Near-zero capacitance effect for phono cables - method

the thin end of the wedge
Post Reply
Guest

Near-zero capacitance effect for phono cables - method

Post by Guest » 10 Jan 2011 19:50

Here's a simple method for avoiding response effects of phono cable and built in phono preamp capacitance , specifically the interaction with MM/MI cartridge inductance. I make no claim as to whether to do so is a good thing or not, and note cartridge manufacturers quite possibly intend there to be a certain interaction and resonance response. However, in the cause of the great ultra low phono cable capacitance pursuit, and avoidance of that well known cable capacitance resonance, the method presented here at least allows one to test how ‘near-zero capacitance’ might perform, in a very non-invasive and straightforward way.

It's a means of 'nulling' almost all cartridge loading capacitance/cartridge inductance interaction in the audio band, and allows near-zero capacitance performance for more or less any normal phono preamp/cable arrangement. It works without modification to cables, phono preamps or cartridges, and can be implemented in an RCA 'loading plug', the same way as R loading is commonly varied, using RCA Y adapters.

I've modelled it for cartridge inductance 400mH to 700mH, and for total loading capacitance 150pf to 600pf, where the standard internal preamp load is 47k. That covers most likely cartridge/cable/preamp combinations. The results are pretty good, as illustrated below.

All one has to do is add this network across to the existing RCA connection :

16090

Where L is the same inductance as that of the cartridge coil, and R depends on specific C (total capacitance loading) but is typically between 68k and 100k. Default values of R=94k for L=700mH, and R=68k for L=400mH are good for normal cables.

Here's are plots of predicted frequency response for 2M like cartridge inductance, versus ‘theoretical 0pf response’, and versus ‘normal’ loaded response. Firstly loaded at 200pf total capacitance :

16092

Predicted response of the modified case fairly closely resembles that of the 0pf loading response. And is quite distinct from the 200pf ‘normal’ case.

Here's a plot with the same 2M like parameters, except for 450pf total capacitance loading:

16091

In this case, resemblance to the 0pf case is still good, and quite distinct from the ‘normal’ 450pf case.

This is a plot for AT95 like inductance, firstly at 200pf capacitive loading :

16094

Again, predicted response of the modified case fairly closely resembles that of the 0pf loading response. And is quite distinct from the 200pf ‘normal’ case.

Here’s the AT95 like case at 450pf total capacitive loading :

16093

Again, resemblance to the 0pf case is still good, and quite distinct from the ‘normal’ 450pf case.

Note that to achieve these predicted matches, it was necessary to optimise the value of R, using values as marked up on each plot, but not very closely. Default values of 94k for 700mH and 68k for 400mH still give pretty good all round results, particularly for C loading around 300pf total.

So there you go, the near-ultimate attainment for zero-pf seekers of sensation !

This is fresh off the simulator, and I confess I’ve not yet actually measured or listened to results. That's physicists for you, caveats apply as usual, but I'm confident :wink: If anyone feels inclined, please beat me to the listening/measuring test !

One last point, it seems very convenient to use a spare cartridge body of the same type (minus stylus assembly) to provide the inductance necessary for the modification. Advantages are that inductance/resistance should match near exactly, and hum cancellation intrinsic to most MM/MI types would avoid the need to shield the inductor, should that prove necessary.

All of the response plots above are for the electrical part of the cartridge, not the overall cartridge response, of course, which is a composite of electrical and mechanical responses.

Interesting ? Comments/questions ?

Guest

Post by Guest » 10 Jan 2011 21:29

Bravo,
Some new thinking on an old problem =D> =D> =D>

pivot
long player
long player
United States of America
Posts: 4979
Joined: 27 Dec 2002 14:31
Location: Albany, NY USA
Contact:

Re: Near-zero capacitance effect for phono cables - method

Post by pivot » 10 Jan 2011 21:39

ld wrote:One last point, it seems very convenient to use a spare cartridge body of the same type (minus stylus assembly) to provide the inductance necessary for the modification.
First - very nice work LD!

Second - the qoute above jogged something in my memory. Didn't some pre-amp designer do something like this in the 1970s? I do seem to recall a review/discription/mention in the hi fi press at the time.

Any old f*rts out there with a better memory then mine?

