Need to be educated re: MC Step-up transformers

I recently purchased the EAR MC-4 Step-up transformer. I was having some noise problems with my analog rig. I spoke with Tim de Paravicini of EAR directly in the UK. He answered all the system matching questions and pronounced the MC-4 a fit with my system. He is very direct, extraordinarily knowledgeable, and seemed very sincere and honest.

I still don't understand the "How" though. Here's what I do understand and relayed to Tim:

1. I use the ClearAudio Stradivari, a MC at 0.7mV output.

2. Currently, my phono stage provides a total of 57 dB of gain: 20 in the MC mode using a JFET and 37 in the MM mode using all tubes. My tube line stage provides 12 dB of gain SE, which is how I run it. So, currently without the MC-4, I have a total of 69 dB gain when running my phono stage in MC mode (20 + 37 + 12).

3. The MC-4 has four taps – I was told by Tim to use the 40 ohm tap which provides a 10x gain in voltage and is compatible with the 32 ohm impedance of my cartridge. This would change the output voltage my phono stage “sees” from 0.7mV to 7.0mV.

4. The phono stage must be run in MM mode, bypassing the JFET in the MC mode.

5. Given all this, then, my phono stage will receive as input a 7.0mV cartridge output from the MC-4. This signal will run through the MM mode and receive 37 dB gain and then another 12 dB gain from the line stage SE for a total of 49 dB gain, down from 69 dB, as we are no longer “gaining” the additional 20 dB from the JFET MC mode.

6. I need to remove the 600 ohm Vishay resistors and get the phono stage back to the stock 47 kohm setting, as the MC-4 will reduce the resistance by the square of the voltage gain or 10^2 or 100: 47,000/100 = 470. That is within the range recommended by Clearaudio of 320 – 900, preferably toward the lower end, though let your ears be your guide.

It was based on this information (which I provided) that the MC-4 was pronounced a fit - I certainly don't doubt that.

Here's my confusion: Am I to understand that 7.0mV of cartridge output from the MC-4 is so much more voltage that all I need is the 49 dB the system provides in its new configuration for low noise and analog bliss?

I just don’t understand the science, I guess. I appreciate the education.

You underestimate yourself! Sounds to me like you understand it perfectly. Basically, you are substituting the 20db of gain provided by the step-up transformer (20db = a factor of 10 gain in voltage) for the 20db which had been provided by the jfet section of the phono stage.

So the overall gain will remain the same. And by removing the resistors you are keeping the loading on the cartridge in the same ballpark, and within specification.

The benefit of the change will be elimination of the noise problems you mentioned, IF those noise problems were due to the jfet section of the phono stage. That is probably a good bet, because since noise generated by the front end of a phono stage is amplified by all of the other electronics that follows in the system, that is typically the most significant contributor to overall system noise.

The only possible concerns that occur to me, which seem unlikely, would be the possibility that the 7mv might overdrive the mm section of the phono stage, or that overall system gain (including your power amp) might result in having to operate the volume control down towards the bottom of its range. But if you haven't had either of those symptoms running the cartridge through the 20db gain provided by the jfet front end, I don't think you will have them with the step-up xfmr that provides similar gain.

-- Al
Some helpful informaton on step-up transformers I had found. It talks about various calculations and stuff. I had to read it a few times (with some note taking) to figure out how to do the calculations.

Post removed 
Saudio -- I would respectfully disagree. In general there is no reason to expect that providing a load to the cartridge which numerically equals its output impedance will provide best results, or even a good starting point. Quoting from the excellent reference provided by Gundam91:

The first rule of thumb is that most (some exceptions exist) MC cartridges like to see a load impedance of 3X to 6X their output impedance value.

And it seems to me, at least as a starting point, that the manufacturer's recommendation for the specific cartridge should supersede any general guidelines.

Also, I'm not sure why placing a load resistor on the primary side should necessarily be given preference to placing a resistor on the secondary side, with the value chosen to reflect an equal value to the primary side when divided by the square of the turns ratio.

-- Al
Post removed 
Post removed 
Hi Al,
you quote Gundam91: "The first rule of thumb is that most (some exceptions exist) MC cartridges like to see a load impedance of 3X to 6X their output impedance value."

This is entirely dependant on the cart DCR in the first place.
Note: The lower the cart DCR, the closer the actual cart impedance to the cart's DCR value.
What you actually try to do is to get an impedance match 1:1 between cart and primary. This way you double the current from the cart in lieu dropping it’s output voltage. The cart is working in: ~ 'current mode'~, which in turn provides for the best and highest voltage output on the secondary.

