phono cart measurement

So I mentioned in the other thread how I bought records for better phono cart measurements. Well, here they are:



I chose those solely based on this great stereoplay post over at ASR comparing different test records and their recommendation:
https://www.audiosciencereview.com/forum/index.php?threads/frequency-response-test-records.30210/

I couldn't resist trying them out tonight before heading to bed and this is what I got for my Lyra Delos on the VPI Classic. Note that I merged measurements from all three records to arrive at this graph:



Now without other carts and arms to compare to this is really just a measurement for the sake of measuring. But I have to say I am very happy with how smooth and extended it turned out. The sweep stops just short of 70kHz and likely rolls off from 50kHz on. I know the 60kHz peak from the Delos is real as it can be seen in all of my vinyl rips.
Likewise I'm impressed by how clear the tonearm resonance came out, though I bet most would prefer a lower frequency. I think the lift at 20Hz is beneficial, however.
I'm also a bit surprised at how well the DIN record still measured despite various clicks and pops. Geneally I think this was a success.

I expected more of an upper treble rise, but this Delos has many hours on it. It doesn't track the two last parts of the tracking test on the Ortofon record at all. Surprisingly, however, this is not something I've really come across when actually listening to music.

I'll get measurements of this caliber of the DV20X2L and the modded DL103 when I have some time. This graph alone probably took me some 2 to 3 hours.


Extracting distortion measurements from the sweep would be great, but I am already happy with how a noob like me was able to get a graph like that. One thing I could improve, I think, is mentioning the sensitivity I calculate from the volume I measure. Should be simple enough.

Edit: Graph said dB SPL which is wrong, but that's how I managed to align the level at 0dB. Attached one with a dBFS axis.

 

I think you could do a lot with this data!
If youre up for it, try running it through a spectrograph so you get to see the frequecy content over time as a (3D) plot. What you'd expect is a clean sweep of frequencies but you'd likely end up with a lot more interesting information about your setup.

 

Yeah, you could definitely get distortion data, for example.
I'd love for REW to support importing .wav files for different sweeps other than the REW specific measurement sweep. The Ortofon record uses a 28s 800Hz to 50kHz log sweep. You can generate exactly that kind of a sweep in REW, however you can't use it as a reference for importing a measurement, sadly.

There's other software to get distortion data. I'll have to have a look at how that works. Even so, I essentially used 4 or 5 different sweeps to arrive at the above measurement. The different records will have different amounts of crosstalk and distortion, etc. Not a super easy task to get accurate measurements and hence why it took me 2 to 3 hours for a single graph.

There are also stepped sine portions on DIN record for accurate single frequency distortion measurements, but I'd prefer using the sweep. However that allows you to view the phase of the harmonics. Phase of the harmonics isn't super accurate for turntables due to wow/flutter, as far as I am aware, though.

Resonances should be easy enough to spot. This VPI arm is kinda ringy and the features around 200Hz and 1.5kHz seem like resonances to me.

As mentioned you could also extract very precise information about the cartridge sensitivity, especially now that I could average over the three records, lol. However I don't find that too important. At the highest gain on the Skoll I get a level that is often comparable to what I get from the 4Vrms outputs of the Gungnir MB. That's good enough for me.


(I have some other crazy ideas like cartridge impedance, however that might best be measured by comparing the 47kOhm vs 10 Ohm setting rather than actually putting a voltage through the cart coils, lol.)

As for turntable speed accuracy, I ordered three belts to compare against the one belt. We'll see.

 

That is a nice curve, @Serious ! Makes me consider taking my old AQ7000 MC (was $1500 WAY back in the day), and give it one more go “just to see”. The stylus and cantilever are fine, I had just “automatically” subbed it out after a measured 1200 hours; perhaps it has a tad more life left in it?

You’ve inspired me on that!

Question: what are those resonances at precisely 50/100 Hz? Power artifacts (IIRC you are in Germany with 50 Hz power)… something else?

 

Yes, I get more hum than I'd like to admit due to how my setup is configured. I could lower the hum (and for measurements I probably should), but then the resulting sound would be worse. For now I just live with the hum.

 

Oddly enough, I found that when I put Skoll into highest gain (70dB) it had less HUM (at the speaker) than I did at 60dB. Did you do this and/or notice any deltas in the hum?

Also, it got better via balanced from Skoll to Kara than single-ended. Almost makes me wanna get a custom, 5pin DIN to dual XLR cable worked up so I can run balanced all the way and get rid of the last of my hum; but that won’t be inexpensive (and I *did* just splurge on Skoll/Kara…)

 

Yes, XLR out is better than SE out.

Welp, for me the balanced connection from the VPI Classic actually made the hum a bit worse. With some luck I can get it as low as the RCA. But it did sound more dynamic and I think the noise floor might also be just a tad lower?

I should really post my impressions comparing 60dB vs 70dB and SE in vs XLR in, but I even got slightly worse SNR at the 70dB gain setting than I did at 60dB. Not that it's bad by any stretch, records have more surface noise than that, anyway. And ultimately there's audible noise with vinyl anyway.

FWIW: I'll try to get more cartridge measurements soon. There are at least two I can measure and two or three arms I can compare them on. Plus I could get speed measurements for the three different belts I have now.

 

Another data point: Dynavector DV20X2L on the same VPI Classic with JMW 10.5 arm. We swapped the cartridges for a bit.



I set azimuth using the stereoplay test record and it seems the alignement is a bit off. This can be seen because the crosstalk isn't symmetrical. On average crosstalk might be a bit worse.

The LF resonance is a bit lower in frequency. Both are fine, though. You can also see that the ultrasonic resonance is a bit lower in frequency, so it should have less bandwidth. Just above 40kHz vs just below 60kHz.
The part above 65kHz is likely a measurement artifact, but I decided to leave it in regardless - I thought it was interesting.

I suspect a Delos with less hours would have even more of a top octave rise, but as it is it doesn't. Subjectively I prefer the Lyra.