Cartridge loading explained

[carlosfm]

I open this thread to explain cartridge loading in detail.
Why? How? How much?
What??? MM cartridges need that too??! Isn't it 47k??!

But first an introduction.
I have a Technics SL-1610 MKII, which I modified many years ago for taking any interconnect that I want to use.
Than meant making two holes on the back, inserting two female RCA plugs, and soldering well shielded coax wire to the arm base.
The interconnect I'm using is very short in length (~40cm), dual shielded 75 ohm coax.
It is important to minimize cable length, because the output signal of a cartridge is very low level, in some cases high impedance, and consequently noise pick up is lower.
It is also important because smaller cables have proportionally lower capacitance.
A typical 1 meter coax has around 150pF capacitance.
I measured capacitance from the wires of an empty headshell mounted on the arm up to the end of the interconnect, and I've got a value of 93pF.
It is hard to get much lower than this, unless you put the phono pre inside the turntable - a viable option, but not practical if you need to switch input loading.
When adjusting input capacitance for your cartridge (as per the recommended values by the manufacturer), you must know your cabling capacitance and count with that.
BUT the recommended values are most of the times very far from ideal.
Say, if the manufacturer of an MM cartridge recommended 10Kohm impedance loading, you would probably not buy that cartridge, because it's not compatible with your MM phono pre, right?
So they recommend 47K, which is a "standard" value for MMs.
By ear (because I don't have the specs, Rega doesn't give them) my first generation Rega Elys - grey body - sounds best with a loading at around 10K. At 47K it sounds incredibly harsh in the treble - it's really too much - and it picks up a lot of surface noise on the record.
I'll give practical examples of other cartridges that I own, wich I have all the data.

I don't use any input shunt capacitance on the phono pre (although I can switch between 4 capacitance values, or none).
So, I always only use the 93pF capacitance of my complete cabling. Let's round this to 100pF.
I only switch impedance.
Why is low capacitance important?
That comes next, but I can only follow with this thread (including simulation graphs of several cartridges) tonight, when I get home.
So stay tuned.

Btw, I suppose that the right place to post those graphs is in the Gallery, right?
The files will be very small, those are frequency response graphs.

[bauzace50]

carlosfm,
thanks for this initiative. Knowing "what", "why", and "what/how to do" goes a long way to lift the veils of half-truths and unsupported beliefs. Thankyou!
bauzace50

[sreten]

Hi,

I'll just point ot that capacitance loading switches on pre-amplifiers are
additional capacitance - zero means zero additional capacitance not
zero input capacitance.

I've not seen a practical phono circuit that does not include some input
capacitance, ones that don't can have horrendous rf pick up issues.
Checking a number of published designs 100pf to 220pf is a reasonable
assumption as to a phono inputs intrinsic capacitance.

/sreten.

[carlosfm]

Hi,

I'll just point ot that capacitance loading switches on pre-amplifiers are
additional capacitance - zero means zero additional capacitance not
zero input capacitance.

I've not seen a practical phono circuit that does not include some input
capacitance, ones that don't can have horrendous rf pick up issues.
Checking a number of published designs 100pf to 220pf is a reasonable
assumption as to a phono inputs intrinsic capacitance.

/sreten.

Sreten, if there is a fixed shunt cap on the phono stage, it can be removed. The additional switch (if any) would add capacitance in parallel (depending on the design), but it could be bypassed too.
This means opening the phono pre and dealing with a soldering iron.
But that's not the point of this thread, the point is understanding what's going on.
Btw input stages of preamps also have some input capacitance, due to the active devices used and the topology. It can be significant, or so low in value that you can ignore.

To unveil something that we will see tonight, capacitance increases the frequency peak caused by the cartridge's inductance. If it's a high-ish inductance, that frequency peak will be in the middle of the audio band (considered as 20Hz to 20Khz). We don't want that.

[carlosfm]

It is very stressing to post files in the Gallery... I've posted a file and then it disappeared. Then I found out that one can't sort by date.
Had to scavenge through there sorting by title...

Well here we go.
I have posted a schematic that simulates my Shure M97, with it's impedance and inductance parameters.

http://www.vinylengine.com/phpBB2/album ... ic_id=3337

R1 and L1 have the impedance and inductance values for this cartridge: 1550 Ohms and 700 mH.
R2 and C1 are the shunt impedance of the preamp, and total capacitante (which must be accounted with the cabling + phono pre capacitance setting).
Shure recommends between 200pF and 300pF for this cartridge. I've settled in the middle value for our example, 250pF.
47K is recommended as loading resistance.

Here's the resulting frequency response:

http://www.vinylengine.com/phpBB2/album ... ic_id=3338

There's a small lift at 7~8Khz, not enough in amplitude to be detectable (less than 0.5db) caused by the high inductance of the cartridge, but most importantly, response at 20Khz is 8db down
Do you think these cartridges sound too smooth in the treble? Here's the answer.
These cartridges have a very decent stylus, capable of much better, and deserve to be better treated than this.
While you appreciate this preliminary data, I'll go have my dinner, I'll come back after it.

