happen. It is important to use rather fast devices in both the second spreader and the clamp to ensure smooth operation (no artifacts) at the higher end of the audio bandwidth.
The pictures below show the voltages at the emitter resistors of the output pair, equivalent to the collector current curves (left - full picture, right - close-to-zero area closeup). You can see, it comes close to zero, but never crosses it (in the normal push-pull EF it always does). You can also see, the limiting is rather soft - no sharp corners - which is also good to have.
I have received the PCBs and starting soldering one channel - we'll see it live soon
Although, Terry can do it quicker, as usual Cheers,
Valery
Attachments
Q3-Q6 are all placed on the main heatsink, under the board, in particular positions.
Attachments
Very interesting!I'm mid way through building a constant current supply for an anodizing tank. I want to make a bunch of chassis to put all these amps into before I end up buried in them like Terry.
Ok that requires some skills. I think "non switching" has no advantage over class AB at higher levels but could have audibly and measurably advantages at low levels. Very few magazines - and DIY - really test at 50mWatt - 150 mwatts. But it is worth.
Yes,amplifiers and preamplifiers everywhere.....no chassis no heatsinks......
This travel is long,too long and expensive,i believe that i must be stop because of the money.
The good news is that i have friends all over the world,they offer to me a lot of parts.
Thank to all of you
Last edited:
Yes,amplifiers and preamplifiers everywhere.....no chassis no heatsinks......
This travel is long,too long and expensive,i believe that i must be stop because of the money.
The good news is that i have friends all over the world,they offer to me a lot of parts.
Thank to all of you
We see the value of a good tester.