Moving Coil Pickups for the Technically Curious

[Joe Gwinn]

I think we are getting a bit far from one important question: Is the SNR better with common base or common emitter. I say they are the same, you said above: "Impedance match is better with the common base configuration, so the resulting SNR is better, even though you are right about the inherent noise of the transistor."

I am not sure if you have changed your thinking on that, so let's review the situation. Start with a transistor stage with a resistor Rc in the collector, biased properly with both the base and emitter at ac ground. This circuit has some noise output. If we put an ac voltage source in series with either the base or emitter, the magnitude of the signal output is the same in the two cases. This situation does not change significantly if a low value resistor is inserted in series with the voltage source. Low means small compared to any resistor in the model of the transistor such as the input impedance looking into the emitter Clearly the SNR is the same in both cases. This is the case that applies to the situation under discussion, and so either type of stage can be used with no significance difference.

Now, consider the question of whether "the best impedance is zero". In demonstrating this, I used a model and some very simple analysis. To refute this, it is sufficient to show that either the model or the analysis is wrong. I believe that you have stated that the model is wrong; that is, that a transistor is too complicated to be represented in this way. Well, this is the model used by the maker of many commercial chips to describe the performance of their product, op amps and so on. If it is wrong, then the noise analysis engineers do is wrong.

Well, if one takes this argument and model at face value, nobody would ever use a common-base or common-collector amplifier, and yet these are fairly common. Common-base amplifiers are most often seen as low-noise RF amplifiers, although there are reasons other than noise to do this. . So, one must either question all those circuit designers or question the adequacy of the model, which may well be oversimplified.

A little quality time with some transistors will clarify the issue.

[Steve Conner]

The thing about "matching" resistors is debunked in The Art Of Electronics, so if you don't believe it, you need to convince Horowitz and Hill.

I also say that the SNR is the same whether you use common-base or common-emitter.

Common-base is used for RF circuits because it gives less feedback capacitance between the input and output of the amplifier. The base is grounded, so there's no Miller effect. The result is an amp that's more stable and easier to neutralise.

Common-collector is just another name for the emitter follower, which is almost never used at RF because it's very difficult to neutralise.

Every MC head amp I've seen has been common-emitter. Douglas Self has some good articles on them in "Self On Audio", and shows designs that push the theoretical noise floor for a 10 ohm source impedance.

You can also see some of the concepts we're discussing in the datasheet for the MAT03 low-noise dual transistor:

http://www.analog.com/static/importe...eets/MAT03.pdf

If the "Super Low Noise Amplifier" circuit shown in Figure 3a doesn't work for the moving-coil pickup, I don't know what will. Unfortunately MAT03s cost several dollars each.

Of course all of this collector current will hardly help battery life in our proposed active instrument.

[Mike Sulzer]

Here are a couple of legitimate responses to the first half of what I wrote:
1. I see; in order to show that the CB and CE have different noise performances, I would have to show that a very small resistor, when moved between the base and emitter parts of the circuit, would make a significant difference in the noise performance.

2. You logic is wrong, and here is why.....

Your response attempts to muddy-up the issue by brining in irrelevant information and mixing up the first and second half of what I wrote. Some particulars:
1. Common base has application at rf; what I wrote has nothing to do with that application.
2. The common collector (emitter follower) circuit is used for high input impedance/low output impedance. It has nothing to do with this discussion.

3. "Common-base amplifiers are most often seen as low-noise RF amplifiers..." As Steve pointed out, the common base is used for good stable high frequency response. Sure, the noise is good, just as good as common emitter.

Your first sentence is thus nonsense. Your second sentence slants the truth. So your third sentence does not follow.

When I get involved in a technical discussion, I expect better than this.

Well, if one takes this argument and model at face value, nobody would ever use a common-base or common-collector amplifier, and yet these are fairly common. Common-base amplifiers are most often seen as low-noise RF amplifiers, although there are reasons other than noise to do this. . So, one must either question all those circuit designers or question the adequacy of the model, which may well be oversimplified.

A little quality time with some transistors will clarify the issue.

[Mike Sulzer]

Douglas Self has some good articles on them in "Self On Audio", and shows designs that push the theoretical noise floor for a 10 ohm source impedance.

.....................

If the "Super Low Noise Amplifier" circuit shown in Figure 3a doesn't work for the moving-coil pickup, I don't know what will. Unfortunately MAT03s cost several dollars each.

