After finishing the Pearl phono I find the gain is too low in my system to work with the Denon 103R and Grado Statement Sonata. One is a 0.25mv and the other 0.5 mv cartridge. It plays at a normal level with the linestage opened all the way up. It will never play George Jones at the proper level after a few Pabst.
The buffered passive linestage and Zen V4 have a lot to do with low system gain. But I will use the stepups used with the Croft phono used prior to building the Pearl.
What I would really like is a balanced input active MC stage powered by the single rail supply of the Pearl. I have seen where members have duplicated the first stage of the Pearl to do this.
But since a phono is a balanced source I would like to use this feature. I wire my stepups this way. The plugs and jacks are 3 pin DIN, the shields and grounds only meet in the phono stage.
Wayne, I saw where you used the Grado Statement to audition the prototype Pearl. Was the linestage gain enough to allow this?
George
The buffered passive linestage and Zen V4 have a lot to do with low system gain. But I will use the stepups used with the Croft phono used prior to building the Pearl.
What I would really like is a balanced input active MC stage powered by the single rail supply of the Pearl. I have seen where members have duplicated the first stage of the Pearl to do this.
But since a phono is a balanced source I would like to use this feature. I wire my stepups this way. The plugs and jacks are 3 pin DIN, the shields and grounds only meet in the phono stage.
Wayne, I saw where you used the Grado Statement to audition the prototype Pearl. Was the linestage gain enough to allow this?
George
> play George Jones at the proper level
Off.
> 0.25mv and the other 0.5 mv cartridge
> a phono is a balanced source I would like to use this
Issues:
A balanced input usually has higher thermal noise than an unbalanced input.
Fractional milliVolt levels with hi-fi S/N expectations really jam you against device noise. Speaking broadly, BJT noise is 1µV so your 0.25mV source gives 48 dB S/N. You can push noise down, and the cart output spec is not the peak level, so it can be a lot better than 48dB. But it is still a challenge.
The transformer is the simple elegant solution. The low-output cart has as much power as a standard cart, it is just at an impedance lower than any practical active device's optimum noise impedance. Transformers work great for that. The bandwidth and distortion of a low-level low-impedance transformer is usually very good, not like working iron at high level with too many turns.
Ah, but why not try active?
> duplicated the first stage of the Pearl
Without even looking at a Pass Pearl: That can work. Because the output level need only be a few hundred mV, it is possible a different topology could be better. National Semi has a LM394 design working very low voltage (like 3V?) so that it can use ultra-low value resistors for minimum voltage noise without becoming a power-pig.
Ah, Pass' Pearl is paralleled JFET. That can work. (A BJT design is a more common and possibly easier path for MC work.) Noise will tend to be lower than a MM cart but higher than a MC cart (though possibly not higher than surface noise; phono amp design is full of little balances like this). Noise voltage could be lower with more JFETs and increased total current. Noise voltage goes down -about- as square-root of number of devices and current: Eight 2SK170BL could be 3dB quieter, sixteen of them could be 6dB quieter. (That's device noise; we soon have coil noise dominating and diminishing-returns starts to bite us.) Sixteen of these JFETs would eat 80 milliAmps of current, over 2 watts of power in this stage. Here is where a lower supply voltage "makes sense", though not if you have to drop it from an existing 40V raw supply or have to go buy another power transformer. Offhand I don't know how low the D-S voltage on a sSK170BL can be and stay linear, maybe Mr Pass knows.
And this is not balanced. A true differential input needs four times as many devices and total current for similar noise resistance. The complication of making it differential is significant. Diff mode can turn even-order distortion into a lower amount of odd-order distortion, but THD should not be a big issue at a few mV. Induced noise from power lines is lower with differential input, but for the short lines and quiet EMI environment used in home phono this is usually not a problem. I'd be inclined to strap the cart single-ended.
Build another "Pass Pearl First Stage". Have a coupling cap before you go into your MM input; 0.5µFd should be ample. Some MC carts want the input resistor much lower than 47K. See how that sounds. If it is "a little noisy" you can try mass parallel JFETs to reduce noise voltage. If it comes out "real noisy", then you need another idea.
W. Marshall Leach Head Amps for Moving-Coil Cartridges
Base LM394 datasheet (good lo-noise BJT reference)
LM394 head pre-amp for moving-coil phono cartridges AN-222
For other very-low-Z and Balanced amplifiers, try microphone preamps. But most of these are way over-kill for a low-fixed-gain job like this, and noise performance at and below 150Ω is generally not "theoretical".
> shields and grounds only meet in the phono stage.
Maybe I misunderstood. Does one side of the cart have a low/zero resistance path to ground?
Off.
> 0.25mv and the other 0.5 mv cartridge
> a phono is a balanced source I would like to use this
Issues:
A balanced input usually has higher thermal noise than an unbalanced input.
Fractional milliVolt levels with hi-fi S/N expectations really jam you against device noise. Speaking broadly, BJT noise is 1µV so your 0.25mV source gives 48 dB S/N. You can push noise down, and the cart output spec is not the peak level, so it can be a lot better than 48dB. But it is still a challenge.
The transformer is the simple elegant solution. The low-output cart has as much power as a standard cart, it is just at an impedance lower than any practical active device's optimum noise impedance. Transformers work great for that. The bandwidth and distortion of a low-level low-impedance transformer is usually very good, not like working iron at high level with too many turns.
Ah, but why not try active?