Very often this coincides very closely (also in experience) with the lowest quoted "recommended" loading.
If e.g. a cart it quoted as >10ohm (at a DCR of ~ 4ohm) or say >30ohm (of DCR ??) etc., then this often turns out to be your point of departure.
Caveat! If a ~ 6ohm cart is e.g. quoted >100ohm then it is VERY probably targeted only at a non-SUT load.
You could see what I mean when visiting some of the Lyra specs (on web) and will see that they actually quote 2 values - one for normal (non-SUT) and one loading for SUT.
Usually some ~ 10:1 difference
Thanks for all the input. Re: loading the MC cartridge. I had been running it at 47 kohms (don't laugh), and found the sound a bit harsh. My original dealer (now out of business) never educated me on matching and I never educated myself until the sound just didn't feel right.

I contacted Garth at Musical Surroundings, and he suggested the range I indicated above as a starting point, but cautioned that I may choose to go lower - let my ears be my guide. His thoughts were based on:

1. His personal experience
2. His contact with other ClearAudio customers and
3. His personal preference to make less of a reduction than more, given I was starting at 47kohms!

He did mention ClearAudio's preferred phono stage settings during our conversation, but he mentioned 800 ohms. Here's an excerpt from his e-mail:

"Yes 10x coil impedance is the rule for minimum resistive loading, thus you can go as low as 350 ohms. Clearaudio phono stages load at 800 ohms, one reason I suggested it. And 350 is a very radical change from your 49k ohm load. If your system is on the ultra bright side, yes go for 350."

My phono stage requires that the resistors be soldered in - so to me it is a bit of a hassle. I chose 600 ohms because I felt my system was not THAT bright and it was in the middle - a sort of compromise/place to start.

I have to tell you to my stone ears, I was satisfied, save for the noise I mentioned.

As I let the MC-4 settle in, I plan to listen and let my ears be my guide. I just may make some adjustments in the future.

Thanks to everyone.

are still talking SUT? Please note that the values you had discussed related to non-SUT, straight into your phono-pre gain stage, the way I understand it, and the values look to me. I do not know your specific cart but ~ 500ohm could be fine to give you some "air" without getting the sound too "fluffed out", loose, chirpy, uncontrolled, what ever you want to label it.

There are not no many carts that go OK with 47k, as it seems to depend on the internal damping of the typical MC rise in the treble.
Only some MC designs are OK with 47k, and some would argue NONE at all.
Again, if you talk SUT, and specifically truly low output ones then the loading values for SUT use change drastically in my experience, but we have already gone into all of that, no?
Flyfish, The others have subverted your post into a discussion of loading. Bear in mind that this is a "loaded" question; there are "front-loaders" (those who urge you to add a resistor to the primary side) and back-loaders (the more common method of loading the cartridge only on the secondary side) in two opposing camps. On this question, I can only add that Tim de Paravicini might know a little bit more about cartridge loading than some of the others here who are telling you that loading the secondary is "wrong". But what I am interested in is whether you got rid of the noise by switching from your MC input stage to using the SUT. I am not clear on what happened to the noise when you made the switch.
Brent -- Regarding using a 47K load, iirc correctly even the estimable Harry Pearson (founder of "The Absolute Sound") wrote many years ago of preferring the sound of some particular mc cartridge when loaded (or perhaps I should say "unloaded") with 47K! But as Axel indicates, that would be optimal only for a small minority of mc's.

Basically, what the typical lower value load does is to dampen what would otherwise be a large high frequency (usually ultrasonic) resonant peak. You can find an explanation of that here:

The degree to which that peak would produce adverse audible consequences, such as brittleness or edginess, would depend on its amplitude, its frequency, the bandwidth of the rest of the system, and the intermodulation and other effects that may be produced in the rest of the system by ultrasonic or very high frequency spectral components, as well as on the mechanical damping provided in the cartridge itself, as Axel points out.

Axel & Saudio -- Thank you for your comments. I don't question your experienced-based comments, and I think we all agree that mc cartridge matching necessitates that listening be the final arbiter even more so than with the rest of an audio system.

I haven't seen any good explanations of why use of a step-up transformer would typically call for heavier loading than would be the case if a high gain phono stage or head amp were used. After giving it some thought, I've developed the following theory:

A real-world transformer can be modeled as an ideal transformer in parallel with an inductor on the primary side, representing "magnetizing inductance," as well as other resistive and inductive elements representing loss mechanisms. See the diagram around the middle of this page:

Referring to that diagram, if we assume that the series leakage inductance Xp is relatively insignificant compared to the cartridge's inductance, then by Thevenin's Theorem the magnetizing inductance Xm can be considered as combining in parallel with the cartridge inductance. That will result in a lower total inductance, and therefore an ultrasonic resonant peak which is higher in amplitude if a heavier load is not applied to damp it.

