Phono Cartridge IMD Measurements, Observations and a Question

Just reading your exchange with Atmasphere, it seems that the readings are repeated under all circumstances and also regardless of VTA and that the L and R channels read the same or possibly identically. Doesn’t this suggest that the measurements might be spurious, and by "spurious" I mean the result of some factor you and we are ignoring (not necessarily RFI). It just seems unlikely to me that two channels of two different cartridges would not at least slightly differ from one another in IMD. On the other hand, you seem to know what you are doing; I am not suggesting that you don’t. I hope we can agree as a separate issue that the differences between the two cartridges might be entirely due to the fact that they are mounted in two different tonearms, one of which is on a pod.

You wrote, "I created an CD IMD test file and played it via my Oppo player through to the preamp out. There is no measurable IMD (probably below the noise floor) except at max gain where the preamp is overloaded. So there’s no IMD being contributed by the Levinson no 38s at my normal listening level and 30dB higher also." Can you say more about the CD? In what way is it an IMD test file? (If it is to test for IMD, then ought there not to be an encoded signal that creates a control for a certain level of IMD?  If so, in your statement are you saying that you see the control amount of IND encoded on your CD, and it is not augmented by passing through the linestage and amplifier or just the linestage. I am only questioning because the topic is interesting and to me worthy of discussion, because maybe I can learn something.

@lewm I found a website called "Audio Check". It has lots of test waveform files. including IMD test waveforms. The one I used is  60Hz/7KHz with a 4:1 ratio 60/7000. This is the SMPTE spec. I downloaded this file and burned a CD. I played that CD on my Oppo 105 through to the preamp out.. This method bypasses the entire phono chain, but keeps the preamp. The conclusion is that there is no contribution to IMD from the digital path to the preamp (included). The IMD test on the phono section involves measuring at the preamp outputs. Since the preamp was tested OK on the previous test with the CD, the source(s) for the IMD have to be in the phono path. This path comprises the cartridge, tonearm, turntable, SUT (if used), and the phono preamp. The source is one or more of these. Or it could be some combination of components.

Atmasphere has suggested an interesting effect may be causing ecxcess IMD. I have no experience with what he's saying, although I do inderstand the physics. I am thinking about how to measure this. I hope Atmasphere replies with some useful suggestions.

My original post stated that the IMD was different for each cartridge. Hana ML was 0.5%. The Shelter 901 was 3%.

Subsequently, I have discovered that adjusting VTA changes the amount of IMD and there is an optimum, but it is like 2.5% for the Shelter, not much of a change. I'm struggling to find what a "normal" amount of IMD is. Maybe these numbers are typical and I shouldn't worry.

@kevemaher You might get a little device called an inverse RIAA filter to put at the output of your CD player. If you can reduce the level properly, you could test the phono section too. The one at the link provides some attenuation.

Thanks for providing more detail. As I guessed, your test CD does encode a control for IMD. Also, I did already understand that you’re seeing different levels of IMD from one cartridge vs the other. That could be real or due to mounting the two cartridges in two different tonearms, one of which is on a pod. Or even u nintentional minute differences in setup accuracy. Now you’re saying that VTA does make a slight difference which is encouraging in that it suggests your method is actually reading IMD due to cartridges. I think this question arose earlier, but have you yet tried swapping the cartridges between the two tonearms? If the cartridge mounted on the pod is consistently exhibiting more IMD than the one mounted on the TT, that may suggest the pod is partly a culprit in causing the differences.

Today, I did a deep dive into IMD definition and measuremet methods. I discovered that the calculation that I created neglected the contribution from the low frequency peak. After that was added, the calculated IMD reduced from 3% to 0.9%, which is more acceptable. This is for the arm on the pod.

I have more IMD measurements to make. I now feel confident that I'm measuring and calculating IMD correctly.

Left unexplaind is why I can't alter the IMD very much.