Here's a better quality copy.
Also the Oliver amplifier manual - they treated it as just another HP piece of equipment. A model of completeness.
People should built this amp!
Jan
you can explain to the client that his desire is contrary to the capabilities of his amplifier circuit, for a larger initial current of the output stages, it is necessary to increase the non-linearity of the Bias circuit, because of which there may be effects that from a "clean" class A, you will get a "dirty" class A at will client.
In the output stage, the triple to make a current per pair of more than 100-150 mA, with Bias control by feedback on the temperature gradient - this is nonsense.
This is done by a current mirror bridge so that the initial current is independent of the temperature of the device.
Tell your client: His amplifier is a design that has nothing to do with clean or musical sound. Whether class A/B or A or whatever;-)
I had high hopes that Oliver had designed a good amplifier, but he blew it with CR1 and CR2 across R5 and R6. But, hey, this was 1973. 😉
Ed
CR1 and CR2 handle higher currents. There is a reason why this approach is no longer used. 😉
Ed
Transistor bias stabiliser circuits were used even in 1968. but it did take a while for some to cotton on to this. Probably related to the prices of transistors vs diodes back then.
It’s really ok. At very low powers, before the diodes conduct, the signal path is through the 8.2 ohm resistors but that quickly transitions to the diodes and the 0.05 ohm resistors.
Papa Nelson did that in F5T too. I didn't hear any complains about it.
The diodes introduce two bends in the transfer curve, relying on negative feedback to straighten them out. The bends were unnecessary in the first place. The approach goes into the "bad" category in my criteria.
Ed
Well, we wouldn’t do that today, as is the case with many things from the 1970’s. However, the amp worked well by all accounts. I’ll just leave it there.
Back in 1966, I think, before Bailey published his 30W complementary design, he had published a Ge-based circuit using a transformer drive. To upgrade that to silicon transistors, he used the same trick to stabilise the quiescent curent as the higher Vbe was harder - as is now well known - to control using just resistors, which Ge devices seemed to get away with thanks to their low Vbe. I don't recall reading any reports of that design, so don't know if people who built it with Ge and compared it with Si transistors found crossover distortion to be worse.
However, with a few additional transistors it is quite possible to design a Vbe compensation circuit that allows silicon transistors to be driven from a transformer without using emitter diode, (as I have done) but in those days, transistors were more expensive than diodes.
However, with a few additional transistors it is quite possible to design a Vbe compensation circuit that allows silicon transistors to be driven from a transformer without using emitter diode, (as I have done) but in those days, transistors were more expensive than diodes.
Does anyone have the circuit, or remember, a 10 W per channel amp published in Practical Wireless or Practical Electrinics c 1970 using the AD161 and AD162 as output devices? I remember looking at that circuit as a kid and wishing I could build it 🙂
No, but I did build a circuit using AD161/162. The Mullard "Transistor Audio and Radio Circuits"book included one with a BC108 input, AC128 driver, and AD161/162 output. 15 ohm thermistor and 22 ohm parallel trimpot for bias, 0.5 ohm emitter resistors.
The circuit I built used a Si driver but I used an AC128 for a Vbe bias regulator, mounted on the heatsink, as I was never convinced about a thermistor/resistor combo.
The circuit I built used a Si driver but I used an AC128 for a Vbe bias regulator, mounted on the heatsink, as I was never convinced about a thermistor/resistor combo.
Well, I dunno, but if you write something and 50 years later they still use it and quote you - you've done something right.
You think people will still qoute you - or me - in 50 years' time ? 😉
My point was simply that progress is inevitable.
I do expect some of my work to be referenced in 50 years' time (in computer architecture).
Ed