Oxford Crystal Reference restoration

[peter]

I still can't get over the fact that many SS preamp designs use overkill of components. I know that Primare have always made good gear, but the complexity of the circuit and the sheer quantity of components is frightening for such a device. I'm sure there are better and more capable transistors available today, let alone IC op-amps?

I am not really up to date and have been looking, but is it not true that the NE5532 and its derivatives are still the OPAMP of choice? That was a 1970s design.

Electronics nowadays is geared to mobile phones, unfortunately.

[peter]

By way of a stutus update of sorts Here is the PCB from my Conrad Johnson Premier 4. This is not short on parts for a tube amp either.



Power supply for input tubes is regulated in these amps, checkout all these diodes that are wired in series. This keeps the semiconducters that do the regulating close to the HT rail and with only modest voltages across them.

The opamp at the top is one of several comparators used for turning on an LED to help bias the output tubes.



After some work I managed to get this going using half of the output tubes while I wait for parts to arrive.



Getting 6CA7s for this is troublesome/pricey, and I have seen a lot of negative things written about this amplifier on the internet. The UK mag review(s) when this came out were IMHO not very fair to it. Hope to redress this a little.

But even with half the valves this amp sounds fabulous.

[Reid Malenfant]

I am not really up to date and have been looking, but is it not true that the NE5532 and its derivatives are still the OPAMP of choice? That was a 1970s design.

Electronics nowadays is geared to mobile phones, unfortunately.

It depends on the impedance of the circuit feeding the op amp in reality

With a fairly low impedance circuit then yes the NE5532 (or single version NE5534) are still two of the top op amps out there, definately for the price

With much lower & conversely higher circuit impedances then others will edge out the above mentioned op amps. However, the cost of some of them takes the mickey

I still can't get over the fact that many SS preamp designs use overkill of components. I know that Primare have always made good gear, but the complexity of the circuit and the sheer quantity of components is frightening for such a device. I'm sure there are better and more capable transistors available today, let alone IC op-amps?

I think you'll find that in reality a discrete pre amp uses vastly less components than one containing even primitive ICs if you take into account the component count of the IC itself. Yes ICs can be good as has been mentioned, however they'll only work in class A up to a few hundred microamps & thermal effects make themselves known over the whole IC where as a discrete stage can be a lot more thermally stable. This may not sound that important but do you really want the output stage altering the input differentials bias & thus noise due to it warming it up with regards to an IC

I think you'll find that this is why some manufacturers stick to discrete components as there is a lot more control over everything going on.

[Reid Malenfant]

Power supply for input tubes is regulated in these amps, checkout all these diodes that are wired in series.

That looks like a hell of a lot of zener diodes in series to me, i may be wrong though. Zeners are noisy things & a load in series doesn't bare thinking about

But even with half the valves this amp sounds fabulous.

It looks very pretty to

[peter]

That looks like a hell of a lot of zener diodes in series to me, i may be wrong though. Zeners are noisy things & a load in series doesn't bare thinking about

The spec for this amp is -96dB hum and noise at full output (100W RMS min to 4 Ohms).

Am I allowed to post schematics on here?

[Reid Malenfant]

The spec for this amp is -96dB hum and noise at full output (100W RMS min to 4 Ohms).

Am I allowed to post schematics on here?

So that would be -76Db @ 1W RMS Frankly i don't know how that measures up against other valve amps.

I'm sure you are allowed to post schematics, only as long as you aren't infringing copyright Try hunting on the net for the schematic & then link us to it Peter, that way causes no problems at all. I had to do similar with my phono pre amp, i could only find the MC section though

[peter]

So that would be -76Db @ 1W RMS Frankly i don't know how that measures up against other valve amps.

Ok, well in the meanwhile, the string of 13 avalanche diodes (32V) has a current source at the top, (by my calculation about 2mA). The top (and top-1) of the string is connected to the input of a discrete darlington voltage follower stage that has 1.0 + 1.0 + 0.15 uF in parallel across the output (one of these for the each voltage used, 410V and 380V ).

Perhaps "regulated" is not entirely accurate description.

The -50V bias supply uses a similar arrangement, except just a resistor not a source.

The 500V plate supplies share a smoothed supply but have their own filter arrangements. The input tube "regulators" are fed from the filtered plate supply.

The output tube heaters are connected to a tap on the mains Transformer. The input tube heaters are fed from a smoothed 12V output.

[Reid Malenfant]

Ok, well in the meanwhile, the string of 13 avalanche diodes (32V) has a current source at the top, (by my calculation about 2mA). The top (and top-1) of the string is connected to the input of a discrete darlington voltage follower stage that has 1.0 + 1.0 + 0.15 uF in parallel across the output (one of these for the each voltage used, 410V and 380V ).

Perhaps "regulated" is not entirely accurate description.

Well it's more 'regulated' than some even if it's not perfect Tbh i don't know a lot about valve amplification, though i do know that what you are describing can be improved on if you so wish. The zeners could be replaced by a TL431 & a high voltage power mosfet & a few resistors & should result in a less noisy regulator.

First up though get the whole amp working as it should, from there on you can modify it if you want to

[peter]

Well, having the electron emitter so close to a 50Hz powered thermal noise source seems a bad plan to me, but I am no expert in tube amp design either.

This power configuration does appeal me though, as effectively it is just a load of impedance in between the PSU and the amplifier. I dimly recall such trickery is what gives the long tail pair good PSRR.

When I get around to it I will measure the noise though, and I thank you for pointing it out as I hadn't really considered it before because, honestly, it is seriously quiet, (The Primare is a howling gale by comparison and that is supposed to be low noise), and lets not even think about the vinyl.

As a point of general interest, I don't know if its commonly known that the noisiest of zeners happen to be the ones with best temperature stability (ie 5.6V) and the more you move up in voltage the less noisy (and faster) they become.

So what is the consensus on noise quality? IOW if you have to have noise, what is the most euphonic kind? (ie thermal vs shot vs 1/f etc).

I am checking out the TL431 as we speak.

[peter]

I think you'll find that this is why some manufacturers stick to discrete components as there is a lot more control over everything going on.

In the case of the Primare, it is as "complemetary" as possible, so both positive and negative swings are dealt with in the same fasion. This isn't the case in most OPamp designs (something to do with "good" PNPs being harder to fabricate or some such not helping).

Whether complementary purity is still a fasionable idea, I don't know.