For the frequency response, it does not matter, an amplifier with a current input or voltage input. Only the complete impedance of the input circuit is important, including the internal impedance of the cart. It seems that there is no reason for the appearance of non-linearity of the frequency response for MC when changing the load resistance. Even zero input resistance will not affect the frequency response in the audio range. It is not clear where the graphs came from, which are the link above.
Gotta say, these graphs, given without even mentioning the brand and model of cartridge tested, are so far outside of anybody else's experience, or any explanation, as to be, let's just say, questionable. But "On the Internet nobody knows that you're a dog." Just don't believe everything that looks like data - sometime it's not real.
All good fortune,
Chris
I think the way the cart behaves with load capacitance//with load resistance in the two modes will affect the response differently. For lower Rcart into a transimpedance stage, you will have more current to charge/discharge load capacitance.
The only correct way to assess the difference would be with a test vinyl and some measurement software. I do know that MC carts work exceedingly well in the current mode regime.
I think there may be arm resonance involved in the graphs plotted earlier - @billshurv and @gapag curated a lot of discussion on this a year or two back - I am not at all expert on the mechanical side. I suspect the RC networks are simply damping the peaky response.
Looks like the graphs are from Tom Evans. I had a look at his website. Lots of product claims but no specs or measurements from what I can see so we can only guess at the provenance.
well it was george and @Hans Polak who did the hard work, I just supplied a pile of cartridges. Let's just say some things that had been taken as gospel were not 🙂
This is true, but for the MC cartridges there should not be a problem with the load capacity (this is mainly cable capacity).
Another thing is for MM cartridges, there the transimpedance amplifier allows you to completely eliminate the effect of the cable capacity. But another problem arises - L/R pole lies in the middle of the sound range.
Yes.
The resonance of the arm lies at very low frequencies. Near the upper edge of the sound range, a different resonance is usually observed - the given mass of the moving system and the elastic modulus of the vinyl. It is surprising not a resonant peak, but a fall in the level after 2 kHz.
Sorry, you are right on the resonance - I meant cantilever resonance and not arm resonance.
I have not looked at transimpedance MM designs - I believe Bob Cordell did one published in Linear Audio a few yrs ago. My only concern with it is that it has to be tuned for the cart. With MC current injection you don’t have this challenge.
I have not looked at transimpedance MM designs - I believe Bob Cordell did one published in Linear Audio a few yrs ago. My only concern with it is that it has to be tuned for the cart. With MC current injection you don’t have this challenge.
I desperately wanted transimpedance MM to be the holy grail solution, but sadly it ends up being just as complex as plain old voltage mode when you take actual cartridge impedance model into account. It did require Hans to beat me over the head with the maths to actually realise I was being deluded!
Yes, a transimpedance amplifier for ММ must be tuned for the cart. Therefore, such a design will hardly be common. But I plan to do as an experiment. There is a topic about a similar Aurak project on this forum.
I would recommend you read the paper here https://www.diyaudio.com/community/...ts-with-individual-transfer-functions.397815/ It goes into why this is more effort that it's worth. I got as far as building a set of boards to do it then realised it was a gigantic pain unless you only ever used one cartridge or kept a VNA on your bench.
TIA requires adjusting only one constant (L/R) for the selected cartridge. The exact equivalent circuit of the cartridge affects only high frequencies, the same for a TIA and conventional amplifier. In reality, everything is bad with the frequency response at high frequencies, since the exact capacity of the cable is unknown.
VinylTrak is not TIA.
VinylTrak is not TIA.
Yes, there is a great difference between TIA inputs and simply deriving the 75us pole from loading LR.
All good fortune,
Chris
All good fortune,
Chris
If we're still talking about a MM cartridge feeding a TIA phono equalizer, seems like there are at least two different possible ways to generate the zero impedance junction, and with different outcomes. Reminds me very much of the issues around the summing junction of a DAC.
One possibility is a simple brute force 10R across the phono input, followed by makeup gain. This will give the raw mechanical response without significant editing by the cartridge's half Henry inductance and various capacitances, all significant within the audio band. Just like a low impedance MC cartridge does. We would see a falling of high frequency response caused by scanning losses and a peak near or often within the audible band caused by the stylus effective mass x vinyl compliance resonance. A Leach style input, without feedback to the "summing junction" would behave like that too.
Another possibility is to bring lots of loop feedback to the input junction, giving a very complicated response involving the cartridge's impedance and the feedback's attempt to both acknowledge that and provide RIAA - quite the task.
