Tone arm cables, what should I choose?

[floppybootstomp]

Well dang, it appears that model is now discontinued, Rapid are listing it but have none in stock. That's a shame cos that model, I noticed, also measures frequency & temperature.

Conrad have it but it's priced at 189 Euros and their postage costs are expensive.

Time to go searching.

[richard]

Andy and Tony,

(Please accept my apology for not remembering precisely who said what; I can't see your names in this text entry mode.) It would be really interesting to take a good meter to these good audiophile cables so we can get a handle upon the actual capacitance per unit of length. Also, allowing for some bravery, with the meter connected, see if capacitance changes when the wire is bent and/or crimped. We want to know how it behaves under real-world conditions.

For me, the real bitchy part of all this is the wiring in the arm. This is the place where so much capacitance is, of necessity, really loaded into this loop. In the arm setup that I'm planning, this may be the most difficult part of the design.

Your cartridge list is quite interesting, and it looks to me like your media interests are quite diverse. I'll bet that the loading needs are different, too.

Werner, thanks for the meter recommendation. Do you happen to know its lowest decent capacitance range? Typically, multimeters that measure capacitance don't go down low enough for phono work. But perhaps this has changed since the last time that I looked.

I bought my own capacitance meter around 25 years ago. I walked into a mail-order electronics importer located nearby in Oakland, California, whose business catered to repair shops (I was handling repair work at the time). So, this is one of those "Brand-X" products from Taiwan, but it's good. This meter cost around $65 at the time, so it seemed pretty expensive, especially considering where it was made. It's digital, and only measures capacitance: nothing else. It's good for our meters to be able to measure really low because it may be useful to go as low as 5 picofarads. But in practice, cruder measurement may be OK.

In practice, we probably don't need pinpoint accuracy, since I doubt that the cartridge makers aimed for such precision. In fact, in thinking about this, I remember that Ortofon didn't specify a single number, but recommended a wide range. I recall it as being something like 275 through 400 picofarads. Naturally, different cartridge makers would not go about this the same way. Some cartridge design aspects, such as Walter Stanton's method of isolating the channels from each other in the common model 500, are very sophisticated and will present their own unique matching requirements. Thus, we can have two very different cartridges from two makers, each one equally fine, but having very different ways of achieving that performance. So, I can see that one manufacturer needing fairly (but not precisely) close adherence to spec (maybe within 25 picofarads), whereas another manufacturer's requirement may be loose (Ortofon's tolerance of 125 picos).

One of the issues that you'll face is that measuring in this range is so touchy that you'll have to zero the meter when you change the position of the test leads, even slightly. This is not a fault of the meter; it's just that working in this range is just so touchy because there is a proximity effect between the meter's own wires. So please just realize that this is going to happen and allow patience.

Having looked at a few graphs over the years, it's been obvious that this loading plays a big part in how a cartridge sounds to us. It would be nice to load a cartridge very wrong on purpose. For instance, let's say that we try some of that really cheap gold-plated cable in the red plastic sleeving: measure it and listen to it. And then load the cartridge right on spec and listen to the differences. I'm very interested in learning what you discover.
-----------
Rewrote the next-to-last paragraph to make better sense.

[rfgumby]

The cartridge is normally an inductance(L) and paring this with a capacitor C will give a reasonably average flat frequency response if you get it right. However this is not the only thing that happens. The LC paring is a filter and will give rise to a phase shift over a lot of the frequency range that we are listening to. This will lead to some smearing in the midrange and upwards. To me that is something I have difficulties living with.

YES!     

Phase shift and smearing (of detail due to phase shift) are big problems in my book.

 And of course they happen more overtly in the XO of the speakers, leading to dispersion abberations, but to have these problems be minimized to begin with is a reasonable goal- and one I think we could mostly agree upon even in this varied perspective environment.

[willbewill]

Richard

What cartridges are you using with capacitors in headshells? In Lenco arms? What tonearm cabling are you using?

[wer]

Also Richard

The lowest range is 40.000nF full scale, so theoretically you can measure 1pF, but god knows with what accuracy. I would imagine 10pF =/-20% to be realistc, but that would certainly be good enough for our purpose.

[richard]

That sounds like an interesting meter, Werner. And I think that for the purpose, what I'm interested in is simply getting the capacitance into the ballparks of our cartridges. I went on a binge measuring most of the cables that I had lying around, and found tremendous differences in their capacitances. So, I began thinking that any reasonable control over this load would be better than nothing, or what the typical audiophile was doing (which was just about nothing).

So, my thinking is simply not to go crazy with this aiming for that last picofarad. Get your technique with the meter consistent. Maybe measure a few low-pico capacitors to get the hang of it, and then go at those wires. If you can get a stereo cable pair within, let's say, 5 picos of each other, that's fantastic.