Thanks for the reference to Self. I read the chapter on head amps (1987 EW), and it is good. I found it on line for free here: Self on Audio - Google Books

On the MAT03, figure 3a in your reference: that is a differential input amp. That should not be needed for a moving coil cartridge. I believe that the differential input degrades the noise performance. Am I wrong?

[Joe Gwinn]

The thing about "matching" resistors is debunked in The Art Of Electronics, so if you don't believe it, you need to convince Horowitz and Hill.

I assume you mean the statement on page 435 of the 2nd edition.

There is an unspoken assumption here, and in most treatments, that flicker noise may be neglected. This may or may not be true at low audio frequencies. Flicker noise can easily overwhelm the Johnson noise of low-value resistors. That said, many of the proposed amplifiers have corner frequencies of 10 Hz, so there is hope. Only a direct test will really settle the issue.

I also say that the SNR is the same whether you use common-base or common-emitter.

I said something related to Mike Sulzer, so I don't understand the point being made here.

Hmm. SNR and noise are not exactly the same thing. More precisely, the internal noise generators of a transistor are the same no matter what kind of circuit the transistor is connected to. But it does not follow that the SNR is unaffected by how the device is used. Both the signal and the noise parts of the equation are affected.

The issue is how best to match to such a small source impedance, at low audio frequencies.

Common-base is used for RF circuits because it gives less feedback capacitance between the input and output of the amplifier. The base is grounded, so there's no Miller effect. The result is an amp that's more stable and easier to neutralise.

That's one reason for sure. Acutally, with transistors packaged for RF use, one may be able to entirely dispense with neutralization.

Another big reason to use common-base amplifiers is to match low-impedance sources.

Common-collector is just another name for the emitter follower, which is almost never used at RF because it's very difficult to neutralise.

This is true, but I didn't suggest use of emitter followers at RF, so I'm not sure what point is being made here.

Every MC head amp I've seen has been common-emitter. Douglas Self has some good articles on them in "Self On Audio", and shows designs that push the theoretical noise floor for a 10 ohm source impedance.

The common-emitter head is intended to drive a long cable without undue loss of high frequencies in cable capacitance.

Self is very good at such things, but has he expressed any opinion on how best to handle a 0.001-ohm source? This would be interesting.

You can also see some of the concepts we're discussing in the datasheet for the MAT03 low-noise dual transistor:

http://www.analog.com/static/importe...eets/MAT03.pdf

If the "Super Low Noise Amplifier" circuit shown in Figure 3a doesn't work for the moving-coil pickup, I don't know what will. Unfortunately MAT03s cost several dollars each.

It's a very good circuit. There are other makers of very low noise audio transistors. If one does not need matched devices, costs are pretty reasonable. In fact, ordinary jelly-bean transistors can be pretty good these days, as transistor manufacture is such a mature art.

Of course all of this collector current will hardly help battery life in our proposed active instrument.

It does not appear that battery power will be practical for such a circuit. This alone may force use of a transformer.


I've also noticed in an analysis I just read equations that imply that a common-gate JFET might be better at near-zero source impedance. (This analysis also neglects flicker noise, but at least admits to it.) I will chase down the cited articles and see if they really say that JFETs would be better - one does not normally think of JFET and low-impedance at the same time.

[Joe Gwinn]

On the MAT03, figure 3a: that is a differential input amp. That should not be needed for a moving coil cartridge. I believe that the differential input degrades the noise performance. Am I wrong?

You are correct. Having two active devices doubles the noise power, and therefore increases the noise voltage or current as appropriate by Sqrt[2].

[Joe Gwinn]

Here are a couple of legitimate responses to the first half of what I wrote:
1. I see; in order to show that the CB and CE have different noise performances, I would have to show that a very small resistor, when moved between the base and emitter parts of the circuit, would make a significant difference in the noise performance.

2. You logic is wrong, and here is why.....

Your response attempts to muddy-up the issue by bringing in irrelevant information and mixing up the first and second half of what I wrote. Some particulars:
1. Common base has application at rf; what I wrote has nothing to do with that application.
2. The common collector (emitter follower) circuit is used for high input impedance/low output impedance. It has nothing to do with this discussion.

3. "Common-base amplifiers are most often seen as low-noise RF amplifiers..." As Steve pointed out, the common base is used for good stable high frequency response. Sure, the noise is good, just as good as common emitter.

Please see my response to Steve Connor.

Your first sentence is thus nonsense. Your second sentence slants the truth. So your third sentence does not follow.

When I get involved in a technical discussion, I expect better than this.

Ad hominem. It's OK to disagree, but it's best to skip the namecalling.

