@mahurkars How are you adjusting the cartridge before doing the recordings? Just with a protractor and a microscope or with a distortion analyzer?
Our experiences might differ and maybe you are better than me with the protractor and microscope. I found that adjusting with a protractor you only adjust the case of the cartridge, but not the needle. A miscroscope comes abit closer but is not sufficient. IMD test tracks used to be a standard and these I found helpful for adjusting cartridges. Until I adjusted with the help of a distortion analyzer I was not listening to differences in cartridges but differences in tracking errors.
schiirrn, using IMD tracks should be more scientific approach and if you can provide muore information the tracks i maybe be able to try and adapt it as well.
Hi Sacheen,
I have matched both tracks in terms of amplitude, turned the second one around and put both together. This gives a better picture of the entire frequency spectrum. From the second part (at 40 seconds) you see a lot more peaks. This is because I had to amplify the signal and therefore also the noise by 20dB.
You can see that the spectrum does not increase evenly. Between 100Hz and 1500Hz the curve is nice and flat but from there it becomes bumpy. Below 100Hz the amplitude is also weaker.
Looking at the Spectral Frequency display, one can again see the amplified noise in the right part of the picture. This is nothing to worrie about.
The strange noise between 14-19kHz is gone, the cause may have been the Sansui, but I still see a band with different noise above 16kHz. I wonder if the mp3 format is at play here.
This is the Spectral Pan Display. The waveform should be in the middle. It is a bit right leaning due to a slight azimuth misalignement. At 2500Hz (at 53 seconds) things get a little out of hand. Hard to tell why.
The Spectral Phase display is pretty good but again something is happening at 2- 2.5kHz. (at 52 seconds)
I would conclude that some problems above 1500Hz need to be adressed.
Hugo
I have matched both tracks in terms of amplitude, turned the second one around and put both together. This gives a better picture of the entire frequency spectrum. From the second part (at 40 seconds) you see a lot more peaks. This is because I had to amplify the signal and therefore also the noise by 20dB.
You can see that the spectrum does not increase evenly. Between 100Hz and 1500Hz the curve is nice and flat but from there it becomes bumpy. Below 100Hz the amplitude is also weaker.
Looking at the Spectral Frequency display, one can again see the amplified noise in the right part of the picture. This is nothing to worrie about.
The strange noise between 14-19kHz is gone, the cause may have been the Sansui, but I still see a band with different noise above 16kHz. I wonder if the mp3 format is at play here.
This is the Spectral Pan Display. The waveform should be in the middle. It is a bit right leaning due to a slight azimuth misalignement. At 2500Hz (at 53 seconds) things get a little out of hand. Hard to tell why.
The Spectral Phase display is pretty good but again something is happening at 2- 2.5kHz. (at 52 seconds)
I would conclude that some problems above 1500Hz need to be adressed.
Hugo
Here's the Pan display. You can see it is a bit leaning to the right side. An azimuth correction should solve this.
Wow Netlist, thats very neat and detailed analysis.. let me review it at peace again and again and then i would like to produce a new cartridge and see if we can address some of the critical aspects of the frequency response. Please feel free to add any other observations or points that should be considered. This is really good. I am so glad to post on DIYAUDIO. Thanks Netlist.
There are far more sofisticated tools to analyse but for now you got some basic points.
There's Scott Wurcer's (R.I.P.) tool here, but I have no experience with it.
https://github.com/FidelisAnalog/SJPlot/
There's Scott Wurcer's (R.I.P.) tool here, but I have no experience with it.
https://github.com/FidelisAnalog/SJPlot/
Thanks, it sounds good I must say, nicely detailed.
I compared it with the same song from YouTube as I don't have it in my collection.
Looking at the FFT in Audition, one can see the bump around the 3kHz range and the elevated highs above 7.5kHz.
One can also see the small dip at +/-2.5kHz
The red plot is the YouTube track, the green one is yours.
The pink/blue plot reflects the highlighted part on the right.
The bump in the 3kHz range gives the recording more clarity, hard to comment on the high frequency part as my hearing stops at around 10kHz 🙂
Both these regions need most attention IMO.
