This is a simple all Symmetric Class AB amplifier.
Distortion is low: THD 0.00045% at 1 Watt into 8 Ohm.
The VAS is a differential stage to increase the gain.
Output is simply EXICON lateral MOSFET.
Max like 15 Watt at +/-20 Volt supply.
Can of course use higher supply with higher output power.
Distortion is low: THD 0.00045% at 1 Watt into 8 Ohm.
The VAS is a differential stage to increase the gain.
Output is simply EXICON lateral MOSFET.
Max like 15 Watt at +/-20 Volt supply.
Can of course use higher supply with higher output power.
Last edited:
Your fully-complementary amplifier has fallen into the usual pitfall: any difference in offset voltage between the two differential pairs (U4-5 and U7-8) will cause the quiescent current in the second stage to vary. In your circuit, the gain is too high. This kind of amplifier needs a lot of emitter degeneration.
Ed
Ed
Or you can plugin a usual bias circuit instead of R8. Just as with usual bipolar output transistors.
Using matched pairs for the input transistors is a good idea, Not just here but in any amp.
I miss RFI protection on the input. I dont think zobel network on the output will make a difference with MOS output and so high idle current.
But i would like to see an idea from you with less expensive transistors. Now i just look and think - to expensive to make a try.
Using matched pairs for the input transistors is a good idea, Not just here but in any amp.
I miss RFI protection on the input. I dont think zobel network on the output will make a difference with MOS output and so high idle current.
But i would like to see an idea from you with less expensive transistors. Now i just look and think - to expensive to make a try.
Yes, small emitter resistances (say, 22R) would greatly reduce the relative difference between the transistors in the LTPs.
Although in my own experiments I have found the KSC1845s and the KSA992s to be remarkably consistent, especially if you take several devices in a row from the same strip. They are now my favourite small signal complementary transistors because of that (and also because of high beta, high Vce rating and very low Cob)
Although in my own experiments I have found the KSC1845s and the KSA992s to be remarkably consistent, especially if you take several devices in a row from the same strip. They are now my favourite small signal complementary transistors because of that (and also because of high beta, high Vce rating and very low Cob)
Like this, on all LTP's. Question remains how big R's to use on first and second pairs.
However I see no point in mirroring - doubling stages. This is proven Hitachi diagram, used and cloned to humongous proportions, always delivers with so little parts. Input and VAS transistors are not possible to find any more, but Im sure your KSA-KSC devices will do superb too.
Only worry here, like in your design, is if wiper on bias pot gets open: it will blow outputs. Best pot or at least some diodes to ensure bias cant go over some limit.
Hehe. Its a bit like bacteria, you know they are there, but you can’t see them. And you don’t like it!
R
R
Since the wiper and the unused end of the pot are tied together, I would assume that if the pot is an appropriate value, it would merely lead to increased dissipation and not an outright failure.
Last edited:
In my eyes the the Hitachi schematic is perfectly symmetric. Input symmetric.
Vas has a differential pair - symmetric followed by a current mirror - symmetric.
It drives the output transistors symmetric from positive and negative rail.
What more can you wish from symmetry?
Vas has a differential pair - symmetric followed by a current mirror - symmetric.
It drives the output transistors symmetric from positive and negative rail.
What more can you wish from symmetry?
The input differential pairs need emitter degeneration. I would get rid of the second stage differential pairs. I have never seen that with this type of amplifier.
The fully-complementary amplifier has some advantages but high gain is not one of them.
Ed
You have to realize that lineup doesn’t build anything. He just design, so he is prone to cough up any kind of designs. Some gets built, some does not.
This is his vocation, I say we let him play.
We are all different and that is how it should be.
R
This is his vocation, I say we let him play.
We are all different and that is how it should be.
R