Learning about SUT's

[rothwellaudio]

Now which will have a lower noise floor? I assume the SUT, but would this be substantially noticeable at normal listening levels?

A SUT will have a lower level of hiss, but a good headamp isn't far behind and the hiss should be way below the level of vinyl surface noise - certainly not an issue at normal listening levels.
On the other hand, you're more likely to get hum problems with a SUT. As long as the tonearm/SUT/phonostage are wired correctly and the SUT is located away from any strong magnetic fields (such as produced by a large mains transformer) the hum should be vanishingly low. Most SUT hum problems are caused by poor wiring or proximity to mains transformers.

 

[Nashou66]

Well I came across a Nakamichi MC-1000 SUT , Finally got another Yamamoto Box wood head shell
for my re tipped Entre EC-20 and it sound pretty great to my ears!!! Also I think it was a good match as the volume
control does not have to be changed from when I switch the Nakamichi from Direct ( MM cart) to Boost ( MC Cart).

If more disposable Income comes my way I will get a Rothwell Head amp for my second system so I can move the Entre to that system and also see which I prefer and use that in my Main system.

Athanasios

 

[AvFan]

Option #2 is what many people do, but it isn't really based on anything sensible. I ask again, why not put loading resistors on the primary side of the transformer to load the cartridge directly? Why does the load have to be on the secondary side of the transformer?

OK, I'm guilty of resurrecting a 4-YO thread, a really good one IMO, and there is a point @rothwellaudio made above I want to get clarified and ask an additional question.

I am using @wyn palmer's preamp modified for MM along with a SUT I assembled using Hashimoto transformers. Wyn at some point clarified that the load resistor in his preamp could be left out without adverse impact to the preamp. If a different cartridge load is desired a second resistor is often added to the secondary side of the SUT in parallel to the MM preamp's load resistor (typically 47Kohm). This can only lower the load but as noted above doing so isn't the only way to achieve a desired cartridge load.

If I remove the MM preamp load resistor in Wyn Palmer's MM preamp and install loading resistors to the primary side of the step up transformers should the resistor value be equal to the manufacturer's recommended load? I know manufacturers typically specify a range (if at all) and the other method is to start at about 10x the cartridges impedance. But if the recommended load is 100ohms is it as simple as adding a 100ohm resistor to the primary side of the SUT? Does the cartridge's impedance play into the calculation?

In this thread it was also mentioned that adding capacitance to the primary side of the step up transformer was OK but may be preferred? If 200pf was typically used in a MM preamp and I wanted to replace that on the primary side of the transformer would it be the transformers ratio squared times 200pf or? I wish I had the skill and equipment to figure out a Zobel network (Jenson shows one in their hookup schematic).

 

[Pio1980]

Where do Mcmartin MT-4 transformers fall in as SUTs?

 

[mkane]

 

[MRL_Audio]

Where do Mcmartin MT-4 transformers fall in as SUTs?

Have these. They out perform my HK XT-2’s. Think @Redboy can provide a better answer though.

 

[wyn palmer]

OK, I'm guilty of resurrecting a 4-YO thread, a really good one IMO, and there is a point @rothwellaudio made above I want to get clarified and ask an additional question.

I am using @wyn palmer's preamp modified for MM along with a SUT I assembled using Hashimoto transformers. Wyn at some point clarified that the load resistor in his preamp could be left out without adverse impact to the preamp. If a different cartridge load is desired a second resistor is often added to the secondary side of the SUT in parallel to the MM preamp's load resistor (typically 47Kohm). This can only lower the load but as noted above doing so isn't the only way to achieve a desired cartridge load.

If I remove the MM preamp load resistor in Wyn Palmer's MM preamp and install loading resistors to the primary side of the step up transformers should the resistor value be equal to the manufacturer's recommended load? I know manufacturers typically specify a range (if at all) and the other method is to start at about 10x the cartridges impedance. But if the recommended load is 100ohms is it as simple as adding a 100ohm resistor to the primary side of the SUT? Does the cartridge's impedance play into the calculation?

