Added an SUT...not sure I understood this

@woofhaven1992 

You remain incorrect, however, about the correct loading with the SP14, which contains neither a head-amp or an SUT built in.  

Wrong. The SP14 has a FET based input at the front end of the phono stage., Its same principle as a head amp. Correct loading is 100ohms minimum as per the manual.

I don't have an issue if you prefer 25ohms, thats your ears. I post so that others reading can get the best out of their cartridge using the minimum loading as recommended by the manufacturer. My own experience with setting up many OC9's of various types in multiple systems is that around 400ohms is optimum and it is clear that you have no interest in even trying alternative loading.

@dover Feel free to dunk, it's apparently important to you.  Yes, you are right that I initially indicated the incorrect sub-species of the OC9. You remain incorrect, however, about the correct loading with the SP14, which contains neither a head-amp or an SUT built in.  

@dover 

This splinter of the OP's tree is in reference to the load a cartridge sees and in particular any variance from the cartridge manufacturers suggestions. 

Lewm was comparing active mc amplification vs SUT's.

No....  Lew was wondering in what world  20Ω = 100Ω. How can 20Ω be unacceptable from an active stage but perfectly OK from a SUT?  I entered my guess at an explanation that the 20Ω value is not an actual load value but the recommended input to use of the AT1000T.  This input happens to load the cartridge at just over the 100Ω manufacture spec.  Simply put a somewhat confusing manufacturers data sheet reconciles the difference and keeps the manufacturers suggestion consistent.  

It is clear from your conclusion that you do not understand how current mode and voltage mode gain stages work

Other than the load value presented to the cartridge, how a gain stage operates has zero bearing on the topic of cartridge loading.  My only reference to current amplification used a specific model (MCCI) since it has a published input impedance number of <3Ω.  To be clear... it is the 3Ω that is the issue here NOT that it is a current amp.  From the most basic perspective I want to know from you how a <3Ω load can be used with a cartridge that has a 100Ω minimum recommended load and possibly sound good?  I can't be the only person who sees this disconnect.

Since you seem unwilling or unable to answer that simple question, let me frame the same situation differently:

Can we agree that the OP's cart into a 1:20 will see just over a 100Ω load and present approximately 8mV to the input of a "generic" gain stage?   Now lets change that 1:20 to a 1:50 and then place an 11Ω resistor across the primary to add an additional load to the cartridge directly.  The voltage output of this combo will be the same 8mV into the same "generic" gain stage. Now the load the cartridge sees will be 7Ω which is well below what you would deem acceptable.  Both the measured and the subjective sonic results of the above experiment surprised me and really got me thinking about what loading a cartridge really does. It also makes me wonder what the manufacturers published values are based on.  

dave

@intactaudio 

I concur with lew and it makes no sense to me how a loading a cartridge can have different "ideal" values for different amplification types.

Wrong. That is not what @lewm said. He said -

I’ve never understood why the minimum load can be acceptably different for the same cartridge with vs without a SUT in the circuit. 

Lewm was comparing active mc amplification vs SUT's.

You really do need to read posts more carefully, unless you are deliberately misinterpreting posts.

It is clear from your conclusion that you do not understand how current mode and voltage mode gain stages work, and I would suggest if you are genuinely interested then you should do some study - there are plenty of high schools that run after school programmes that are excellent.

Dear @woofhaven1992  ; This was my very first post in your thread:

And this your answer to that:

" 

Yes I have been using the SP14 with no SUT for some time, and the sound is great except for a touch of hum just at the level of the noise floor. I decided to fiddle with this SUT as a way of addressing that problem, which otherwise has been  persistent despite lots of attempts to address it, including having ARC work on the pre-amp twice.

You are correct that through the SUT I lose some resolution,  although I was blaming that on the crappy built-in output cable. I figured I could find a way to replace that but maybe I should, as you say, forget it.  "

You answer should been the end to your thread. Don't you think?

R.

@dover 

I cannot fathom how you cant see the difference between increasing gain and reducing gain.

again... not the topic at hand.  The cartridge doesn't magically see a sut and say 20Ω is fine and scoff at that same 20Ω if it is a load resistor to ground before a Jfet.

