Cartridge Alignment: Setting Azimuth with a PC and Soundcard

Discussion in 'Vinyl Nutjob World: Turntable and Related Gear' started by purr1n, Apr 17, 2016.

  1. lehmanhill

    lehmanhill Almost "Made"

    Contributor
    Joined:
    May 3, 2018
    Likes Received:
    545
    Trophy Points:
    93
    Location:
    Home of Jiffy Mix
    I ran across the blog at the link below. I'm not sure I understand or completely agree with all of his arguments on azimuth, but the 3 blog posts on the subject end with their setting azimuth by equal distortion rather than minimum crosstalk. Check Azimuth parts I. II. III.

    http://korfaudio.com/blog36

    The rest of the blog is pretty interesting too, especially if you are like me and like thinking about the various design choices in a tonearm.
     
  2. MellowVelo

    MellowVelo Friend

    Pyrate Contributor
    Joined:
    Sep 5, 2016
    Likes Received:
    913
    Trophy Points:
    93
    Location:
    Chicago
  3. ohshitgorillas

    ohshitgorillas Friend

    Pyrate
    Joined:
    Nov 27, 2015
    Likes Received:
    681
    Trophy Points:
    93
    Location:
    Sacramento, CA
    Does anyone know of any good, relatively cheap torque screwdrivers with a range that would work with headshell screws? Alternatively, can someone tell me the pressure values that such screws usually have so I can find something that has the correct range myself? I am going to be fitting my turntable with an Ortofon Bronze cart/needle, and would obviously like it to be as well aligned as possible. I really appreciate the posts on this forum that demonstrate how to do this quantitatively. I'll be using an old Behringer UCA202 as the ADC. I've also installed a VTA adjustable arm base (Rega P3 w/ RB300), have a blank record, still need a cheap test record with 1kHz or something similar, and have a protractor specifically made for the RB300 family of tonearms. Between all that, I should be able to nail the alignment with a little patience.

    I can definitely appreciate the idea that having the cartridge mounted perfectly perpendicular to the arm with the same tightness on each screw as NOT being the ideal left/right alignment, but I feel like starting from a place of equal pressure and then trying to adjust in one direction or the other would be much easier than estimating from the beginning.

    Between the Ortofon Bronze and my upcoming Eros 2 build, I think I'm going to be happy with my vinyl rig for a very long time to come.
     
  4. JoshMorr

    JoshMorr Friend

    Pyrate BWC
    Joined:
    Oct 4, 2015
    Likes Received:
    3,066
    Trophy Points:
    113
    Location:
    Portland, ME
  5. Pogo

    Pogo Friend

    Pyrate
    Joined:
    Jun 14, 2018
    Likes Received:
    494
    Trophy Points:
    93
    Location:
    SoFla
    You are over thinking this, your own hand/eye coordination is more than adequate.
    Feel the force,Master Luke!
    In 45 years of mounting cartridges to head shells, my biggest issue has been cantilever and stylus alignment vs the cartridge body. Your eyes,your ears, and a good test record are the most important factors here.
    There is no such thing as a perfectly aligned cartridge for all of your current and future vinyl!
     
  6. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Thank you for posting this. Really helped me understand what gets influenced by setting azimuth (and I have to admit that I never thought about alignement between the generator and stylus, just thought of stylus/generator unit vs cartridge body).
    I just recently set azimuth with a small 7" record that was included in a German stereoplay magazine a couple months ago. On one side it includes 1kHz and 3.15kHz sines and lateral and vertical pink noise, as well as individual left and right channel pink noise tracks, perfect for setting azimuth in REW (aswell as measuring FR and crosstalk).

    I got much more precise and stable imaging (especially center imaging) by setting the carts with the RTA function in REW than by doing it by hand. I wouldn't try to do it any other way now.

    As far as the measurements go, I found that with an improper azimuth setting not only the distortion and crosstalk become unbalanced, the FR does too. While the individual left and right channel pink noise was at a very similar level before, the mono signal was quieter and the out-of-phase signal was louder in the left channel (though that's not a surprise considering mismatch in crosstalk).
    The difference for the 2nd harmonic for the 3.15kHz tone was 8dB in the beginning, the difference for the mono signals was 1.5dB for visually aligning it. Both definitely audible - pretty bad in hindsight.

    I'm not a fan of setting distortion or crosstalk with just a sine, though. Cartridges can also have smaller spikes and dips in their crosstalk and distortion curves, sometimes only in one channel and as such setting it at 1kHz might make it less even for most of the frequency range. I think pink noise is a much better idea and it was quite simple setting it with the record and REW. In the end I managed to get the channel matching for each of the signals to well within 0.1dB for most of the frequency range for both of the carts that I aligned on their respective turntables. Averaging the two sine wave distortion results, they also looked very similar.

