Those that know me, know that I'm a complete neophyte, noob, boob, numbskull when it comes to understanding some of the audio measurements and how they "matter". What things are quantitative only within a given measurement setup / operator? What things are meaningfully quantitative consistent/repetitive to some level of confidence of precision and accuracy between "labs" and operators? I recently read a fascinating paper by member @PMA that indicates why measurements from a certain website re: phono stages may not be ideal. It doesn't directly relate to this, but it's a point of context re: "whose measurements to trust?"
What things are even meaningfully qualitative within and between "labs / operators"? I've done a lot of measurement calibration / operator training / statistical analysis for standards bodies over the years, and I still can't make sense of a lot of it. I know AES has measurement standards, but I've never been able to find a study (like we did with ASTM / AATCC) that shows expected error within / between / overall for a given test method run with X repetitions across operators with various types of products. Certified labs / operators running known test methods seem to be non-existent.
Also, our DIY builds likely differ more between samples than "production" products.
That's a long way of saying... when Operator Patrick creates a report from Patrick's lab for something Patrick built, take it with a pile of salt. I'm here to learn and have fun, but critique is welcome.
So, here goes nothing.
I always try to take 'real world' experiences and see if there's a measurement that can either detect it / predict it / make sense of it.
I had an amazing time building my first phono pre. I took great care (b/c it is for my brother) to build it "as good as I could build it". However, I don't have a turntable. All I had at hand was my measurement gear. So, I took a stab at doing some measurements a couple weeks ago.
It seemed wonderful, but something was nagging at me, so even though I wanted to get it to him as soon as I could, I held onto it.
WHY?
I kept looking back at the measurements and doing some back of the matchbook math and thinking... is this thing going to hum wildly, and why is the mains fundamental frequency changing in amplitude across measurements to a degree that I'd consider "not normal"?
Here's what I found out... If you want to skip to the end and just shout "Well, DUH!", go right ahead. I deserve it.
I had some spare time this morning, and I tried various methods to figure out WTH was going on with my measurements and/or my build. Did I miss a ground point? Was the wire routing proper? Was the wiring twisted properly etc. etc. THEN, I noticed something as I was banging my head against the wall and ran a quick test.
If we compare this...
To this...
That made all the difference in the world. In my rush to do the measurements before a trip and get out the door, I was sloppy and I didn't control the cabling of the test rig. My gut (and memory) says that the input interconnect(s) were too close to the PSU. Testing confirms it.
It taught me a practical (perhaps) lesson / application. In the rack, I won't drape the interconnects over the PSU or get them anywhere near the PSU for that matter, and I'll of course tell my brother to take care with cable management. Also, there's a reason why having the PSU in a separate box makes a (probably very practical) difference in noise particularly when we're running such high gain.
Maybe someone else learns from my dodoness ... maybe a few folks get a laugh... either way, I'm happy. 🙂
What things are even meaningfully qualitative within and between "labs / operators"? I've done a lot of measurement calibration / operator training / statistical analysis for standards bodies over the years, and I still can't make sense of a lot of it. I know AES has measurement standards, but I've never been able to find a study (like we did with ASTM / AATCC) that shows expected error within / between / overall for a given test method run with X repetitions across operators with various types of products. Certified labs / operators running known test methods seem to be non-existent.
Also, our DIY builds likely differ more between samples than "production" products.
That's a long way of saying... when Operator Patrick creates a report from Patrick's lab for something Patrick built, take it with a pile of salt. I'm here to learn and have fun, but critique is welcome.
So, here goes nothing.
I always try to take 'real world' experiences and see if there's a measurement that can either detect it / predict it / make sense of it.
I had an amazing time building my first phono pre. I took great care (b/c it is for my brother) to build it "as good as I could build it". However, I don't have a turntable. All I had at hand was my measurement gear. So, I took a stab at doing some measurements a couple weeks ago.
It seemed wonderful, but something was nagging at me, so even though I wanted to get it to him as soon as I could, I held onto it.
WHY?
