neiljadman
Senior Member
But if I'm right, them PDs design avoids this need...and the same board can be set up not only for MM/MC implementation, but also fine tuned to specific cartridges.
Phono board options
- Thread starter PigletsDad
- Start date
But if I'm right, them PDs design avoids this need...and the same board can be set up not only for MM/MC implementation, but also fine tuned to specific cartridges.
PigletsDad
My intelligence test came back negative.
Yes, the Naim design does. In this case, we can tune performance by change the compents that set the operating points.
If the 4 board solution (or the 2 stereo board one) is chosen, there is the option to use different configurations in front end, which becomes a good idea for some fairly extreme corners of the cartridge world.
For ultra low output devices (say below 200uV), I have an alternate front end that does not have differential pairs. This is 3dB quieter, but has some electrolytics in the signal path.
I have an alternate MM design using FETs, that has some different trade-offs to the bipolar one; marginally less noise in the treble, but slightly worse distortion (still pretty good).
The one big board, or board per channel, options make it harder to play round with alternate front-end stages if they are needed.
If the 4 board solution (or the 2 stereo board one) is chosen, there is the option to use different configurations in front end, which becomes a good idea for some fairly extreme corners of the cartridge world.
For ultra low output devices (say below 200uV), I have an alternate front end that does not have differential pairs. This is 3dB quieter, but has some electrolytics in the signal path.
I have an alternate MM design using FETs, that has some different trade-offs to the bipolar one; marginally less noise in the treble, but slightly worse distortion (still pretty good).
The one big board, or board per channel, options make it harder to play round with alternate front-end stages if they are needed.
Hi PD
After reading this from Post #22: "For ultra low output devices (say below 200uV), I have an alternate front end that does not have differential pairs. This is 3dB quieter, but has some electrolytics in the signal path."
I am interested in seeing the alternative front end design as I have a Troika which has a quoted output of 100uV and may well benefit from the 3dB lower noise floor, opinion?
Thanks in advance.
Kind Regards
Peter
After reading this from Post #22: "For ultra low output devices (say below 200uV), I have an alternate front end that does not have differential pairs. This is 3dB quieter, but has some electrolytics in the signal path."
I am interested in seeing the alternative front end design as I have a Troika which has a quoted output of 100uV and may well benefit from the 3dB lower noise floor, opinion?
Thanks in advance.
Kind Regards
Peter
PigletsDad
My intelligence test came back negative.
http://www.stereophile.com/phonocartridges/987linn/index2.html suggests the Troika is about 200uV, so not exceptionally low.
For the amusement of the assembled crowd (both of you), I'll sketch the alternate front end, and do a simulation or two of it, over the next couple of days.
For the amusement of the assembled crowd (both of you), I'll sketch the alternate front end, and do a simulation or two of it, over the next couple of days.
PigletsDad
My intelligence test came back negative.
Linn's own document at http://www.33audio.com/enter/data/Linn.pdf says 300uV on page 8.
PigletsDad
My intelligence test came back negative.
OK, I have a very rough netlist for the alternate stage, but not yet a usable schematic.
Simulation shows that under artificial conditions, the noise floor may be as low as 0.35nV/rt Hz; roughly the same as a 7Ohm resistor. The artificial conditions consist of ignoring the Johnson noise from the cartridge itself, which is comparable to this, except for the very lowest resistance types (like the Kondo).
The stock version has about the same noise as a 15 Ohm resistor.
Simulation shows that under artificial conditions, the noise floor may be as low as 0.35nV/rt Hz; roughly the same as a 7Ohm resistor. The artificial conditions consist of ignoring the Johnson noise from the cartridge itself, which is comparable to this, except for the very lowest resistance types (like the Kondo).
The stock version has about the same noise as a 15 Ohm resistor.
Thanks PD and Paul
Using the S boards (Asaka) and now K boards (Troika) in my pre-amp, I have not had any problems with noise, hence my original question.
PD amuse away, I am really looking forward to the alternative low noise front end (will try both options). Thanks for a great design and some wonderful threads full of patient explanations, this is all very interesting indeed - thank you for sharing your knowledge with us!
Kind Regards
Peter
Amazing how the specs and measurements vary from one source to another. I have copied the published specifications from the Linn Arm & Cartridge Service Manual (April 2002) below.
CARTRIDGE SPECIFICATIONS
Moving Coil: Troika, Karma, Asaka
Frequency response: 20 Hz to 20 KHz +/- 1 dB
Separation at 1 kHz: 30dB, 30dB, 27dB
Tracking weight range*: 1.5-1.7g, 1.5-1.7g. 1.5-1.7g
Channel balance @ 1 kHz: within 1 dB
Recommended load: greater than 10 ohms; Nominal 470 ohms
Capacitance loading: not critical
Stylus type: Vital, Vital, Vital
Tracking angle (deg): 20, 20, 20
Weight (grams): 7, 7, 7
Compliance (cu): 10, 10, 10
Output voltage: 100uV, 100uV, 100uV
* Tracking is normally set to 1.6 - 1.7 for moving coil cartridges and
1.7 - 1.8 for moving magnet cartridges. Exact tracking weight should
be set by listening at normal room temperature.
