Capacitive Loading Imposed by the Preamp

[richard]

I'm starting this thread in the cartridge subject because loading is so essential to the sound of a cartridge. If we don't apply the correct resistive and capacitive loading to a cartridge, we really don't know how it sounds and what it's capable of.

Resistive loading is pretty easy for normal magnetic stereo cartridges. It's 47,000 ohms. We are discussing capacitive loading, especially in the cables, in other threads right now. Loading is additive. In most cases, we are applying excessive capacitance to our cartridges. The total capacitive load is the arm wiring plus the leads running from the turntable to the preamp. Also included is the load of any wiring in the turntable base (soldering block, etc.) and the headshell (usually negligible). And, of course, the capacitance of the preamp's input circuit.

The stumbling block that i see is, in most cases, an inability to know the capacitive load is that's imposed by the preamp.
- The capacitance meters that I know of cannot properly measure the input circuit due to the influence of multiple devices within the preamp.
- Whereas a few manufacturers will tell us what this value is, most do not.

So, in most cases, we're in the dark about how much capacitance the preamp is adding to the load imposed by the wiring.

If the manufacturer won't tell us, how can we know the capacitive load that the preamp imposes upon the cartridge?

[ecosprog]

Well, all this talk of loading MM cartridges got me digging around in my workshop this afternoon. I figured I had the bits need to put a simple switch box together that would let me try loading my Stanton 681EEE with some different values. This was the end result.





Unfortunately, I only had 100pF capacitiors, so only one value to load with at the moment. More values on order, but these will take a couple of days. My rough calculations give me a total capacitance of approximately 300pF. Need to wait for a new meter to arrive before I can get accurate readings of the whole signal path through to the phono preamp.

[louie3]

Richard I suspect most phono preamp makers these days don't care much about the capacitive load the cartridge will see from their product.

This is either through ignorance, which I find hard to believe, or the assumption that their larger market is for users have moving coil cartridges and are not particularly concerned about the capactive load.

Back in the day, many preamps and preamp sections in integrated amps had phono sections that included some way of changing the capacitive load.

My suspicion is the moving coil thinking won.

It is quite possible most of us have ever heard what a Stanton 681EEE actually sounds like properly loaded.   I am sure you have, but I doubt I can say that for sure.

[RCAguy]

Reese - Looks like you'll have 6 selections of capacitance with that 2-pole (2-channel) rotary switch.  Glad to see the cast aluminum box for shielding - and it looks like insulated jacks that can ultimately be grounded through a single path to the box's ground lug.  For the benefit of others who might rig something similar, either remove or take into account the capacitance inside the following preamp, as well as the preceeding phono cable - these are constant, possibly already too large, and will add to the value selected by the box.  You said you have assorted of caps on order, but remember that in a pinch, you can make many values from the 100pF you have, including by series and series-parallel, the 50pF and 150pF values in my 3-position selector at http://www.lencoheaven.net/forum/index.php?topic=14975.30

To louie3's point, do you, Richard/Gene/KentT or anyone else, know what the popularity/acceptance has been of the moving coil cart v MM/MI?  While the rather small audiophile community might lean toward MC, my sense is that broadcasters and other mere mortals went the MM/MI route.  Even the Library of Congress and other restorers!  If so, why would manufacturers favor a MC approach to (no) capacitive loading?  Did the industry "decide" that all C should be in the phono-to-preamp interconnects?  Or at some point did they think vinyl was too small a market for concern?

[louie3]

It is not so much, as I claimed earlier, that manufacturers ignored the moving magnet user...all phono stages have capacitance at their inputs...it is just, as Richard points out, that most don't don't tell us what that figure is. 

Much as I would like to assign thinking to this, I have no idea why other than to assume the manufacturers either think capacitance does not matter in the overall scheme of things, or that ignorance is bliss.

[richard]

First, I'd like to express my gratitude to all the people who have jumped into this discussion so quickly! I'm feeling very encouraged.

I'm going to reply to all of you from knowledge that I've accumulated during the decades, from what I've noticed and observed. This also includes hunches and perception of trends. I want to say, too, that I am not an expert about this issue: I'm learning along with the rest of you, and it's really nice to have your company.

Robin, I'm especially grateful that you ran the graphs that confirmed the importance of this issue. One thing that I think is evident from looking at your graphs is that we can relax: we have some wiggle-room and don't need to hit the loading right on the nose to get meanfully-accurate music reproduction.

