Okay, here are entire posts:
Anyhow, comprehension of what's written may improve SNR of the threads here.
No reason not to do both. A popular magazine has an interesting quandary- it could measure en and in vs frequency and report that. With this information, any reader with a basic engineering background can calculate the noise that he'll get from a particular cartridge. The problem is that the audience is dentists, investment bankers, and lawyers, who in general wouldn't know what to make of that info.
If it were my magazine, I'd report the shorted input, but then repeat the same measurement with a standard source (say, 400R + 0.5H or something) representing a MM and/or 5R representing a MC. Then show the standard source data compared to an ideal noiseless preamp with the same gain and frequency response curves on the same graph so that the reader can see what the preamp is actually adding, compared to theoretical limits determined by the source. These are easy measurements to do with the sort of setup that Stereophile has and give a readily understandable and graphically dramatic indication of what the real world noise of the preamp is likely to be.
John,
My concern is with DC current through the coils of the cartridge. If we want to keep this to a minimum with MM Bipolars are out. You can allow a little more with MC, but even here you really have to watch it.
Why not use a J-Fet? 2SK246 will do fine for MM only...
I do not consider any MM to be "Sota", though some can be charming.
Sorry John, but it does not, when compared to a pure resistor.
There is always a rise in noise (self noise and if bipolar's are used a rise in current derived noise) compared to resistor...
Honestly, I would NOT use Bipolar Op-Amp's (or bipolar transistors) in a position where they couple into a MM Cartridge (or a volume control at that) while there is still a low capacitance J-Fet or Mosfet in some garage foundaries sortiment...
But to each their own.
Ciao T
My concern is with DC current through the coils of the cartridge. If we want to keep this to a minimum with MM Bipolars are out. You can allow a little more with MC, but even here you really have to watch it.
Why not use a J-Fet? 2SK246 will do fine for MM only...
I do not consider any MM to be "Sota", though some can be charming.
Sorry John, but it does not, when compared to a pure resistor.
There is always a rise in noise (self noise and if bipolar's are used a rise in current derived noise) compared to resistor...
Honestly, I would NOT use Bipolar Op-Amp's (or bipolar transistors) in a position where they couple into a MM Cartridge (or a volume control at that) while there is still a low capacitance J-Fet or Mosfet in some garage foundaries sortiment...
But to each their own.
Ciao T
Pavel,
The definition is for power, which is meaningless in the context of line level electronics though technically correct.
This is a nice attempt at muddying crystal clear waters, as this stages precisely the same as stating the signal amplitude in dB with reference to a known reference level and the noise amplitude in dB with reference to a known reference level, as in -145dBV noise and -66dBV signal.
So what you post underlines precisely why we specify noise we do, as the signal level is essentially unknown and may vary widely with cartridge (from 0.1mV for some of the lowest out MC Cartridges to nearly 10mV for some higher output MM types)...
Well, I'm afraid I disagree on usage. In fact, input noise is VERY commonly used to specify noise in circuits with variable gain and low signal levels, precisely because of the difficulties to come up with any other sensible specification.
See for example the following links. As they are pro-audio related the measurements are stated as Ein in dBu rathe rthan dBV, which will give a touch under 3dB difference...
Measuring Mic-Preamp Noise | Benchmark Media
http://www.avisoft.com/test/noisefloors.pdf
Selecting Mic Preamps
Rupert Neve Designs – Microphone Impedance and Noise
APHEX Systems 1100 Class A Tube Mic Preamp +A/D Conv.
All of these use input referred noise (Ein) in dBu for specifying noise, so not only is this kind of specification regular, but in fact common (lest you wish to imply that the folks at Aphex, Benchmark, Rupert Neve, Rane and others all do not know what they are doing).
Ciao T
The definition is for power, which is meaningless in the context of line level electronics though technically correct.
This is a nice attempt at muddying crystal clear waters, as this stages precisely the same as stating the signal amplitude in dB with reference to a known reference level and the noise amplitude in dB with reference to a known reference level, as in -145dBV noise and -66dBV signal.
So what you post underlines precisely why we specify noise we do, as the signal level is essentially unknown and may vary widely with cartridge (from 0.1mV for some of the lowest out MC Cartridges to nearly 10mV for some higher output MM types)...
Well, I'm afraid I disagree on usage. In fact, input noise is VERY commonly used to specify noise in circuits with variable gain and low signal levels, precisely because of the difficulties to come up with any other sensible specification.
See for example the following links. As they are pro-audio related the measurements are stated as Ein in dBu rathe rthan dBV, which will give a touch under 3dB difference...
Measuring Mic-Preamp Noise | Benchmark Media
http://www.avisoft.com/test/noisefloors.pdf
Selecting Mic Preamps
Rupert Neve Designs – Microphone Impedance and Noise
APHEX Systems 1100 Class A Tube Mic Preamp +A/D Conv.
All of these use input referred noise (Ein) in dBu for specifying noise, so not only is this kind of specification regular, but in fact common (lest you wish to imply that the folks at Aphex, Benchmark, Rupert Neve, Rane and others all do not know what they are doing).
Ciao T
Hi,
So, what do we standardise upon?
Some of us use J-Fets or Tubes and we would probably like a lot of L and R like the old Shure and Pickering MM carts to make our designs look especially good next to those inferior bipolar designs.