[Mike Sulzer]

It depends on how you use the devices. Paralleling devices improves SNR when noise voltage rather than noise current is the issue. But I believe it is worse in the case of the diff. amp.

You are correct. Having two active devices doubles the noise power, and therefore increases the noise voltage or current as appropriate by Sqrt[2].

[Mike Sulzer]

Please see my response to Steve Connor.

Ad hominem. It's OK to disagree, but it's best to skip the namecalling.

Unbelievable. I show that you are using nasty rhetorical tricks to try to make your case, and your response is to refer me to the response you made to Steve. Yet another nasty trick.

Then you turn it back on me!

ad hominem |ˈad ˈhämənəm|
adverb & adjective
1 (of an argument or reaction) arising from or appealing to the emotions and not reason or logic.
• attacking an opponent’s motives or character rather than the policy or position they maintain : vicious ad hominem attacks.

You think I attacked you or your motives? You think I am the one avoiding logic? Not at all. It is how you go about making your case that is the problem. Refute what I wrote, if you can, by directly dealing with it.

[Steve Conner]

Hey, be nice to each other!

The circuit I referred to has differential inputs, yes. It's a super-low-noise op-amp, and requires a feedback network around it the same as a regular op-amp before it'll do anything.

So you just use it as you would an op-amp, only with very low value resistors in the feedback network, so the Johnson noise of the resistors won't degrade the SNR.

Douglas Self did something bizarre that might be relevant. All of his audio power amp designs ended up optimized for a source impedance of 50 ohms, because he wanted the lowest measured noise performance, and that was the output impedance of his Audio Precision test set.

As far as I know, JFETs are better for high source impedances, bipolars are better for low source impedances, but there are low-noise dual JFETs now available that give bipolars a good run for their money.

Nobody actually knows where 1/f ("flicker") noise comes from. But as far as I'm concerned, it's caused by manufacturing imperfections and therefore an issue for the device manufacturer to worry about. If he sells it as a low-noise device for audio, I'd expect to see a 1/f corner frequency below 20Hz in any conceivable circuit.

[Mike Sulzer]

The circuit I referred to has differential inputs, yes. It's a super-low-noise op-amp, and requires a feedback network around it the same as a regular op-amp before it'll do anything.

Exactly. But my concern is that the noise voltages from both sides of the differential amp are present. I think Self's circuit with this device would make the quietest preamp. One would use two differential devices all in parallel; that is, a total of four devices in parallel. The fourth does not make much of an improvement, but you have it; so you might as well use it.

[Joe Gwinn]

The circuit I referred to has differential inputs, yes. It's a super-low-noise op-amp, and requires a feedback network around it the same as a regular op-amp before it'll do anything.

So you just use it as you would an op-amp, only with very low value resistors in the feedback network, so the Johnson noise of the resistors won't degrade the SNR.

The question was if the use of two active devices in a differential pair would increase noise, and the answer is that it will. Use in a differential pair is not the same as electrically paralleling devices, but still there will be a price to be paid somewhere.

Douglas Self did something bizarre that might be relevant. All of his audio power amp designs ended up optimized for a source impedance of 50 ohms, because he wanted the lowest measured noise performance, and that was the output impedance of his Audio Precision test set.

I recall that, and while I understood why he did it, it still seemed a bit like the drunk looking for his car keys under the streetlight, because that's where the light is.

As far as I know, JFETs are better for high source impedances, bipolars are better for low source impedances, but there are low-noise dual JFETs now available that give bipolars a good run for their money.

The summary rule that bipolar for low impedance and JFET for high impedance assumes common-emitter or common-source respectively. The question is if use of common-base and common-gate respectively changes this rule of thumb.

Nobody actually knows where 1/f ("flicker") noise comes from. But as far as I'm concerned, it's caused by manufacturing imperfections and therefore an issue for the device manufacturer to worry about. If he sells it as a low-noise device for audio, I'd expect to see a 1/f corner frequency below 20Hz in any conceivable circuit.

I think most device-physics people agree that flicker noise is caused by manufacturing defects, especially those at a surface, and/or insufficiently pure materials. For manufacturers selling devices intended for low-noise audio operation, I would agree that they will have to have a pretty low corner frequency, with 20 Hz being the highest allowed. But, oddly, this data is very often omitted from datasheets.


