Now, what are the goods, bads, and 'uglies' regarding phono stages? How are they similar to a line stage, and how are they different?
Well, first: The phono stage has to have a great deal of gain, a minimum of 60dB at 50Hz and up to 100dB for some moving coil cartridges. You can just imagine how easy it is to get hum, noise pickup with thus kind of necessary gain.
The input has to be very low noise, 0.4nV/rt Hz or so. Or the equivalent noise of a 5-25 ohm resistor.
The output should be as high as 10V without clipping in order to handle excessive levels that might be on certain records. Distortion at 3V, should be in the region of .01%. Of course, better is sought, if it does not require added compromise. (more later)
Well, first: The phono stage has to have a great deal of gain, a minimum of 60dB at 50Hz and up to 100dB for some moving coil cartridges. You can just imagine how easy it is to get hum, noise pickup with thus kind of necessary gain.
The input has to be very low noise, 0.4nV/rt Hz or so. Or the equivalent noise of a 5-25 ohm resistor.
The output should be as high as 10V without clipping in order to handle excessive levels that might be on certain records. Distortion at 3V, should be in the region of .01%. Of course, better is sought, if it does not require added compromise. (more later)
Last edited:
Check the runout for Bob Ludwig's mastering mark. I have a dozen or so great Nonesuch LP's that I can't wait to play again.
Now, what are the components that are similar in both a line stage and a phono stage?
First, the primary power supplies can be similar: Same voltage, current, noise, etc.
Most of the same jfets can be used for both, and even the mosfets, as well. Same resistor quality, direct coupling, servos, etc.
One critical factor in phono playback, more than a line amp, is the QUALITY of the RIAA EQ caps. I use Reliable RT Styrene caps or Reliable Teflon caps for my EQ, in the Vendetta, Constellation, and the JC-3 phono stages.
The difference between Reliable Styrene and Teflon is very subtle, and may not be worth the extra cost. (more later)
First, the primary power supplies can be similar: Same voltage, current, noise, etc.
Most of the same jfets can be used for both, and even the mosfets, as well. Same resistor quality, direct coupling, servos, etc.
One critical factor in phono playback, more than a line amp, is the QUALITY of the RIAA EQ caps. I use Reliable RT Styrene caps or Reliable Teflon caps for my EQ, in the Vendetta, Constellation, and the JC-3 phono stages.
The difference between Reliable Styrene and Teflon is very subtle, and may not be worth the extra cost. (more later)
One feature I haven't seen yet, but that would be worthwhile on phono equalizers, is a non-sticking soft (?) clipper. Needs to be adjustable for level sensitivity unless the equalizer itself has adjustable gains separate for each channel (which would itself be a huge improvement on classic designs). Then it could be fixed-set at maybe +25db re 5cm/sec 1KHz.
Thanks,
Chris
Thanks,
Chris
As I may have mentioned, I intend to provide an adjustable clean clipper as an option on at least some models in a line of phono preamps. When it is defeated, JA and others can blast the thing with huge overloads and report high margins, but which if passed on to the rest of a given system might entail >10kW at the loudspeaker.
Why FETs for an MC stage? I assume you're still talking MCs- for an MM, it's a logical choice, but has its own design issues which you might wish to discuss.
Of course it's easier to get to ultra-low voltage noise with bipolars, especially if some of the alleged low rbb' still-available parts are as good as claimed, but there is a comfort in tiny input bias currents from JFETs, even if one has to string a plethora (SJ Perelman would quip and insist on, mocking the then-contemporary cliche "veritable plethora") of them together.
It is somewhat more complicated than whether a low Rbb' bipolar transistor is available.
Difficulty in biasing, is one reason that bipolars are not so good.
With either jfets or bipolars, you can get the same input VOLTAGE noise. It just takes more current to achieve the same noise with jfets, and the input capacitance can be VERY high.
It has been known how to use either jfets or paralleled bipolars since the mid '60's. The problem then was the available product. Most bipolar transistors have fairly high Rbb' from 40-200 ohms, only a few have Rbb' of 2-4 ohms, and I don't know how available they are today.
All else being equal, the jfet is more linear, equally low noise, and easily self biased to allow doing without any coupling cap.
Difficulty in biasing, is one reason that bipolars are not so good.
With either jfets or bipolars, you can get the same input VOLTAGE noise. It just takes more current to achieve the same noise with jfets, and the input capacitance can be VERY high.
