Bill, cheer up. I don’t drop Shure as a cartridge for reading music vinyl.
My opinion is that it tracks very well on complex musical passages and this is a difficult task for a cartridge, especially so for a cheap one.
It’s good mechanics help it to behave very well with vinyl artifacts (scratches, dirt) too, another plus when listening to music.
Waveform reading recorded passages show that the control of motion with such impulsive stimuli is exemplar.
Happy Easter celebration everyone.
George
My opinion is that it tracks very well on complex musical passages and this is a difficult task for a cartridge, especially so for a cheap one.
It’s good mechanics help it to behave very well with vinyl artifacts (scratches, dirt) too, another plus when listening to music.
Waveform reading recorded passages show that the control of motion with such impulsive stimuli is exemplar.
Happy Easter celebration everyone.
George
Likewise, enjoy the holiday and thank you, GeorgeBill, cheer up. I don’t drop Shure as a cartridge for reading music vinyl.
My opinion is that it tracks very well on complex musical passages and this is a difficult task for a cartridge, especially so for a cheap one.
It’s good mechanics help it to behave very well with vinyl artifacts (scratches, dirt) too, another plus when listening to music.
Waveform reading recorded passages show that the control of motion with such impulsive stimuli is exemplar.
Happy Easter celebration everyone.
George
You are right, of course, and excellent trackability is perhaps the most important feature to get right in a cartridge IMO. My reading of the way history unfolded is that Shure vigorously pursued trackability at low VTFs, probably with the aim of making small radius styli feasible at low wear. There were other penalties though, as seems often the way in these things.
I've been thinking about the effect of coil current load on core flux, and changed my mind: I now think that increasing coil current load reduces total flux changes in the coil core. This is because the direction of flux induced by coil load current opposes the generator's flux changes. And, unlike a transformer primary, the generator's originated flux changes can't increase because they are simply set by motion.
If this is true, increasing coil load current should reduce losses, which are proportional to rate of change of flux.......if the current induced flux change is at all significant.........
LD
George,Bill, cheer up. I don’t drop Shure as a cartridge for reading music vinyl.
My opinion is that it tracks very well on complex musical passages and this is a difficult task for a cartridge, especially so for a cheap one.
It’s good mechanics help it to behave very well with vinyl artifacts (scratches, dirt) too, another plus when listening to music.
Waveform reading recorded passages show that the control of motion with such impulsive stimuli is exemplar.
Happy Easter celebration everyone.
George
As shown in posting #615 your Cart is within +/-1dB from 50 Hz to 15 Khz, which is excellent, no reason at all to drop your Cart.
Could you still respond why FR drops so sharp at 20Khz.
Is this because of the pink noise being band limited, is it your A/D convertor or is it the Cart ?
Hans
Read those top to bottom little or nothing on digital RIAA's. Two bi-quads (10 coefficients) does RIAA to .0001dB or better I don't see the point in using anything else.
For phono I would now use the SSM2219 or a THAT equivalent.
Du.uh! It might be in his Phono Transfer System thread ... Nope! as far as I can see. I'll ask him directly.
That has good stuff on rewiring arms for balanced too.
I couldn't find them either in the 2 threads I linked to. Sorry Bill. But that new thread might be better if you wanted to build something.
Can't find AD SSM2219. Do you mean 2019? THAT's (pun intended) not very good for MM (2pA/rtHz). Your AD624 would give better 'real life' noise with MM. Wayne found an INA using NJM2068 best with his MM cartridges.
Last edited:
You can't use a cheapo or expensive L meter to get anything other than a ball park figure for cartridge inductance.
If your expensive L meter has 2 test frequencies, you'll get 2 different values.
You have to measure with a frequency sweep and 2 channel analyser to get the impedance & phase.
Then you fit your 'linear' modelled 'cartridge impedance' to these parameters and you'll find some of your discrepancies go away.
The remaining errors are your 'non-linear' bits which might include eddy currents & flux saturation stuff. (I don't think flux saturation should be a problem with cartridges in use ... but it might with measurements
Julian did all this before he gave up and introduced his 'fudges' for speaker voice coils.
