I have one amplifier which has gain of 67.8 using OnSemi JFETs, one with SSM2019 for 100x and the one Groner 1000x described in Linear Audio which uses 8 BF862. You can do much the same thing with 2SK209s. I have some boards for the Groner amplifier and could send one to Italy if you pay the postage (or wait til I get to Rome in November).
The 1K Ohm resistor "sanity test" -- I picked it up from the Quantech user manual. I no longer have the Quantech 5173
I don't know if there is a library of AP Macros -- there are a bunch on the distribution CD-ROM which came with the unit. PM me.
There is a linked-in group "Classic AP Audio Systems Group LA"
Jack
The 1K Ohm resistor "sanity test" -- I picked it up from the Quantech user manual. I no longer have the Quantech 5173
I don't know if there is a library of AP Macros -- there are a bunch on the distribution CD-ROM which came with the unit. PM me.
There is a linked-in group "Classic AP Audio Systems Group LA"
Jack
thank you again. I should update my profile in here, but I am too sentimental lol !
I now live in orange county California. I have meant to sign up for the LinkedIn group, but I have been so busy working that I didn't even yet have a chance to open up an account and join.
Funny enough, I am going back to visit my family supposedly in November too. My home is in Perugia, if we are in Italy during the same time and you are touring, I can offer to tour you around my region. it's an hour and half north of Rome. BTW I will also go to Rome for an audio fest I believe first week of November, but it is all contingent to me finishing up a project.
I am going to PM you with my information.
I now live in orange county California. I have meant to sign up for the LinkedIn group, but I have been so busy working that I didn't even yet have a chance to open up an account and join.
Funny enough, I am going back to visit my family supposedly in November too. My home is in Perugia, if we are in Italy during the same time and you are touring, I can offer to tour you around my region. it's an hour and half north of Rome. BTW I will also go to Rome for an audio fest I believe first week of November, but it is all contingent to me finishing up a project.
I am going to PM you with my information.
Just wanted to share some results on the 70dB balanced fully discrete transimpedance phono amplifier I am working on, in specific the noise density measurement using AP.
Below the plot of the SE output.
Not sure why I have that rise above 10KHz (balance noise flattens out). Noise measured at 1KHz 330-340nV/sqrtHz
SE noise density output of the simulated circuit:
Simulated noise 237nV/sqrtHz @1KHz
I do love AP and how reliable it is, the match is striking. Thanks for sharing the macro with everyone.
Below the plot of the SE output.
Not sure why I have that rise above 10KHz (balance noise flattens out). Noise measured at 1KHz 330-340nV/sqrtHz
SE noise density output of the simulated circuit:
Simulated noise 237nV/sqrtHz @1KHz
I do love AP and how reliable it is, the match is striking. Thanks for sharing the macro with everyone.
I would be interested if you can share. BTW QA403 is a great budget tool that offer an incredible bang for the buck, small great performer all round.
Going back a few pages re: 555 vs 555B. Samuel's measurements on this page could explain why the B version was quietly ushered out: https://www.nanovolt.ch/resources/oscillators/pdf/low_distortion_oscillator_comparison.pdf
The thread has reminded me to give my 2722 its yearly boot up. I can't remember when it was last used...The UPV-B1 may have around 2dB more THD+N, but they get turned on every day (I got 3 of them, but have never lusted after another 2722...). I keep thinking about selling the 2722... Maybe I need to bite the bullet? The DScope probably has cobwebs in its fans...gets even less love than the 2722.
The thread has reminded me to give my 2722 its yearly boot up. I can't remember when it was last used...The UPV-B1 may have around 2dB more THD+N, but they get turned on every day (I got 3 of them, but have never lusted after another 2722...). I keep thinking about selling the 2722... Maybe I need to bite the bullet? The DScope probably has cobwebs in its fans...gets even less love than the 2722.
why? is it because you don't work as much on audio projects or what?
Your 2722? is truly is a reference machine for debugging and measuring audio components.
I too have for example a Sound Technology ST1710 which is a great machine and I turn it on for quick measurements especially because of the absence of a spinning fan, but whenever I want a reliable and precise measurement, I turn on my AP on, though I wished I had the 2722 instead of the 2522 which has 2dB less, nevertheless offers me peace of mind that the results I am getting are consistent and real.
The UPV-B1 is a great machine too, but why did you get 3 of those and only one 2722? What is that you dislike so much about your AP?
Your 2722? is truly is a reference machine for debugging and measuring audio components.
I too have for example a Sound Technology ST1710 which is a great machine and I turn it on for quick measurements especially because of the absence of a spinning fan, but whenever I want a reliable and precise measurement, I turn on my AP on, though I wished I had the 2722 instead of the 2522 which has 2dB less, nevertheless offers me peace of mind that the results I am getting are consistent and real.
The UPV-B1 is a great machine too, but why did you get 3 of those and only one 2722? What is that you dislike so much about your AP?