I concur with lew and it makes no sense to me how a loading a cartridge can have different "ideal" values for different amplification types.  I think much of the confusion here stems from AT expecting you to pair their cartridge with their SUT.  The AT1000T has a 20Ω input setting that claims a 26dB gain and specifies a 47kΩ termination.  Simple math tells us that 26dB = 1:20 which reflects back 117Ω to the input labeled 20Ω.  In this convoluted  case 20Ω really does = ~100Ω!  Assuming there are not other loads or networks inside the AT1000T to truly provide a 20Ω load,  this small part of the discussion now makes sense.    This labeling system is typical to the Japanese SUT's and the Denon that started this whole thread uses a similar input labelling method.  Essentially the inputs are labeled  with the approximate impedance of the cartridge they expect you to feed it with with no hard and fast rules as to how those numbers were devised.

Now back to the OP and his finding that a 20Ω actual load sounds best.  Sure he possibly got to that value by following a bunch of conflicting / confusing data but ultimately that is where he stayed.

dave

I thought it was ok to mention current driven phono stages after the OP himself mentioned it.

what we found was that the SP14 had a 100 ohm resistance across its phono inputs. That coupled with a 1:10 SUT resulted in the cartridge seeing a 1 (one) ohm load. Which is clearly a bad thing and explains the bad SQ.

Per Dover’s point about the two different AT cartridges, I’ve never understood why the minimum load can be acceptably different for the same cartridge with vs without a SUT in the circuit. (Maybe not applicable here if indeed we’re talking about two different cartridges.) But you see this among those who use for example the DL103.

@intactaudio 

Lets get back to the simple question I asked....  If going well below a cartridge manufacturers load is not recommended, how can we reconcile the use of transimpedance amplification or in your case a load that is 1.7X the cartridge internal impedance?

It is my belief that when specified by the cartridge manufacturer the load value range would relate to the behavior of their cartridge and not in response to the unknown capabilities of what follows. 

This is where you go wrong. Loading down a MC produces more current and less voltage. ( explained by Carr above ).

Therefore whether the mc phono input is voltage mode ( as in SP14 ) or current mode  is vitally important because in voltage mode loading down the MC reduces voltage going into the voltage based phono and reduces overall gain.

With a current mode phono, loading down increases current and hence increases overall gain.

I cannot fathom how you cant see the difference between increasing gain and reducing gain.

@woofhaven1992 

Looks like you've got the wrong cartridge.

You claimed to have an ATOC9XML

Paoerwork here

Note the recommended minimum load is 100ohms as I posted.

The paperwork you have posted appears to be from the ATOC9ML/II which is a different cartridge. 

However the 20 ohm recommended minimum loading referred to is the minimum recommended loading when using a step up transformer. The recommended loading for a head amp is 100 ohms. See see up instructions #6.

Here is the paperwork for the ATOC9ML/II

In your case your SP14 has a FET based front end, not a step up transformer, and the recommended minimum load is 100ohms.

@intactaudio I misunderstood the intention behind your post then.

Unfortunately, the question you have asked has a rather uninteresting answer. Everyone on this thread has been assuming I determined that the best sounding load for my cart was a large deviation from spec, apparently because one participant loudly and repeatedly felt the need to insist I did not know how to read the instructions that came with my cartridge. I have attempted to clear up that confusion by asserting that my cart in fact has a 20 ohm spec, but apparently I don't know whatever secret handshake a person needs to know in order to be taken at one's word on these forums.

But none of that changes the fact that I am loading my cartridge at spec, which might render your question moot. I do agree with you, though, that if I achieved better sound at a different load than that recommended, I would use what sounds best. 

You are correct in your conclusion, but IMO asking the wrong question.

I am the only one here asking the relevant question in this situation. This all goes back directly to the insistence that something was broken in your system if your cartridge sounds better with a 20Ω load.  The only question I asked (and has yet to be answered) is how does one account for the use of a load well outside a manufacturers recommendation (ie a current amplifier as an extreme example) and still adhere to the wishes of the manufacturer?

In your particular case I trust your assessment that the sound is good and suspect that anyone that insists otherwise based on a published spec is in error.  Who knows I may be he only one with this belief.... it wouldn't be the first time.

I did not  drag this off topic into how current amps differ from voltage amps, others took it there.  I really think this is a simple question that apparently nobody seems to have a good answer to.

dave

from the cartridge POV, how can 3Ω be different than 3Ω?  I am speaking primarily about the effects of the load on the electromechanical behavior of the cartridge and what impact that may have on the sound.

