LOL! Sorry I didn't realise you were Richard's "intern."
Good luck with that. If Boggit chooses to deploy this I wish him great commercial success. And if it leads to more sales for Sonic Imagery then good on them as well.
Perhaps you will also pop back when you have useful content of your own.
Good luck with that. If Boggit chooses to deploy this I wish him great commercial success. And if it leads to more sales for Sonic Imagery then good on them as well.
Perhaps you will also pop back when you have useful content of your own.
Last edited:
No I don't work with Richard, but he makes an excellent discrete opamp, and designed an awesome buffer circuit for free. And it's not just specific for use with his opamp, it will work with any that can provide enough current to the NC-500.
And if it wasn't for my contributions I've shared on this forum, none of this buffer talk would be happening in the first place. As I recall, and is on record here, (although much was deleted for some reason) Colin's original plans were to just use the Hypex OEM buffer. I already ran all of these discrete opamps through the paces with the NC-500's long before Colin came along to announce he had a new amp in the works that will use the NC-500 module.
Here's Richard's buffer circuit. If you can contribute a better circuit design combined with measured results, please share schematic and measurement.
So is this new circuit compatible with Burson, Sparkos, LM4562 and Sonic Imagery and with the Hypex voltage regulators. Also can it be driven by a standard output of 2.0 volts unbalanced or 4.0 balanced from the likes an Audiolab MDAC
Indeed. And Colin might well be rueing the day he listened to you. Just imagine the difference if he had simply provided a very good buffer and allowed his customers to enjoy their new equipment. Providing for amateur fiddling isn't often good business sense.
You were also the guy spreading falsehoods with respect to the output current demands of the buffer's op amps.
You were also the guy banging on about Class A discrete designs for lower noise and yet you've shown no understanding of where noise originates in the circuit.
I merely raised a few questions in response to nerdman's post. In particular, I challenged the statements that the 47k resistors were the "dominant noise noise and that the "Hypex Amp needs current drive > 70mA", this latter one being a topic dear to you. I later provided analysis to support my questions. and invited critique of such. Neither you nor anyone else has found error in the analysis I did. Again, if you are able to correct the analysis for any error you have identified please do.
I'm not even sure you fully understand Richard's proposed changes, especially Cfb and why it might be necessary for the SI op amp or its impact on UGF, PM and GM. I wouldn't want to roll off the feedback for all op amps anyone might try and I don't think its use is necessary for an LM4562-based circuit (in fact I would argue it is detrimental).
Achieving the results posted by Richard isn't hard and it certainly doesn't require expensive, discrete op amp designs.
Hi, Bjørn,
Sorry for the slow response, but just got home from an Easter weekend in Germany (a combination of track day at Nürburgring and wine tasting in the Ahr valley - not the best combination).
I did reply by email, but let me respond here as well (despite this not really being hypex-related).
Personally I am a bit too busy with my two daytime jobs (and a bunch of hobby projects). I did discuss this with a friend who has done electronics assembly work and has the time, but he is rather reluctant - seems he has had his fingers burned by golden ear audiophiles in the past. He is concerned that someone will complain about the sound quality based on arbitrary subjective opinions, no matter how good the measurements are.
I also discussed this with Arny Krueger (who pretty much invented the modern ABX testing box), and his opinion is that to really do a good job, you need to use some pretty expensive components (especially the relays), so the price comes close to what the commercial products cost, even if you don't allow much for labor.
This stuff is of course totally off topic, but interestingly it seems that it is the hard core subjectivists who have hit the wall in that thread. The last post (#566) actually states "Good ol' subjective assessment is not considered good enough here. Especially as the final arbiter of whether fidelity has changed or not. That must have been clear to you by now, but you insist on proceeding with definitive proclamations based on subjective assessment".
Colin offers the standard buffer option as well. But the buffer everyone is raving about is the upgraded buffer. And all they have heard is the noisy one! Just wait till they hear the new one with Sonic Imagery 994!
If you have doubts about Richards circuit, then when you build your buffer, and share measured results, things should look even better then. I can't wait to see the end result.