The UPV is a workhorse. It has a sucker punch that the 27-series does not: live FFT. Although the UPV is not without its foibles - to me - it's a lot less clunky than the 2722. The UPV has features such as live FR to 20KHz (very accurate), which is extremely useful for centring knobs on pots. For daily duties, the UPV kicks the AP's butt in pretty much all areas. If I want to spec something, I turn on the 2722 because the oscillator has ~2dB better THD than the B1 in the UPV. If I feed the 2722 oscillator into the UPV it gives pretty much the same THD+N reading as the 2722. I've been meaning to get in touch with Victor, as one of his oscillators will make the 2722 truly redundant for me. I guess I only keep it in case a clients wants the 'AP' logo on their plot in truth. I am not in the DAC arms race against the likes of Okto, so the UPV suits my needs well. The sad thing is that I don't think R+S have the expertise to service them correctly anymore (they no longer make AAs...apart from AP, who makes a real AA? I mean: floating I/O and autoranging - if you don't have those features, it's not a real AA).
the live feature not gonna lie it's really useful. AP could do that as well with an upgrade on their SW they just would acquire less points but they don't. I do agree the AP you need the computer so it's more clunky but there is a reason AP is now the standard for distortion measurements and also why customers want AP logo on the plots as well.
If I need live FFT for some reasons, I use on a controlled setup a soundcard sampling at 192KHz, this will give me a good idea if I am trimming the second or third harmonics. I wonder if I can have live FR on the soundcard, but that is less important than a live FFT IMHO.
AP makes full AA with floating I/O, autoranging and precise measurements throughout the entire usable range of amplitude and frequency. I would use those 2dB more on my 2522, I want to, at one point upgrade to 2722. Granted also an external gen with superior distortion performance would be acceptable solution, but so far I cannot find anything that beats AP.
I also wonder what is the difference between 2522 generator and 2722's, I don't expect it to be a different design but perhaps a change in some specific relays or some other parts or a different tuning on the circuit which purposely is tuned to achieve progressive performance throughout their system 2.
Would you happen to know what's the difference between 192K and 96K? is it the sampling freuqency i.e. you can plot up to 90KHz on your FFT while I do half?
If I need live FFT for some reasons, I use on a controlled setup a soundcard sampling at 192KHz, this will give me a good idea if I am trimming the second or third harmonics. I wonder if I can have live FR on the soundcard, but that is less important than a live FFT IMHO.
AP makes full AA with floating I/O, autoranging and precise measurements throughout the entire usable range of amplitude and frequency. I would use those 2dB more on my 2522, I want to, at one point upgrade to 2722. Granted also an external gen with superior distortion performance would be acceptable solution, but so far I cannot find anything that beats AP.
I also wonder what is the difference between 2522 generator and 2722's, I don't expect it to be a different design but perhaps a change in some specific relays or some other parts or a different tuning on the circuit which purposely is tuned to achieve progressive performance throughout their system 2.
Would you happen to know what's the difference between 192K and 96K? is it the sampling freuqency i.e. you can plot up to 90KHz on your FFT while I do half?
The 96K SYS-2722 is rumoured to have THD+N that's a fraction of a dB lower than the 192K version. I stress that this is something I've not confirmed for myself and may simply be BS.
The UPV has live FFT to 250KHz. The live FR feature is similar to the one in the DScope, but has way less lag.
The UPV has live FFT to 250KHz. The live FR feature is similar to the one in the DScope, but has way less lag.
wow live FFT up to 250KHz? It has sampling frequency of 500KHz? Ok that is a neat! I always thought 96K on my 2522 stands for 96KHz sampling frequency while 192KHz for the 2722.
I can only have FFT up to 32KHz on my unit that is a little limiting as I would like to explore higher frequencies as well, though less important, mostly for stability I could use a broader view.
Who is Victor and what oscillator does he have available?
The AP unit is not only a spectrum analyzer but it's more like a measuring tool where you can execute precisely and fast many types of measurement, however nowadays, many SW packages handle plots of impedance vs frequency THD vs amplitude or frequency, SNR, noise density etc etc etc...
Personally, I like to publish the measurements using AP people interested in my products are more likely to feel reinsure the performance of the circuit are indeed real.
It would be interesting to see if having newer opamps with better performance compared to the ones available when AP was originally designed, would improve the distortion performance.
I can only have FFT up to 32KHz on my unit that is a little limiting as I would like to explore higher frequencies as well, though less important, mostly for stability I could use a broader view.
Who is Victor and what oscillator does he have available?
The AP unit is not only a spectrum analyzer but it's more like a measuring tool where you can execute precisely and fast many types of measurement, however nowadays, many SW packages handle plots of impedance vs frequency THD vs amplitude or frequency, SNR, noise density etc etc etc...
Personally, I like to publish the measurements using AP people interested in my products are more likely to feel reinsure the performance of the circuit are indeed real.