You are correct in your conclusion, but IMO asking the wrong question.  Rather than looking at it from the cartridge's perspective, consider the pre-amp's perspective.  The pre-amp is the device that will be passing along a signal to the next component in the system.  Both IV and traditional preamps "see" the entire "electromechanical behavior of the cartridge" but neither pays attention to the full set of that behavior.  One type of circuit pays attention primarily to the current, and the other, voltage.  Thus the two methods can look at a cartridge behaving in very different ways but end up paying attention to--and passing along--signals that are similar in character; i.e., if the cartridge's current under heavy load is similar in character to its voltage under a light load, expect an IV preamp to pick up a similar signal from the cartridge through its low impedance input terminals as a traditional preamp would through its high impedance (and lightly loaded) input terminals.

Ralph...  we have been over this.  I made a point above to never mention ground and simply referred to the cartridge as a two terminal device feeding a two terminal load.   If the goal of that load is to maximize current draw then by necessity it must be low in value.  Since both Ohm and Kirchoff must be obeyed.... all of the current leaving one terminal of the two terminal cartridge must traverse the load and return to the other terminal.  The application of Ohm to this situation clearly defines all of the parameters at hand.

dave

the industry standard of 47kΩ was adopted from the MM cartridge world and applied to the MC realm because is first and foremost "does no harm".... until you throw a SUT into the mix that is.  Once a SUT is added the load the cartridge can see can get vanishingly low to the point where you actually start losing gain as you increase turns ratio.

Most SUTs are meant to drive a 47,000 Ohm load since they are also built to work with a specific cartridge. Generic SUTs (like Jensens) might also be built to drive 47K although with some additional loading to allow it to express the turns ratio correctly with different cartridges (which have different source impedances). Some Jensen transformers are meant to drive lower impedances but IIRC that's still about 6.8KOhms.

Lets get back to the simple question I asked....  If going well below a cartridge manufacturers load is not recommended, how can we reconcile the use of transimpedance amplification or in your case a load that is 1.7X the cartridge internal impedance?

A transimpedance input is a special beast. It requires an opamp and when you set up an opamp circuit with feedback, you get something called a 'virtual ground' where the input signal meets the feedback signal. Its not actually ground ('virtual' means 'almost or nearly as described, but not completely or according to strict definition.'). This is why "virtually 100%" is not in fact 100%...

Anyway, in a transimpedance amp the cartridge is substituted for the input resistor. The significance here is that if the virtual ground were actually ground, the signal going thru the input resistor (or coming from the cartridge) would be snubbed at the virtual ground- the signal would stop dead. But it doesn't!

The lesson here is that virtual ground isn't in fact ground. Thus while its impedance might appear to be quite low, it does not function as if its impedance is actually that low. IOW it does not have all the qualities of an actual ground!

Its confusing yes, but in actuality the cartridge isn't driving an impedance that is at or near ground (which would be 0 Ohms or very nearly that). Its actually driving something quite a lot higher.

The gain of an opamp circuit is defined by the input resistor vs the feedback resistor. For example if the input is 100 Ohms and the feedback is 1000 Ohms, the gain will be 10. Since the cartridge is a variable source impedance from cartridge to cartridge, the gain of the circuit will vary from cartridge to cartridge too- more gain with a lower source impedance, less gain with a higher source impedance. This means the output will not vary that much as you change out the cartridge, but it also means that the circuit inherently has limits (due to the amount of feedback applied) and won't work with every cartridge.

This is a second take on woofhaven’s post, nothing to do with what Intactaudio posted above, or only tangentially. Woof, you seem to ignore the fact that none of the available current driven phono stages actually presents a "perfect" zero impedance load to the cartridge. Each of the different units has a finite input impedance that is above zero and usually less than 10 ohms. Case in point is my BMC MCCI that I bought recently to experiment with current drive in my home system. (Otherwise, I would have to depend upon others’ opinions, which I don’t find helpful to do.) The MCCI has an input Z stated by the maker to be <3 ohms. I guess that’s why Intactaudio uses that value in his thinking process. In other words, none of these devices is "perfect" when it comes to current drive, because they really cannot be perfect; zero ohms is a short circuit to ground. My experience with two or three very good LOMC cartridges driving the MCCI is that you cannot predict in advance what you’re going to get out of the phono stage in terms of either gain or tonal balance. I’ve had to think of the cartridge and the MCCI as a single device where a major determinant of the output SQ and db is taking place at the interface between the cartridge and the MCCI I/V stage. The cartridge internal impedance, its voltage output (because that affects current output), and probably its inductance along with the ~3ohm input Z and the current amplifying devices of the MCCi would seem to be the important variables. So, there’s no free lunch, and there is nothing inherently superior about current drive vs voltage drive, at least so far in my experience. This is not to say that the results cannot be superb.