I compared it with the same song from YouTube as I don't have it in my collection.
Looking at the FFT in Audition, one can see the bump around the 3kHz range and the elevated highs above 7.5kHz.
One can also see the small dip at +/-2.5kHz
The red plot is the YouTube track, the green one is yours.
The pink/blue plot reflects the highlighted part on the right.
The bump in the 3kHz range gives the recording more clarity, hard to comment on the high frequency part as my hearing stops at around 10kHz 🙂
Both these regions need most attention IMO.
Super thanks Netlist. Will start work on new one... Though its very difficult to directly target these specific areas. Will analyse wherever improvements possible and incorporate with best intensions.
For a manufacturer investing in the Analog Magik (www.analogmagik.com) package might be worth it. It gives you frequency response information, etc and makes proper adjustment based on actual output of the cartridge instead of some geometric assumptions possible. I bought that when I started building cartridges from scratch and learned a lot from it.
An alternative would be any test record with an IMD track and a distortion analyzer of your choice.
Netlist did a great job analyzing your data. A problem with targeting specific areas in a cartridge is that whatever you change will influence at least three other things. Setting up your own measurement gear will help to figure out what does what and how things influence each other.
One of the obvious issues with testing or using the Dual here is its tonearm effective mass is too low for the low compliance of the typical MC cartridge. That tonearm was specifically designed for high compliance low tracking force MM cartridges of its day. Typical MC needs a higher mass arm to put the low frequency resonance of arm/cartridge to the ideal ~10Hz -- above the warps, below than lowest musical tones.
Beyond that, I can attest from experience that 99% of Dual tonearms have bearings that cannot be tightened enough to eliminate chatter (small degree of undesired free play) without impeding its movement. It's the same problem found in a large proportion of stock Thorens tonearms of that era. And the cartridge mounting plates often get mechanically sloppy, the contacts to the arm getting corroded.
Then there are the mechanical linkages on the underside of the arm to enable the automatic cueing & arm return/shut-off. They are guaranteed to resonate, rattle, etc. This may occur at microscopic levels only -- but that's the same level where the stylus operates in the groove!
The idler drive in the dual is not a bad system per se; idlers have been shown to have better torque than belt drives and suffer much less dynamic stylus drag in highly modulated passages. BUT, the motor mounting in Duals was never great. The mechanical motor isolation from vibration is poor. When the motor is running, it is easy to feel its vibration in the top plate of the TT. This vibration will find it's way into the stylus & negatively affect data retrieval from the groove. I routinely use a stethoscope to check for vibration in TTs. The floating sub chassis systems employed by Thorens, Linn, AR, etc are tops; I rank the Dual far lower.
To reiterate, for all the above reasons & more, a Dual 1229 at this time is a seriously compromised test platform for developing a high performance MC cartridge.
Beyond that, I can attest from experience that 99% of Dual tonearms have bearings that cannot be tightened enough to eliminate chatter (small degree of undesired free play) without impeding its movement. It's the same problem found in a large proportion of stock Thorens tonearms of that era. And the cartridge mounting plates often get mechanically sloppy, the contacts to the arm getting corroded.
Then there are the mechanical linkages on the underside of the arm to enable the automatic cueing & arm return/shut-off. They are guaranteed to resonate, rattle, etc. This may occur at microscopic levels only -- but that's the same level where the stylus operates in the groove!
The idler drive in the dual is not a bad system per se; idlers have been shown to have better torque than belt drives and suffer much less dynamic stylus drag in highly modulated passages. BUT, the motor mounting in Duals was never great. The mechanical motor isolation from vibration is poor. When the motor is running, it is easy to feel its vibration in the top plate of the TT. This vibration will find it's way into the stylus & negatively affect data retrieval from the groove. I routinely use a stethoscope to check for vibration in TTs. The floating sub chassis systems employed by Thorens, Linn, AR, etc are tops; I rank the Dual far lower.
To reiterate, for all the above reasons & more, a Dual 1229 at this time is a seriously compromised test platform for developing a high performance MC cartridge.