Retro - A fully symetrical phono stage with RIAA filter

[Russ White]

OK, I went back to the drawing board for MC cartridges.

This is the MC version with servo/rumble filter.

It works very well.

Differential offset is < 1mv.

I borrowed a couple low output MC carts, and they both worked very well.

Still open to suggestions.

Cheers!
Russ

[PigletsDad]

Russ,

I'm afraid it still has a massive noise penalty; R18 and R19 give 1.3nV/rt Hz of noise each, in total they will put the front end noise floor at about 1.9nV/rt Hz, without any allowance for the active devices in IC2. Unweighted, across 20kHz, this comes to about 260nV in total. Referenced to typical 500uV input level, this gives ~ 66dB SNR.

A good design can do about 0.7nV/rt Hz, the best are round 0.5nV/rt Hz, giving 74 or 77dB SNR respectively.

My other concern is that you can't raise the input impedance for cartridges that are specified for loads above 200Ohms, without lowering gain and making the noise worse. For example, some of the Linn cartridges want 470 Ohms. I don't view fiddling with MC cartridge loading as big concern, but it matters to some people.

Getting the full SNR potential of an MC cartridge is hard enough single ended; balanced is even more of a problem!

One possibility is just to give up on the load matching for the cartridges, and run them into a virtual earth; I think the Dynavector phono amp offers this as an option, and it is known as Dr T mode. This would be done by shorting out R18/R19. You would need to reduce R1/2 R7/8 to keep the gain down. The noise floor would then be set by the en of Ic2.

[Russ White]

Hi PD,

Yes I envisioned shorting R18/R19 for low output carts, and using them only for high output MC carts where DNR should still be very good. My only concern there is that the response of the cart may suffer into 0R load.

The MM/MC version (it should work for either) is pretty much the same but uses the instrumentation amp front end. I am thinking I will do both versions. Then people can use whichever cct suites their particular wants.

I think all of the challenges are ones that can be overcome.

FWIW as I had it tested last night (mostly with high output MC) it sounds quite nice. But the DNR would have been quite a lot better than with the low output carts.

[PigletsDad]

Another corner case, which may be no problem at all, is the Decca London series of cartridges. They are MM (I think strictly moving iron), with sum and difference coils; so unlike almost every other cartridge known to man, the two channels are not electrically isolated from one another. They have somewhat higher output voltage than most MMs.

I can't see any problems for the Retro design, but it is worth a quick think.

[Joshua_G]

I don't get why have a balanced phono stage when phono cartridges are SE.

[Russ White]

Originally Posted by Joshua_G

I don't get why have a balanced phono stage when phono cartridges are SE.

This is only partly true.

If you run the both ends of the cartridge into a virtual ground it is truly differential signal.

Using it into a fully differential op amp via XLR allows you better reject any cable or wiring related artifacts. Though using standard RCA wiring works very well too. Just let the cart - side float. Nothing is connected to anything but the +/- inputs of the differential input.

In the case of XLR just the shield is connected the GND.

At also allows for the nice simple passive RIAA filter, and a balanced output which is good for getting to the next stage.

There are also all the usual low distortion benefits inherent in fully symmetrical amplifiers.

Cheers!
Russ

[Joshua_G]

Most record players have RCA output and the cartridge minus is internally connected to the player's chassis. Changing it to balanced output will require modification of the player's wiring.

[Russ White]

Sure, this is actually not a problem, as making one end of the differential input GND still works perfectly, and still has all the other benefits of the design. You would only lose some of the input noise rejection.

Also, lifting that chassis GND from the RCA's is not a hard thing to do at all in most cases. It took me about 10 minutes.

Cheers!
Russ

[BrianL]

DNR -- what's that? It appears you are talking about noise floor or dynamic range??

There have been a number of publications on using transimpedance (I-to-V) stages to amplify moving coil outputs. The coil is very low impedance, and when you think about the physics, a current is what is being induced into the windings by wiggling within a magnetic field. As mentioned, Dynavector has had various preamps over the years using this circuit configuration. I believe I first saw this in an article by R.N. March in Audio Amateur in 1982 (?). This was also advocated at one time by J. Peter Moncrieff (International Audio Review) who claimed that this connection of the mc cartridge ('driving' a 0 Ohm load) did a better job of damping the cartridge response.

I haven't checked the RIAA EQ (equalization) closely, but I assume you've consulted Stanley Lipshitz's paper on the topic.

[Russ White]

Originally Posted by BrianL

DNR -- what's that? It appears you are talking about noise floor or dynamic range??

I haven't checked the RIAA EQ (equalization) closely, but I assume you've consulted Stanley Lipshitz's paper on the topic.

Hi Brian,

I personally was talking about dynamic range, but I would like to optimize both as much as is practical.

So far I have only checked the RIAA equalization by simulating it with an AC sweep. I will see if I can find the paper. I have not read it.

I have always been more into DACs, so this is my first foray into the wild world of Vinyl.

Cheers!
Russ