In this thread it was also mentioned that adding capacitance to the primary side of the step up transformer was OK but may be preferred? If 200pf was typically used in a MM preamp and I wanted to replace that on the primary side of the transformer would it be the transformers ratio squared times 200pf or? I wish I had the skill and equipment to figure out a Zobel network (Jenson shows one in their hookup schematic).

I'm not entirely certain what you are trying to do, or ask.
As has been said before, the impedance at the Input of the transformer is transformed to the output by the square of the turns ratio, and the inverse going the other way. The voltage is transformed by the turns ratio. So, for a 20:1 ratio 100 ohms at the transformer input is "Seen" by the circuitry at the output as 40kohms. If you remove the 47k from the input of the mm preamp and the transformer is not ac coupled (no series cap) then the preamp will be correctly biased as far as dc is concerned- the transformer output winding provides a dc path. So, yes, you can, under these circumstances, replace the 47k with c. 120 ohms across the input winding. If the input is ac coupled then removing the board input cap will cause the preamp to no longer be operable.
The cartridge impedance is not intrinsically important in the sense that you seem to mean it. No zobel network is required, or even desirable. You do know what a zobel network is, and what it does, right?
The 200pF is there to properly load the MM cartridge, so why would you want it on the primary side of a SUT?
200pF would have essentially zero effect on a LOMC cartridge, and in fact 400x 200pF would have zero audible effect.
Remember, SUTs are also imperfect. They add distortion, degrade the frequency response, and some , albeit very low, noise.
Also, load resistance changes do not have the miracle properties that some claim. Manufacturers always recommend loads, when they can be bothered to do so, that flatten the frequency response, trading off the electrical characteristic of the input resonance and that of the mechanical cantilever resonance.
These characteristics generally have pretty wide tolerance ranges where loads are within specs.

Sometimes people claim that an extreme loading value has some miraculous benefit. What it generally means is that the frequency response has been so altered from the nominally flat situation that it colors the response in ways that they like.

 

[AvFan]

@wyn palmer : Thanks for the response and you can tell by my rather odd question that, no, I don't have a fundamental understanding of how these things work together. Think Civil Engineer who got a barely passing grade in my two mandatory circuits classes. I have seen a Zobel Network but I'm not familiar with how they work.

I had the question because there was discussion about the wisdom of loading the secondary side of a SUT and that it was better to do it on the primary side. As I understand it, the typical MM load of 47Kohm is on the secondary side when using a SUT so I wondered if it could just be removed entirely and load the cartridge with a resistor on the primary side of the SUT. If this was true would the value of that resistor be equal to the recommended loading for any the cartridge? This occurred to me after (I thought) it was said in another thread that the MM load resistor could be removed and not degrade the performance of the preamp. I was hoping the math was simpler by having a load resistor on just the primary side of the transformer.

I understand that capacitance has virtually no impact on the sound of a MC cartridge I just wanted to make sure there wasn't some reason to have it when the load was on the primary side of the SUT. Apparently not. However, does there need to be a capacitor on the MM board at the load position? "If the input is ac coupled then removing the board input cap will cause the preamp to no longer be operable."

For example today I put a 100ohm resistor on the primary side of my SUT (Hashimoto HM-7), removed both resistor and capacitor from the load position on the MM V2.0 board. It operated but was odd; the left channel was noticeably more prominent like the anti-skate was way off. I have since installed a 100pf capacitor on the MM board and the left channel is less pronounced but will need more listening to know whether the problem has been resolved by adding a capacitor. Does this make any sense based upon the preamp design?

Thanks again!

 

[rothwellaudio]

But if the recommended load is 100ohms is it as simple as adding a 100ohm resistor to the primary side of the SUT? Does the cartridge's impedance play into the calculation?

