Pre-amplifer for magnetic phono pick-up.

[bobhowe]

Hi David glad your turntable has a Shure Cartridge as it will provide a good warm sound & it you have any 78 s you can get a stylus for these records . If you have to build a separate power supply why don't you use a a valve rectifier ? & what choice would it be ? should we go over the top & use a 5U4G or similar or go the easy way & use a EZ80 or Ez81 should we have a poll ? & you will have to use one of your famous ADAPTER PLATES on this pre - amp regards Bob

[Kalee20]

I designed a valve phono amplifier, see attached diagram. It's a tad short on gain, but (to me) works surprisingly well.

The first stage is configured as a voltage-to-current converter, operating into the virtual earth of the second stage grid. So the signal voltage on the anode is peanuts. It's linearised by the unbypassed cathode resistor (so a DC heater would be nice for ultimate lowness in hum, and a floating heater line is a definite no-no!). The second stage is a transconductance current-to-voltage converter, with frequency-selective feedback around it.

[Murphyv310]

Hi David glad your turntable has a Shure Cartridge as it will provide a good warm sound & it you have any 78 s you can get a stylus for these records . If you have to build a separate power supply why don't you use a a valve rectifier ? & what choice would it be ? should we go over the top & use a 5U4G or similar or go the easy way & use a EZ80 or Ez81 should we have a poll ? & you will have to use one of your famous ADAPTER PLATES on this pre - amp regards Bob

Ha ha.
David will be in his element with an adaptor plate and a 5U4. Perhaps add an EF37A too

[Murphyv310]

I designed a valve phono amplifier, see attached diagram. It's a tad short on gain, but (to me) works surprisingly well.

The first stage is configured as a voltage-to-current converter, operating into the virtual earth of the second stage grid. So the signal voltage on the anode is peanuts. It's linearised by the unbypassed cathode resistor (so a DC heater would be nice for ultimate lowness in hum, and a floating heater line is a definite no-no!). The second stage is a transconductance current-to-voltage converter, with frequency-selective feedback around it.

Hi.
Could you explain the theory around both stagesc please.
In my understanding the first stage uses an bypassed cathode to introduce degenerative negative feedback, the values and type of triode also to me look like a standard voltage amplifier. I also don't see how the second stage is a virtual earth amp, put me right though if I'm missing something.

[Mark Hennessy]

The output from a valve is a current. The load resistance turns it into a voltage.

But a virtual earth has zero ohms input impedance. That's why you don't see a voltage on the inverting input of an op-amp wired that way. But while there is no voltage, there is current. The input resistor in an inverting op-amp amplifier simply converts the input voltage into a current.

In this case, the virtual earth is at C3. Obviously, as far as the signal is concerned it's in parallel with R4, but R4 is about DC conditions, not signal.

A lot of transistor amps work this way. It's also Tim de Paravicini's favourite topology for a phono stage, whether implemented with valves, transistors, op-amps, or a mix of them.

Another example of a virtual earth in a valve stage is attached:

   

[Kalee20]

Thanks Mark, saved an explanation, I've been offline a few hours!

[Murphyv310]

Hi Chaps.
I've encountered many circuits and applications using "virtual earth" mostly discreet semiconductor and op amps. I have also come across this type of circuit as well and have raised the query before (not here though) I also have replicated the circuits and built them. The thing that happens is they often dont behave in the way a virtual earth circuit should.
The coupling and circuit design is no different to me than umpteen other designs, all the difference I see is the lack of a first stage cathode bypass cap that I understand and the feedback network from anode to grid of the second stage for RIAA equalisation. There is no DC path from the RC network to either anode or grid of stage 2 . The input of stage 2, in fact both stages are high impedance and not at zero impedance, if they were zero impedance then this would damp the output of the first stage.
Surely to get a "current" amplifier we need to be thinking of grounded Grid or Base in a transistor amp?

Sorry but I still don't get it

[Mark Hennessy]

To save talking at crossed purposes, do you mind if we approach this in steps?

First of all, do you understand how an inverting op-amp amplifier circuit works?

If it helps, there's an explanation here (about a third of the way down): http://www.markhennessy.co.uk/articles/op-amps.htm

It's easier to start with this because the op-amp has oodles of gain, so the actual behaviour is closer to the theory.

The 2 key questions to answer from that are:

1. Are you happy you know why the inverting input is called a "virtual earth"?
2. Are you happy that this amplifier works with currents, not voltages?

[Murphyv310]

Absolutely Mark. I have no issues with "Virtual earth" theory and practise on OP amps.

[Mark Hennessy]

OK

So let's look at the second stage first of all.

And to keep it really simple, let's remove the feedback network (C4, C5, R8, R9) completely.

That leaves a conventional valve stage with a bypassed cathode resistor. So that stage will have a pretty high voltage gain. Not as much as an op-amp, but high...

But what happens when the feedback network is introduced?

In fact, let's just use a single resistor in place of those RC networks - just for understanding.

With that done, C3 (either end - this is signal) becomes a virtual earth. If you try to lift the voltage there, the anode voltage falls, and via the feedback resistor, it attempts to counteract this lifted voltage, hopefully pulling it back down to where it initially was.

Likewise, if you try to lower the voltage at C3, the anode voltage will rise, and the feedback resistor will try to counteract the fall.

So whatever signal you apply at C3, the negative feedback fights it. If the valve stage had infinite gain, it would win all the time (until the anode voltage clips). With a 'scope attached to C3, you'd see hardly any signal, if any.

Does all that make sense?