@woofhaven1992 

 The reason a cartridge sounds different under a heavy load is because the low load impedance suppresses output voltage.

so add more gain.  I'm not trying to be cheeky here but this becomes a problem related to what happens downstream of the cartridge and not with the load the cartridge sees.  If you load a cartridge severely and do not have a capable low noise good sounding stage after it things will indeed go south.... this is not a problem with the cartridge load per se but with the behavior of what follows.  

In the case of the IV preamp, the low input impedance (not quite the same as load impedance)

how can the two be separated form each other?  from the cartridge POV, how can 3Ω be different than 3Ω?  I am speaking primarily about the effects of the load on the electromechanical behavior of the cartridge and what impact that may have on the sound.

 but those changes cannot be attributable to the effect of loading on voltage output since the IV preamp effectively ignores that effect.

Sure the IV preamp can ignore that its low input Z effectively reduces the output voltage to 0.  My question is how can the cartridge ignore that it is seeing a load that is possible a fraction of its internal impedance?

Lets get back to the simple question I asked....  If going well below a cartridge manufacturers load is not recommended, how can we reconcile the use of transimpedance amplification or in your case a load that is 1.7X the cartridge internal impedance?

It is my belief that when specified by the cartridge manufacturer the load value range would relate to the behavior of their cartridge and not in response to the unknown capabilities of what follows.  In any case... Like you I feel that number is only a suggestion and people should feel free to use whatever load sounds best to them.

dave

Yes, JCarr sort of contradicts himself a bit and ends up being half right.  The reason a cartridge sounds different under a heavy load is because the low load impedance suppresses output voltage. But a preamp that is uninterested in output voltage would not be affected by this phenomenon.  In the case of the IV preamp, the low input impedance (not quite the same as load impedance) results in high current flow through the preamp.  It does for the IV preamp what a high load & input impedance does for a traditional preamp--maximizes the electrical signal that the preamp is using.

Changes in sonic character between the two approaches may be based on inherent differences in the architectures, or based on differences between the way the cartridge's--let's call it "frequency response"--as expressed by its current output varies from its "frequency response" as expressed by its voltage output, but those changes cannot be attributable to the effect of loading on voltage output since the IV preamp effectively ignores that effect.

JCarr makes the exact point I am ultimately trying to get to when he says the below.

 Also, IV phono stages sound qualitatively different to voltage amplification phono stages., and part the reason is that undoubtedly the cartridge is forced to operate into a zero-ohm load (or some other values that is quite close to the cartridge's internal resistance).

I don't necessarily agree with his use of the term "forced" but that is for another day.  He attributes  "part of the reason" for the difference  in sound to the load the cartridge sees and I fully agree with that thought. Since this branch of the  topic is about the load the cartridge sees and not how that load is obtained, discussing the difference between current and voltage amps is not at issue.  It just so happens that a current amp gives an extreme example of "non-traditional" loading.

 A cartridge is a two terminal device that sees a two terminal load downstream.  It doesn't know or care what form of amplification that load takes, it just does what the load tells it to do.  

dave

@intactaudio 

Obviously you haven't got a clue so I'll explain it to you - actually here's a quote from JCarr

The loading resistor value is placed across the phono cartridge's output terminals, which means that any output voltage produced by the cartridge will be forced to flow through the load resistor in the form of current. The higher the load resistor value is, and the farther the net impedance is from the cartridge's internal resistance, the less current will be produced. Conversely, the lower the load resistor value is, and the closer the net impedance is to the cartridge's internal resistance, the more current will be produced.