If there's no load at all on the secondary side of the transformer, ie it is open circuit, the cartridge will see no load either, at least in theory. In that case, any resistor across the primary winding will become the load impedance, ie a 100 ohm resistor will give you a 100 ohm load impedance.
If you have a 47k load on the secondary side, which is usual, and the turns ratio is 1:15, the load impedance seen by the cartridge will be 209 ohms. In that case, putting a 100 ohm resistor across the primary would cause the cartridge to see a load impedance of 209 ohms in parallel with 100 ohms, ie 68 ohms. The cartridge's impedance doesn't come into the calculation.
If you want the cartridge to see a specific load you could put the load on the primary side of the transformer or the secondary side, or you could split it and have some of the load on the primary side and some on the secondary side. If the transformer was a text book theoretically perfect one, ie it has zero winding resistance and infinite primary inductance and zero leakage inductance and zero inter-winding capacitance etc., it would make no difference which side the load was on. However, real transformers don't behave like the simple model of an ideal transformer. The advantage of having the load on the primary side is that it loads the cartridge directly without any worries about the transformer's real-world complexities. The advantage of having the load on the secondary side of the transformer is that it can lower the low frequency cut-off point. It can also damp high frequency ringing. However, it can also lower the high frequency cut-off point. These effects are all dependent on the specific parameters of the transformer in question and there's no magic formula which can be applied to any and all transformers, ie you don't really know effect a load on the secondary side will have without measuring your transformer. Bear in mind that these are effects on the transformer, not effects on the cartridge.
Personally, I think the importance of the load impedance seen by a LOMC cartridge has been wildly exaggerated on many forums, and claims that a particular cartridge performs best with a particular load impedance should be taken with a pinch of salt. For example, say somebody claims that cartridge X sounds best with a load of 80 ohms - how have they come to that conclusion? If they placed a 10k resistor across the secondary side of a 1:10 SUT, that would give the cartridge an 80 ohm load, but it may have changed the transformer's performance much more than it changed the cartridge's performance. If the transformer had a 1:20 turns ratio and was from a different manufacturer, a load on the secondary to create an 80 ohm cartridge load might have much less effect on the transformer's performance, or much more. Who knows? Or was the "80 ohms is best" claim made by feeding the cartridge into a solid-state mc phonostage with adjustable load impedance? If so, what loading options were available? If the only options available were 20 ohms, 80 ohms, 400 ohms and 2000 ohms, is the "80 ohms is best" claim valid for someone using a SUT? I doubt it.

Remember, SUTs are also imperfect.

Yes, they certainly are imperfect, which is why a load on a transformer affects its performance. When altering the load on the secondary side of a transformer, are we to affecting the cartridge's performance or the transformer's?

 

[mkane]

Maybe I should go back to bed. In other words, one day I'll leave well enough alone.

 

[MRL_Audio]

Maybe I should go back to bed. In other words, one day I'll leave well enough alone.

What fun would that be...

 

[wyn palmer]

Time to get back on the soapbox for a little while...
My apologies in advance. It would seem that I'm feeling a tad pompous/maudlin this morning.

It is possible to model the cartridge/SUT/preamp interface with extreme precision. I've done it, and I'm certainly not alone in this. You can make the model replicate the distortion, frequency response and noise with very close compliance to measured results. You need to include all of the commonly accepted parasitic components and impairments associated with the transformer (at least a sixth order system), the cabling, the input impedance of the preamp, and the electrical characteristics of the cartridge (at least a third order system). You even need to add non-linear elements to replicate the distortion, although I ignore the need for this when I model the transformer Barkhausen noise.
You can even make measurements and include the mechanical components of the cartridge (also at least a third order system). However, it's a pretty complicated, highly iterative, process, and not for the faint of heart.
However, why bother?
My phono stage design measures as close in frequency response to the ideal RIAA characteristic as a CD is to a flat response- errors in mdB- has essentially zero measurable distortion by any metric you choose to apply, meets all of the accepted psycho acoustic metrics, has arbitrarily adjustable input impedance, extremely low input noise, extremely high and frequency independent overload characteristics, etc. etc. It's just about as ideal as an RIAA stage can be, and people STILL insist that this other design sounds better or this one will sound better if only you make this change, and also insist on using strange load values on their cartridges that are guaranteed to alter the overall response in ways that deviate from the optimal, as defined by the manufacturer of the cartridge.
Heck even the designer of the cartridge I use- the Miyajima Madake- informed me in an email exchange that I had with him that even though a certain set of load components gave the flattest response (and I confirmed that with measurement/simulation) and was necessary to meet the published specs, that he preferred a sub optimal load, because it sounded better to him...
Just do what you feel and stop trying to justify it, or get really curious, learn to use SPICE, buy test gear to allow measurements to be made, learn how to extract component values to match transformer response curves (I use an optimizing extraction program "borrowed" from ADI's modelling group a number of years ago), etc. etc.
Frankly, this all seems like too much bother (unless you just happen to be me, or some other OCD driven technical person, for example).