I kept looking back at the measurements and doing some back of the matchbook math and thinking... is this thing going to hum wildly, and why is the mains fundamental frequency changing in amplitude across measurements to a degree that I'd consider "not normal"?
Here's what I found out... If you want to skip to the end and just shout "Well, DUH!", go right ahead. I deserve it.
I had some spare time this morning, and I tried various methods to figure out WTH was going on with my measurements and/or my build. Did I miss a ground point? Was the wire routing proper? Was the wiring twisted properly etc. etc. THEN, I noticed something as I was banging my head against the wall and ran a quick test.
If we compare this...
To this...
That made all the difference in the world. In my rush to do the measurements before a trip and get out the door, I was sloppy and I didn't control the cabling of the test rig. My gut (and memory) says that the input interconnect(s) were too close to the PSU. Testing confirms it.
It taught me a practical (perhaps) lesson / application. In the rack, I won't drape the interconnects over the PSU or get them anywhere near the PSU for that matter, and I'll of course tell my brother to take care with cable management. Also, there's a reason why having the PSU in a separate box makes a (probably very practical) difference in noise particularly when we're running such high gain.
Maybe someone else learns from my dodoness ... maybe a few folks get a laugh... either way, I'm happy. 🙂
Last edited:
Heh. I'm with you here, Patrick. I've been doing a lot of measurements with my AP SYS-2700 and QA403, and have found, particularly with the latter, that there are many small variables that have a huge effect on the test results. Not the least of which is the settings in the software. Lately, I've been looking at how test results from various preamps are affected by input signal changes vs. gain settings, vs. attenuation, and many other software settings. I can tweak everything for the best results, but they certainly aren't what would likely be considered "standard", though I have yet to find any standard settings for measuring preamps, only amplifiers.
Moving on to noise and hum, cable type and positioning has had huge impacts on my test results, as well as DUT positioning - e.g. vertical, horizontal, rotated around each axis, distance to the laptop, to the test equipment, to any chargers, noise sources, etc. I recently realized that my "lab" is way too noisy to take measurements without turning off the cable modem and router - a serious inconvenience to me and the household to lose internet and TV access.
Regarding PSU and interaction with cables and the preamplifier, I learned this many years ago in practice with some of our phono preamps. In particular, our Threshold FET Ten/PC was made with a really long cable to the separate PSU for a reason! Hum was audibly affected by positioning of the PSU. Some units have very good PSSR, others don't. Testing with good instruments really points that out. Today I'm re-testing my M2 after some repairs, trying to figure out how to lower the LF noise and hum, which I took great pains to minimize in the build, as well as test procedures.
I still haven't found the best approach to measure MC phono preamps. I've tried direct input signals, and using a reverse-RIAA signal, with quite various results.
Moving on to noise and hum, cable type and positioning has had huge impacts on my test results, as well as DUT positioning - e.g. vertical, horizontal, rotated around each axis, distance to the laptop, to the test equipment, to any chargers, noise sources, etc. I recently realized that my "lab" is way too noisy to take measurements without turning off the cable modem and router - a serious inconvenience to me and the household to lose internet and TV access.
Regarding PSU and interaction with cables and the preamplifier, I learned this many years ago in practice with some of our phono preamps. In particular, our Threshold FET Ten/PC was made with a really long cable to the separate PSU for a reason! Hum was audibly affected by positioning of the PSU. Some units have very good PSSR, others don't. Testing with good instruments really points that out. Today I'm re-testing my M2 after some repairs, trying to figure out how to lower the LF noise and hum, which I took great pains to minimize in the build, as well as test procedures.
I still haven't found the best approach to measure MC phono preamps. I've tried direct input signals, and using a reverse-RIAA signal, with quite various results.
^ Thanks!
It gives me a little comfort to know I'm not alone. I still have a blast with it, but at times it drives me batty.
When I was wondering to myself... does this level of noise even matter... I think it might. It could be a considerable annoyance, particularly if someone's using a higher gain pre-amp or amp and/or more likely if they're running the high gain on the Pearl 3.