Scott and Decca, the effect that you've both brought up is interesting, and I wonder what the professional-trade people thought about the inductance/capacitance phase shifting. Of course, speakers and phono cartridges share one thing in common: they're both transducers that are connected to circuitry. Is this effect as pronounced at the input end as at the output end? Both types of professional preamps that I have contain switchable loading capacitors. This is a sure sign that capacitance loading does indeed matter and makes a difference. And, I'll assume that the negative effect is insignificant. I could be wrong, of course. I really don't know. My broadcast preamps are by Stanton and McCurdy.

You've both brought up an interesting concern, and I know exactly who I want to ask about this: engineers at Stanton Magnetics who I used to talk with 29 years ago. The reason that I want to talk with these people is because they made not only the cartridges, but they also made professional preamps (of which I have two). Who better to intimately understand this entire loop!

Well, you already know what I'm going to say next: I can't talk with these people because:
- When the company was changed to all disco all the time, these people were scattered to the four winds.
- Most of the old timers are no longer with us. I've not been able to find them.
However, on Lenco Heaven, we have two engineer-members who may be able to shed light on this issue, so Robin and Kent: if you're listening, please start talking!

Finally, Malcolm: Wow! I don't want to write such a huge reply, so I'll try to keep this short (Hah! Lost cause.) Well, when I had my small custom stereo business, I obtained a sample from Belden of their super-duper ultra-low-capacitance shielded cable. Can't recall the stock number. It's fairly thin (not thin enough for tonearm wire), stereo paired. I made 12-foot runs to use with a Weathers arm (with stock wire). The cartridge was a low-impedance Stanton. At the other end of these cables was a Stanton head amplifier. However, note this: the stock Weathers wire is custom stuff, 2-conductor in shield for each channel, high capacitance. And neither the original Weathers cartridge nor the newer Stanton low-impedance pickup are very capacitance-sensitive. Neither of these are standard magnetics. The Stanton can work into a load up to 1,000 picofarads!

About the other wire: one day, my circuit technician phoned me, all excited. "Richard; I just got this wire from MCM. It's fantastic; really low-capacitance. You've got to get some of this." So, I did. I bought 50 feet of it. My tech was a serious amateur cellist, so like myself, a person with good, analytical hearing and built-in live instrument references. And a capacitor junkie. Alas, he is no longer with us.

At this moment, my in-service turntable is an ordinary Lenco L75 with stock arm wire. Wire to the preamp is the un-named MCM stuff from Taiwan. It's really fat, like ignition wire, 7 or 8 mm (can't remember). There's a mid-cable junction to allow me to manipulate the wiring within a nasty cabinet. But I've cut everything to give me an approximate 235 picofarad load to the cartridge within all of the wire, measured long leads + tonearm leads + headshell.

I have not added any capacitors in the headshells for the simple reason that all of the cartridges that I'm using are Stantons and Pickerings, so every one takes the same loading: 275. But I laid in a small stash of little caps just in case. You can see that I've standardized on cartridges that take the same loads for this turntable; I think of this Lenco machine more as a production system than as an audiophile setup. A second machine, either an L75 or Rek-O-Kut, or both, will be more aimed at audiophile-grade listening with finer arms.

I've long been curious about the Shure V15 IV, so that cartridge awaits. And I have a few classics from Micro Acoustics that are all waiting, too. I intend to set up different audiophile-type arms to use with the highest-end Stantons, the MAs, and probably the Shure. Well, then, I'll have to eat my own words, huh?

Dunno: have I given you a decent answer?

[RCAguy]

Capacitive loading profoundly affects the frequency response of the cartridge, and should not be taken lightly.  For example, a Stanton 880/881 (top curves in chart below) was designed with a broad resonance that flattens both its frequency and phase responses when its inherent inductance is loaded with the intended capacitance of 275pF and resistance of 47kOhm.  Flatness means "smearing" has been compensated.  If cabling in particular is too capacitive or too long and C were to approach 1,000pF, a Stanton 680/XV15/681 (bottom in chart below) would noticeably peak +2dB between 2,000 and 6,000Hz, followed by a drastic falloff in response of -7.5dB at 10kHz.  Following the mfgr's loading recommendation within +/-20% precision is much better, although some might prefer the effect of scratch filtration using 500pF, where the response is flat in the mid-range, but falls to -3dB at 10kHz and -7.5dB at 20kHz.