Look, the way I approach these things is basically a world away from what the usual audiophile does. I can't say that I don't believe in any of the usual audiophile bull****, as there is, and always has been, some grain of truth in what is said, and much of that truth has become apparent in the last 50 years.
However, having said that, much of what is said IS bull****, nevertheless.
My system IS objectively flat within a few dBs- 16Hz to 20kHz +/-3dB. It is extremely low distortion; it creates a realistic phantom image field with great detail. It does all of the things that audiophiles crave, yet small changes in frequency response in certain areas- well within the +/- 3dB tolerance significantly change the sound/image profile.
Also, the present speakers employ DSP in the dual woofer LF to create a much more dipole like bass response that better matches the dipole characteristic of the electrostatic panel.
Yes, it does sound different than the essentially identical predecessor that I had (ESL11A now, Montis then) that did not do this, and there are those that claim that they can "hear" the change in the radiative pattern due to the change from dipole (the electrostatic panel) to diffraction limited omnidirectional (the bass driver boxes) sections in the speakers. My current observations sort of confirm this, as the difference is subjectively obvious, but when others have been asked to judge the speaker quality in the region subject to these changes (c. 500Hz and below) the jury has been split on which one is better/more realistic.
Interestingly enough, false image theory suggests that the "best" speaker is one with a horizontally omnidirectional, vertically coherent, frequency independent, radiation pattern (i.e. a very thin, tall speaker with a width less than a wavelength at 20kHz, or about 1.7cm (my apologies if my math is wrong) that operates seamlessly down to c. 100Hz where we no longer can detect directionality in the sound field.
How many of us have one of those, and even if we did, do we want to listen in an anechoic chamber and use DSP to shape the acoustic field pattern appropriately to cancel reflections etc.
Anyway, enough rambling. The bottom line is, don't sweat the details. Just do what generations before you have done, use your ears and convince yourself that what you have sounds fine, and all will be well- at least until you hear something "better", or someone tells you that something is "better", and because you have no objective "reality" to fall back on you then go further down the rabbit hole...

 

[AvFan]

Thanks for the detailed reply! Let me separate some of the key points:

If there's no load at all on the secondary side of the transformer, ie it is open circuit, the cartridge will see no load either, at least in theory. In that case, any resistor across the primary winding will become the load impedance, ie a 100 ohm resistor will give you a 100 ohm load impedance.

The advantage of having the load on the primary side is that it loads the cartridge directly without any worries about the transformer's real-world complexities

Hopefully @wyn palmer will chime in and let us know if his preamp, revised for MM use, behaves that way. For the lesser technically inclined loading only on the primary side of the SUT is the most straight forward way to apply a cartridge load.

However, real transformers don't behave like the simple model of an ideal transformer. The advantage of having the load on the primary side is that it loads the cartridge directly without any worries about the transformer's real-world complexities. The advantage of having the load on the secondary side of the transformer is that it can lower the low frequency cut-off point. It can also damp high frequency ringing. However, it can also lower the high frequency cut-off point. These effects are all dependent on the specific parameters of the transformer in question and there's no magic formula which can be applied to any and all transformers, ie you don't really know effect a load on the secondary side will have without measuring your transformer. Bear in mind that these are effects on the transformer, not effects on the cartridge.

Without measuring equipment and skills to interpret the results I'm left with trying out the various ways of applying loading (either to the cartridge or transformer) to determine what sounds best to me with my system's components. That explains why I have a pile of resistors at various values that I can slip into Wyn's preamp or a load plug into the primary side of my SUTs. If I can get in the ballpark with the formula to calculate the load then I can use that as a starting point for changing resistor values.