It would be interesting to see if having newer opamps with better performance compared to the ones available when AP was originally designed, would improve the distortion performance.
The 2722 and UPV were released over 20 years ago. I very much doubt that changing opamps will improve performance for the A-D side. Even with older opamps, there are tricks that can be done to cancel distortion and reduce noise (Self shows us numerous ones in his books). My feeling is that the ADCs are the limiting factor. For the generator, I am not so sure.
Victor is a member here: https://viccc42.wixsite.com/uld-audio
NB - I've read that changing out 5532s for LM4562s on the Dscope gives a useful improvement, but the 2722 and UPV are on a totally different level to a DScope.
Victor is a member here: https://viccc42.wixsite.com/uld-audio
NB - I've read that changing out 5532s for LM4562s on the Dscope gives a useful improvement, but the 2722 and UPV are on a totally different level to a DScope.
PM me- the The QA 5173 file is too big for email or to be posted here. I have one on my bench now in for service. Its a pretty clever design for its vintage.
On most modern analyzers the limitation is not opamps. Its fundamental noise on the input and nonlinearity in the agc and tuning elements. Opamps won't change that much and sometimes make things worse.
At some point there is no return on percieved audio quality from lower distortion. Great bragging rights but "full of sound and fury, Signifying nothing".
At some point there is no return on percieved audio quality from lower distortion. Great bragging rights but "full of sound and fury, Signifying nothing".
The lack of live FFT with any System Two model is really a bummer. I usually connect the channel monitors or the "reading" monitor to a PC running REW, with a good ADC (ADI-2 Pro, and for some time now the Cosmos ADC, and both cover 200kHz and beyond). I also have a 1kHz (quite old version) from Viktors (this is how his name is spelled) and Ivan's APU and also Tek TM500 racks with the trusty SG-505 Gen and the other usual suspects. I've never found the time to replace the AC supply tranny in the SAG-505 with a proper SMPS to reduce the large mains pollution in the output.
At my former day job, the workhorses both for daily work and production-related stuff have always been UPL, UPV and dScope3, but of course we had to have a 2722 as well.
Though, as a circuit designer, I've learned more or less the hard way that the most important rig in practice -- besides a good scope -- is the network analyzer, not the audio analyzer ;-)
The extra analogue buffers for the 2722's analogue output + soundcard input makes the noise floor increase significantly. If your DUT has a perfectly flat noise floor, that sits around -130dBV (for example), even if the monitoring card has low noise, the live FFT will move it up by a few dB, maybe -124dbV. If you are used to seeing the company's DUT at -130 and in order to use the 2722 with live FFT it takes a hit by 6dB, it can be a little disconcerting... If you have a UPV nearby, you'll simply power that up!
why would you say the Vector analyzer is the most important? AP is also a vector analyzer.
And yes, the lack of live FFT on system 2 is a big bummer and could easily be corrected with a new SW release data are all there. Unfortunately, they don't support that anymore because otherwise it would be a matter of updating the control SW. I wrote them if they were going to release a new update even paying for the update, but they never responded.
Can anybody tell me what 96K and 192K stand for on AP system 2s? I thought it was the sampling frequency, but why can I only go up to 33KHz on the FFT on my system 2?
yes, the limitation is on the ADC used at the time is not possible to improve, but the spectrum analyzer sits at -160dB I don't think more is needed.
Conversely, it is missing the higher FFT frequency analysis and the lack of live feature. There has to be a reason why they went for non-live and only a depth of 32K for the bin. Perhaps at the time the HW would have legged too much, now with faster computers you can sample higher Fs larger depth and average and still have a live feature. But I do remember 15 years ago, I could already have live FFT with 192KHz sampling, 46K dept and 2-4 average with the EMU card and the dual core PC, so I don't think it is necessary because of the HW limitation, maybe they thought it gives you a more precise and repeatable result, I don't know.
I was however referring to the possibility of improving the Analogue generator with better opamps and other parts, but obviously this is not something I would personally attempt on my unit at the moment, I was just entertaining the thought.
The bandwidth is the important thing. I have an old Wandel&Goltermann SNA-2 that covers 100Hz to 200MHz (so about 100MHz realistically) and that allows me to make impedance measurements, both of supplies (right at the consumer's load point) as well as of opamp circuit output, regulators/references etc. It's also OK for some pre-compliance EMC testing in the lower ranges. And measuring transfer functions, open-loop gains etc.
All that offers insight you don't get with restricted frequency range instruments. With an AA you may find that something didn't work out as expected, but with a NA you have chances to find out what exactly went wrong... It was a real eye-opener to be able to see what the actual supply impedances and spectra are, right at the chip's pins, say for a DAC chip. I learned so much about proper decoupling by this. Before the NA, it was a blind spot (even though pulse loading and looking at the response with a scope also gives pointers).