Here's a quick shot with the interconnect over the PSU (something that could happen in the real-world). P3 at high gain with an output of 250mV. I'm fairly certain that would be annoying. Why'd I pick 250mV? It was a nice round number. It happened to be ~30dB above the highest levels of the mains noise, and at those levels... 250mV into an Iron Pre BAL at 6dB of gain... => 500mV into a 'typical' amp with 20dB of gain or so... => 5V with almost any speakers at 1m that's plenty loud.
The 'noise' at that level (-30dBV) is ~0.03V => 0.06 => 0V6. I know if any of my amps see 0V06 at the input; 0V6 at the output with any of the speakers I own, I can hear it.
So, assuming I did my math correctly... that's why I thought it was kinda nifty / practical.
Who knew learning about "record players" would be so fun?
Edited to fix voltage of noise into / out of amp. Also, the voltage in this sample of the output is 0V125, but the mains noise is correct. Apologies.
It gives me a little comfort to know I'm not alone. I still have a blast with it, but at times it drives me batty.
When I was wondering to myself... does this level of noise even matter... I think it might. It could be a considerable annoyance, particularly if someone's using a higher gain pre-amp or amp and/or more likely if they're running the high gain on the Pearl 3.
Here's a quick shot with the interconnect over the PSU (something that could happen in the real-world). P3 at high gain with an output of 250mV. I'm fairly certain that would be annoying. Why'd I pick 250mV? It was a nice round number. It happened to be ~30dB above the highest levels of the mains noise, and at those levels... 250mV into an Iron Pre BAL at 6dB of gain... => 500mV into a 'typical' amp with 20dB of gain or so... => 5V with almost any speakers at 1m that's plenty loud.
The 'noise' at that level (-30dBV) is ~0.03V => 0.06 => 0V6. I know if any of my amps see 0V06 at the input; 0V6 at the output with any of the speakers I own, I can hear it.
So, assuming I did my math correctly... that's why I thought it was kinda nifty / practical.
Who knew learning about "record players" would be so fun?
Edited to fix voltage of noise into / out of amp. Also, the voltage in this sample of the output is 0V125, but the mains noise is correct. Apologies.
Last edited:
I would expect those readings to suggest the noise would be audible, yes. There are a heck of a lot of harmonics there, also.
BTW, what are you using for interconnects? Are you using a QA40 something? There are no settings shown on your screen shots.
BTW, what are you using for interconnects? Are you using a QA40 something? There are no settings shown on your screen shots.
I tried two just to see if one type was more 'immune' than the other for picking up the EMI from the PSU.
Standard Monoprice and some Analysis Plus.
QA403. It's odd, I just downloaded the new firmware / software the other week. When I used to 'right click' / 'save image' it would capture all of my measurement settings (similar to below). I don't know what I've done....
"File, Save Bitmap As" will do the trick.
PM me if you want some suggested settings for noise measurements, so we don't sidetrack this thread.
PM me if you want some suggested settings for noise measurements, so we don't sidetrack this thread.
@ItsAllInMyHead 🍿 I'm enjoying your time dedicated towards discovery. I hope to get a scope at some point (this year???) and get my feet in the water.
Ive finally finished my build and currently I'm enjoying the first hours of listening to it. truly impressive, so much detail! It fired up without troubles. Thnx Wayne and diyAudio for this lovely project! Next will be to finish my F5T V3.
Also note:
The resistor values that you are changing to control gain are in the feedback loop. The distortion performance will also change when you vary the feedback resistor values.
DT
Yes, but what?
That is what I pointed at. As designed, the resistors are in the feedback loop.
Surely I've done something stoopud. Did the "smoke test"...no smoke. Yay. However, neither the on-off switch on the Schurter nor the front panel switch seem to have any effect on on-off. I've tracked wiring, and even though I can't see well anymore (really, almost blind) it looks like what I see in the build guide. Perhaps I've got the SPST wired incorrectly?
Nice thumb, huh? Slammed it with a car door...that'll come off soon. Anyhoo, advice appreciated!!
Nice thumb, huh? Slammed it with a car door...that'll come off soon. Anyhoo, advice appreciated!!