    Like @purr1n I found that both carts had to be tilted a bit to be aligned properly. You definitely wouldn't have set them up like this without the measurements, but the end result is much better now. It seems that as per Korf I got lucky twice that while the stylus and generator weren't precisely aligned with the cartridge body, the stylus was aligned nearly perfectly with the coils. Both the distortion at the two frequencies and all the FR and crosstalk curves were very closely matched, as they should be. I can attach images of the cartridges as they are aligned now and of the FR curves for those curious. The stereoplay magazine always includes cartridge FR and crosstalk measurement curves for their tests, but that's the only source with a large database that I can think of at the moment. I'm sure not too many people know how their phono carts measure when buying them.
     
  7. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,975
    Trophy Points:
    113
    Location:
    Vancouver BC
    @bazelio 's oscilloscope technique for channel balance and phase can be faked in REW etc. by appropriately scaling the impluse window. Here are 3 examples of many from yesterday's fumbling around with a cart install:

    [​IMG]

    To get started I used an arc protractor made from the Conrad Hoffman ( @Ruby Rod ) custom arc template generator to set alignment, and left the azimuth alone. This gave the red and green curves, which are a little out of phase and have a small amplitude difference. Crosstalk was essentially identical, meaning azimuth was probably good.

    Repeatedly tweaking cartridge alignment to try for better phase coincidence and amplitude balance got me a few results like the green & tan pair - either or both phase or amplitude getting worse. It also screwed up the crosstalk, with some results being as much as 6 dB different between channels. Amplitude imbalance might be inherent to the cartridge but phase should be able to come pretty close to overlap because the outer null point for the alignment I chose (Lofgren A) lay within the test track. In the end I settled on the blue & purple alignment because phase offset was the smallest I could get to, amplitude imbalance is almost zero, and with a tiny tweak on the headshell, crosstalk is about 0.5 dB different between the two channels (black is the amplitude in the 'other' channel):

    [​IMG]

    [​IMG]

    Although even, crosstalk is still pretty high ( = separation low at just under 25 dB). This wasn't any better on the first try, and on the intermediate trials one channel would improve but the other would get worse. 25 dB (at 1kHz) is the minimum spec given for the cartridge, but I would have hoped for better than this; a 'typical' graph on the spec sheet suggests it ought to be better than 30. The cart is however not new.

    But it sounds better than the cartridge it replaced - and it sounds better on this table than it did on my old one.
     
  8. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I have a feeling it could be the test record. I measure roughly -25dB at 1kHz with my own record aswell, despite the German Stereoplay magazine consistently measuring crosstalk close to the manufacturer's specification.
    BTW here are some of the measurements I have taken that I mentioned in my post above:

    These are entirely uncompensated and unsmoothed. I will try to get better looking graphs in the future.
    The record seems to have some LF rolloff, a 10kHz bump and is severely lacking in ultrasonic response compared to real measureements.
    Dynavector DV-20X2 L pre alignement.jpg
    Dynavector DV-20X2 L post alignement.jpg
    I have a feeling the record which I am using has poor crosstalk, especially at higher frequencies. As mentioned it's a smaller 7" record which may could play a role aswell.

    I chose the DV-20X2L as an example here because it is the one cartridge of the 4 I have measured so far where matching crosstalk at 1kHz would give a worse FR evenness. Depends on the cart.

    There was a roughly 0.7dB channel imbalance for mono content before. This was definitely audible, even on speakers! Since the crosstalk is mismatched the left channel was louder for mono (horizontal) signals, while the right channel was louder for out of phase (vertical) signals.
     
  9. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,975
    Trophy Points:
    113
    Location:
    Vancouver BC
    I'm indeed suspicious about the record I'm using. It (Popular Science Monthly Stereo Test Record 1967) wasn't designed for this sort of precision, it's more about testing whether your speakers are wired out of phase, etc. But at least all its 1kHz tracks (both channels and each channel) are actually 1kHz, not 1013 Hz like another I have, so I felt ok using it as a refinement of, rather than an alternative to, the arc protractor. Interesting that yours is only giving you around -25 dB as well.

    I don't have a pink noise track, so I can't check a full-range signal as you've done.
     
  10. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I have this same problem with a test record I got for near free aswell. Pitch is way off for the frequency it is supposed to be. Imagine someone used that record to seit their pitch and now suddenly all the records are out of pitch!
    Another one I have has more wow&flutter than the cheapest turntables. I wonder how they managed to achieve that. Fortunately most of the older records seem to be centered properly. I've had to return modern records because they weren't even centered right (causing pitch fluctuations).

    It's just a freebie from the Stereoplay magazine some months back. I'm sure there are better ones, but it seems to do the job for me for now. I think you could find it online for cheap-ish. Maybe at some point I'll look for a better one.

    Something I find interesting is that we are still getting 20dB of crosstalk in the bass. As far as bass typically getting summed to mono on vinyl, it seems that test records are an exception. And it doesn't kick the stylus out of the groove, either despite being pink noise (not white noise).
     
  11. bazelio

    bazelio Friend

    Pyrate
    Joined:
    Nov 16, 2015
    Likes Received:
    3,417
    Trophy Points:
    113
    @Biodegraded v cool.