Using the S boards (Asaka) and now K boards (Troika) in my pre-amp, I have not had any problems with noise, hence my original question.
PD amuse away, I am really looking forward to the alternative low noise front end (will try both options). Thanks for a great design and some wonderful threads full of patient explanations, this is all very interesting indeed - thank you for sharing your knowledge with us!
Kind Regards
Peter
Amazing how the specs and measurements vary from one source to another. I have copied the published specifications from the Linn Arm & Cartridge Service Manual (April 2002) below.
CARTRIDGE SPECIFICATIONS
Moving Coil: Troika, Karma, Asaka
Frequency response: 20 Hz to 20 KHz +/- 1 dB
Separation at 1 kHz: 30dB, 30dB, 27dB
Tracking weight range*: 1.5-1.7g, 1.5-1.7g. 1.5-1.7g
Channel balance @ 1 kHz: within 1 dB
Recommended load: greater than 10 ohms; Nominal 470 ohms
Capacitance loading: not critical
Stylus type: Vital, Vital, Vital
Tracking angle (deg): 20, 20, 20
Weight (grams): 7, 7, 7
Compliance (cu): 10, 10, 10
Output voltage: 100uV, 100uV, 100uV
* Tracking is normally set to 1.6 - 1.7 for moving coil cartridges and
1.7 - 1.8 for moving magnet cartridges. Exact tracking weight should
be set by listening at normal room temperature.
simonbrown
The Dogfather
Likewise
PigletsDad
My intelligence test came back negative.
I have a project in my tube but the time seems not to be enough.
http://sjostromaudio.com/smf/index.php/topic,106.0.html
http://sjostromaudio.com/smf/index.php/topic,106.0.html
orrinoconnor
pfm Member
I'm wanting to build a phono-stage at the moment so what ever the outcome or arrangement of boards, as long as it will do low output MC like DL103, you can count me in!
Ray
Ray
PigletsDad
My intelligence test came back negative.
Well spotted - there actually a couple of other mistakes too, now I look. But it should be clear enough to show the basic idea - rather than a differential pair, the input transistors are single ended, and the feedback path contains a coupling capacitor.
I would prefer a single board but a wire link between the two sections or similar for tinkering.
power supply should be included on the same board, but laid out to allow easy connection of off board supply for those mad buggers using a power amp supply. having two possibilities includes those that want a small unit with no large caps for fitting into a plinth with offboard psu and those who want simplicity.
power supply should be included on the same board, but laid out to allow easy connection of off board supply for those mad buggers using a power amp supply. having two possibilities includes those that want a small unit with no large caps for fitting into a plinth with offboard psu and those who want simplicity.
PigletsDad
My intelligence test came back negative.
Schematic above now corrected.
PigletsDad
My intelligence test came back negative.
OK, yet another variant for the front-end circuit.
I've simulated this from a netlist, but not drawn a schematic yet.
This is MM only. Well, a high output MC would be just fine, but it is too noisy for low output MC. It use a pair of N-channel JFETS as input devices (or I suppose a dual, if we can source one at a sane price). Distortion is low for anything like a level that even the hottest MM cartridges will produce; depending on the gain setting, it can handle about a volt input without clipping.
I've done the simulations assuming 2SK117 input devices. 2SK170 would be even better, but not so easy to get. 2SK389 or LSK389 are nice duals, but hard to get, and would need a group buy.
The noise will be limited by the fundamental Johnson noise of the cartridge impedance and load resistor; the circuit adds negligible amounts to that.
PSRR is fundamentally very good. Combined with the low gain, this means that mV levels of noise on the rails don't matter, so any simple arrangement will do. A VBE is fine or an LM317 will do nicely. I have simulated the circuit with a simplistic VBE, and and volts of ripple on the raw supply are rejected.
Schematic for this variant coming soon.
I've simulated this from a netlist, but not drawn a schematic yet.
This is MM only. Well, a high output MC would be just fine, but it is too noisy for low output MC. It use a pair of N-channel JFETS as input devices (or I suppose a dual, if we can source one at a sane price). Distortion is low for anything like a level that even the hottest MM cartridges will produce; depending on the gain setting, it can handle about a volt input without clipping.
I've done the simulations assuming 2SK117 input devices. 2SK170 would be even better, but not so easy to get. 2SK389 or LSK389 are nice duals, but hard to get, and would need a group buy.
The noise will be limited by the fundamental Johnson noise of the cartridge impedance and load resistor; the circuit adds negligible amounts to that.
PSRR is fundamentally very good. Combined with the low gain, this means that mV levels of noise on the rails don't matter, so any simple arrangement will do. A VBE is fine or an LM317 will do nicely. I have simulated the circuit with a simplistic VBE, and and volts of ripple on the raw supply are rejected.
Schematic for this variant coming soon.