I suspect you gentlemen may be thinking about these very different possibilities, perhaps causing confusion.
All good fortune,
Chris
One possibility is a simple brute force 10R across the phono input, followed by makeup gain. This will give the raw mechanical response without significant editing by the cartridge's half Henry inductance and various capacitances, all significant within the audio band. Just like a low impedance MC cartridge does. We would see a falling of high frequency response caused by scanning losses and a peak near or often within the audible band caused by the stylus effective mass x vinyl compliance resonance. A Leach style input, without feedback to the "summing junction" would behave like that too.
Another possibility is to bring lots of loop feedback to the input junction, giving a very complicated response involving the cartridge's impedance and the feedback's attempt to both acknowledge that and provide RIAA - quite the task.
I suspect you gentlemen may be thinking about these very different possibilities, perhaps causing confusion.
All good fortune,
Chris
I was under the impression VinylTrak was TIA - that said I have not read Bob’s article, just what Ive heard.
Standard voltage amplifier for MM gives plenty of flexibility and means any cart can be used which is why I stuck with it. For TIA approaches, you have to have a setup for each type of cart.
For MC, you don’t have the high cart L and R so TIA make a good alternative that brings noise benefits.
Standard voltage amplifier for MM gives plenty of flexibility and means any cart can be used which is why I stuck with it. For TIA approaches, you have to have a setup for each type of cart.
For MC, you don’t have the high cart L and R so TIA make a good alternative that brings noise benefits.
Last edited:
VinylTrak is an improvement, made to optimize S/N ratio, to an idea from (at least) the 1970s National Semiconductor Handbooks, where the cartridge's inductance and the phono preamp's load resistance make the 75us RIAA pole. Load resistance needs to be different for each cartridge model's inductance, and usually ends up somewhere in the several K Ohm range. Could be done with a pot I guess, which would also be a shelved treble control. Give the pot a fancy name, and you're famous.
Cordell improved that by setting the LR pole around 8kHz and incorporating a correcting pole/zero later in the amplifier, for best weighted (and that's a very interesting part of the article) S/N. But none of this is really related to what you're doing - these are still high Z voltage sensitive inputs but with most of the cartridge loading sensitivity removed.
These give the raw mechanical response of the cartridge, like a MC does, and lack the fine tuning for flatness possible in the "old-fashioned" high Z high touch method, as you say, so are a matter of taste. Everything's a tone control.
All good fortune,
Chris
Cordell improved that by setting the LR pole around 8kHz and incorporating a correcting pole/zero later in the amplifier, for best weighted (and that's a very interesting part of the article) S/N. But none of this is really related to what you're doing - these are still high Z voltage sensitive inputs but with most of the cartridge loading sensitivity removed.
These give the raw mechanical response of the cartridge, like a MC does, and lack the fine tuning for flatness possible in the "old-fashioned" high Z high touch method, as you say, so are a matter of taste. Everything's a tone control.
All good fortune,
Chris
Such a load will give a strong drop in the frequency response due to the action of the pole of the L_cart/(R_cart+R_load). This will also give a strong drop in the signal level and the fall of the signal/noise ratio to the unacceptable.
If the zero input resistance is provided with a feedback, it will be a current-to-voltage converter. The frequency response will be the same as in the case of a low R load.
The "old-fashioned" adjustment method has very limited possibilities. The pole of the L_cart/(R_cart+47kOhm) and the cartridge resonances are trying to compensate for the resonance chain formed by L_cart and C_load (C_cable plus C_input). This can only be done approximately, and usually it is done blindly, switching the input C. Moreover, nothing is really noticeable by ear, because we are talking about frequencies above 10 kHz.
A good option that Nick Sukhov proposed is to increase R_input, which will shift the pole above the sound range. But the capacity of the input circuit should not be large, which requires the installation of the RIAA-amplifier inside the turntable. And this is a serious restriction. Another very good idea is the so-called "active shell", where the preamp is located next to the cartridge.
Last edited:
I don't think so. The problem is that the transimpedance approach, when you take everything into account you end up with something more complex than old fashioned voltage amplification. The gotcha is the rising response above 10kHz.
Vinyltrak does look nice, but when you put a real cartridge model on it becomes more tricky AND it is noisier.
We need to make a clear distinction here between MC and MM. MM seems to me to be a much more complex beast when not used in the manner in which the industry has settled on i.e. 47k load with some cable capacitance.