Am I talking about same total power?
Anyway, we can build the circuit(s) and test it later to our hearts content. But, you see where I am going with this.... the chance to learn something new besides just building yet another opamp... To me Scotts circuit offers the most flexibility, as I indicated long before.
Now, I wont have time to make one for awhile... unless someone offers pcb's of it. Going to asia in a week for R&R and wont be back for a month or so. This DIYAUDIO is just too stressful.
BUT, I'll still check in from time to time. Ciao !
Last edited:
Thanks Wayne, that is a great compliment.
I have some more in store in the "fun with 2SK2145s" category, but as well I need to follow through with the "linuxguru's buffer and related" thread.
New thread for IT comments
http://www.diyaudio.com/forums/analogue-source/220542-riders-storm.html#post3181699
George
Its all good stuff
This is all good stuff that might not have come about if we had rushed to build. And, thx for this new forum idea on opamps. --Ciao
This is all good stuff that might not have come about if we had rushed to build. And, thx for this new forum idea on opamps. --Ciao
I would suggest having a suite of defined, different power supplies on hand to power the device while testing performance parameters, and see what the variances in behaviour were. As an example, 3 supplies: first, a totally standard non-regulated, then a straightforward regulated, and finally a Jung style super reg. or similar. Could be interesting ...
Frank
Am I talking about same total power?
Going to asia in a week for R&R and wont be back for a month or so. This DIYAUDIO is just too stressful.
BUT, I'll still check in from time to time. Ciao !
Have fun. Just ran a math exercise 2 tones vs 4 tones and a simple square law non-linearity the problem scales directly. Each doubling of the tones needs a 3dB reduction in amplitude and the total power in all the intermods only goes up 2dB per doubling.
Crossover issues are more complicated, it depends on density and slope of zero crossings not a simple relationship to number of tones.
Last edited:
I know these guys are gifted. I just wish they would cut to the chase and design and stop worrying about proprietary issues. If a design emerges, then, it will either be bought off by some fat cat that wants to monopolize it or mushroom it or take 6 months to a year to optimize by changing parts. Every man has his price! And these guys are not being paid to do this so it will probably be slow. Also it would be nice rather than to read all 133 posts, to know where the dominant designs are. I have Wurcer so far after reading half of them before going to work. Thanks for listening. Ray
Haven't used them but have seen a lot of other amp designs using them. They have fairly constant, if not particularly high beta with collector current over a decent range. I'm skeptical of that 10 ohm number as I think we would have heard about that if accurate.
The old National databook curves are significantly different than the Toshiba ones, not too surprisingly. Toshiba says "low noise" and lists 8dB max for the PNP, 10dB for the NPN, over a BW of 10Hz to 15.7kHz at 250uA and Rs of 1k. There are no noise figure contour plots at different frequencies, unfortunately. So who knows how much 1/f contributes, and there is also no "typical" NF listed.
I don't think anyone is concerned about proprietary issues in here. Some lurk hoping to lift designs they might manufacture, I suppose. But once presented here there is nothing to be bought off by some "fat cat" --- it becomes public domain, unless already patented. As well, I doubt anyone in here can be "bought" in the sense of revealing trade secrets of their employers. Most who know are here to instruct, and principles can't be patented.
Does that also mean that we cannot duplicate well enough what the ear is exposed to because of test limitations. What on earth is A-P doing..... smoke and mirrors? They use a lot of digital/DSP to make it work for them. can you explain thier process and if it can be used to figure if the total unwanted/generated harmonics would reach a threshold of detection (near .05%) for the DUT. IMO, that would go a long way towards answering the question once and for all about what is or isnt audible in amps re. harmonics with music listening. -RNM
Last edited:
I have only used them as cascodes.
Bob C is using RB=60 and RB=92 in his models for the 5401/5551, to improve the accuracy of noise simulations.
See his models here: Bob Cordell spice models
Stein
This is not rocket science we do it all the time. The 31 tone test would be trivial to implement on a sound card my 100 tone test was in the noise with loop-back. 0.05% is -66dB at that level people could not hear a marching band buried in the music.