Whitneyville
long player
long player
Posts: 2276
Joined: 13 Feb 2009 19:08
Location: Tulsa, OK

Post by Whitneyville » 11 Jan 2011 02:25

Ham operators have used variable capacitors for a century to "tune" the reactive "L" from antennas, so this seems practical that it should worth at AF. We've also used LC networks to limit frequency responses to what we consider "worthwhile" and to reduce adjacent channel interference. I do wonder what effect(s) this might have on reducing high frequencies with "real" cartridges and "real" phono stages, or if you'd make a fine RF receiver at "X" Khz.

wordwizard
senior member
senior member
Posts: 466
Joined: 02 Sep 2010 02:24
Location: Los Angeles, CA

Post by wordwizard » 11 Jan 2011 02:53

rocker65 wrote:Bravo,
Some new thinking on an old problem =D> =D> =D>
Not really "new thinking"; more like a new implementation of an old RF solution applied to an audio frequency circuit.

Just "out of the box" thinking.

wordwizard
senior member
senior member
Posts: 466
Joined: 02 Sep 2010 02:24
Location: Los Angeles, CA

Post by wordwizard » 11 Jan 2011 03:00

Whitneyville wrote:Ham operators have used variable capacitors for a century to "tune" the reactive "L" from antennas,...
Not just "Ham operators" but RF designers from day one ;)

However, there are only three basic principles in electrical circuit design as explained to me on my first EE semester:

a) -if you have too little, put some more in
b) -if you have too much, take some some out
c) -don't re-invent the wheel.

dlaloum
long player
long player
Posts: 3895
Joined: 29 Dec 2009 05:21
Location: Melbourne, Australia

Post by dlaloum » 11 Jan 2011 14:19

heheh

This I like - It would allow me to stop tearing my hair out trying to get my capacitance any lower....

So do we have a formula that does not require a circuit simulator that we could use to derive the value of R? Or in this case perhaps just a simple table? Table of total C and related R...

The curves are also perfect for a Digital RIAA setup, they droop gently in the HF area, providing more headroom for the ADC, and if the real life response is so smooth, then standard digital linear EQ can compensate for it easily.

I'm definitely up for trying this out. (my RTA should be back up and running fully licenced within the next day or two...)

Having said that... 500mH inductors are not that common (nor are they cheap).... and won't fit into an RCA plug either .... so the exercise starts to get more complicated.

Also is an inductor likely to generate interference? (or act as an antenna) - and would it therefore need to be shielded?

By next week I'll also have a couple more Shure bodies to try out the SAS on - which will allow trying the varying loadings and strategies in a situation where only the Generator is variable (cartridge L & R)

Bebé Tonto

Post by Bebé Tonto » 11 Jan 2011 18:02

So, ld, let me see if i understood it: What you are doing is adding a low pass filter so the curve matches the same curve that you would got if capacitance was 0?

Guest

Post by Guest » 11 Jan 2011 23:20

Thx for responses and encouragements, appreciated !

The mod models as rf stable. Phase response most closely resembles 'normal' C loading for the vast mojority of f range, and notably above the audio band, specifically rf. It's pretty boring at rf actually, dominated by C loading. Also impedance of the phono 'hot' connection most resembles the 'normal' C loading case, in all but a narrow range of frequencies within the audio band. It has advantages over true 0pf models in this respect !

I suppose it has operating principles that are a cousin of antennae loading, but never thought of it as such. And it's not a straightforward low pass filter because of the intentional resonance with C loading. But hey, the overall amplitude and phase response look bona fide ............

It's interesting that previous preamp designers might have trodden this path...........

I will audition the mod, of course. Probably on an AT95 or OM body, because that's what I have duplicates of ! I think that gets around hum pickup because AFAIK those carts are bucking. We'll see !

Guest

Post by Guest » 12 Jan 2011 12:49

dlaloum wrote:So do we have a formula that does not require a circuit simulator that we could use to derive the value of R? Or in this case perhaps just a simple table? Table of total C and related R...
Hi dlaloum. Good idea, I'll post a table in due course. Default values of 94k for 700mH and 68k for 400mH work pretty well for all total C between 200pf and 500pf, and values in between can be obtained by proportion, for different L values.

Good news on the RTA !