It has been known how to use either jfets or paralleled bipolars since the mid '60's. The problem then was the available product. Most bipolar transistors have fairly high Rbb' from 40-200 ohms, only a few have Rbb' of 2-4 ohms, and I don't know how available they are today.
All else being equal, the jfet is more linear, equally low noise, and easily self biased to allow doing without any coupling cap.
This is used most of the time since decades. Makes-me smile with people allergic to Digital
You guys are mean
Do you know you can buy plug-ins for your digital audio workstation to emulate vinyl sound? I wonder what the algorithms are...
They do take in account a LOT of parameters, like:
iZotope VinylThe ultimate lo-fi weapon, iZotope Vinyl uses 64-bit processing and advanced filtering, modeling and resampling to create authentic "vinyl" simulation, as if the audio was a record being played on a record player.
You have complete control over the following parameters:
Mechanical Noise: The amount of turntable motor rumble and noise
Wear: Control how worn out the record is, from brand new to played a few thousand times
Electrical Noise: Internally generated electrical noise, such as 60 Hz grounding hum
Dust: The amount of dust on the record
Scratch: The number and depth of scratches on the record
Warp Depth: The amount of warping and the warp shape for the record—from no warp to the edges totally melted and warped
jan
Do you know you can buy plug-ins for your digital audio workstation to emulate vinyl sound? I wonder what the algorithms are...
They do take in account a LOT of parameters, like:
iZotope VinylThe ultimate lo-fi weapon, iZotope Vinyl uses 64-bit processing and advanced filtering, modeling and resampling to create authentic "vinyl" simulation, as if the audio was a record being played on a record player.
You have complete control over the following parameters:
Mechanical Noise: The amount of turntable motor rumble and noise
Wear: Control how worn out the record is, from brand new to played a few thousand times
Electrical Noise: Internally generated electrical noise, such as 60 Hz grounding hum
Dust: The amount of dust on the record
Scratch: The number and depth of scratches on the record
Warp Depth: The amount of warping and the warp shape for the record—from no warp to the edges totally melted and warped
jan
Last edited:
Might be pertinent to mention this little 'expose': http://www.soundonsound.com/sos/feb10/articles/analoguewarmth.htm. They've been fooling us for years ...!!
Frank
Frank
Thanks, John. Let me follow up with a few more:
1. Does that input capacitance really matter, again restricting ourselves to low output MCs?
2. How important is the linearity for the first stage considering the tiny amount of signal involved at that point in the circuit?
3. What about gain comparisons? At the end of the day (as they say in MBA-speak), we're interested in s/n. More gain = higher s/n, with all else equal.
4. Is it useful to make the s/n better than the limiting s/n due to the cartridge itself (i.e., the output versus Johnson noise)? If so, why?
Jan, have you ever tried those vinyl plug-ins? I have - in the hope that emulating some of the defects of vinyl might give it "that sound". I.E., is it the flaws of vinyl that we like?
I found the effects awful. Maybe I just wasn't good at using the software, but it never sounded better than a crude gimmick to me. Was wondering if anyone else had better luck.
I found the effects awful. Maybe I just wasn't good at using the software, but it never sounded better than a crude gimmick to me. Was wondering if anyone else had better luck.
No I never tried it, and it's just one example I found.
But in general, with these type of effects and others like room reverb simulation, you must tread VERY softly.
Adjust it to the point that you think you switched it off, and with longer listening you do hear the effect. Don't overdo it; I can very well understand it can sound terrible.
But it's an interesting technology, like the plugins that emulate tube sound, where you can sometimes even select different tubes and/or the biasing point and overload.
jan
But in general, with these type of effects and others like room reverb simulation, you must tread VERY softly.
Adjust it to the point that you think you switched it off, and with longer listening you do hear the effect. Don't overdo it; I can very well understand it can sound terrible.
But it's an interesting technology, like the plugins that emulate tube sound, where you can sometimes even select different tubes and/or the biasing point and overload.
jan
You can remove them by hand and you would be surprised how inaudible the result is. At 24 bit I was surprised at how few actually hit the rails. A soft clipper still leaves the pop and my preamp recovered very fast, so fast that you usually can see the grove wall sag and tip resonance on the falling edge.
I noticed on the 24/96 files I recorded in Austin there were a few activations of the Fostex's soft clipper, only during the start of SY's VERY entusiastic applause (he was sitting directly next to the mic).
- Status
- Not open for further replies.