If trivial then we should use It instead of 1-2 or 3 tone tests. So if we use it for DUT (opamps or other) we can see how good a circuit is with 30 or 100 tones. And easy to empliment is better news than i hoped for. How to do it -- where do i look - did you publish the method using a sound card? Software used? I can use an outboard E-MU 0204/usb?
In other words -- I am looking for near turn-key info. -Dick
Last edited:
the soundcard loopback multitone test will have gone thru some monolithic op amps in the analog domain - since Scott's results are in the noise floor it stands to reason that those monolithic op amps already perform "perfectly" on the multitone test at the soundcard signal levels - likely some Al electolytics in the signal path too
the "noise fill" multitone tests don't appear to be the Holy Grail - just awkward to do in the past
the "noise fill" multitone tests don't appear to be the Holy Grail - just awkward to do in the past
MultiTone tests are better than ---
I'd rather not spectulate and make conclusions based upon such. I'd rather do a multitone than a single tone or IM as it is closer to actual app. use. Only if it cant be reasonably done by most people here who can do thd/harm/im tests with thier equipment. Rather than do a bunch of individual tests... that would be a throw-back way... reminds me of point by point pencil plots. Just automated. The multitone may not be the holy grail but it is not worse - IMO it would be a better standard test.
IC opamps have become sooo good that there is little to fault them for audio apps. Discrete opamp circuits are better in current and voltage level of output and possibly cheaper than premium IC's. But the distortion is unquestionably extreamly low.
Power amps might be another story and I, for one, would like to see the multitone test applied to someone's whole chain of audio gear... Then the recording side of the process. That is what we are listening to/hearing... the whole chain. So, how to get a turn-key multitone test going using a PC? Quick, easy and not too expensive. Without the rocket science approach, pls. Just want to do some tests and not reinvent anything. Thx -RNM
I'd rather not spectulate and make conclusions based upon such. I'd rather do a multitone than a single tone or IM as it is closer to actual app. use. Only if it cant be reasonably done by most people here who can do thd/harm/im tests with thier equipment. Rather than do a bunch of individual tests... that would be a throw-back way... reminds me of point by point pencil plots. Just automated. The multitone may not be the holy grail but it is not worse - IMO it would be a better standard test.
IC opamps have become sooo good that there is little to fault them for audio apps. Discrete opamp circuits are better in current and voltage level of output and possibly cheaper than premium IC's. But the distortion is unquestionably extreamly low.
Power amps might be another story and I, for one, would like to see the multitone test applied to someone's whole chain of audio gear... Then the recording side of the process. That is what we are listening to/hearing... the whole chain. So, how to get a turn-key multitone test going using a PC? Quick, easy and not too expensive. Without the rocket science approach, pls. Just want to do some tests and not reinvent anything. Thx -RNM
Last edited:
RNMarsh,
The question is if you can possibly use the internal sound card in a typical computer to do this. I am probably wrong but it doesn't seem like most sound cards would have the precision that you are asking for, and what are you using as a reference to calibrate such a system? I assume that there is a reason for the Audio Precision testing equipment to exist besides looking nice.
The question is if you can possibly use the internal sound card in a typical computer to do this. I am probably wrong but it doesn't seem like most sound cards would have the precision that you are asking for, and what are you using as a reference to calibrate such a system? I assume that there is a reason for the Audio Precision testing equipment to exist besides looking nice.
Harmonic vs Intermod measurements
I did this in da old days and had a little table showing what equivalences you'd expect for 2nd, 3rd Harmonic vs various Intermod measurements.
But it was really hard work and I don't think I've got enough brain cells left to do it again.
It was scribbled in my B&K 1902 manual.
Scott, have you worked all this out for simple 2nd, 3rd, 4th ... etc non-linearity.
I did this in da old days and had a little table showing what equivalences you'd expect for 2nd, 3rd Harmonic vs various Intermod measurements.
But it was really hard work and I don't think I've got enough brain cells left to do it again.
It was scribbled in my B&K 1902 manual.