Pardon if this sounds weird. (Newbie here
) How about having a stereo tone track with one channel out of phase* then we add left and right channel what will remain is mechanical resonance.
* Like anti noise, anti phase. which can cancel signal
Regards.
There are probably only three significant contributors to the audioband f response profile of MM carts; or 2 for MC.
1. The generator's transfer function, Vout or Pout versus frequency for a given velocity.
2. The mechanical top resonance, and perhaps an associated mid-band dip.
3. Electrical resonance LCR, of coil leakage inductance and loading CR, which isn't an issue for MC.
However 3 is an awkward number, because to examine any one contribution really means observing it in isolation, and two of them are tough to eliminate or characterise.
Electrical LCR resonance for MM can be eliminated by transimpedance loading. Generator transfer function tends to have 1st order roll-off, if it's there at all in the audioband. Then, for better carts with good generators and transimpedance or MC loading, what one is left with is, by elimination, the mechanical resonance.
I think the mechanical resonance is irreducible, in that it's impossible to eliminate from realistic tests involving movement of a stylus. It follows that the generator's transfer function is tough to isolate from mechanical resonance too.
My experience of the better AT and Ortfon MM carts is that their 1st mechanical resonance appears to lie in the 20-30kHz range, and have a reasonably low Q, perhaps a peak of 3-6dB or so. But, in my archive, I can't find any plots that show this though there were some posted on the web a few years back.
LD
1. The generator's transfer function, Vout or Pout versus frequency for a given velocity.
2. The mechanical top resonance, and perhaps an associated mid-band dip.
3. Electrical resonance LCR, of coil leakage inductance and loading CR, which isn't an issue for MC.
However 3 is an awkward number, because to examine any one contribution really means observing it in isolation, and two of them are tough to eliminate or characterise.
Electrical LCR resonance for MM can be eliminated by transimpedance loading. Generator transfer function tends to have 1st order roll-off, if it's there at all in the audioband. Then, for better carts with good generators and transimpedance or MC loading, what one is left with is, by elimination, the mechanical resonance.
I think the mechanical resonance is irreducible, in that it's impossible to eliminate from realistic tests involving movement of a stylus. It follows that the generator's transfer function is tough to isolate from mechanical resonance too.
My experience of the better AT and Ortfon MM carts is that their 1st mechanical resonance appears to lie in the 20-30kHz range, and have a reasonably low Q, perhaps a peak of 3-6dB or so. But, in my archive, I can't find any plots that show this though there were some posted on the web a few years back.
LD
Last edited:
6dB at 30kHz I can live with. For me the important part is levelling the playing field. Looking at the Miller results its clear that standard MM loading is just not good. And if the AT440 distortion figures are as good as they seem, we are tipping the balance away from MCs with their rare wood or machined from solid bodies and shiny bling .
Still confused about that double peaked distortion tho...
. Still confused about that double peaked distortion tho...
Apart from as IM products, harmonic distortion above 10kHz isn't audible in any event.
Apparently elevated measured hf distortion, such as the double peak perhaps, is probably a measurement artefact due to the cartridge actually being able to reproduce 2nd harmonic distortion at ultrasound. Esp if one of the resonances cropping up in ultrasound, eg 25-30kHz which would be 2nd harmonic of 13-15kHz, or 3rd harmonic of 8-10kHz.
I previously posted measured OM40 harmonic distortion up to 10kHz, which is impressive, into a transimpedance preamp. I figure that's a good indicator, and I don't sweat about harmonic distortion above 10kHz. If there's a hf gross-out, which there isn't IME, it would show up as mistracking which is obvious for its 4-8kHz signature, since that is the typical range of mechanical resonance for a free swinging stylus at the end of a cantilever - 1/4 wavelength mode which is only available in mistracking.
Hope this makes sense !
LD
It would, were it not for the fact that the lower peak in distortion is right in the midrange.
(I would post the graphs but they are copyright). For the AT440 you bumble along well below 1%d distortion until 3kHz. 1.8% at 5kHz, 1% at 12kHz then it rises back up again.
