Accession for Moving Coil

@ Richard et al, from a manufacturing point of view it is sensible to make the MM and MC inputs as plug-in cards, or at the very worse (but good for connectivity) modular cards with solder-in flying leads.

The former would make it "user configurable", being able to swap one type of card for another. We're only talking about a small section of circuitry, but it is an important part because it does all the required gain to lift the signal out of the noise as well as the cartridge EQ.

Another thing it will require IMO is section level matching for those wanting to do direct comparisons between cartridge types - thus requiring a potentiometer per input (after each gain stage). In my opinion these would feed a summing junction, and have a mute switch each, such that the unused input noise would be removed.

By combining these mutes into a single switch it would be cross-mutable such that with a two arm turntable AB comparisons could be made.

The savvy reader will now be able to jump to the conclusion that with a centre-off "cross-mute" the thing becomes a turntable mixer, albeit an expensive one.

Richardl60 wrote: Sounds perfect Graham.

All of Graham's products sound perfect

ICL1P wrote: Graham, how is this project progressing?

Quite a few projects took a back seat this summer whilst building work took place, and summer became autumn, and very soon it will be winter (in approx. 3 days...)

Frustratingly the build process didn't fall into the neat little boxes I would have preferred, and the finishing off jobs became awkward. The listening test "environment" changed from having stud wall dry-lining to dot and dab on masonry, and I can hear the differences. Even the chair I bought isn't doing me any favours. Hopefully in the next few days all the niggling tools and materials will be tidied away to make way for progress to happen.

The differences between the MM and MC input stages are considerable, and I still don't understand how anybody can make a "one circuit does all" - technically it doesn't make sense (although to a simple mind it may do).

The MC input has taken some getting right. It being an inverting integrator having an inverse frequency response to that of a cartridge - plus input impedance (which the Dynavector patent doesn't have), it should be unique (let's hope the patent engineers agree...). And because it's unique it requires solutions nobody has done before. It isn't what the technical bystander's mind-set might think, and so there have been new challenges to overcome.

Technically we can hark back to the OP37 op-amp data sheet which acknowledged the possibility of cartridge magnetisation because of input bias current: the silicon valley engineers incorporated bias current cancellation circuitry to minimise the issue - but even so some bias current must happen, so cancellation is the wrong word - it is vastly reduced but not non-existent.

That is what DC coupling brings: bias current flows in the cartridge full-time. AC coupling, so frowned upon by the misguided, at least removes DC current from the cartridge.

The integrator uses an FET input op-amp because the large signal gain of the integrator requires a high impedance feedback circuit, and so the bias current at the input is minute. Therefore the essential input coupling capacitor which must be large in value to pass undistorted low frequencies, has to be electrolytic. A GP electrolytic has more initial leakage than the input bias current and so will never switch on. Those suitably qualified will know how I worked around that.

What I have to do now is replicate the existing circuits in a overall system which also includes an output stage which can swing voltages in excess of what most op-amps can do.

Why? Well, because quite a lot of my customers insist on valve amplification (and have never heard the Proprius, and won't, because it doesn't have valves...), and valve amps need such a high input voltage (having these days been designed for CD player hot outputs). To obtain that signal voltage swing cleanly, so it never clips, requires more headroom than a +/-18 volt rail will allow.

This might well be the stumbling block. The power substructure required to bootstrap an op-amp output stage to deliver the voltage swing might dictate DC coupling, which is a nightmare in itself. Added to that is the intermediate voltage regulation necessary to reduce the power supply voltages for the input stages - voltage regulation always relies on zener diodes one way or another, and I don't want any because I can hear what they do to the sound.

Unfortunately zener diodes exist in op-amps and voltage regulators, and it's my belief that as far as op-amps are concerned, that it is the degree of zener isolation which makes one op-amp sound different to another, and not the crazy talk you hear on some forums.

Discrete transistor circuits can be made zener free, but will never perform at preamp level as good as the op-amps I use. Therefore there will be quite a bit of contemplation on the matter to obtain the right solution.

As always, the sound dictates all.


Edited by Graham Slee - 27 Nov 2016 at 6:49am

There will always be a bit more to do