• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

The Phono Stage

What is the funktion of it. When I compare to other phonoclones they just have a single resistor to ground on the noninverting input at U3. You have added one resistor from output of U2 to the noninvering input at U3 and then a 10uF to ground. Why?
Hopefully I am specific enough ;)
/Staffan
 
The network in question from what I see in LTSpice forms a high pass filter with -3 dB about 5 hz. I am not able to find models for the OPA627/637 so I am using AD746 in the model. I would think for these purposes it is pretty close.

Whether this is worth the trouble I have yet to find out for myself.

I am using a moving magnet semi-version of this phonostage.

Using the Clearaudio Charisma and combining ideas from Pete Millett and Bob Cordell I think I have found something that works very well.

Pete Millett's LR article was an inspiration. But this was tuned to work best with an input transformer and a MC. I bought two of Mr. Millett's boards to use for the stage. Nicely made with a large ground plane. All but two of my connections are with wires above the board and those two are just to ground points. The power supply is wired as configured though I am using all of the power supply positions for each mono board.

From Mr. Cordell: the idea of using the inductance of the cartridge to form the 75 uS pole. This requires a 5000R resistor in series between the cartridge and the - pin of the 627. + to ground. 120kR feedback resistor.

Then to a 610mH inductor (from Dave Slagle) with a 1580R resistor in parallel - a 88R resistor in series to the - pin of 637. 17kR feedback resistor. Lots of gain. There is also 100 pF of air trimmer capacitance at the input.

With LTSpice this gives an astonishingly flat RIAA curve with the model for the cartridge included which is admittedly nothing more than a 660R resistor in series with 440mH of inductance. Shows being only 4 dB down at 100kHz.

Of course, I think it sounds great. I had listened to this cartridge with the SALAS phono stage and then the MILLETT as configured by Millett- I had tried the AD797s and did not like it. I installed the OPA627/637 combo and I knew something interesting was possible. Much tinkering followed.

Mr. Cordell with his VINYL TRAK phono was the inspiration for using the inductance of the cartridge. For some reason unspecified he decided to not go fully with the cartridge performing the EQ but used a higher frequency (much larger input resistor) and then added a zero in his passive equalizer between the two amplification stages.

I have no "engineering" idea why one would choose one over the other and there is no question he knows lots more about this stuff than I do but I tried it both ways and I find using the cartridge alone sounds better in my system.

Seems all one has to do is select a input resistor that resonates at 2120 Hz with your cartridge. You can then play around to find the absolute best value but you know where to begin. At least on TLSpice each section acts as two independent hinges. The rest of it should be fine. Of course if one is determined to use a line stage you will need to lower the gain. I use a version of Ed Schilling's TRUTH VC buffer - it shares no parts with Ed's but it is the same idea. Instead of the BUFs I use the TOSHIBA 2sk170/2sj74s before and after the photoresistors, which are two in parallel with a much larger format. I use incandescent light instead of LEDS. There is a capacitor between the phono and the buffer. My offset voltages were much too high and I do not want to fool around with servos. I got the feeling the phono amp likes the buffer. This device eats up about 3 dB of level hence what might seem too much gain.

I have wondered about that high pass filter. I will give it a try but was hoping that Peter might say why he included this. I have no idea if this was part of the original circuit.

I hope this is not considered too far afield for this forum. But I am an admirer of Mr. Daniel and have learned lots from him. I know folks had wondered about using his idea with MMs, and so had I, along with curiosity about using inductors for correction so ...
 
The network in question from what I see in LTSpice forms a high pass filter with -3 dB about 5 hz.


I think the 3k3 and 330k resistors are just there to minimise offset. The cap introduces a useful hi pass filter.

Last year i spent a few months playing with all sorts of LCR and LR topologies. For some reason, the LR never sounded any good to me compared to LCR. Inexplicable and disappointing. Same opamps, same regulators, same riaa accuracy.
 
You take those resistors away and you lose the high pass filter.

I found a superior inverse RIAA generator yesterday - from Stephie Bench - found it at Dave Slagle's forum.

This made changes to my values and the sound greatly improved. If you play around with LTSpice you should insert it!

If anyone cares: the resistor in parallel with the 610 mh choke is now 1830R and the input resistor into the second stage op amp is now 155R. No comparison in the sound. I was getting a 3 dB rise from 100 hz on down! The system sounds greatly more relaxed now and it is obvious that LTSpice did predict this correctly. One of these days I will make some real measurements.

I think the "Bench" just generates the needed curve where before I was using some very complicated collection of resistors and capacitors between the generator and the input.

I think I damaged an OPA627. It is quiet with no signal playing but once I place the stylus on the record (looking at the RTA screen in RWE) there is this spike at 27 hz. Until I get it corrected I can get rid of most of it with a notch filter but certainly not a permanent solution. I made so many wiring changes with these - I think this is the one I connected V+ to one of the offset adjusting pins. Damnit. This is a TO99 package. COuld not resist the allure ... Will probably go with DIP for the replacement. What is interesting is the thing seems to sound fine other than this strange oscillation.

I am speaking only about MM cartridges. I do not doubt with MCs those caps are necessary. But I think the simplicity of this approach might have something going for it but I make no pretense at knowing much about what is available out there having no opportunity or interest in hearing them.

Per the Bench generator and LTSpice I am with .5 dB from 10 hz to about 40KHz. 100kHz down about 1.5 dB. MMs can sound very good in the highs with this approach. I had a TRANSFiGURATION Proteus before it lost a channel and the fellow who was TRANSFIGURATION dies of a heart attack so I have an unfixable cartridge. Got the CLEARAUDIO based upon the recommendation of a trusted audio kook and have been very pleased.
 
How can a balanced version of this phono section be implemented?

I have two independent boards, and only 1/2 of each board is populated since each board is running one channel.

On the cartridge side, I know wiring pins 2 and 3 across the coil is correct. The tonearm uses a 5 pin DIN.

The shield is supposed to float at the cartridge side, and is connected to pin 1 which is connected to the case. A separate ground wire grounds the tonearm.

If I were to populate the other half of each board, is balanced implemented by wiring input pin2 to one channel and pin3 to the other on the same board?
 
Specifically for one channel:

Cartridge coil hot -> Input Pin 2 -> Left Channel Hot Input -> Output pin 2
Cartridge coil gnd -> Input Pin 3 -> Right Channel Hot Input -> Output pin 3
Input Pin 1 -> case
Output Pin 1 -> case

All audio signal grounds remain as built. Obviously, you need to have two complete phono sections to do this, as you will use one per channel.

Comments? Is this correct?