He encountered this when trying to measure speaker impedance accurately to do xover optimisation. As his model and fudges are still used in the better speaker design & measurement packages, it looks like no one has produced a better model in the last 2.5 decades.
George, if the aim of the exercise is to isolate & simulate the electrical loading of a cartridge, you still need to measure the complex impedance of the cartridge directly.
I have some comments on Wayne's 2016 design but will put that into the digitizing vinyl thread to try and tempt Richard over there 🙂.
I am expecting the twisted pair boards in soon. I bought two in case someone wanted to try it or measure it. (9 weeks left so I won't get that much done). Given the somewhat edge case of balanced transimpedance MM I guess I ought to check if there is such a person out there before I solder it up.
I am expecting the twisted pair boards in soon. I bought two in case someone wanted to try it or measure it. (9 weeks left so I won't get that much done). Given the somewhat edge case of balanced transimpedance MM I guess I ought to check if there is such a person out there before I solder it up.
Err.rh! This pic is from Holman's 1977 re-packaging of the SHURE data. It shows vinyl playback has HUGE slew and level dependencies including response, phase, THD bla bla ... AT LEAST 20dB above any electrical stuff including the cartridge core. Remove RIAA EQ to your taste but the picture doesn't really change.
I'm not sure agonising over purely electrical parameters is productive ... except for making sure you get sorta flat frequency response.
__________________
As I said, measure the whole shebang with your flat (or not) preamp and apply EQ as required. In da days of V15 II Improved (which had improved trackability at the expense of a mid droop) it was common to reduce the RIAA 75us to 50us to compensate.
For this you need a trustworthy Test Record. We used the CBS and B&K. Not sure about modern ones today but I would suggest those from cutter & record makers like the Ortofon & JVC would be more trustworthy.
____________________
On pink noise .. it was usual to brickwall bandlimit pink (or white) noise used for audio measurements to the top of the '20kHz 1/3 8ve'. This was cos the noise was used for sensitivity specs. The bottom of the '20Hz 1/3 8ve' was supposed to be limited too but this usually didn't make much difference (possibly cos most 'pink' noise generators had to level of below 20Hz)
Can't comment on George's other plots except that he must be analysing & averaging very looo.oong Pink Noise tracks to get such smooth results .. much smoother than his 'swept sine' measurements. 😱
Attachments
Last edited:
Yes 2019, every cart is different I use MI you need to do the integrated noise probably with some perceptual weighting for each case before you decide. I used the AD624 with MC for years at 4nV the surface noise usually was still dominant (or at least the SNR was a don't care), similar experience to folks using paralleled 6DJ8's for MC.
Richard I still have not gotten a clarification on your ancient MC head amp, for some reason I associate your noise claim with something like 250uA of supply current, this is physically impossible. I am probably mixing up other versions of this topology.
Last edited:
Actually the flux supports voltage.
If you feed into a virtual earth transimpedance amp, your voltage is very small and THD due to flux is minimal.
Great Guru Baxandall used this in a couple of specialised balanced line applications to use very small transformers for very high 'levels' & low THD.
Things get out of hand and sensible consumption limits are traditionally trashed early on in an Easter fest. Spin lamb and kokoretsi are the serial killers.
Tzatziki is the most benign of all the consummated [incidently (?) garlic forces the blood pressure remain low]
Don’t ask on the level of reflex responses among the guests late in the afternoon.
https://www.thespruce.com/greek-easter-foods-1706214
You’ve mentioned a Greek neighbor in the past, you may have direct experience. 😀
I'm trying to recover, post with technical content is postponed for later.🙄
George
I'm envious "Greek haggis" never turns up around here. Visiting Chicago we found a very nice family owned restaurant with good versions of the usuals like lamb shank, etc. and some modernist Greek wines that were pretty nice. When in college it was nothing at all but Mavrodaphne and Retsina.
