Michael Fidler Spartan 20 Phono Stage Review

[Multicore]

I will never understand why people prefer MC over MM.
Don't all these measured Phono stages perform better with MM?
Don't we want better SINAD at all times?

Perhaps it is analogous to the preference for vinyl itself. Fussy, inconvenient, expensive, and delicate equipment with bad sound quality and nostalgic vibes.

 

[mike70]

Perhaps it is analogous to the preference for vinyl itself. Fussy, inconvenient, expensive, and delicate equipment with bad sound quality and nostalgic vibes.

people listening at hi-res streaming says the same about CDs ... what a coincidence

let me guess something ... maybe personal preferences aren't always guided by technical properties or what we think it must be ... something horrendous. I'm glad you're in the right way, be proud.

 

[Mnyb]

I think mr fidler has his own ideas about mm cartridge loading ? Hence why there are no capacitance settings.

Could @Michael Fidler perhaps elaborate ?

 

[levimax]

I will never understand why people prefer MC over MM.

The biggest advantage of MC carts is that they have very low inductance so the FR is very insensitive to cartridge loading. MM can have very flat response but it usually takes some fiddling around with the loading and taking measurements where as MC are much more "plug and play" (relatively speaking).

 

[welwynnick]

The biggest advantage of MC carts is that they have very low inductance so the FR is very insensitive to cartridge loading. MM can have very flat response but it usually takes some fiddling around with the loading and taking measurements where as MC are much more "plug and play" (relatively speaking).

Yes, but against that you often have to tune the performance of a MC cart with the impedance. I tend to like a low input impedance.

 

[Michael Fidler]

I think mr fidler has his own ideas about mm cartridge loading ? Hence why there are no capacitance settings.

Could @Michael Fidler perhaps elaborate ?

I don't think it's a good idea to include these, as they do confuse people a bit, and can also be conflated with the total system capacitance which would lead to poor results.

The small capacitor that yields most of the input capacitance of 120pF is 100pF, and has to be connected after the RF stop resistor of 100 ohms or so to prevent radio-frequency interference in the VHF band from getting onto the input where it would be envelope-detected as audible hash. If it were to be routed through a switch, the extra inductance would make it a less effective shunt.

A lot of the preamplifiers advertised as 100pF are in fact 120pF or more, as you've got to factor in the PCB track-to-ground capacitance, op-amp input, and so on...

Adding more capacitance can be done quite easily, but adds complexity to the design, and only really affects the response above 10kHz by a few dB at most. Most modern cartridges work best with about 200-300pF, so I don't really see a need for it if we have 120pF from the preamp and then another 80-150pF from the tone-arm/cabling. Adding it would just increase the size/complexity of the board, enclosure, generate long e-mail exchanges about setting it, give people the opportunity of setting too much of it and complaining etc.

If it's not necessary, then it's best avoided. Keep the phono-stage simple so it can be plugged in and work straight away with the vast majority of devices, with the loading question completely out of mind.

The same goes for MC loading, which doesn't even really affect the frequency response until it drops below the coil resistance or less. You need a low enough resistance value that lets the resistor dissipate the resonance of the cartridge inductor against the load capacitance. The load capacitor selected for its highest value such that the appropriate resistance is about ten times the coil resistance to keep insertion loss under 1dB. If you design the input amplifier correctly, then the thermal noise of coil resistance will dominate if it increases beyond 30 ohms, such that the noise penalty from insertion loss is mitigated for higher coil resistance.

If you let the user select the MC load resistance value, then they'll either increase insertion loss and harm the signal-to-noise ratio of the device, or create an underdamped RF peak at the coil-inductance/load-capacitance resonance point that will be detected on the input and lead to complaints. Changing the loading alters the level at the input slightly, but perhaps enough for people to convince themselves they have a 'preferred' amount that will probably provoke trouble... Most preamps on the market use pretty brutal (and rather noisy) resistive series RF stopper networks on the input amplifier, preventing the resonance from getting through, at a massive noise penalty which is only acceptable because the market has such low expectations with the intuitive (but incorrect) explanation that the level is lower and therefore noise is correspondingly higher.