IMHO, the only time that forcing the cartridge to produce more output current could be justified is when using it into an IV phono stage. Since this kind of phono stage converts input current into output voltage (and is more or less oblivious to input signal voltage per se), it would make sense to feed an IV phono stage with as much input current as possible. An IV phono stage, however, possesses de facto a low impedance input node that receives the output current from the cartridge as a series element (without needing a separate load resistor). This is not how a normal loading resistor works with a typical voltage amplification stage, since there the separate load resistor bleeds the current from the cartridge into ground (thereby wasting that energy). Also, IV phono stages sound qualitatively different to voltage amplification phono stages., and part the reason is that undoubtedly the cartridge is forced to operate into a zero-ohm load (or some other values that is quite close to the cartridge's internal resistance).

In other words comparing the resistive load in a current mode mc input to the the resistive load in a voltage gain mc input is like comparing apples and oranges.

If you cant understand what JCarr wrote, and don't understand the difference between voltage and current then I cant help you any further.

The maximum output of the cartridge will be with little or no load. 47K is the industry standard in this regard.

the industry standard of 47kΩ was adopted from the MM cartridge world and applied to the MC realm because is first and foremost "does no harm".... until you throw a SUT into the mix that is.  Once a SUT is added the load the cartridge can see can get vanishingly low to the point where you actually start losing gain as you increase turns ratio.  My best guess is that the min load value specified for a cartridge is a value that still assures nearly full specified cartridge output and has little or nothing to do with the sonic behavior of the cartridge.   

What I want to know is why the minimum load value that has to be strictly adhered to in the case of a traditional phono stage can be completely ignored in the specific case of the MCCI with a published input impedance of <3Ω.  

This is completely ON topic since it relates directly to the OP's insistance that a 20Ω load sounds best on his AT-OC9XML.  

dave

Do you not think the cartridge designer and manufacturer knows what they are doing when they recommend a minimum of 100ohms ?

They know that the designer of the phono preamp may not have taken the RFI generated by the cartridge and tonearm cable into account. The 100 Ohm resistor detunes the resonance that they create and so prevents RFI from messing with the preamp.

If I had to guess that spec is set to assure near maximum output of the cartridge. 

The maximum output of the cartridge will be with little or no load. 47K is the industry standard in this regard and qualifies in this regard.

Dear @dover  : He already did it in this thread and in the cartridge loading one too. I can't understand why came here again with the same topic but is up to him:

R.

@intactaudio 

Then why don't you start your own thread on "loading - voltage vs current mode amplification of moving coil cartridges".

Do you understand the difference between voltage and current ?

If you start a new thread there are many knowledgeable folk on the forum who can help you.

@dover 

the last post in this thread from the OP is below:

@dover The sheet that came with my cart says 20 ohms. AT has not been consistent with their specs on the OC9 series.  More importantly, it sounds better around 20 ohms than at 100.  Plenty of people agree with me on this.  I'm not going to load it at 100 just because you won't stop yelling at me.  I'm going to load it at whatever achieves the best sound in my system. 

to which you responded with the text quoted in my first post here where you insisted there must be something wrong with his system.  

I applaud and support the OP's POV that what sounds best is best. I brought current injection into the topic because it is a perfect example on how an amplification method that many find appealing provides a load that is in stark contrast with what any cartridge manufacturer suggests as optimal.  I for one would like to see more discussion about how to reconcile this discrepancy.

dave

@intactaudio 

Once again, please read my posts more thoroughly - here is what I posted initially

What I would suggest is get a pair of 470ohm resistors and try running the OC9 straight into your SP14 loaded at 470ohms. This will give you a direct comparison with and without the step up with the same loading. 

The SP14 does not use current injection mode. Your insistence on bringing into the discussion another variable ( current mode ) that is irrelevant to the OP's issues only serves to add more confusion to an already confused OP. 

apologies to all for veering off topic here....

@dover 

I have yet to se actual measured frequency sweeps from a MC cartridge that show the ""rolled off" top end you suggest above.  Can you point me to some?

The only reason I took the time to write my post with regards to published specifications and recommended load was to inform the thousands of readers of these forums whom do not post, but might rely on what they read in the thread.