War story: I'm involved with some time-obsessed radio amateurs in the design of a Dual Mixer Time Difference instrument, which can measure frequency differences in the parts per trillion and path length differences in the picoseconds. It's a long and arcane story, but the key is the generation of a 1 Hz beatnote and the precise time measurement of its zero crossings. Flicker noise is a big issue. And protecting against 1 Hz leakage is very difficult, as ordinary shielding is almost totally ineffective at so low a frequency. At least it's a 50-ohm system. Anyway, this effort is what led me to the audio opamps made by That Corp and competitors.

[Mike Sulzer]

The summary rule that bipolar for low impedance and JFET for high impedance assumes common-emitter or common-source respectively. The question is if use of common-base and common-gate respectively changes this rule of thumb.

Do you have a reference to support that first sentence? My understanding is that it is a function of the physics of the devices, not how they are used.

Again, for the question considered in your second sentence to be answered as "it changes", the commonly used model using a voltage and a current noise source must be seriously deficient.

[Joe Gwinn]

Do you have a reference to support that first sentence? ("The summary rule that bipolar for low impedance and JFET for high impedance assumes common-emitter or common-source respectively.") My understanding is that it [the input impedance] is a function of the physics of the devices, not how they are used.

It's a common rule of thumb, not a scientific law, and need not be perfect to be useful. I have seen it in one form or another in many publications over the years. Perhaps it's in Horowitz and Hill as well.

Consider the lowly bipolar transistor. In the Common-base configuration, the circuit input impedance is in the tens of ohms, while in the common-collector (emitter follower) configuration, it can be tens of millions of ohms. In the common-emitter configuration, the circuit input impedance is hundreds to thousands of ohms. Yet the transistor is the same in all cases. What differs is the topology of the circuit surrounding that transistor.

Again, for the question considered in your second sentence to be answered as "it changes", the commonly used model using a voltage and a current noise source must be seriously deficient.

The context of the second sentence was circuit impedance, not noise.

But in any case, those datasheet plots showing how noise level and/or noise figure varies with source impedance certainly imply that how a transistor is used in a circuit very much affects how much noise is generated by that circuit (versus the device within). If the circuit made no difference, such plots would not be provided, as they would be pointless.

[Mike Sulzer]

It's a common rule of thumb, not a scientific law, and need not be perfect to be useful. I have seen it in one form or another in many publications over the years. Perhaps it's in Horowitz and Hill as well.

Oh, I was not questioning the rule; I have used it for almost four decades. When I read this sentence: "The summary rule that bipolar for low impedance and JFET for high impedance assumes common-emitter or common-source respectively.", I assumed that you meant that the rule is only intended for common-emitter or common-source topologies. I am still having trouble reading it any other way.

In any case, the rule also works for a BJT in common-base and a FET in common-gate. When you ground the base and connect a low source impedance to the emitter, the noise current source is shunted and has little if any effect. Thus, one is left with the noise voltage source. This is less than the noise voltage of the FET in general, as the rule implies. And so in situations where the rule applies (maybe not so much anymore as FETs get better) the transistor is better for this low impedance application as well as for common-emitter vs. common-source.

The context of the second sentence was circuit impedance, not noise.

The sentence is this:
" The question is if use of common-base and common-gate respectively changes this rule of thumb."

Given the first sentence, I must interpret the second one as talking about noise.

But in any case, those datasheet plots showing how noise level and/or noise figure varies with source impedance certainly imply that how a transistor is used in a circuit very much affects how much noise is generated by that circuit (versus the device within). If the circuit made no difference, such plots would not be provided, as they would be pointless.

Yes, the device and the rest of the circuit determine how much noise results. The key thing for a BJT is that there is current flowing into the base of the transistor due to its physics; it has noise fluctuations as any such current must. (The noise behavior of this current is represented by the noise current source in the common model I have mentioned before.) When one increases the source impedance in the common-emitter mode, a bigger noise voltage develops across the base and emitter. That is, when the noise source that is part of the device is shunted by a larger external resistor, a larger noise voltage develops between the base and the emitter.

All this is correctly predicted by the commonly used model which has a voltage and current noise sources. I think what you write now agrees with this model, while what you wrote above contradicted it. That is, the model predicts that with low source impedance, common-base and common-emitter have the same SNR.

Here are the links to a couple of files professor Marshall Leach at Georgia Tech uses to illustrate some of these concepts:
users.ece.gatech.edu/mleach/ece4391/noise_a.pdf
users.ece.gatech.edu/mleach/ece4391/noise_b.pdf

The second one is a two page file computing the noise in a BJT and a FET. See, it is not so hard when you make a good model of what is happening and then apply it.