> duplicated the first stage of the Pearl
Without even looking at a Pass Pearl: That can work. Because the output level need only be a few hundred mV, it is possible a different topology could be better. National Semi has a LM394 design working very low voltage (like 3V?) so that it can use ultra-low value resistors for minimum voltage noise without becoming a power-pig.
Ah, Pass' Pearl is paralleled JFET. That can work. (A BJT design is a more common and possibly easier path for MC work.) Noise will tend to be lower than a MM cart but higher than a MC cart (though possibly not higher than surface noise; phono amp design is full of little balances like this). Noise voltage could be lower with more JFETs and increased total current. Noise voltage goes down -about- as square-root of number of devices and current: Eight 2SK170BL could be 3dB quieter, sixteen of them could be 6dB quieter. (That's device noise; we soon have coil noise dominating and diminishing-returns starts to bite us.) Sixteen of these JFETs would eat 80 milliAmps of current, over 2 watts of power in this stage. Here is where a lower supply voltage "makes sense", though not if you have to drop it from an existing 40V raw supply or have to go buy another power transformer. Offhand I don't know how low the D-S voltage on a sSK170BL can be and stay linear, maybe Mr Pass knows.
And this is not balanced. A true differential input needs four times as many devices and total current for similar noise resistance. The complication of making it differential is significant. Diff mode can turn even-order distortion into a lower amount of odd-order distortion, but THD should not be a big issue at a few mV. Induced noise from power lines is lower with differential input, but for the short lines and quiet EMI environment used in home phono this is usually not a problem. I'd be inclined to strap the cart single-ended.
Build another "Pass Pearl First Stage". Have a coupling cap before you go into your MM input; 0.5µFd should be ample. Some MC carts want the input resistor much lower than 47K. See how that sounds. If it is "a little noisy" you can try mass parallel JFETs to reduce noise voltage. If it comes out "real noisy", then you need another idea.
W. Marshall Leach Head Amps for Moving-Coil Cartridges
Base LM394 datasheet (good lo-noise BJT reference)
LM394 head pre-amp for moving-coil phono cartridges AN-222
For other very-low-Z and Balanced amplifiers, try microphone preamps. But most of these are way over-kill for a low-fixed-gain job like this, and noise performance at and below 150Ω is generally not "theoretical".
> shields and grounds only meet in the phono stage.
Maybe I misunderstood. Does one side of the cart have a low/zero resistance path to ground?
PRR said:
Go for the transformer. Sounds good and less greif. Lundahl LL9206 or 1678, depending on how much extra gain you want.
http://hometown.aol.com/kevinc927/myhomepage/business1.html
GRollins said:
Thanks everyone! I will keep my AN transformer for now. The George Jones has to stay also.
The input only is balanced. The trasformer and lead do not ground either leg of the cartridge. All the grounds meet inside the preamp, along with one side of the cartridge. Then it is SE on.
I rewired all my analog rig years back to use 3 pin DIN connectors. They do sound better than RCA's. Plus balanced is a breeze.
This is the good thing about using three pin connectors. The cartridge can be wired up without tying the ground to either polarity of the cartridge..
Analog SA recommended simple MC stage. It is only SE, but I will try it. If I prefer the AN step up nothing is lost.
George
It wouldn't be difficult to make that JFET front end balanced if you wanted.
I've got something similar here, only using 2N5457s (another N-ch JFET) with a 30V rail. Nuthin' to it.
Balanced phono stages don't get troublesome until you get to the RIAA filter, and even that's not too much of a problem. Just try to use well matched parts on both phases.
Grey
I've got something similar here, only using 2N5457s (another N-ch JFET) with a 30V rail. Nuthin' to it.
Balanced phono stages don't get troublesome until you get to the RIAA filter, and even that's not too much of a problem. Just try to use well matched parts on both phases.
Grey
Balanced phono EQ
"Balanced phono stages don't get troublesome until you get to the RIAA filter, and even that's not too much of a problem. Just try to use well matched parts on both phases."
Grey,
You post is a little short for me to understand exactly what you are saying. Do mean that you put a RIAA network on each phase (like two of the passive networks in the Pearl phono preamp for example, one for each phase) that the RIAA networks for each phase should be closely matched to the other one for part value tolerances?
How about a simple schematic or block diagram.
Thanks,
Fred
"Balanced phono stages don't get troublesome until you get to the RIAA filter, and even that's not too much of a problem. Just try to use well matched parts on both phases."
Grey,
You post is a little short for me to understand exactly what you are saying. Do mean that you put a RIAA network on each phase (like two of the passive networks in the Pearl phono preamp for example, one for each phase) that the RIAA networks for each phase should be closely matched to the other one for part value tolerances?
How about a simple schematic or block diagram.
Thanks,
Fred
Just to assist and confirm what has been said here.
Who is better qualified than the designer for a recommendation?
If you do a search you will find the designer Wayne Coburn has already specified a solution.
I asked this querry yonks ago when I ordered the Pearl semi kit and boards from Karen at Passlabs.
Wayne promptly responded advising duplication of the 1st stage and amending one resister in the last stage.
As I recall an extra set of Jfets arrived in the parcel from ForestHill.
That's what I call service.
Ian
Who is better qualified than the designer for a recommendation?
If you do a search you will find the designer Wayne Coburn has already specified a solution.
I asked this querry yonks ago when I ordered the Pearl semi kit and boards from Karen at Passlabs.
Wayne promptly responded advising duplication of the 1st stage and amending one resister in the last stage.
As I recall an extra set of Jfets arrived in the parcel from ForestHill.
That's what I call service.
Ian
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.