Careful about a C meter measuring 40nF full scale.  Nano-Farads are 1,000 times as much as pico Farads, so a Shure cart specd at 400pF will occupy only 1% of that meter's full scale.  I prefer a capacitance bridge, an old tech contraption that measures easily to 10pF, but mine is an antique and something similar might hard to come by.  Conrad Hoffman has instructions for rolling your own - http://conradhoffman.com/cap_bridge.pdf

Assuming an accurate way of measuring C, first disconnect the cart (or remove the headshell), unplug either L or R at the preamp input, and measure C between pin and shield of the cable's male phono connector - this is the capacitive contribution of the tonearm wiring and cable (without removed headshell, the C of which is much smaller and likely can be ignored).  Second, measure the preamp C between the female center socket and shield - typically preamps have 150pF capacitors in parallel with 47kOhm resistors at each input.  Adding the C of the cabling and the C of the preamp gives the total C loading of the cart.  If more than 20% less than recommended, lengthen the cable, or add a capacitor somewhere.  If more than 20% higher, shorten or improve the quality of the cable.  Good coax for audio (not RF) is about 10pF/ft - and not outrageously expensive.

C loading is important, and relatively easily controllable at either the headshell or the preamp - Richard's post #23 is a good headshell solution, and 50~150pF ceramic NPO/COG disc caps are tiny, 5% accurate, very good audio performance, and cheap.  Alternatively, add a switch to the preamp.  My custom preamp has a simple three position C selector that adds 100, 150, or 250pF to my measured cart-to-preamp wiring C of 125pF for a total C of 225, 275, & 375pF respectively - covering A-T carts (recommended 100~200pF) and Shure models (250~400pF).

[richard]

Thanks, Robin.

I can immediately think of two Lenconians who probably would hear these differences acutely: Sk&ter and Scott. I'm saying this because I have some understanding of what they can hear and the acuteness of their evaluative skills. I think that most others here would hear them, too, but more as indistinct colorations, or they'd probably say, "I like Cartridge A more than Cartridge  B." They will definitely hear these differences.

Robin, did you run these tests yourself?

What sort of meter do you add the bridge circuit to, Robin? Otherwise, I'd like to suggest that the board have a communal capacitance meter to lend out to people, since most people don't have a need to do these measurements often. Some of our members are not wealthy, so this could be a nice solution for everyone. I can almost assure everyone reading this that you'll get some real surprises when you measure a few of your own cables!

[willbewill]

I built one of these for matching caps - mightbe near enough to do the job

But also just spotted this:

http://radiohobbystore.com/high-resolution-capacitor-meter-diy-pro.html

[RCAguy]

Malcolm aka Scott, that $35 C-meter looks very interesting on paper, certainly more modern and possibly a lot more ergonomic than my old Leader LCR-740 impedance bridge.  My kinda project, but no time now to build & evaluate.  Do you or anyone else wanna step up?

Richard asked me privately what tolerance is acceptable for capacitive loading?  Looking at Stanton's chart above, you can visually interpolate what would happen at other than their recommended 275pF.  For example for an 880/881, halfway between 275 & 500, say ~375pF, might actually flatten the mid-range, but drop -2dB at 15kHz - is that what you'd consider acceptable?  I might.  Stanton might have had the good engineering sense to anticipate it being more likely that users would err on the side of higher accumulated capacitance?  In your interpretation, remember that dips in frequency response sound benign, where peaks are malignant, to be avoided.

And yes, I run these kids of tests all the time (I "trust but verify"), tagging my patch-cords and raw cable intended for C-sensitive applications with their total C or C/ft, respectively.  Many run-of-the-mill cables are no problem for line-level interconnects, where impedances are much lower and so line capacitance has less effect.  For more critical tonearm-to-preamp runs, you really owe it to yourself to know - get your head out of the sand, so to speak.

[richard]

Agreed: it's those peaks that stand out as sonic ugliness. Even with what I know about these cartridges, I was impressed with the smoothness of those curves that you pulled when the capacitance was on-spec.

I agree completely about the listening results produced by cable quality at phono levels vs. line levels. In fact, there have been times when I've mused that cables from the dollar store (at the cost of, well: one dollar!) might be  adequate between components (except for the workmanship quality in connectors and fastening them to the cable, and not for phono).

What's different, in my mind, between Stanton and most other companies in the larger-scale cartridge business is that Stanton was an engineering-driven company (as opposed to sales-driven). These people were serious about understanding the needs of professionals who used their products. This is a particular slant, a mindset. Note that I am not vaunting them above, let's say, some of the excellent cottage-industry audiophile designers, but pointing out the company's focus on practicality along with performance. It's as if the company was designing for professionals; audiophiles and other home users were along for the ride. With the death of Walter Stanton and the move to Florida, this orientation was flipped on its head. Meanwhile, Shure began to move into this territory with the late SC-35 models.