I have the same belief.... kinda...  I was just trying to point out to the 'thousands of readers of this thread' that the whole concept of "current injection" flies 100% in the face of blind adherence to the manufacturers minimum load spec.  ie...I see the natural conclusion of your logic to be:  if 100Ω is the minimum <3Ω cannot sound good so it need not be tried.

dave

@intactaudio 

I know of several people who routinely load MC carts at 3-5X the internal impedance.  They do this out of sonic preference

Yes - I have seen this phenomenon myself when I was a high end audio distributor in the 80's. Usually it is psychological, they like an effete sound or rolled off top end because their mother told them off too much when they were young.

Odd... it seems to me that insisting one trust a published suggested load over their ears is just the opposite of open minded.

If you read my earlier posts correctly I suggested he should try the higher loads. The OP has chosen not to try higher loads.

The only reason I took the time to write my post with regards to published specifications and recommended load was to inform the thousands of readers of these forums whom do not post, but might rely on what they read in the thread.

@perkri 

"The reason the hum goes away when the SUT is added is because the DC in the signal isn’t passing through the transformer."

Please explain where DC would come from. The signal entering the SUT is generated by the cartridge and is a AC signal, there is no way it can generate DC.

BillWojo

@dover 

Do you not think the cartridge designer and manufacturer knows what they are doing when they recommend a minimum of 100ohms ?

If I had to guess that spec is set to assure near maximum output of the cartridge.  Changing the loading value makes subtle but easily audible changes for many.  I know of several people who routinely load MC carts at 3-5X the internal impedance.  They do this out of sonic preference and not out of ignorance or to thumb their noses at "accepted practice".  I have yet to see a published minimum load that comes anywhere near the <3Ω of the MCCI which precludes their use in general if one were to adhere to the manufacturer's suggestion.

but for those who are more open minded

Odd... it seems to me that insisting one trust a published suggested load over their ears is just the opposite of open minded.

I am not saying that the OP is correct in their 20Ω choice or exactly why that value was chosen.   It may very well be that the 4dB reduction in cartridge output from that cartridge-load ratio may be partially responsible for some additional noise.  The point I was / am trying to make is the disconnect between minimum recommended load and the use of a stage such as the MCCI.

dave

@intactaudio 

He hasn't solved his problem, although he now knows why the SUT did not work ( there was a 100ohm loading resistor already in the SP14 ),  I believe the last post I saw was that he still had hum issues.

He is not using a current injection stage.

Apparently he is going back to loading at the tonearm 25ohms, which means that given there is already a 100ohm loading resistor in his SP14 that he is effectively loading his 12ohm cartridge at 20ohms. If you think this is a good idea - good luck to you.

I really don't care what the OP does - but for those who are more open minded, I have set up plenty of OC9's over the years - in most instances where customers were using solid state phonos ( excluding current mode ) I found optimum loading usually around 400-470ohms.

Do you not think the cartridge designer and manufacturer knows what they are doing when they recommend a minimum of 100ohms ?

@dover 

The recommended loading for your cartridge is Min 100 ohms.

Min means minimum

and

If it sounds better at 20 ohms, there must be something wrong with your system.

since the OP has solved his issue I have to ask if 100Ω is minimum and 20Ω means something is wrong, what would happen if he used the MCCI current injection stage that would provide a less than 3Ω load and liked the sound?

dave

@rauliruegas

Just finished building a Hiraga Super 30 watt Class A amp

500VA transformer, followed by 192KuF of capacitance in the power supply

Toroidal transformer hums on start up, had to instal a Hypex Soft Start board so as to manage the inrush of current on start up as the magnetic field stabilizes.Humming stops once warmed up.

In an SUT, traditional transformers are used, and the discussion had to do with electrical hum, and not physical hum.

The reason the hum goes away when the SUT is added is because the DC in the signal isn’t passing through the transformer.

Dear @perkri  : Please read this link about DC in the signal. Was instructive for me:

https://www.diyaudio.com/community/threads/toroidal-transformer-noise.279794/

R.

Dear @atmasphere  : " Failing filter caps have a way of eating power transformers in older gear and you really don't want that happening!  "

Good to know that because this year I changed the filter caps in my 35 years old amplifiers.

Every day is a learning day.

R.

Seems to me, you have DC in the signal. DC will not pass through the transformer(s), which is why the hum is going away

In a 30 year old unit, hum is not surprising and it need not have anything to do with the subject. Your logic makes some sense regarding phono stage gain. Good point about phono gain at tape outputs being 46db. The fact remains that a SUT is not needed.