Robin, to your question about the acceptance of moving coil vs. normal magnetic cartridges (moving magnet/moving iron): The majority of audiophiles have used normal mag cartridges. Of the moving coils, most sales have been the classic low-output products. A few have been the high-output models.

Reese, please allow me to recommend that if you don't have a good capacitance tester on hand, that you get one and check your cables soon.

Regarding the US Library of Congress, which is the leading sound archive in the USA, they have long been firmly in the Stanton camp. Stanton used to make custom-profiled styli for them, the same way that they made them for anyone else. Stanton used to grind almost all of their own diamonds, so they had the capability to do this. This was done by either the Custom Products or Broadcast division. The company was so deficient at blowing their own horn that almost nobody knew that they offered this service. In recent years, the LOC has been having this work performed by Expert Stylus. I'd be surprised if they didn't have their capacitive loading spot-on. I'm certainly not clear what the LOC is going to do in the future except that I expect that Expert Stylus will be there for them.

Regarding the component industry, I don't think that they gave much thought to this issue beyond some manufacturers who were thoughtful. I suspect that most manufacturers nowadays take an attitude of, "Look, we've given you a phono preamp, alright? Go away and leave us alone."

I have a Luxman preamp designed by Tim de Paravicini. There's no cap loading spec. I asked him via email. I received no answer. (I was in the process of becoming a small-time Creek dealer during the 1980s. I wrote them asking for some tech information. I received no answer, which is why I became an ex-Creek dealer very quickly. I hope this isn't typical of UK manufacturers.)

Robin, I think that it's always been the case that capacitance loading has been mostly overlooked by the equipment makers. Some people have said to just look at the capacitor that's across the phono input and use that value. To me, this method just seems so hit-or-miss. I mean, logically, it would seem that the first stage of gain would probably impose its own capacitance, too.

I also suspect that the preamp makers could easily overlook the requirements of phono cartridges. These are circuit people; they are not transducer people. They'd probably think to themselves, "Well, we should have some sort of capacitor across the phono cartridge. What value should we use?" "Dunno, boss." "Well, what is Fisher using?; what's Sherwood using? Do we have a Sherwood receiver in the lab? I think I saw one in the corner."

Regarding hi-fi writers, I'll guess that some paid attention to loading capacitance and some didn't. This might explain why some reviews make sense, and others do not, to people who own the products and are analytical.

Louie wrote,

It is quite possible most of us have ever heard what a Stanton 681EEE actually sounds like properly loaded.   I am sure you have, but I doubt I can say that for sure.

Actually, I'm pretty sure that the opposite is true. And in most cases, the cartridge will be facing excessive capacitance. Most wiring that people have been using for phono leads is typical high-capacitance stuff. That's what's easy to find.

At the "electronics store" where "...we've got answers."
"This cable is the best. You can tell when wire is really good: good wire is called 'interconnects.' We're all using it for our CD players. Phono's so much easier, this'll be a piece of cake for you. Look: it's got gold connectors; that's how you know it's the best. Gold is the best for music; that's why flutes are made of gold. Bassoons, too!"

What I'd like for us to come up with is a few different approaches for getting this under control in our own systems.

1. How much variation from spec is acceptable?
2. How can we tell what the preamp's load is, really?
3. Were there any trends or universal standards within the preamps?
4. Are there any methods that we can use to measure preamp capacitance that don't involve unsoldering devices?

[GP49]

For the sake of discussion, here is an excerpt from the General Electric data sheet for its 12AX7A vacuum tube,
showing its specified inter-electrode capacitances.  The 12AX7 is commonly used in tube phono stages.

The GE specifications are typical of those for other manufacturers' 12AX7-type tubes.

[mimoser]

It is quite possible most of us have ever heard what a Stanton 681EEE actually sounds like properly loaded.   I am sure you have, but I doubt I can say that for sure.

… run 1,2 m canare starquad (150pf/m) into a trichord Dino (100pf) and you are pretty much at the 275 pf Stanton claims in their datasheets. You will have to add some capacitance for the 50cm of cable in the tonearm …

sounds great; there is a lot of detail and what I consider uncoloured sound.
)
Michael

PS: what it looks like:

[Turbon]

Good reading...

http://www.tnt-audio.com/sorgenti/load_the_magnets_e.html

Regards

[Rotsch]

Now that i bought the 103 from Chris, i'm also thinking about
switching between different loads.....
First choice i this case is something like THIS !!
http://de.rs-online.com/web/p/dil-und-sil-schalter/6821064/
We'll see.....