For the OM black left and right channels are so different I suspect a problem somewhere, but it also has a peaks at 5 and 15kHz. MCs don't show this midrange hump. I'll look at some more reports to see if I can find a 3rd data point but for now there is something odd I don't have even a conjecture for.
EDIT: more digging done and have found a cartridge that doesn't have a circa 5kHz distortion peak, which is the OM blue. That rises to around 4% at 11kHz then drops like a stone. Now I would have expected the blue and black to be very similar in that respect, but clearly not.
For me 7% THD at 5kHz (clearaudio concept) is a bit high, but not a patch on the grado prestige gold which goes over 10%
(I would post the graphs but they are copyright). For the AT440 you bumble along well below 1%d distortion until 3kHz. 1.8% at 5kHz, 1% at 12kHz then it rises back up again.
For the OM black left and right channels are so different I suspect a problem somewhere, but it also has a peaks at 5 and 15kHz. MCs don't show this midrange hump. I'll look at some more reports to see if I can find a 3rd data point but for now there is something odd I don't have even a conjecture for.
EDIT: more digging done and have found a cartridge that doesn't have a circa 5kHz distortion peak, which is the OM blue. That rises to around 4% at 11kHz then drops like a stone. Now I would have expected the blue and black to be very similar in that respect, but clearly not.
For me 7% THD at 5kHz (clearaudio concept) is a bit high, but not a patch on the grado prestige gold which goes over 10%
Last edited:
Methinks it can't be real, because such things would sound awful for programme material but they don't, quite the opposite.
Here's a guess: the test record was RIAA encoded, so 0dB@5kHz ref 5cm/s@1kHz would be extreme in terms of groove angle, curvature and pinch effect.... and would only crop up rarely in programme material at peak levels ?
Perhaps the test record then had a step down in level above 5kHz, and another at 10kHz? Alignment then critical ?
See also my OM40 harmonic distortion measurements with transimpedance preamp, which is for gentle levels, and shows none of this malarkey !
LD
PS: If the f response really were as poor as those measurements suggest, in terms of hf roll-off, how could measurements of HD (relying on measuring content well above 18kHz) be made ? I wonder if there might be some kind of 'correction' for hf roll off which is mucking up HD figures ?
But I shouldn't lose sleep.......... I believe the 2M Black is a close cousin of the OM40, and the AT440Mla is also very fine into a transimpedance preamp IME, both by measurement and sound, IME
LD
If the measurements were comparable between MM and MC I would agree completely, but MCs don't show the same shape. I don't want to mistrust his tests as he's being doing it a long time but some of the graphs should be 'woah, that's wrong, better check'. I need to go look at your measurements again.
BTW found your tracking analysis as to why the Serato might be a good thing the other day.
BTW found your tracking analysis as to why the Serato might be a good thing the other day.
Thank you bondini for the measurements.
Here are some Impedance sweeps (22Hz-22KHz) using Speaker Workshop SW.
Data imported into Excel and smoothed exponentially with a D.F. of 0.9
Impedance measuring standard jig
Ref Resistor: 1KOhm
R coil measured for each cart.
Loading is the input impedance of the audio card, which together with the jig wiring is 10KOhm//110pF
George
Here are some Impedance sweeps (22Hz-22KHz) using Speaker Workshop SW.
Data imported into Excel and smoothed exponentially with a D.F. of 0.9
Impedance measuring standard jig
Ref Resistor: 1KOhm
R coil measured for each cart.
Loading is the input impedance of the audio card, which together with the jig wiring is 10KOhm//110pF
George
Attachments
One needs to consider electrical impedance to be 'complex' in the mathematical sense, when either or both inductive and capacitive elements are in play, if one is to describe them so as to meaningfully work with them here. For example, to consider the ratio of voltage to current, ie impedance, which is a 'complex' parameter, or so as to work out power transfer function of the cartridge generator.
So be careful out there if dealing with complex variables as magnitudes. This approach has limitations which might otherwise mislead............