While selectable loading is very easy to implement, it is not necessary in my opinion and actually highly undesirable from a customer-facing point of view.

 

[levimax]

Yes, but against that you often have to tune the performance of a MC cart with the impedance. I tend to like a low input impedance.

It takes a big change in impedance to affect the FR of an MC cart and is generally not recommended as you can run into other issues by taking things to extremes. Running low input impedance makes the cart "work harder" (it takes more mechanical force to move the stylus) which in general is a "bad" idea for a stylus but it could possibly sound better to some people in some cases. With MC carts it is best to use something close to the "recommended loading" and it you don't like the sound get a new cart as they do not respond very well or predictably to loading changes unlike MM carts.

 

[Michael Fidler]

It takes a big change in impedance to affect the FR of an MC cart and is generally not recommended as you can run into other issues by taking things to extremes. Running low input impedance makes the cart "work harder" (it takes more mechanical force to move the stylus) which in general is a "bad" idea for a stylus but it could possibly sound better to some people in some cases. With MC carts it is best to use something close to the "recommended loading" and it you don't like the sound get a new cart as they do not respond very well or predictably to loading changes unlike MM carts.

That's a very intuitive explanation, but in practice it's less than a 0.004% of the mechanical resistance of the cartridge, even with the outputs shorted...

https://www.audiosciencereview.com/...preamp-headroom-why.56265/page-7#post-2066165

From a year or so ago, this topic keeps popping up, so maybe I should just write my own thread on moving-coil cartridge loading? I really dislike this topic as it's so perennially repetitive. Maybe it would be best to get all the info and calculations down in one place!

 

[welwynnick]

120 ohms sounds just fine to me.

 

[levimax]

That's a very intuitive explanation, but in practice it's less than a 0.004% of the mechanical resistance of the cartridge, even with the outputs shorted...

https://www.audiosciencereview.com/...preamp-headroom-why.56265/page-7#post-2066165

From a year or so ago, this topic keeps popping up, so maybe I should just write my own thread on moving-coil cartridge loading? I really dislike this topic as it's so perennially repetitive. Maybe it would be best to get all the info and calculations down in one place!

Thanks for busting the "stylus harder to move" myth for me

Can you comment on MC loading in general. My experience is that loading at ~100 ohms works fine in most cases. My experience is also that changing the load impedance of an MC cart does little if anything to the FR even if taken to relative extremes.

 

[Cbdb2]

Frequency response errors are still common in phono stages as is channel differential. This can be far more audible than results of distortion and noise tests.

Of course, I was talking specificaly about noise specs.

 

[The_Fall]

Any thoughts on the X-Feed affecting the channel separation?

 

[Michael Fidler]

Thanks for busting the "stylus harder to move" myth for me

Can you comment on MC loading in general. My experience is that loading at ~100 ohms works fine in most cases. My experience is also that changing the load impedance of an MC cart does little if anything to the FR even if taken to relative extremes.

From the manufacturers perspective: easy to implement with little electronics knowledge - you just need to switch in the components that amount to whatever you want to write on the switch face. It adds perceived value for little effort/allows the manufacturer to claim it's necessary for 'fine tuning' or some other dubious reason that would put their product on the right side of the fences of 'high end'.

From the customer's perspective: we could say that it's entirely a placebo that lets them adjust in what they believe to be right and thus confirm their own presuppositions about it. Or they're making small changes to the level at the input (but not frequency response or distortion), with moderate adjustments to the load resistance and perceiving those level changes as sounding 'better', 'more direct', 'harder', 'softer', etc.