[carlosfm]

Here is the M97 at 47K and 100pF, just about as little a capacitance as we can get (have you read the first post? Good).

http://www.vinylengine.com/phpBB2/album ... ic_id=3339

What we see here is that the low treble peak is no more and there's a smooth rolloff towards the high frequencies.
Also, there is a much lower loss at 20Khz, now we have -5.5db.
The excessive capacitance was making a reactance with the cartridge's inductance (the peak), and also made a very steep rolloff at high frequencies.
But we can still do better.
The shunt resistor damps the peak (believe me, it could be much higher on the first graph, if a higher value resistor was used, with that capacitance).
But now we have lower capacitance. No peak.
We just need to get that treble right at up to 20Khz, even if for that we have a very small peak (undetectable) in the same low treble region.

Ladies and gentlemen, the M97 with 68K and 100pF shunt resistance and capacitance:

http://www.vinylengine.com/phpBB2/album ... ic_id=3340

This is about the best that can be done with this cartridge.
Now we have just a little under 2db at 20Khz, and the -3db point is at 23Khz.

[bauzace50]

Carlos Filipe,
This is enlighening! If I get my hands on a bargain price on another M97xE, I would experiment with it. However, another thing that bothered me about it is the average tracking ability, even after a special re-tip with a long-line low-mass diamond. One less-expensive unit I own tracks circles around it, and sounds better right from the box...and it is discontinued in 2006 (Stanton 500 E Mk II). Shame that some enthusiasts did not experiment downwards in price, to get a true gem.
Knowledge empowers, as Blue Angel sagely says.
bauzace50

[carlosfm]

Let's go to the other extreme now, a low inductance cartridge.
Glanz MFG-31E has the following specs:
- Impedance: 900 Ohms
- Inductance: 110 mH

Glanz recommends 47K impedance loading and 100pF capacitance -> you're doomed here, this is about as low as you can get, better cut those caps and that interconnect.

Here's how it looks, with the standard loading:

http://www.vinylengine.com/phpBB2/album ... ic_id=3341

Look as the lower inductance makes the peak appear at a much higher frequency. There's more than +3db at around 41 Khz.
There is also +1db at 20Khz.
Treble will (and does) sound fierce, aggressive, too much.
The cartridge in underdamped, we need to change the input impedance to a much lower value.

Here's how it looks at 23k + 100pF:

http://www.vinylengine.com/phpBB2/album ... ic_id=3342

Amen
What we have here is flat response through the audioband, -1bd at 30Khz and the -3db point is at 45Khz.

[carlosfm]

Carlos Filipe,
This is enlighening! If I get my hands on a bargain price on another M97xE, I would experiment with it.

No, don't bother, I'm not telling you to buy an M97 whatever, I'm posting this for you to understand, keep reading.

[carlosfm]

Bauzace (Carlos),

My M97 is old, I have it for many years.
I think that these parameters are still the same for the current M97xE, although, even if Shure doesn't spec compliance, it is a high compliance cartridge.
Mine is the one with integral (plastic) headshell, very lightweight, and with easy to set overhang. It works well with my arm because of this.

The Glanz I posted above is also old, although I bought it some months ago.
I love it's sound, very extended, natural treble, gorgeous midband, it only lacks the bass dynamics and slam of the Shure.
One can't have it all...

The conclusions here, even if you don't want to make a simple schematic on a free simulator software, are:

- Inductance is your enemy
- Capacitance should be as low as possible
- If the cartridge sounds very explicit, agressive in the treble, it needs a lower value shunt impedance.
- If the cartridge sounds too smooth and shut in, with lack of high treble, it needs a higher value shunt impedance.

Leave capacitance alone, low, the cabling is enough, and just change the resistors.

[carlosfm]

And now a request from several families (me): that wonderful cartridge database is missing a required field: cartridge inductance.
To be filled if and when possible.
Now you know why it is important.

[bastlnut]

hallo,

respect for all the work you have done!
is it not also possible to just try different cables to see which works best?
different cables have different inductance and capacitance values and will effect the sound differently, as will changing phono stages and tonearms.
maybe, just changing the speaker cable will make the sound right.

many roads lead to system nirvana. let us not forget this!

regards,
bastlnut

[carlosfm]

is it not also possible to just try different cables to see which works best?
different cables have different inductance and capacitance values and will effect the sound differently, as will changing phono stages and tonearms.

Yes, you will have different parameters by changing cables.
But notice that the most important is low capacitance (inductance is insignificant for a normal cable length), and you are limited to shielded (coax) cables.
Even then there are some variations, like semi-balanced (two inner wires and shield, connecting the shield on one side), but these are more capacitive than normal coax.
I don't recommend unshielded cable, it will pick up every noise, it would be just like an antenna.
Precious metals? it's up to you to try it out, but some basic cable parameters are more important here, and keep the cables as short as possible.

[Axon]

I've been looking at these issues myself, while looking into building a preamp.

How did you derive the resistance values for the cartridges? For most carts, I was under the impression that resistance wasn't broken out, and you had to derive it from impedance + inductance.