I like to try to get inside the heads of the designers. So, keeping this practicality context in mind, it's easy to see where they were coming from. As the principle supplier to the broadcast industry, they could see that broadcasters typically put the preamp very close to the turntable. Therefore, a relatively low cable capacitance could be a realistic design parameter, and you've described the resulting benefit in flatness through the highs, especially in the smoothing of the ubiquitous resonant peak of magnetic designs. So, this explains this company's philosophy and context for their particular capacitance goal. Do you think that I have this right, Robin?

I've provided a context, above, for working with cartridges that have a 275 pico design goal. Of course, many people here on Lenco Heaven will prefer different cartridges, many with higher capacitance load figures. These are a bit more forgiving about cable capacitance and longer cable runs. But it only takes a little more cable length to go well above a designer's goal of 400 picos. This is the sobering part.

Malcolm, that's a really interesting-looking meter, and the price is right, considering that it's half of what my cap meter cost me 25 years ago. My meter is pretty easy to use. This one's leads are shorter, which should not be a problem. So, you need this kit and a cabinet for it, and you're in business.

What I'd particularly like would be some safe method of measuring the input capacitance of preamps. I need to know that the meter's output won't blow the preamp's circuitry. I'd like some thoughts about this. I may have a spec for my meter about what its measuring signal is.

[willbewill]

I built one of these for matching caps - mightbe near enough to do the job

But also just spotted this:

http://radiohobbystore.com/high-resolution-capacitor-meter-diy-pro.html

I forgot the link to the first meter:

https://www.sparkfun.com/products/9485

[RCAguy]

I admire engineers who not only design for performance, but who anticipate the variables that users will experience in common practice, and build in a margin for that error.  Beyond his/er control, the sources of error are the necessarily non-ideal wires, and the preamp load presented back to the cartridge.  Imagine that cartridge's electrical engineer, working alongside its mechanical engineer (for all the stylus & cantilever stuff), with only passive components to work with, having to begin with a practical coil, smallish so as not to add a lot of mass, with both the necessary inductance L to "read" a moving magnet or change in reluctance due a moving slug of iron, and with a certain resistance due to many turns of its fine wire.  Now this impedance has to be "bridged" by a load of 10~20 times higher, not so high that sensitivity to hum or RF would be too bad, compromising with the 47kΩ based in the load resistor R.  Now add a lumped capacitance C that in combination with L & R produces a broad resonance just where you want it for flatness within the audio band, but with falling response higher, then try-and-err a few more times, and finally voila!  

While other load combinations might work, 275pF in parallel with 47kΩ, within 10% or so, seems to have become a de facto standard among manufacturers, possibly both for the benefit of interchangeability as well as good science.  We can't say whether these engineers merely reverse-engineered each other, or formed a consensus - when it benefits all, engineers generally don't avoid this until less forbidden to.  But a savvy user would note that a Shure V15-IV is quite flat loaded with 275pF, but less forgiving than either Stanton above when loaded with 375pF, brightly bumping up to +1.5dB at 8kHz before falling off 3.5dB at 15kHz to -2dB.  The Shure is more forgiving of load capacitance less than 275pF, but that condition is less likely, and easier to correct.  An A-T 14S loaded with 275pF has a broad -2dB deficit in the mid-range, rising abruptly 4dB at 14kHz to +2dB before falling off rapidly; it's smoother loaded at 60pF total (wiring & preamp), but who can achieve that?

[richard]

When I last casually surveyed cartridges, I found that the 47k ohms was definitely universal, but that capacitance loading varied up to 400 depending upon the product. My circuit technician, now deceased, disliked Shure cartridges, pronouncing them "shrill," but he never knew about the importance of capacitance loading!

Malcolm, thank you very much for posting those URLs for the affordable capacitance test meter kits. This is a treat. And, thanks to you, they're already in the Supplement to my Handbook!

[RCAguy]

As promised, here's my DIY "Phono Capacitive Load Selector" circuit consisting of a toggle switch and 6 disc capacitors made for cheap (<$15 if retrofitted inside preamp enclosure by drilling one 1/4in hole).  Within acceptable ~10% tolerance, it would accommodate cartridges looking for between 200 and 400pF capacitive load when used with 125pF wiring between cartridge and preamp - otherwise add or subtract the difference to the middle capacitor.  (Also include any C not eliminated within the preamp.)  Note: make nearly any value of cap you want, adding their two values when paralleling, or making a smaller value by putting two caps in series, with the resulting C= (C1 x C2) / (C1 + C2).  The six (6) COG/NPO ceramic disc caps have very good audio characteristics and are tiny enough to solder inside a headshell, if adding a single fixed C is your preference.