@lewm From specs listed in the SP-14 Sterephile review:

Phono stage gain: 66dB (to main out), 46dB (to tape-out). Phono sensitivity: 0.25mV at 1kHz for 0.5V output. Line stage gain: 20dB (to main out), 0dB (to tape out). Line sensitivity: 50mV RMS at 1kHz for 0.5V output.

I think the clue is that they list phono stage gain: "66dB (to main out)" BUT only "46 dB (to tape-out)". To me, that indicates they’re adding the line stage’s hearty +20dB gain to get that 66dB figure. If the phono stage were doing 66dB before its line stage, then that should have applied equally to the tape-out.

It seems like it was a thing for C-J and ARC full-function preamps of this era: standard MM phono stage, but with GOBS of downstream line-stage gain to cover users who wanted to run MC cartridges anyways. I think I recall seeing spec’d phono stage gain on some of these preamps worded like "including line stage gain". It was a weird era. Fortunately this practice has fallen out of favor, as it’s suboptimal. The extra gain for MC carts should be before the RIAA stage, where it belongs - and not in the line stage, where it just increases noise floor for modern digital sources that are 2V - 4V or more.

And I agree with @atmasphere , the hum in direct connection is concerning, and should be looked into! Also the capacitor age, though op did mention it was recently serviced...

Again:

getting to the bottom of the hum thing might be a good idea too. SP-14s are older ARC preamps if memory serves (and a quick search says it was introduced in 1989...) - when was the last time the filter capacitors in it were replaced?

If the answer is 'never' then it would be a good idea to get it serviced out. Its old enough that filter caps (including those in the DC filament supply) can be failing. Failing filter caps have a way of eating power transformers in older gear and you really don't want that happening! It smells terrible and you can expect a transformer like that to be really expensive if you can even find it.

mulveling, Perhaps you know this unit much better than I do.  My info was taken from the "specifications" section of a 1990 review of the SP14, published in TAS. Based on those data, it appeared to me that the phono section alone has a gain of 66db and that the linestage per se adds 20db of gain.  But I certainly could be wrong.  This is from a second source:

• Preamp Output: 2V (50V max)
• Frequency Response: 5Hz to 50kHz
• Distortion: 0.01%
• Gain: 66dB (mm), 20dB (line)
• Input Sensitivity:
• Signal to Noise Ratio: 78dB (mm), 98dB (line)
• Line Output:
• Dimensions: 480 x 134 x 260 mm
• Weight: 5.5 kg
• Year: 1989

Apparently, if you wanted to use this unit with an MM cartridge, you could do so via a selector on the face plate that selects for various levels of attenuation of its phono gain or perhaps of its linestage gain.  Kind of strange. Hence the inclusion of another selector that allows the user to add phono input capacitance.  So if you really really must have a SUT in the signal path with an MC cartridge, you would need the 47K ohm load resistors, not 100 ohms, and you could select for attenuation of phono gain on the front panel. That's a minefield for the novice.

@mulveling  : " and resume using it with the SUT! "

With all respect certainly you don't know what you are talking about in this OP specific issue.

R.

@woofhaven1992

Glad we got that mystery solved!

As for the MC vs. MM - perhaps semantics but I wouldn’t consider the SP14 as having an MC stage. An MC stage would have a ~ +20dB headamp or SUT up front *before* the RIAA stage. This SP14 has a 44dB RIAA stage up front, and then a high gain 20dB line stage AFTER this. Maybe on paper that’s fine, but personally I haven’t found good results feeding too low a signal level into an MM RIAA stage and then "making up" for it with line stage gain. This suboptimal arrangement may partially explain why OP liked it with such a low loading impedance (20 ohms, or even less if in parallel with 100 ohms), in an effort to change the sound.

I disagree with ARC’s decision to solder 100 ohm resistors on the phono input. That absolutely kills any flexibility. With 47K fixed input, a user could always get a loading kit (like from DB Systems) if they needed to change loading for MC direct-input.

Also agree with others - a 12 ohm coil like this AT cart should be loaded at LEAST 100 ohms or greater. As long as this 100 ohm resistor is fixed on the phono input, that means absolutely no SUT, and no additional loading should be used. You’ll start losing dBs and modifying frequency response as you move lower.

