Accession for Moving Coil

I've just caught up with this thread - and am sad to see the AccMC abandoned. But I respect Graham for sticking to his principles and not releasing what he considers an inferior product not up to the universal standards of the rest of the range. 

Presumably the abandonment of the Accession MC kills off the development of a twin stage box. I had been hoping for a twin MM, switchable between two TTs to feed my Majestic. I suppose I'll have to continue with a switch box and if I want better and prettier than I have (rek-o-cut) I'll probably have to take Bob's advice and buy something from Dodocus.

Of course I would prefer a Slee switch. Anyone else out there needing to feed multiple analogue sources into a Majestic?

Seeing I've abandoned it I thought I'd have another 'play'

Last post I said "Maybe my electronics are too good?", and I promptly found out they weren't because I spotted the possibility of some oscillation due to the inductance of a capacitor used to stabilise the reference voltage of the input stage op-amp.



The thing about simulation is it will tell you what you want to know, and if you don't model things accurately it won't give you accurate results.

The only way around this was to add series resistance, but too much and you have zener noise modulating the signal which sounds bad too.

Having done that my attention turned to noise gain stability again. The stage was stable as the earlier curves show - see curve two. However, the degree of negative feedback to the input - this being an inverting stage - wasn't much up at the 2MHz tuned frequency exhibited by MCs.

OK, there is no audio signal up there, but there is something, and by reducing noise gain such that there is more negative feedback which means the virtual earth is more 'zero' up at 2MHz resulted in a positive audible improvement, especially with the Hana cartridge.

By the way, I'm back on the ADA4627 op-amp seeing it must be the last op-amp on the planet capable of doing this work. Although there is no schematic illustration of it on the data sheet I am sure from its characteristics it is an FET input OP27, and by adjusting the noise gain for a GBW of 8MHz, which is what the OP27 does, improved it further. At its rated 19MHz the phase margin is quite slim at around 58 degrees (this design), but manipulating noise gain for 8MHz gives 80 degrees. OK, it's not the 90-plus degrees I like to go for, but as I said above, it looks as if the virtual earth or ground needs to be "more zero" at 2MHz, and pushing it out to 8MHz ensures that.

80 degrees is stable, and the op-amp manufacturers insist 45 degrees is stable but I beg to differ, but 80 degrees will I suppose do. Gain margin is about 30dB which is better than the 20dB you'd get without noise gain.

So then I had a workings out session on slew rate. 8MHz using a non-degenerated bipolar input works out at 2.4 volts per micro-second (SR = 0.3 x Ft), which would be right for an OP27, but for the fact the manufacturer claims 2.8V/uS. So it looks like the ADA4627 works best manipulated like an OP27.

The equation should really state SR = 0.3mFt where m is the degeneration, so there must be some slight degeneration in an OP27 (degeneration is mainly good by the way).

An FET input also has some degeneration which is factored into m, and m can be obtained by manipulating the equation, but I will not bother you with that. It is approx. 13 or 14 for an FET input, so if we got 2.4V/uS, we multiply that by say 14 and we get 33V/uS, and that is the input slew rate we now have.

If you remember a few posts ago I said an MC cartridge would need around 11V/uS, then 33V/uS is going to be good.

The only thing about noise gain is it adds noise (hiss) so the result isn't going to be silent. It isn't going to be in your face either, but there will be some. Putting it on the analyser I found S/N measured 63.5dB, which if you add the gain gives about 125.5dB equivalent input noise (EIN), which isn't bad, but isn't the best you'd get. The best is around 130dB so we're around 4.5dB worse. If you consider 3dB is double the power, then 4.5dB is going to be noticeable, but I can live with that.

Have you ever heard a live band? Maybe not these days (because it's all digital) but not so long ago you could hear the hiss when they set up and got the PA ready. Well, the hiss isn't going to be that loud, unless you use a 0.1mV output cartridge.

It now sounds a lot better. MC has been tamed a different way, and without an input filter because input filters kill the cartridge's inherent slew rate.

So has it been un-abandoned? We'll see. I'm not rushing into anything.