UCD180 questions

[Pierre]

Thanks, Tiki.

I have played a little bit with RMAA thinking on using it for testing amplifiers. It is quite nice for that! Although my soundcard is more limited and has no balanced i/o.

Of course results should be better using an alluminium power resistor and then a couple of small resistors or a good potentiometer as the voltage divider. This way you can test it to higher powers as well.

The "good" way to do tests is using a quite complex lowpass filter to avoid getting hf transients into the soundcard input, that could deteriorate results.

Please read AudioPrecision's app.note on how to do these measurements. However, I think that you can get quite approx. results without that filter, as you did.

Best regards,
Pierre

[serengetiplains]

What function does the 8620 perform? Does it provide voltage gain? Can it be bypassed with a step-up transformer?

[classd4sure]

Well there's two of em, they buffer the inputs and provide some gain... 4.5 I think.

That goes through the coupling caps and connects right to the comparator differentially. Think of a three op amp instrumentation amp with the third amp being the comparator, that's it.

Transformer?? Interesting.. it works in my simulation. Wonder what it would do to EMI, and what EMI would do to it? Might not work well if the feedback isn't mirror like.

[serengetiplains]

Hi Chris, I'm not sure what you mean by your reference to feedback. Transformers tend to do well with EMI, at least the higher frequency variety as they have a limited (almost audio-only, ahem) bandpass. If the comparator is at ground potential, a transformer could replace the two op amps, the coupling caps both before and after the op amps, and give *true* differential drive while blocking DC at the same time. The transformer, in fact, could be a transformer volume control to give level attenuation. Very elegant. And sound quality? Excellent, if you ask me. Ask Rowland, even.

[ghemink]

Originally posted by serengetiplains
Hi Chris, I'm not sure what you mean by your reference to feedback. Transformers tend to do well with EMI, at least the higher frequency variety as they have a limited (almost audio-only, ahem) bandpass. If the comparator is at ground potential, a transformer could replace the two op amps, the coupling caps both before and after the op amps, and give *true* differential drive while blocking DC at the same time. The transformer, in fact, could be a transformer volume control to give level attenuation. Very elegant. And sound quality? Excellent, if you ask me. Ask Rowland, even.

But the input impedance could be a problem. The input impedance of the negative UcD input (so after the input opamps) is only about 1kOhm when driven pure symmetrically. When you use a transformer to get gain, the input impedance that you see at the input of the transformer would be less than that, with a factor 5 gain, input impedance would go down to 200Ohm or so, can your preamp drive that?

Gertjan

[classd4sure]

Hi,

Yeh, seems like a bright idea. I mentioned feedback because if the differential feedback loops aren't identical (I don't know if they are or not) I thought it might introduce an offset, and wouldn't be entirely common mode? Just a thought, not gospel, good for provoking discussion

There's only coupling caps after the buffers/amps, at the comparator inputs, to prevent DC offset being amplified.

Anyway, I simulated it with mirror image feedback loops and inputs, an ideal transformer with secondary grounded center tap, .99 coupling, and at 1:1, 1:2, 1:100...

I can tell you it worked great, but what I have would differ from the real thing, also from non ideal simulator land, by a great deal in this case I would think.

The transformers for this kind of app change impedance with gain right, by an order of magnitude for the one I looked at. At some point it might load down the amp.

At the input of the comparator portion I also don't think the impedance is the same, the old pdf say's ~1.8k inverting input, ~10k non inverting. I'm sure Jan-Peter will fill us in if it can work or not.

[serengetiplains]

1K-ohm is a lowish impedance. My DAC outputs a 24dBu signal, so I think it could drive the comparator via a 1:1 line-level coupling transformer.

Jan-Peter, would appreciate any comments. If the UCD sounds as good as some say it does, the opamp and coupling capacitors would seem to be the weak links in this amplifier---ridding the amp of them should provide a healthy upgrade. For those of you who have not heard a high quality coupling transformer, it can be a jaw-dropping experience.

[serengetiplains]

Chris, thanks for your simulations. A good transformer provides precisely-mirrored opposite-phase signals the precision of which is unobtainable by active devices, which tend to bounce around, even if accurately matched.

Jensen Transformers probably sells a 600ohm:600ohm transformer that would work in this circumstance.

By the way, a transformer would also offer galvanic isolation, reducing ground-carried noise from the oh so sensitive input circuitry.

[serengetiplains]

Jan-Peter, to be a little more clear about the information we're needing, if you could indulge us we need the drive requirements and input impedance of the comparator (can the comparator be driven either single-ended or differentially---if so, we need drive requirements for both drive modes), and on a related question the gain provided by the opamps.

Thanks!

[serengetiplains]

FuriousD, any comments re the transformer suggestion? It was your posts re the opamp that got me thinking of a solution.