Changing the loading on a 12 ohm cartridge coil from 200 ohms to 60 ohms would increase insertion loss by 1dB, enough to cause a perceived change in sound quality that may not be interpreted as the level change that it is. In extreme cases, where the preamp has very poor overload margin (if passive RIAA equalisation is used, for instance), then increasing the loading to bring the nominal level at the input down could also mitigate preamp overload, leading to improvements in sound quality.

Like you say, 100-150 ohms works perfectly for almost all LOMC cartridges (a few very-low-output outliers are best dealt with using step-up transformers into MM inputs IMO), and I would agree with you 100%. 100 ohms is enough resistance to keep insertion loss below 1dB with a typical 12 ohm coil (Audio Technica), low enough to absorb RF peaks to prevent resonant detection, and allows a sufficiently large load capacitor to shunt away HF/VHF/UHF radio-frequency interference.

 

[Michael Fidler]

Any thoughts on the X-Feed affecting the channel separation?

Looks about right to me. A nice first-order function at 6dB/octave as intended. At 100Hz, you get about 26dB at 1kHz, which is better than half of the cartridges out there.

By digital-audio standards, it's obviously unacceptable, but within the limitations of the format it works very nicely. You can, of course, switch it off entirely to get channel separation orders of magnitude greater than mechanical cartridges .

 

[DesertHawk]

I don't think it's a good idea to include these, as they do confuse people a bit, and can also be conflated with the total system capacitance which would lead to poor results.

The small capacitor that yields most of the input capacitance of 120pF is 100pF, and has to be connected after the RF stop resistor of 100 ohms or so to prevent radio-frequency interference in the VHF band from getting onto the input where it would be envelope-detected as audible hash. If it were to be routed through a switch, the extra inductance would make it a less effective shunt.

A lot of the preamplifiers advertised as 100pF are in fact 120pF or more, as you've got to factor in the PCB track-to-ground capacitance, op-amp input, and so on...

Adding more capacitance can be done quite easily, but adds complexity to the design, and only really effects the response above 10kHz by a few dB at most. Most modern cartridges work best with about 200-300pF, so I don't really see a need for it if we have 120pF from the preamp and then another 80-150pF from the tone-arm/cabling. Adding it would just increase the size/complexity of the board, enclosure, generate long e-mail exchanges about setting it, give people the opportunity of setting too much of it and complaining etc.

If it's not necessary, then it's best avoided. Keep the phono-stage simple so it can be plugged in and work straight away with the vast majority of devices, with the loading question completely out of mind.

The same goes for MC loading, which doesn't even really affect the frequency response until it drops below the coil resistance or less. You need a low enough resistance value that lets the resistor dissipate the resonance of the cartridge inductor against the load capacitance. The load capacitor selected for its highest value such that the appropriate resistance is about ten times the coil resistance to keep insertion loss under 1dB. If you design the input amplifier correctly, then the thermal noise of coil resistance will dominate if it increases beyond 30 ohms, such that the noise penalty from insertion loss is mitigated for higher coil resistance.

If you let the user select the MC load resistance value, then they'll either increase insertion loss and harm the signal-to-noise ratio of the device, or create an underdamped RF peak at the coil-inductance/load-capacitance resonance point that will be detected on the input and lead to complaints. Changing the loading alters the level at the input slightly, but perhaps enough for people to convince themselves they have a 'preferred' amount that will probably provoke trouble... Most preamps on the market use pretty brutal (and rather noisy) resistive series RF stopper networks on the input amplifier, preventing the resonance from getting through, at a massive noise penalty which is only acceptable because the market has such low expectations with the intuitive (but incorrect) explanation that the level is lower and therefore noise is correspondingly higher.

While selectable loading is very easy to implement, it is not necessary in my opinion and actually highly undesirable from a customer-facing point of view.

Thank you for such an in depth and we'll explained response.

 

[EERecordist]

This is a review and detailed measurements of the Michael Fidler Spartan 20 Moving Coil and Magnet phono preamp. It was sent to me by the company and costs £500/€600/$670.
View attachment 442291
Is it me or is the On/Off upside down? I am used to On being up, not down. That nit aside, there are LP noise mitigation facilities such as mono conversion and dual frequency cross feed. Unlike many other phono preamplifiers, there are independent moving coil and magnet inputs:
View attachment 442292
I appreciate the upside down additional labeling when you are trying to change things and doing it from above. Nice to see the range of input voltages as well.

Best part of the design is what is on Michaels's site:

That is one pretty looking PC board! The box badly needs a clear top for this to be appreciated in use.

ICs are socketed but company mentions that is only for repair and not silly "op amp rolling."

Spartan 20 Moving Magnet Preamp Measurements
As usual we start with our dashboard with fair bit of optimization on my part to minimize ground loops:
View attachment 442293
Nice to see no distortion and only hint of power supply noise. This easily lands the Spartan 20 in our upper tier ranking of phono stages:
View attachment 442294

What is absolutely superb is the frequency response:
View attachment 442296
It doesn't get better than this folks. Not only the equalization perfect but so is channel matching. We also have a proper high pass filter to get rid of LP rumble.

There is a crossfeed function for noise reduction but I am unclear as to whether this is the optimal implementation or not:
View attachment 442298

Another excellent aspect of this phono stage is the way it almost doesn't care what the frequency is when it comes to overload:
View attachment 442300
Most phono stages lose substantial amount of their headroom at high frequencies which is where you need it when it comes to ticks and pops. Not here. You have that nearly 100mv at all frequencies.

Distortion is admirably kept low:
View attachment 442304

Spartan 20 Moving Coil Measurements
Let's look at our dashboard:
View attachment 442305
Due to much higher gain, it is hard to completely minimize mains noise. What we have is respectable anyway.

Same excellent overload characteristic is provided for moving coil:
View attachment 442307
I already showed you the distortion vs frequency in previous section.

Conclusions
I don't know any other phono stage designer that is so focused on engineering excellence as Michael. Spartan 20 is another evidence of that with execution that borders on perfection. You pay some premium of course but considering the device is manufactured in the UK, I think that is justified.

I am happy to recommend Michael Fidler Spartan 20.

------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

We appreciate you developing a repeatable test suite for phono preamps.

Before consulting ASR I purchased the low performing NAD which I will throw in for free when I sell my Revox turntable.

 

[mike70]

I don't understand why the preamp capacitance cannot be 50pF or so.
If the user needs higher values, is easily reached with higher capacitance cables (as an example)

you cannot lower it

 

[The_Lecht_Rocks]

And now I understand why @Michael Fidler quizzed me about my turntable setups prior to selling me my rather exquisitely manufactured and sounding Spartan 15 mk2 !

 

[Frank Dernie]

Oh, and are most cars in the UK manual?

They used to be but less so now and most Ubers that aren't electric are often Toyotas which have an ingenious (and patented) E-CVT transmission which is light and efficient.
We calculated a CVT was a worthwhile performance advantage over a manual transmission for Formula 1 racing in the mid-80s, developed one but it was banned (probably because all the other teams would be beaten until they developed one of their own).

I bought a Toyota in 2005 just to experience the CVT myself, not expecting to like the car much and not keep it long but have had one ever since...

I have 4 cars a 6-speed manual sports car, one E-CVT hybrid, an old V8 with sludge pump auto transmission and an EV, so only one manual myself now.

 

[Michael Fidler]

And now I understand why @Michael Fidler quizzed me about my turntable setups prior to selling me my rather exquisitely manufactured and sounding Spartan 15 mk2 !

Mainly because I get often order requests for MM phono preamps when the user has an LOMC cartridge and vice versa . Or 'I have a problem with hum in the setup and want you to fix it for me' - usually caused by wiring issues in the turntable or ground path and not solve-able by changing the phono preamp, but will certainly result in a return when it doesn't and I can't fix the issue via e-mail exchange. Very time-consuming...