    I've always wanted to play with AnalogMagik, but the price had always been prohibitive. TL;DR about a month ago, I was able to pick it up at a very reasonable price on the up and up, so I did. Playing with this software, adjusting the cart by the numbers, and listening to results and changes along the way has taught me more about cartridge alignment than I could have ever learned by reading the usual suspects that are out there. None more important than the interrelated nature of all parameters. I should do a write-up at some point. But probably the most surprising takeaway has been the fact that zenith angle is by far the most touchy of any alignment parameter, and that relying on common sighting techniques only gets you so far. What looks dead on through the loupe or in the mirror can very easily be well off from ideal. And a small twist of the cart can send IMD from 1% to 6%. I've managed to achieve 14% IMD at a zenith setting that only looked slightly off by eye.
     
  12. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Zenith adjustment also made more complex if we don't have a linear tracker, much less a test record where the entire side is 1kHz signal. One thing for sure, vinyl is so screwed up if going by just the numbers.
     
  13. bazelio

    bazelio Friend

    Pyrate
    Joined:
    Nov 16, 2015
    Likes Received:
    3,417
    Trophy Points:
    113
    It would be useful if they provided test tracks at both null points at least. If the numbers aren't petty close at both points, it'd be a clue that something else is off. I'd also like min() and max() functions across an entire side now, too. Given how touchy zenith seems to be, I really wonder how much variation there actually is across the entire arc. That could make a believer in tangential arms outta me, yet.
     
  14. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,975
    Trophy Points:
    113
    Location:
    Vancouver BC
    That's what most surprised and frustrated me about this experience. The tiniest twists to the zenith, ones that I couldn't visually detect had changed the relative alignment of the pencil lead I had taped to the side of the cart vs the nice long lines on the template, would send channel balance and/or phase and/or crosstalk to hell.

    I wished I'd just accepted the visual alignment and the good crosstalk result I had at the outset - but because the test track included a null point, once I started with balance & phase I figured I couldn't stop... :rolleyes:
     
  15. bazelio

    bazelio Friend

    Pyrate
    Joined:
    Nov 16, 2015
    Likes Received:
    3,417
    Trophy Points:
    113
    I hate that feeling of "god damn it, I should've just left it where it was.". Probably through a lot of f'ing around, I've gotten a lot better at this stuff. Somehow I know how to get back to where I was now and this tool can easily verify across all parameters. I decided to try Baerwald vs Loefgren B. I'd done this once before but was always uncertain if any differences I heard were due to unrelated setup aberrations. Now I feel fairly confident that, by the numbers, things are otherwise very close to equal and the alignment geometry is the most significant difference. And I'm pretty certain now that I prefer Lofgren whereas I'd always been using Baerwald in the past.
     
  16. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Regarding Zenith, I found the stereophile article @OJneg posted on changstar very illuminating. It should visually be the easiest parameter to set since the relevant parameter (stylus vs perpendicular line) is clearly visible.
    Something that annoys me about it is that the angle you see when putting the cart down on the protractor isn't the angle the stylus is in during playback. When the record isn't turning the cart body will tilt to one side after setting it down, while the record pulling on the cantilever is going to make it closer to straight during playback.

    So I usually hover the cart slightly during alignment and set it that way. After setting it like that I set anti skate accordingly, so that the stylus doesn't move to either side when lowering the cart onto the turning record. FWIW I use a custom alignment based on the aforementioned article. Overall my method has worked well for me so far, but I haven't done any measurements yet.
    Usually it just sounds clearer when aligned properly, although when it's very off distortion should be obvious. I somehow find it most obvious on bass drums and other LF content, despite the RIAA curve.

    Something that helps me a lot in making alignment easier is a small mnemonic: Set the outer null point first. Then move the cart to the inner one. If the end of the stylus is aligned with the straight line and the front is a negative angle (stylus tip is towards the left, clockwise rotation) the overhang is too small. Obviously now if it is a positive angle overhang is too large. Has helped make the process a lot less random for me.

    FWIW my dad (despite technically having better eyesight than me*) always lets me adjust it because I am super accurate with these things. The downside is that it's not something that can be done quickly.

    *I wear glasses to correct my nearsightedness, while he often doesn't for his age related farsightedness. Without my glasses I couldn't even use my phone. But with glasses I probably do have better eyesight than him.
     
    Last edited: Jun 24, 2020
  17. bazelio

    bazelio Friend

    Pyrate
    Joined:
    Nov 16, 2015
    Likes Received:
    3,417
    Trophy Points:
    113
    Do you have a link to the article?
     
  18. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
  19. bazelio

    bazelio Friend

    Pyrate
    Joined:
    Nov 16, 2015
    Likes Received:
    3,417
    Trophy Points:
    113
    I haven't seen that article before. I'll read it.

    Someone pointed me to a lecture by Wally this morning. He said the Baerwald alignment had a 2º error at the outer edge of the record and that for some reason it is mathematically “provable” that distortion from improper Zenith at the outer groove is less than at the inner groove for the same angular error. He just says it in passing and doesn’t offer anything more. It is in the first minute of this video.

     
  20. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,975
    Trophy Points:
    113
    Location:
    Vancouver BC
    Presumably because of the correlation with linear groove speed, same as distortion vs linear tracking error.
     

Share This Page