Although the peak will be higher in frequency as well, and it will also be affected by the bandwidth limitations of the transformer, all of which sufficiently muddles the situation such that, again, listening would seem to be the only way reach any meaningful conclusions.

-- Al
Thanks again for the posts. In my second posting, I was discussing loading my phono stage, or what I have just learned is referred to as "back loading". I never even knew one could "front load". I apologize to Axelwahl for any confusiuon. In my phone conversations with Tim, he felt I could use the MC-4 "as is", just remove the two 600 ohm resistors I use to "back-load" and I should be good to go.

Did I remove the noise - can't answer that, yet. Neither the US importer nor EAR carry any inventory on the MC-4. Each is made as an order is placed. From e-mails, I understand my MC-4 just made it out of testing in the UK and should be shipped to the US either Monday or Tuesday of next week.

I will post again with results.

Again, this is wonderful info for me.

Flyfish, I am going to stay clear of all the technical discussion here. As valuable as it can be, the bottom line is how it all sounds to you.

I have been running with a Clearaudio Accurate for several years now. With this previously straight into an Aesthetix Io phono stage, I ultimately found the "best" load to be between the 500 and 1kohm jumpers. When I changed to the Aria WV preamp with Sowter SUTs, the 200, 300 and 500 ohm jumpers were so close that any of these would have been fine. Running without any of these defaults the input to 1k and this was just a touch too "hot". And now with the cartridge running into a rebuilt Counterpoint SA-2, I have found that a 500 ohm resistor in the load terminals is a great balance. So your drop way down to 700 ohms is close to my own experience here.

Thanks, John. When the MC-4 finally is in my system, the load will be at 470 ohms due to the SUT "loading down" the cartridge by the square of the volatege gain (47000/10^2). Thanks for posting your experience - it helps to give me some confidence in my starting point.

Today is August 5th - the EAR MC-4 finally arrived! Hoorah!

I had a chance to insert the MC-4 into my system and change my phono stage by re-setting the resistance back to the stock 47kohms and switching to MM mode. The results - dead quiet.

Now, the line stage still needs to be up around the same gain value for similar volume levels as before - I didn't "gain" any additional gain, if that made any sense. I swapped out 20 dB of JFET for 20 dB of EAR MC-4.

The difference is that I replaced a 1994 year old JFET (Bill Thalmann completely modded the phono stage, but, of course, did not replace the 15 year old JFET) with a 2009 SUT. Very quiet.

Another very welcome, but unexpected benefit is that the sound is markedly improved. I realize it is trite and overused, but it is as if a "veil" has been removed. Listening to just a few albums suggests a more natural sound, as if I was present at a concert.

While I had no intention of buying an SUT and did not want to spend the $$, I am very pleased with the purchase and cannot speak highly enough of EAR's work and Tim de Paravicini's absolute attention to customer service and build quality.

One month later...

I have settled in with the EAR MC-4 more. I have adjusted the set-up. I now use the 12 ohm tap instead of the 40 ohm tap on the MC-4. The 12 ohm tap provides 18x voltage gain instead of the 10x of the 40 ohm tap.

While the EAR MC-4 is MUCH quieter than the JFET in the MC stage of the phono stage (actually this is an understatement - the EAR is dead quiet), the 40 ohm tap presents two problems:

1) It gives me the same 20dB gain as I get when I used the MC stage - I don't gain any gain. With my line stage providing only 12dB in SE mode, I still had to turn the line stage to the same volume level for similar gain. At "fun" listening levels, this produces surface noise even on immaculately clean LPs; and

2) My Clearaudio cartridge likes to see 150 - 200 resistance from the phono stage. Clearaudio actually sets their phono stages to a factory stock of 200ohms. Using the 40 ohm tap, I am at 499; using the 12 ohm tap, I am at 154 - a much more musical presentation for my ears.

I realize that with a cartridge impedance of 32 ohms, I am "breaking" the rules - I was told to use the tap nearest the cartridge impedance. In addition, the all tube MM stage is seeing a cartridge output of about 12.6mV - a tad high (though Tim de Paravicini told me you had to get to around 14 before you typically had any issues).

To date, I have had no problems and I use less gain at the line stage, producing a vastly quiet system even at "fun" volume levels.