As far as I'm concerned, measurement data only matters to the engineer of the product. As far as the end user is concerned, as long as the industry standard measured data meets or exceeds acceptable levels, the final evaluation must be preformed subjectively. No matter how much you're against this concept, this is simply how things are done on planet earth.
Well he provided the noise floor measurement. More than I've seen from anyone else so far. But I can imagine if any OEM's decide to use the circuit, we will see comprehensive measurement data from them. Since there's so much competition building amps with this concept now, the only ones with a hope of survival, are the ones that do provide the measurement data to back them up.
I guess that means beg borrow or steal an APx-555 🙂
You can rent them in the U.S for $1500 a week.
Last edited:
"Must"? One of the laws of the golden ear business? 🙂
The planet earth I live on has listening test specifications such as ITU-R BS.1534 and BS.1116. Listening tests are useful when done properly - and they complement the measurements.
But yes, as many manufacturers (especially "high-end" ones) fail to provide any test results, "end users" are unfortunately often forced to rely on just subjective, casual listening. Are you saying you see that as a good thing?
In reality 99.9% of any end user I've seen or heard of evaluates gear by listening to it. If you are expecting end users to preform lab rat style testing to evaluate your products, you'll likely find sales will be quite weak. If someone enjoy's a particular piece of audio gear for any reason, he doesn't need to justify ownership to anyone.
Last edited:
That's the only measurement that was provided! Any other measurement's we will see from the OEM's who decide to use the circuit. For comprehensive measurement data on the 994 discrete opamps, look here:
http://www.sonicimagerylabs.com/pro...e_HD_OpAmp/994Enh_DiscreteOpAmp_Datasheet.pdf
Sure. And some people buy gear just based on how it looks, or because they like the brand. It is our own money, and as you write, we don't need to justify it to anyone.
Making recommendations to others, or making claims about the superiority of one piece of gear over another is a different story.
I don't expect end users to perform that kind of testing, just as I don't expect a car buyer to have to put a car in a test bench. Fortunately the car industry has good, solid standards for providing comparable and verified test data, so end users don't have to (a certain emissions test is a notable exception 🙂 ).
While organisations such as IEC and ITU have specified a bunch of useful test standards, it seems the high-end industry is mostly ignoring them- reflecting the positioning of "high-end" gear as lifestyle/luxury products rather than technical equipment.
As far as I'm concerned, if the measured results of the combination of SMPS1200A700, buffer board, and NC-500, meets or exceeds all of the measured data provided on the NC-400 spec sheet, then any tests beyond that should be preformed subjectively.
The NC-400 can be used as a benchmark.
Were we talking about op-amps or a buffer circuit that has "much lower noise floor" etc.?
Indeed. Too much inscrutable behaviour like arguing falsehoods such as the LM4562 doesn't have enough drive current to work in an NC500 buffer.
BTW you don't need an APx-555 to do this sort of testing...
No. If they heard the Sonic Imagery slotted into the standard buffer they listened to a circuit in which the op amps were unstable...
It's worth noting that the LM4562 has a GBW of 55MHz. It won't be hard for anyone to test the standard buffer for instability. And it won't be hard for anyone to simulate the circuit and take a look at phase and gain margin for an LM4562-based circuit. My bet is they will be just fine without any further pole added by an additional compensation capacitor. Unnecessarily rolling off the feedback can only hurt distortion at high frequencies.
BTW you don't need an APx-555 to do this sort of testing...
No. If they heard the Sonic Imagery slotted into the standard buffer they listened to a circuit in which the op amps were unstable...
It's worth noting that the LM4562 has a GBW of 55MHz. It won't be hard for anyone to test the standard buffer for instability. And it won't be hard for anyone to simulate the circuit and take a look at phase and gain margin for an LM4562-based circuit. My bet is they will be just fine without any further pole added by an additional compensation capacitor. Unnecessarily rolling off the feedback can only hurt distortion at high frequencies.
Last edited:
The THD+N measurement data Richard provided was of his discrete opamp installed in his buffer board circuit.
Last edited: