Actually, post 1639 is incorrect in asserting CMRR drops if source impedances remain balanced. Those interested in the specifics can refer to THAT's AES paper on the topic. Also, three op amp instrumentation amplifiers routinely exceed the bootstrapped input impedance of a THAT 1200 by three orders of magnitude. The same can be accomplished with intelligent use of audio op amps. So the bootstrap network and fourth op amp in the 1200 isn't taking things farther in any meaningful sense.
What makes the THAT 120x series difficult to improve is they've the highest loop gain of any difference amplifier on the market with competitive CMMR. 90dB typ is pretty routine by instrumentation amplifier standards but the 22MHz GBP is 17dB above the next best part I'm aware of. In "discrete" implementation it's easy to increase GBP but it's an order of magnitude more cost for resistor quads offering CMRR that's not as good.
What makes the THAT 120x series difficult to improve is they've the highest loop gain of any difference amplifier on the market with competitive CMMR. 90dB typ is pretty routine by instrumentation amplifier standards but the 22MHz GBP is 17dB above the next best part I'm aware of. In "discrete" implementation it's easy to increase GBP but it's an order of magnitude more cost for resistor quads offering CMRR that's not as good.
I agree with most of what you say, but Bruno Putzeys calls this the 'imbalance illusion' in his article 'The G-word ...', published in Linear Audio. Very worthwhile reading All articles | Linear Audio[0]=field_author%3A379
Okay, guys, I'll look into it. I know I saw some in-depth description of that circuit many years ago in some book. I thought it was in the first or second edition of the Ballou-edited "Handbook for Sound Engineers" but I just looked and only found a short, cursory comparison with transformers.
I've seen the Linear Audio blurbs with interest, but they seemed rather expensive, but the first six issues are on sale as a package, and the mention here was enough to prompt me to order them.
The 5th Edition is now available:
"Handbook for Sound Engineers"
5th Edition
Edited by Glen Ballou
Focal Press: Handbook for Sound Engineers: 5th Edition (Hardback)
**************************
The Bill Whitlock chapter 11:
"Audio Transformers"
Handbook for Sound Engineers, 3rd Edition
Jensen Transformers Application Notes | Jensen Transformers
"Handbook for Sound Engineers"
5th Edition
Edited by Glen Ballou
Focal Press: Handbook for Sound Engineers: 5th Edition (Hardback)
**************************
The Bill Whitlock chapter 11:
"Audio Transformers"
Handbook for Sound Engineers, 3rd Edition
Jensen Transformers Application Notes | Jensen Transformers
I skimmed over all this thread, though I may have missed it: Tom, did you ever happen to get around to measuring the NCore monoblocks? Just curious how they look.
That sounded interesting, but when I checked the site all I found was the pay to play version. Am I missing something here?
Mike
I did. I was able to borrow a pair of Hypex NC400 powered by the Hypex SMPS600.
They measure very well. I don't have access to the plots right now, but I recall the measurements revealing a very well-behaved amplifier with good performance. Not quite as good as the MOD86, but close to.
In terms of sound quality, I liked the Ncore as well. They had a good firm grip on the bass and good imaging. I did find the highs straining, though. The highs on the MOD86 are smooth and natural. The highs on the Ncore were rather harsh by comparison, which I found tiring in the long run.
The Ncore seems more sensitive to ground-related interference than any other amplifier I've tried. One channel had a slight buzz in it. Not very loud, but loud enough to notice with the ear against the speaker. I tracked this down to be related to my preamp, but I've never noticed it with other amps, and I've had quite a few amps connected to that preamp. I'm not sure I got all the way to the bottom of this issue as I was running out of time.
There's no doubt the NC400 is a nice amplifier module. I tortured it pretty good and wasn't able to cause it to misbehave. Sadly, the same cannot be said for the SMPS600 power supply. After running the amp at the rated output power into 4 ohm for a good 15-20 minutes, the power supply overheated and shut down. I think that's rather disappointing. That was the only significant disappointment, though.
Having tried both, I still favor the MOD86, but I'm obviously biased.
Tom
Last edited:
Were you using the speakers with Alpair 10p driver? I assume that the 10m was not to your liking.
I briefly browsed through the G word paper, seems that I might have read it before because the general ground arrangement considerations are similar, one thing that I like more is to avoid connecting two or more reactive components to the same node the unless the design specifically benefits from doing so. This may result in more resistors than normally used.
Last edited:
Hi
It is well known that the class D has problem in the reproduction of high frequencies, in the sense of what Tom has said. Sound spectacular but tiring ..
Today, class D for subwoofers and woofers is fine but not for tweeters, which is better class AB or A.
The novelty is about the reliability of the power supply when it is pushed to the limit (which is very high). First news about it.
It is well known that the class D has problem in the reproduction of high frequencies, in the sense of what Tom has said. Sound spectacular but tiring ..
Today, class D for subwoofers and woofers is fine but not for tweeters, which is better class AB or A.
The novelty is about the reliability of the power supply when it is pushed to the limit (which is very high). First news about it.