Jim Lesurf wrote:
In article , Don
Pearce
wrote:
Jim Lesurf wrote:


Odd that - I would expect crosstalk performance to depend mostly on the
accuracy of the right angle between the two coils with some secondary
effects from the cantilever mounting. No real reason for either of those
to be so poor.

Indeed. I am also puzzled (as well as dissapointed) by this. Seems to me to
be a sign that something isn't right. I spent more time on the M97s than
anything else, trying to see if the results could be improved. But in the
end I decided the values I was getting were about it. Nor did the crosstalk
seem to be due to distortion. Maybe the stylus is 'rotating' because the
mass and force are off the cantilever/suspension axis. Dunno. Gave up
puzzling over this and published the results.

Also surprised by the stark differences shown by the photos of the
stylus of the 97xE compared with the old HE and MR. Chalk and cheese.


I spy cost-cutting.

Maybe. Again this was quite a surprise for me given the superb V15.


FWIW one 'feature' I didn't comment on in the mag is that if you look
at the responses you can see a 'kink' in most of them. This seems to
depend on the headshell and the level of the cartridge's effective
compliance, etc.


I noticed that kink (you do mean the one around 200Hz?).

Yes.

I can't find any trace of it on my system in normal use, but there is
just a hint of it on the crosstalk test. I do wonder just how good the
cutter that makes test discs is compared to the cartridges that we use
to reproduce them. I suspect that for many tests the answer is - not
much better.

I also wondered about that. But it showed up to a different extent (and
frequency!) when I changed from one cart+headshell to another. My guess is
that the V15 has high compliance and tiny mass, so delivers minimal
vibration into the arm. But the other carts aren't as good in this respect,
so excite more in spurious arm or headshell resonances. The 97 does look as
if it has a much bigger effective tip mass than the V15, so a static
compliance value may not give the game away.

For convenience I used more than one headshell, so they may also differ.

May sometime do more precise plots with noise and time-averaged spectra.
But I didn't do that here as they need to be corrected for the recorded
response[1], and I also wanted to extract THD values. The stepped sinewaves
on the 'Ultimate Analog' test LP were excellent for this. But the stepped
waveforms were more tedious to use than noise. Was meaning to also do
crosstalk versus frequency, but again ran out of time and patience! Another
day, maybe.


Here you go, got that for you - the AT OC9 again:

http://81.174.169.10/odds/croddtalk_oc9.gif

All those odd resonances are there on the crosstalk channel, but it does
manage 30dB midband. Could do with a good long noise track, a minute or
so, to make better plots.

By then end of that article I wanted a break from 4 in a row about LP. So
ended up doing at least four on speaker cables! Must be mad. 8-]

Been told that the second cables article is now out, but not seen it as
yet.

Slainte,

Jim

[1]I have done some responses with various 'noise bands' on test LPs, and
they seem to give a wild variety of results. Results generally neither
pink, nor flat, not anything else that would make obvious sense. Prompt for
a latin tag about who guards the guardians... :-)


That would be Quis custodiet ipsos custodes.

d