IIRC, I noticed previously that typical MM carts appear to behave as a generator as though inductance was somewhat different (smaller?) than that measured by simple meters..... perhaps a factor of 10-20% from memory.... that might fit with some of these recent measurements ?
Otherwise, IME the output impedance of a typical MM cartridge within the audioband closely follows an LR law for practical purposes. Generator transfer function, which typically follows a 1st order roll-off, and which, for better carts does not appear in the audioband or reasonably above it, is super-imposed on this.
LD
So be careful out there if dealing with complex variables as magnitudes. This approach has limitations which might otherwise mislead............
IIRC, I noticed previously that typical MM carts appear to behave as a generator as though inductance was somewhat different (smaller?) than that measured by simple meters..... perhaps a factor of 10-20% from memory.... that might fit with some of these recent measurements ?
Otherwise, IME the output impedance of a typical MM cartridge within the audioband closely follows an LR law for practical purposes. Generator transfer function, which typically follows a 1st order roll-off, and which, for better carts does not appear in the audioband or reasonably above it, is super-imposed on this.
LD
Agreed. Any complication might arise if one were to treat magnitude or modulus as 'real', which isn't nec so, of course. But George hasn't done that in any event. I was trying to get my head around, and interpret Bondini's posts, when I thought to mention the general point.
LD
The AT440MLa is a well-designed, well-made cartridge.
For what it's worth, my take on the impedance measurements for the AT440MLa up to 20 kHz are very consistent with modelling it as a generator, resistance and inductance, all connected in series.
In my view, neither the impedence data nor the Miller Research output data show artefacts of mechanical resonance in the bandwidth up to 20kHz. That is not to say that there aren't such things above 20kHz, but that is beyond the capacity of my hobbyist instruments.
Slightly off-topic: The AT440MLa's elevated output at high frequencies is well known. It can be ameliorated somewhat by altering the loading on the cartridge (yes, I have tried). But, as Werner Ogiers showed in his excellent 2008 TNT Audio article (Load the Magnets!!! - [English]) it is a process of compromise. As he puts it:
For what it's worth, my take on the impedance measurements for the AT440MLa up to 20 kHz are very consistent with modelling it as a generator, resistance and inductance, all connected in series.
In my view, neither the impedence data nor the Miller Research output data show artefacts of mechanical resonance in the bandwidth up to 20kHz. That is not to say that there aren't such things above 20kHz, but that is beyond the capacity of my hobbyist instruments.
Slightly off-topic: The AT440MLa's elevated output at high frequencies is well known. It can be ameliorated somewhat by altering the loading on the cartridge (yes, I have tried). But, as Werner Ogiers showed in his excellent 2008 TNT Audio article (Load the Magnets!!! - [English]) it is a process of compromise. As he puts it:
Both electrical and mechanical aspects must be incorporated, and this means measuring the real thing, or failing that, listening to it. Or vice versa of course.
Agreed, and well put.
Did you get as far as trying transimpedance loading, bondini ? That cart's a good candidate, and the boxes you identify are those that need to be ticked for this to work out well. After that, TI is the way to fly with MM carts that can do it, it's sort of rude not to IMO.........
LD
Last edited:
Which is more or less where I came in. I bought an AT150MLx when they announced they were phasing them out, then did some research and discovered they were finiky over capacitance. Luckily SY had done his EO balanced MM stage with only 22pF input capacitance for exactly this cartridge so I got some boards. Flat to 20kHz no problem.
Then I noted the work of Cordell and his loading the cartridge to provide the 75uS pole (or part of it depending on your taste). This then lead to discovering the work of Barney Oliver and Van Raalte. Then just as I had it straight in my head LD goes and throws a curve ball wiht Transimpedance loading. Too many choices... arghh.
But I do now have 2 of the best MM cartridges of recent years (AT150 and SuperOM). Whilst I'll probably never enjoy a V15V or one of the beryllium technics I think I have all I need for ongoing musical pleasure and experimenting. It's also been interesting (and sometimes saddening) researching this. After 30 years of only MC, discovering the web of confusion and downright untruths has been enlightening. Saddening because its clear nearly all manufacturers of MM phono stages (and there are a lot) are just following the same wrong path. No one has done their research leaving DIY to pick up the challenge. Doubly sadding because earnest people with more enthusiam than scientific knowledge are changing cartridges to get around limitations in their electronics. Every week you read threads like that on other forums (fora?). people are living with dips in the midrange and overblown treble because they don't know better.