[louie3]

Thanks to all...I already know more than I did, or have confirmed what I suspected.

If I may, let me add some capacitance to a statement I made earlier about hearing a properly loaded 681EEE...what I meant to say was that most of us have NEVER heard the cartridge with a proper capacitive load, and what I probably should have said was:

"With a capacitive load we could verify."   I don't doubt the cartridge has sounded good in many systems, but at least some of this was happenstance.

Forgive the misuse of the language...it was late here.

And to Richard, maybe the answer to why manufacturers don't state capacitive loading at the input of their various phono stages is that they don't know??

Last night I went back and forth between a Denon DLS1 and a Stanton 881E (which I have learned is the uncalibrated version of the cartridge).

Same tonearm, same table, same phono stage...Tube (12AX7s---Luxman), running the Denon through a Hagerman Piccolo, the Stanton straight in, of course.

What was really surprising to me was how close in sound the two cartridges were, considering their significantly different generating systems.

The Denon was maybe a little more "Hifi", but not much...the Stanton at equal volume was just slightly "crisp" in the upper registers...I turned it down one step.

Interesting, indeed.

[RCAguy]

Roger - If you're going to be changing carts or doing C-loading experiments, I'd be wary of that "DIP-switch" as it's only spec'd for 1,000 operations, intended for set-and-forget use on PCBs, because the contacts are gold flashed, which wears with switching, and the mechanism is not as robust as a toggle or rotary type.

louie3 - Yes, it's better to know than not to know! - I like to "Trust but verify." (by measurements, then of course by ear, "retraining" my perception if warranted).  An inexpensive  C-meter, sufficiently accurate in the pF range, is fine - I use a capacitance bridge because it also measures dissipation that affects performance for audio purposes.  I haven't seen documented any "standard," but many preamps I've opened (or built as Dynakits, etc.) have used a 150pF capacitor in parallel with 47kΩ resistor for phono loading (half that R and twice that C when paralleling cart channels for mono!) - and because that capacitor trumps any stray capacitance, as Gene's negligible 1.6pF shows for a 12AX7 (chip op-amps are even less), in most designs, it likely can be relied upon without desoldering anything as the preamp's C load, well within acceptable tolerance, as follows...

Richard - I'm amending slightly what I've said elsewhere: for carts including less forgiving Shures & A-Ts, tolerance for load capacitor of +10%/-0% is probably quite good enough, maybe even +20%/-0%.  At least Stanton's thoughtful engineering designs combined their cart's internal R & L with external total load C to resonate in a broad peak, so that intentionally none of these component values is too critical, although savvy users do need to know that they are well in that ballpark of precision.  Too low a C-load causes peaking, which is an audible no-no; dips in response using higher C-load are audibly more benign, and more likely in practice as the capacitances of tonearm, cables, and preamp add up.  While the Stanton 681 is not as robust at so much higher than 275pF, perhaps only to 325 (+18%) where it is -3dB at 15kHz, the 881 particularly is robust, inaudibly affected by up to 375pF (+36%) where it down -1.5dB at 15kHz, interpolated visually from their frequency responses v C-load at http://www.lencoheaven.net/forum/index.php?topic=14975.30

However, do not be satisfied until you attend to a more important issue demanding better than 10% precision: that loading and thus frequency response matches channel-to-channel.  If rolling your own, don't just buy a pack of two ceramic NPO/COG caps; buy a dozen (they're cheap from Mouser) and measure all to select two that are equal even if not as near one closer to the value you want, favoring a pair that might be slightly higher (<+10%) than say 275pF.  Or if in measurements you have discovered some inequality in your cabling, use this opportunity to tune for overall matching of C-loading.

[GP49]

Good reading...

http://www.tnt-audio.com/sorgenti/load_the_magnets_e.html

Regards

The point I take away from this article is that flattening the electrical frequency response of a MM/MI cartridge by optimizing the theoretical
response curve for its inductance, DC resistance, and the load impedances into which it operates, is not enough.  A cartridge manufacturer
could have tailored the coils' inductance and DC resistance, and specified the load, to compensate for non-linearity in the MECHANICAL
response of the cartridge's stylus, cantilever and suspension.  If a stylus-cantilever-suspension has a drooping response at the 9kHz-15kHz
end, creating an electrical peak centered at around 12kHz can approximately compensate for it by acting as a sort of "tone control."  There
are known examples of this (Shure's M44 is one).  