When (most) test records only produce pink noise up to 20 Khz, they are not well suited for the experiments in this thread, because especially the region above 20Khz is where the Aurak should show it's strength.
And the region from 10Khz to 20Khz will be affected in amplitude because of this filtering and therefore less accurate. So the question is, is there really a peak at 20Khz or is it an artefact of the noise filter.
Looking at the pink noise plots showing a filter bandwidth of ca. 2Hz, FFT timeframe will have to be 0.5sec and to get results as smooth as he showed, averaging over at least 20 FFT's.
So the track must be at least 20*0.5sec = 10 seconds. In practice you will need twice this time, which seems more than O.K. for a noise track and not even so very long.
Hans
Hans, I'm reminded that you had knowledge of a DBLT conducted this Millenium of Bandwidth Limiting where those who could tell the difference (ie not deaf) preferred the wide band version.
As far as I know, ALL other tests, in this & the previous Millenium show the true golden pinnae (ie not Audiofools) prefer the bandlimited versions.
Can you post links to your tests please?
Apologies for being out of thread but it has some bearing on the subject. Cartridge resonances might be a major factor in this preference. 🙂
As far as I know, ALL other tests, in this & the previous Millenium show the true golden pinnae (ie not Audiofools) prefer the bandlimited versions.
Can you post links to your tests please?
Apologies for being out of thread but it has some bearing on the subject. Cartridge resonances might be a major factor in this preference. 🙂
Pink noise from HFN record (HFN Side 1 Band 3) is 60s long
Pink noise from UAT record (UAT Side 2 Track 2) is 120s long
Full range freq sweep from HFN record (HFN Side 2 Band 2) is 18s long
1KHz to 20KHz stepped freq sweep from UAT record (UAT Side 1 Track 6) is 33s long.
Hans I think it is the bandlimiting of the pink noise (hence the sharp peaking before the cut off).
If it was due to any of the components, it would show equally at the full range frequency sweep, but in there, the cutoff is more benign and no sharp peak.
George
Sorry I missed this Scott.
The original July 81 Wireless World circuit was 0.5mA for 52nV over 15.7kHz optimised for 20R eg a Denon MC but pretty poor devices.
The 0.28nV/rtHz version was 3mA and optimised for Ortofon MC20/10, 4R3 or less.
Beyond 3mA, there was no improvement and that's quite a good indication of rbb' for noise.
Wayne says there's no improvement with his circuit beyond 5mA with ZTX851/951 which means their rbb' is significantly less than the Hitachi/Toshiba/Rohm devices from Ye Olde Unobtainium Shoppe
No magic .. just VERY close to what's theoretically possible.
The 15.7kHz bandwidth is cos that's the brickwall equivalent of 6dB/8ve -3dB @ 20kHz
Sort of, I think.......
Flux is a consequence of either permanent magnetism, as in the case of a cartridge, or (optionally) current in its coil. In a cartridge generator, an armature moves so as to cause flux changes for coil, and this invokes an emf.
The emf is internal to the coil, so remains even if a short circuit is applied to the coil pins: an external current then flows, determined by the coil's impedance, and a mechanical force is applied to the armature which opposes motion, albeit typically small in the scheme of things.
External current, in turn, invokes a flux within the coil which opposes the change that caused it, ie motion of the armature. Assuming the armature continues to move in the same way, the net flux in the coil is therefore reduced.
Inductance of the coil L can be defined as dΦ/dI, the instantaneous ratio of flux change to current change, BTW.
Since coil losses such as eddy currents mostly follow net flux change in the coil's highly permeable core, increasing current through transimpedance loading, for example, reduces losses. Since these losses also mostly follow slew rate, ie programme level slew rate, in principle transimedance preamps offer reduced distortion and extended frequency response, especially for high slew rate programme material.
So that is the thesis, and the claim is benefit is not just in f response above the audioband by elimination of the LCR resonance, but also to slew rate performance and distortion by changes to magnetic operating point.
LD
Great Guru Baxandall used this in a couple of specialised balanced line applications to use very small transformers for very high 'levels' & low THD.[/QUOTE]