Which leads to the 'load the magnets' article. Its wrong in a number of places, but I cannot blame the author as the truth is hard to pan from the dirt and few have the easy access to AES papers to realise this had all been sorted 40 years ago and then forgotten. I've certainly been forced to go back to first principles on a few things on this thread, which is good as I'd been forgetting how to think
I do think, if properly interfaced, The 440MLB is arguable the best value neutral cartridge out there. At £140 I have not found anything that will come close. And I understand that you can get a modders delight DL-103 for the same money.
And once I get a suitable phono in place the 150MLx may end up on the Roksan...
Is there a separate thread with details of your Transimpedance MM pickup preamp?
This was fairly common in da 60's with valves but of course has to be matched to the cartridge.
Transimpedance is one of the classic LN configurations that can approach the theoretical limits for noise.
__________________
Finally managed to get registered on milleraudioresearch.com
I'm suspicious of his THD plots of phono cartridges.
THD of a vinyl playing system goes up at about 12dB/8ve with freq. especially at high level. The old gurus were well aware of this.
That's why the better valve preamps like LEAK & QUAD had variable hi-slope HF filters. The idea was to cut out more THD than signal.
If the THD goes up and then goes down with freq., there is another bandwidth limiter in the system.
As a final check on this, I've conjured a controlled way to get rising THD with frequency & level.
It re-introduces a lot of vinyl sound on my own nice clean digital recordings.
Last edited:
Thanks again Hans and Lucky for the comments.
With post #270 I wanted to show the Impedance modulus trend and the observation that the modulus is not affected almost a iota from the finger damping of the cantilever or from the total absence of the moving magnet (another indication of the lack of reciprocity due to very weak magnetic coupling)
I withheld the complex data for a bit later. But since you asked…
I have only “touched” the phase data below 300Hz (too rough variations due to only few sampling points –thirteen- below 300Hz). Otherwise, data is legit.
I opted for log vertical scale, so that differences btn the modulus, real and imaginary parts are clearly visible.
The drawback with the log scale is that it can’t display the negative values of the imaginary component at high frequencies.
Shure turns negative above 17780Hz, AT above 19550Hz and Stanton stays healthily positive up to the measurement limit of 22KHz
George
With post #270 I wanted to show the Impedance modulus trend and the observation that the modulus is not affected almost a iota from the finger damping of the cantilever or from the total absence of the moving magnet (another indication of the lack of reciprocity due to very weak magnetic coupling)
I withheld the complex data for a bit later. But since you asked…
I have only “touched” the phase data below 300Hz (too rough variations due to only few sampling points –thirteen- below 300Hz). Otherwise, data is legit.
I opted for log vertical scale, so that differences btn the modulus, real and imaginary parts are clearly visible.
The drawback with the log scale is that it can’t display the negative values of the imaginary component at high frequencies.
Shure turns negative above 17780Hz, AT above 19550Hz and Stanton stays healthily positive up to the measurement limit of 22KHz
George
Attachments
Luckily SY had done his EO balanced MM stage with only 22pF input capacitance for exactly this cartridge so I got some boards. Flat to 20kHz no problem.
You might want to check out the Leach JFET preamp. Dr. Leach advertised 20 pF input capacity. I bought one in the 1970s for my Shure V15VMR when I was trying to minimize the capacitive load.
http://www.diyaudio.com/forums/solid-state/51367-leach-wide-bandwidth-preamp.html
You might want to check out the Leach JFET preamp. Dr. Leach advertised 20 pF input capacity. I bought one in the 1970s for my Shure V15VMR when I was trying to minimize the capacitive load.
http://www.diyaudio.com/forums/solid-state/51367-leach-wide-bandwidth-preamp.html