Good Phono preamp circuit?

[nuvistor]

Hope you're using an MM or high output MC, that's what your phono pre schematic is best suited for despite the MM/MC switch.

2nd stage could be inverting to eliminate common-mode issues there, but you'd need to increase 2nd stage gain, R11 would double and 2nd stage input resistance would equal new R11 value.

As soon as the MM cartridge's stylus meets the moving groove, choice of bipolar vs FET input op amps or passive vs active EQ based on noise at 1kHz becomes an academic discussion. I'd be more concerned about 1/f noise noted by jcx.

I have an MM phono pre similar to your schematic using OPA637 or AD8610 for first stage, LM6171 for second stage, cascoded FET current sinks on both outputs. It sits on the shelf, replaced by a tube phono pre with 2 triodes, first stage FET CCS-loaded, second stage FET/bipolar "mu follower" loaded, active EQ with 2-pole compensation.

The op-amp pre is compact, noise is low enough, and sounds clean and detailed, certainly better than a Dyna PAS, but the tube/CCS pre seems better at revealing low-level detail and ambience.

You can't go too wrong with your design for MM, give it a try.

[PeterMoreton]

hi, thanks again to all those who have contributed to this thread - i have learned a great deal.

Could some explain what an FET current sink on an opamp output is used for? - does it bias the opamp into class-a mode or something?

should i add current sinks to the opamps in my design?

[nuvistor]

Could some explain what an FET current sink on an opamp output is used for? - does it bias the opamp into class-a mode or something?

Yes, current sinks connected to op amp outputs bias the output stage to class A. Op amp outputs are already biased class A to a certain current output, the current sinks increase the output current for class A operation and can reduce distortion with low impedance loads. A good source of information on this subject is here. Your preamp will measure well without them. Some like them and some don't, they're simple to add so try one and listen for any difference.

[amplifierguru]

The design you have settled on should be fine - although, by providing full gain wideband from the first stage, it suffers from a 20dB reduced signal handling capability in the HF. It will clip on a 10 times lower HF signal, whereas, had you chosen the full 'around one stage' RIAA 10x more HF signal could be accomodated.

For lowest noise a discrete FET front end could be used with 1 op amp AND provide >40dB extra OL gain into the bargain.

Topology shown below for Nch pair. ( for your 2SK146's mlloyd!)Flip cct for Pch FETs. RIAA as per standard (see Andy C's WWW).


jcx, It's arguable which is a better chip while surface noise is some 20dB+ higher as soon as the needle drops.

Cheers,
Greg

[janusz]

Hi Peter,

I have an interesting discrete phono preamp design, namely AEM6010. It is 20 years old and was designed by Dave Tilbrook. I have never built this preamp but I listened to it and in those times it was probably one of the quietest and detailed preamps available.

I have a copy of the article with PCBs etc and associated PS and control unit schematics as well. A zip file with its schematics is attached. Of course it could be redesigned to accept different components including low noise fets. It may not be what you are looking for but some food for thought.

Cheers,

[PeterMoreton]

Janusz, wow that circuit is complex!

I like to follow Einsteins advice:

"keep everything as simple as possible, but no simpler"

In my research I have found a huge number of Phono/RIAA circuits out there on the 'net, and they appear to fall into a few distinct categories (IMHO!)

1/ Too simple, and not technically great.
(eg: single Opamp)

2/ Simple, and potentially great, with careful component selection.

(eg: FET frontend + Opamp, Dual Opamp etc)

3/ Complicated because they date from an era when good sound dictated complex electronics. Probably still sound good, but modern components can do the same with less.

(eg: John Linsley-Hood, AEM6010)

4/ Complicated and very good

(eg: some commercial designs)

5/ Complicated because they are snake-oil :-)

I am looking for Option 2, and I think I'm close to settling on a dual-opamp design, although "amplifierguru's" topolgy using a front-end built from 2SK146's is making me thing again.....

This is fun...

[andy_c]

Originally posted by PeterMoreton
I am looking for Option 2, and I think I'm close to settling on a dual-opamp design, although "amplifierguru's" topolgy using a front-end built from 2SK146's is making me thing again.....

As well it should. One thing I'd like to mention regarding some of the info on my WWW as it relates to ampllifierguru's design. The analysis on my site assumes a single-pole compensation of the op-amp. In this configuration, it's necessary to have an op-amp with a very large DC open-loop gain for accurate low-frequency equalization. For accurate high-frequency equalization, an op-amp with a large gain-bandwidth product is needed. There are very few op-amps that satisfy both requirements at once. Amplifierguru's circuit uses two-pole compensation, which elegantly sidesteps this problem. The additional FET stage at the input provides a higher DC open-loop gain, allowing much more flexibility in the choice of op-amp for the second stage.

[PeterMoreton]

Andy_c,

I'm having trouble finding a UK supplier of 2SK146's - they appear to be quite a rare beast over here. Are there any more commonplace alternatives that you can offer...?

thanks, peter moreton

[andy_c]

Sorry, I don't know. Maybe a separate thread in Electronics and Parts would turn up something?

[Panelhead]

Originally posted by PeterMoreton
Andy_c,

I'm having trouble finding a UK supplier of 2SK146's - they appear to be quite a rare beast over here. Are there any more commonplace alternatives that you can offer...?

thanks, peter moreton

This is going to be expensive. Erno Borbely stocks them at www.borbelyaudio.com. But they are 25 euros each.
MCM shows they can be ordered. I have found they can always deliver items like this. Do not know their sources, but all kinds of rare old transistors are available at reasonable prices. Look at www.mcminone.com to get an idea of cost and lead time.
Usually the wait for the parts like this that are not stocked at MCM is 3 - 4 weeks. They must have access to a transistor graveyrd somewhere.
They came through with all the transistors to build an original Le Monstre a year ago or so. Some of those were hard to find ten years ago. Same with the original transistors to build the Krell KSA-50 clones.


George