Personally I'd clip out those 100 ohm resistors and resume using it with the SUT!

There's nothing per se wrong with having those 100 ohm load resistors on an MC input, most of the time.  So, I would not see the point of removing them except if one wanted to load an MC cartridge with a higher resistance.  The problem seems to have been connecting a SUT to a pair of MC inputs in the first place, unless I have missed something.  I finally broke down and looked up the specs of the SP14. It seems to have only an MC or high gain (66db) pair of phono inputs.  The nominal load is said to have been 47K ohms, but the reviewer (in TAS) says his unit was factory loaded with 100 ohms, like the OP's unit.  So, no SUTs allowed or needed here, hum or no hum.  Especially since the linestage section is said to add 20db of additional gain.

Dear friends: This is what posted the OP:

"" @rauliruegas Yes I have been using the SP14 with no SUT for some time, and the sound is great except for a touch of hum just at the level of the noise floor. I decided to fiddle with this SUT as a way of addressing that problem, ....""

So, he already knows that in that way the QS is excellent but that hum. Nothing is perfect and always with trade-offs but that SUT the OP choosed is a bs of SUT, it's a wrong path not only for that cartridge but for any cartridge in any audio system.

In the other side if that hum is at the system noise floor then it's not really a problem to die for.

R.

@lewm is right and so is @dover . At this point IMO the thing to do would be to get someone to remove those 100 Ohm resistors. SUTs are usually designed to drive 47KOhms (although Jensen makes one designed to drive opamps that expects to see 6.8K).

By the way, when we speak of MM inputs vs MC inputs, those are euphemisms for low phono gain (usually 40 to 50db) vs high phono gain (usually >60db). Also with your SUT set at its 40 ohm mode(1:10 ratio), and into those 100 ohm resistors, the load seen by the cartridge is 1 ohm! 

You came here for help, and several knowledgeable people have tried to help. You ought to be open minded about accepting advice, since you clearly do not have an in depth understanding of what you’re doing. There’s nothing wrong with that, unless you continue to insist on many of your misconceptions. Dover pasted the spec sheet here, so you can see that the minimum load R is 100 ohms. At 20 or 22 ohms the cartridge ought to sound weak and dull, because a large fraction of its output would be lost to ground and high frequencies would be rolled off. The fact that you found 100 ohm load resistance in your unit proves you’ve been connecting the SUT to its MC inputs, because there’s no way ARC would have installed 100 ohms at MM inputs. So if you put your SUT away, just directly drive your MC inputs and sayonara. Like Raul said.

@woofhaven1992 

The sheet that came with my cart says 20 ohms.

I don't believe this. All OC9's are recommended 100ohm minimum. I used to sell them. If it sounds better at 20 ohms, there must be something wrong with your system.

Perhaps you could post a picture of the sheet you are referring to.

@dover The sheet that came with my cart says 20 ohms. AT has not been consistent with their specs on the OC9 series.  More importantly, it sounds better around 20 ohms than at 100.  Plenty of people agree with me on this.  I'm not going to load it at 100 just because you won't stop yelling at me.  I'm going to load it at whatever achieves the best sound in my system. 

@woofhaven1992

the cart sounds better when I load the tonearm with 22 ohms than when I load it at 20 ohms, which is the spec for this cartridge.

NO

How many times do I have to post. That is not the spec.

The recommended loading for your cartridge is Min 100 ohms.

Min means minimum

It is not 20 ohms

It is not 22 ohms

It is not 40 ohms

Look at the Audio Technics website if you don’t believe me

Well this has been interesting, and now the mystery is solved.  The prize goes to @mulveling for correctly deducing that when ARC last serviced the SP-14, they soldered 100-ohm load resistors in there.  They did know I was using an MC cartridge, but wow it would have been nice if they had said something.

Anyway, this explains a lot.  Not only does it explain why the SUT behaves as it has, it also explains why when not using the SUT, the cart sounds better when I load the tonearm with 22 ohms than when I load it at 20 ohms, which is the spec for this cartridge.

I am definitely not going anywhere near that circuit board with a soldering iron, so there will be no more experimenting with this SUT for me.   I'm going to try 25 ohms on the back of my arm, and if I like that sound, start tracing everything I can to find a way to eliminate that last remaining hum.  Thanks to everyone who contributed.