Meanwhile the manufacturers who use Bipolar inputs may wish to not use any MM's at all as model and instead concentrate on high output MC's which are often between 100 and 200 Ohm to show how their excellent designs are superior to those inferior Fet or Tube designs.
I'd give it up as a bad job in any case.
Most modern Phono stages are sufficiently quiet to not have noise constitute any issue when playing vinyl. And that is what really matters.
Ciao T
So, what do we standardise upon?
Some of us use J-Fets or Tubes and we would probably like a lot of L and R like the old Shure and Pickering MM carts to make our designs look especially good next to those inferior bipolar designs.
Meanwhile the manufacturers who use Bipolar inputs may wish to not use any MM's at all as model and instead concentrate on high output MC's which are often between 100 and 200 Ohm to show how their excellent designs are superior to those inferior Fet or Tube designs.
I'd give it up as a bad job in any case.
Most modern Phono stages are sufficiently quiet to not have noise constitute any issue when playing vinyl. And that is what really matters.
Ciao T
That's an intriguing suggestion. I have pondered for a long time if I should load the preamp input with something that resembles a real MM or MC cartridge - Noel Keywood in the UK used to use an actual cartridge - but the question then becomes what would be most representative; a Grado or Shure MM, for example, which are very different electrically.
The shorted input becomes the default condition because of its consistency.
And thank you for your comments on Stereophile's incorporation of measurements in its reviews. I have always believed that reviews must be anchored with measured data since I first started reading audio magazines in the mid-'60s. I have made no secret of the fact that I have modeled Stereophile's reviews on those by John Crabbe and Ralph West in the 1960s-vintage Hi-Fi News, which did a great job of balancing the two worldviews.
John Atkinson
Editor, Stereophile
" the Krohn-Hite amp schematic is available on Dick's site at:
http://burwenaudio.com/Krohn-Hite_UF-101.pdf
Regards
Giorgio
Wow, that is weird, i have not run into anyone who knew what those were.
I have a pair of them that "someday" i plan to clean up.
One worked fine, the other had a problem.
Anyway, thanks for the memory jolt.
Best,
Tom Danley
Danley Sound Labs
http://burwenaudio.com/Krohn-Hite_UF-101.pdf
Regards
Giorgio
Wow, that is weird, i have not run into anyone who knew what those were.
I have a pair of them that "someday" i plan to clean up.
One worked fine, the other had a problem.
Anyway, thanks for the memory jolt.
Best,
Tom Danley
Danley Sound Labs
John,
I use 10 Ohm as "default MC" and a dead Shure V15 with it's cantilever assembly removed as "default MM" for measurements that are not shorted.
I would suggest that as the V15 is representative of most real MM's and one of the few "High End MM's" it fits the bill well, the 10 Ohm for MC also hit a spot somewhere in the middle...
Ciao T
I use 10 Ohm as "default MC" and a dead Shure V15 with it's cantilever assembly removed as "default MM" for measurements that are not shorted.
I would suggest that as the V15 is representative of most real MM's and one of the few "High End MM's" it fits the bill well, the 10 Ohm for MC also hit a spot somewhere in the middle...
Ciao T
How about using both shorted input and R +jwL simulations of MM and MC. Similarly as dummy load when testing power amp loading.
The problem is that you've gone from a simulation that satisfies some to a simulation that satisfies none. If you pick a model (the numbers I used in my earlier post were extracted from my hindquarters) that's somewhere in the middle of the pack, it will not be dead-on for anyone, but it will be Close Enough for most. As an intro to the new measurement methods, you can do a one-time measurement of several cartridges that are on the high and low side of the chosen resistances and inductance and show those as an envelope- this way, a reader can understand where his cartridge falls, and how to use the "standard" measurement to predict whether or not he'll hear the preamp noise over the basic thermal noise of the cartridge.
There's an analogy to the standard speaker load you use for amplifier tests ("But what if I have Quads?")- that simple addition to your test suite is superbly useful and was a great choice on your part. I have confidence that you could be just as clever in measurements on the other end of transduction.
I'll try these suggestions out. (I am sure Michael Fremer must have a long-unused V15 somewhere in his cupboard!)
John Atkinson
Editor, Stereophile
Looking last night at various phono cartridges that MIGHT be used with the JC-3, I think that 500mH with 500 ohms might be a good reasonable worst case. The VERY OLD Shure V15's might have had more inductance, but this limits the frequency response too much for anything of quality. This is because the electrical resonance has to be balanced with the mechanical impedance resonance to create a flat output. In the old days, 1H was typically used, but rarely, today.
For the record, however, the Ampex AG440 analog tape reproduce used an equivalent reproduce head inductance of 1H, and we used high quality ($0.12) bipolar transistors on the input, with more difficult reproduce noise conditions than the RIAA reproduce might provide. Thousands of successful recordings were made with that combination.
For the record, however, the Ampex AG440 analog tape reproduce used an equivalent reproduce head inductance of 1H, and we used high quality ($0.12) bipolar transistors on the input, with more difficult reproduce noise conditions than the RIAA reproduce might provide. Thousands of successful recordings were made with that combination.
- Status
- Not open for further replies.