For completeness, here's amplitude and phase response pots up to 10MHz, to support the claim that the arrangement is rf stable. You can see above the audio band, amplitude and phase responses simply return to 'normal' C loading :

16111
16112

It's neat, don't you think ?

Again these are electrical response plots only, not total cartridge response (which is a composite of mechanical and electrical responses).

dlaloum
long player
long player
Posts: 3895
Joined: 29 Dec 2009 05:21
Location: Melbourne, Australia

Post by dlaloum » 12 Jan 2011 14:41

hmm the only 500mH inductors I could find were for wah pedals and are around $30 each... so this becomes a relatively expensive exercise!

Once the circuit is in place, and cancelling out existing C loading.... how does one apply load - R only?

Can one then apply electrical damping to offset the cantilever resonance? (as discussed in the loading thread)

I have to reiterate that entire encyclopaedias could and have been written with what I do not know about electricals and electronic circuits.... my last serious contact with them was in 1st year university.....in the days when dinosaurs ruled the earth.

Bebé Tonto

Post by Bebé Tonto » 12 Jan 2011 15:11

ld wrote: 16112
Aha!! Busted!! GOTCHA ld, gotcha!!

My suspicions were confirmed!! (evil laugh)

You can correct the freq response but you cannot achieve the more stable phase response achieved with near zero capacitance. Although the difference is not great.

I'm all for minimizing phase shifts (that's why i prefer the panasonic semiconductor cartridge btw), and thus i'd still go the "minimize capacitance" route, namely, mounting the preamp under the tonearm or inside the turntable, etc.

Bebé Tonto

Post by Bebé Tonto » 12 Jan 2011 17:25

dlaloum wrote:hmm the only 500mH inductors I could find were for wah pedals and are around $30 each... so this becomes a relatively expensive exercise!
... Or you can wind your own coils. They are $30 because the guitar effect people is willing to pay that price.

Guest

Post by Guest » 12 Jan 2011 17:28

Ahahahahaha.......not necessarily ! Much depends on your view of what 'optimum' phase response is, it could well be the 'normal C loaded' case rather than '0pf' case This is because the plot is just for the electrical part of the cartridge, not the combined electrical/mechanical model. Mechanical response, e.g. cantilever, has it's own phase response to add in.

In any event, here's a close up of the various phase responses in the relevant part of the audio band. You can see how the modified phase response most resembles the 'normal' 200pf case. This is electrical response only, of course :

16113

In practice the variance is not large. But there's no free meal ticket, that's the price of getting the amplitude response to match the 0pf case by just using bits of bent wire !

In the case of a MC or strain gauge cart, where there is no electrical response, one is left with the cantilever/mechanical response alone that there is no means to compensate.....is that a good thing ? I saw the article you scanned in recently which touched on this topic, Bebé Tonto, and I thought that to be the crux question, that was not raised or addressed !

However, throughout the whole vinyl production path , phase is probably thoroughly well mangled in any event. For example, cutterhead response, eq and RIAA compensation applied before cutting etc.

Though it's good practice to preserve phase, I think this mod might well do the right thing in the audioband, as best as that can be judged !

Guest

Post by Guest » 12 Jan 2011 20:22

dlaloum wrote:hmm the only 500mH inductors I could find were for wah pedals and are around $30 each... so this becomes a relatively expensive exercise!
Hi dlaloum. I intend to try it out using a spare cartridge body of the same type to provide the inductance, in a non-destructive way. Need to remove stylus assembly to prevent microphony though.
dlaloum wrote:Once the circuit is in place, and cancelling out existing C loading.... how does one apply load - R only?
It's based on a 47k load, typically inside the preamp. If a different R load is used, then the mod's own R value will need to be adjusted. But as a first try, 47k seems a good place to start.
dlaloum wrote:Can one then apply electrical damping to offset the cantilever resonance? (as discussed in the loading thread)
In the TL model, there is some interaction between electrical loading and mechanical damping. If this bears up in reality, then yes, R loading provides mechanical damping. But that's yet to be proven.
dlaloum wrote:I have to reiterate that entire encyclopaedias could and have been written with what I do not know about electricals and electronic circuits.....
You know the saying, 'the more you know, the more you know how much you don't know'. :wink:

Post Reply