I guess the lesson is read, learn, ask questions and then experiment.

Thanks everyone.
Ive been experimenting with sut's and loading on my stradivari for a little over a year now. The clearaudio rep I spoke with recommended around a 400 ohm load with a transformer. Im now using a 1:10 step up into 47k which yields 470 ohm load which in my system seems ideal. I think the down stream components will make a big difference in determining optimal loading, especially your speakers.
Thanks. Garth Lear of Musical Surroundings had suggested 300 - 400 for the Stradivari.

My speakers are 85dB sensitivy and 4 ohm - may be the culprit.

Good listening.
Hi Flyfish2002,
seems you are NOT the only one that uses a somewhat 'wacky' SUT setup e.i. going over the top with the input voltage, etc.

I finally got my NEW Ortofon Windfeld back and was trying it, something just didn't gel when I used the SUT as I had it with the old one, which had an 'VTA error' according to Ortofon --- it sat very low, etc.

Following a hunch, I removed the 13ohm primary loading resistor from the SUT --- just to see what would happen. I be darned, it works VERY WELL without. In theory I'd be overloading the MC input but it looks like I don't. Also I recall that with the old cart, voices seemed too recessed in the stage, none of that either now.

My current thinking is, that since an SUT is an 'impedance-converter', it would AMPLIFY not only the voltage but also any problem with a cart.

Go figure - and enjoy the music.
Sorry for jumping into the discussion, the main subject was never clear to me, even after reading many posts and detailed explanation on vinylengine article for example.
I have Fidelity Research AGT-5X SUT, which can accept loading of the carts from 3 to 10 ohms, with respective gain from 31 to 29 db. If I look at the carts produced at the time SUT was made, carts by FR, they all fall into 3 to 10 ohms load impedance (with 2 carts as exception), which would make a lot of sense to me - mfrg making SUT to fit their carts, why would they make incompatible SUT at all (I presume that they making SUT primarly for their carts, not just for the market).
Now I have Transfiguration Phoenix, which has impedance of 7 ohms, and requires loading higher than 7 ohms. That should almost perfect match between SUT and cart.
But I am reading not for the first time about 3X or 10X rule of thumb ratio between cart impedance and SUT loading, which makes me confused again.
My Phono stage is Audion Phono (which requires impedance of 47 kOhm which SUT has), volume of playing at the same level as say from CD, not different.

Can someone clarify?
Hi Salavat,
your Phoenix has 0.4mV output, yes?
A 30 dB SUT has a 31.6:1 voltage ratio. Therefore 0.4mV*31.6 = 12.64mV output of SUT secondary to your phono-pre ---- and that will pretty surly OVERLOAD your phono-pre input stage, usually starting to overload already if the above calculation gets to ~ 8.5mV

So, before you want to work on the impedance side of things, the first thing is to make sure you will not get into OVERLOADING the phono-pre input --- all else comes next.
Mid spec for phono-pre input is actually 4.7mV. Check any MM/MI cart output and it will give you a pretty good idea what seems an acceptable range: 3mV - 6mV, so 12.6mV is too much by far.
Only if you consider SUT primary or secondary resistive loading will this value it come down somewhat, but not enough by far for the above example.

After reading this post I swapped my 1:10 sut for 1:18 with my strad. This should make the output to hot for the phono stage, also it reduces the load to around 140 ohms. I hadnt been using this xfrmr (4722) because the math didnt work for this cartridge (I use this sut for my denon). Man was that wrong. This is my new preferred setup. There is no sign of overload with the several phonopres I tried it with and the sound is beautiful. I tried this set up very briefly when I first got the cartridge and didnt like it but at that time the cart was not broken in and it is a completely different cart now. This brings up another point for strad buyers or new owners. This cart takes a full 100+ hours to break in and changes dramatically during that process. If I hadnt discovered that earlier with several other CA carts I may have bailed on it early on but it is definitely worth waiting. CA suggests breaking it in with a 47k ohm load which only exacerbated its shrillness during the break in period. Its worth the wait, Im very happy now and thanks to Brent Ive discovered something new for it.
Rccc - glad it worked out for you. As I mentioned above, Tim de P told me that 12.6mV is a bit high, but a decent phono stage in MM mode won't get overloaded until you hit 14 or so. I didn't believe him, at first.

My future plans are to replace my monoblocks so I can run the line stage to monoblocks balanced - currently I am SE. The gain on my line stage is 20 balanced and 12 SE. When I do this, I might not need the gain of the 12 ohm tap and I may need to go back to the 40ohm tap.

Future plans...