Personally, I think the importance of the load impedance seen by a LOMC cartridge has been wildly exaggerated on many forums, and claims that a particular cartridge performs best with a particular load impedance should be taken with a pinch of salt.

Absolutely. From a subjective view (or listen?) I do find that loading above a certain tight range of values does not change the sound but below that range things can get wonky pretty fast. And what that load range is depends on whether I'm using a MC preamp or a MM/SUT combo that is further impacted by how the cartridge load is applied. Lots of variables at work here!

 

[wyn palmer]

Thanks for the detailed reply! Let me separate some of the key points:





Hopefully @wyn palmer will chime in and let us know if his preamp, revised for MM use, behaves that way. For the lesser technically inclined loading only on the primary side of the SUT is the most straight forward way to apply a cartridge load.



Without measuring equipment and skills to interpret the results I'm left with trying out the various ways of applying loading (either to the cartridge or transformer) to determine what sounds best to me with my system's components. That explains why I have a pile of resistors at various values that I can slip into Wyn's preamp or a load plug into the primary side of my SUTs. If I can get in the ballpark with the formula to calculate the load then I can use that as a starting point for changing resistor values.



Absolutely. From a subjective view (or listen?) I do find that loading above a certain tight range of values does not change the sound but below that range things can get wonky pretty fast. And what that load range is depends on whether I'm using a MC preamp or a MM/SUT combo that is further impacted by how the cartridge load is applied. Lots of variables at work here!

The MM version of the preamp (for those who are uninformed- because of the unusual architecture of the phono stage you must have different MM and MC designs) uses FET input devices.
The input impedance of the two devices in parallel is c. 10^13 ohms DC and 12pF, common mode, so in essence, the input looks like a 12pF cap, plus whatever board parasitics, interconnect wiring parasitics, and SUT to phono stage cabling parasitics exist. The total is likely in the order of 20pF, plus whatever on board cap you choose to add.
Assuming that the SUT has a 20:1 gain then this translates out to c. 10nF input capacitance. The output winding cap of the SUT will also be scaled up.
Remember though, that although the DC input resistance is very high the input "bias" current for the FETs still has to have a path to ground, so an SUT with a non-resistively terminated input must not have a blocking cap and should always be connected to avoid unwanted issues. Under these circumstances, if the SUT is removed the output from the phono stage will slew to one or other power supply rail which could have catastrophic effects on your speakers.

The load R on a LOMC acts to both damp the electrical resonance of the coil/load capacitance and to attenuate the output if it gets too close to the DC impedance of the coil- just because it acts as a resistive divider.
For example, the Madake has a DC resistance of 16ohms and about 15uH inductance with a few pF capacitance, The capacitance is often dominated by the interconnect cabling. With the recommended 200ohm R load and no additional cap there is c 0.5dB of attenuation, the resonance is way, way above the audio range, in the MHz area, and the peak is well damped.
Increase the 200 ohms to 47k and the resonance shows many dB of gain, making the cartridge very sensitive to EM pickup.
Put 500nF of cap on the cartridge (c. 1.2nF on the secondary of the 20:1 SUT) and the resonance moves into the audio range, interacting with the high end response.
In fact, as there is a several dB peak in the HF response of the cartridge due to the cantilever resonance, adding 500nF-1uF gives the overall flattest response, but depresses the response a bit in the 5kHz region, changing the audible dynamics- even for old codgers like myself and Miyajima-san.
Reducing the load R damps the resonance progressively more, causing the HF response to fall off as well as reducing the overall broadband output.
With 60 ohms load the broadband loss is c 2dB, but the loss in HF somewhat counteracts the cantilever resonance peak without degrading the c. 5kHz response too much, so some claim that 60ohms is the ideal load...
However, these changes are not abrupt, and obviously are highly cartridge dependent.
I generally run simulations of the loads when I get a new cartridge, or if a friend is getting a new one and asks. Generally, the simulated results and the manufacturers recommendations are in agreement, and I can provide a range of acceptable values and a rough description of the effect as the components vary in value.
I can't say that I've noticed that these conclusions were appreciably different when I used to use an SUT.