This also explains many early moving-coil cartridges.  They had a mechanical resonance which, as all such, was determined by the
mass of their moving system and suspension compliance.  Mass vs. compliance resulted in a peak just above the audible range, which
was, in most cases, less-than-critically damped by the suspension (damping and compliance are SEPARATE functions), so a rise in the
high frequencies resulted, extending well down into the audible range...remember that 10kHz is only one octave down from 20kHz, and
the effect of a 20kHz mechanical resonant peak will extend well beyond one octave from its center frequency.  

The low inductance and DC resistance of a low-output moving coil means that even a high load capacitance will not roll off that HF rise,
so it becomes part of the overall system response.  A lot of audiophiles called it "detail" or "air" and this resulted in their accepting
moving coil cartridges as their "norm."  Who knows?  What they may have been compensating for could have been their weak
tweeters...or their own rolled-off high frequency hearing.  More than one means to an end, then?  Perhaps the phase anomalies
resulting from the HF droop and the HF peak might have even been complementary!

[RCAguy]

Gene - Nice deductions.  If the overall response turns out flat (and related phase response restored), as in the Stanton 681 and especially 881 curves I posted that were drawn from data with each cart tracking a CBS STR-100v3 test disc, it is evidence that they worked toward compensating all the electrical and mechanical factors.  Similarly by using complementary first-order filters, the RIAA or other characteristic imposed during lathing becomes flat again in frequency and phase once the inverse curve is applied in the replay preamp.  

Furthermore, ever wonder why all Stanton carts are spec'd at 275 pF C-loading?  I think they accepted this de facto standard by users, then worked backwards to design all the electrical and mechanical stuff to produce the best outcome.  275pF became accepted by other mfgrs (I don't know who was first?), as it's reasonably achievable with decent phono cable of a typical 4ft length, and with ~150pF typically inside the preamp trumping well any stray capacitance, therefore 275pF seems a well-reasoned target.

Since 275 is not a round or sexy number, we infer that the marketing dept lost this one to the scientists!  So we owe it to such thoughtful engineers - and ourselves, maximizing our investment in, and ultimately our enjoyment of, the products they designed for us - to implement using them correctly.  Matching channels to 275pF +10~20% is not that much to ask of us; you only need to do it again if you change something.  And as flatness of response indicates compensation of a cartridge's various mechanical & electrical systems is complete, we should not do our equalization for speaker or hearing deficiencies by mis-loading the cartridge - that will lead to unintended consequences!  And we wouldn't attempt RIAA equalization by cartridge loading!  Much better to address the speaker issues independently, and if you haven't adapted sufficiently to aging ears, use proper EQ sparingly to compensate.

I do not understand what audiophiles mean by "air"; looking at typical discs on a spectrum analyzer, what "details" other than preamp hiss could they be hearing in recordings containing little >8kHz, and no signals >15kHz?

[richard]

I agree with your assumption, Robin. Until their descent into a hyped-up marketing blowtorch on crack, Stanton was a disciplined engineering-driven company. I think that your logic is correct and that's how these people worked.

I also concur that we should never try to alter the sound of a system by doing any tuning at the cartridge level. Tone controls can be our friends, especially if we are listening with studio monitors and someone's blowing a close-miked trumpet right into our head.

Regarding 275 picos as an industry-standard load, I'm not sure about this. I've seen enough cartridges with a higher spec to think that this isn't the case. But I have not surveyed this, just noticed a few that are up around 400. Ortofon has stated the figure as a range, rather than a single number. And that range is pretty broad. I recall it something like 275-400. I could be wrong.

Do you think that it's reasonable to assume that 150 picos is a common load within the preamp?  Having measured a bunch of garden-variety shielded cables a few years ago, I found them to be pretty high in capacitance, and tagged them as "not for phono." I haven't surveyed the recent "interconnect" offerings.

Look, guys: I know that I started this capacitance thing, but I've picked up a very nasty virus on the main computer. I'll have to spend major time attending to that (I'm coming at 'ya on system #2). So, I'll have to take a back seat here for a while. Please carry on without me with my appreciation!