 

[AvFan]

Remember though, that although the DC input resistance is very high the input "bias" current for the FETs still has to have a path to ground, so an SUT with a non-resistively terminated input must not have a blocking cap and should always be connected to avoid unwanted issues. Under these circumstances, if the SUT is removed the output from the phono stage will slew to one or other power supply rail which could have catastrophic effects on your speakers.

Thank you for the warning that speaker damage could occur under certain circumstances however I need some help with the terms you've used relate to how I have hooked up a SUT to the V2.0MM. First, I turn everything off before I change any connections or swap out a component like an SUT. Second, I was using a 15:1 SUT with the V2.0 MM with only a resistor on the primary side of the SUT, no resistor on the MM circuit board but did have a 200pf or 300pf capacitor on the MM circuit board. The cartridge, Ortofon Cadenza Red, has an impedance of 5ohms and an output of 0.45mv. Is that method going to potentially cause speaker damage?

 

[mkane]

^^^close to what I've been doing and all of our speakers are still running strong. Speaker damage would have happened long ago.

I think what wyn is saying is don't run the mm side without a load resistor whether it be when using a SUT or not.

 

[wyn palmer]

^^^close to what I've been doing and all of our speakers are still running strong. Speaker damage would have happened long ago.

I think what wyn is saying is don't run the mm side without a load resistor whether it be when using a SUT or not.

Either the cartridge or the SUT without blocking cap will be fine without the input R as long as they are connected. For safety purposes I'd strongly recommend keeping a resistor in place at the input to the phono stage. If the input bias current path is removed the phono DC offset correction servo loop output will be overwhelmed and the output can slew at 10s of volts per second to either rail, which might be catastrophic. Remember, not any old opamp can be used in this design. It requires a number of different characteristics, both DC and AC, to be satisfied for everything to work properly.
If you don't want to use a fixed 47k on the board (or the desired load R multiplied by the SUT voltage ratio squared), then use something about 1-2 Mohm. The input bias current magnitude for the OPA1641 or 1656 is a max of 20pA, or 40pA for two devices in parallel.
The opamps are +/-1mv offset so the bias current contribution to the input offset should be much smaller for the servo loop to retain most of its full range- say 50uv.
40pAx 1Mohm= 40uv. For a 20:1 transformer a 1Mohm translates to 2500 ohms at the input to the SUT, leaving lots of opportunity for a lower value R to be added at the primary.
However, I'll repeat my advice. For many years I used a SUT with the transformed load R on the secondary side of the SUT (either pushed into spring sockets or soldered in place) and never had an issue. I think that the benefits of having the load R on the primary side is minimal at best- more theoretical than actual.

 

[AvFan]

@wyn palmer : Thanks for the safety tip based upon your design. Just to clarify, the formula for determining the cartridge load is for parallel resistors. So say with a large value (say 1Mohm) resistor at the input to the MM stage and an additional resistor on the secondary side of the SUT it will be: 1/desired load = 1/1Mohm + 1/added resistor, correct? And again with a 1Mohm resistor at the input of the MM stage but with the resistor on the SUT's primary side the equation would be (assuming a 20:1 SUT) 1/desired load = 1/2500ohms + 1/added resistor, correct? Thanks!

 

[wyn palmer]

@wyn palmer : Thanks for the safety tip based upon your design. Just to clarify, the formula for determining the cartridge load is for parallel resistors. So say with a large value (say 1Mohm) resistor at the input to the MM stage and an additional resistor on the secondary side of the SUT it will be: 1/desired load = 1/1Mohm + 1/added resistor, correct? And again with a 1Mohm resistor at the input of the MM stage but with the resistor on the SUT's primary side the equation would be (assuming a 20:1 SUT) 1/desired load = 1/2500ohms + 1/added resistor, correct? Thanks!

Yes.

 

[mkane]

Thanks[/QUOTE]