I have no background in testing ADC/DAC until recently - so I dont know all the in's-n-out's of them... I start with simple basic tests and get deeper and deeper into it as I see something to explore and understand....
with that prelude-- So far the thd vs. input level on low cost ADC is pretty bad; the undistorted dynamic range is more limited than i thought it would be; You cannot get anywhere near the 0dB level (1v) and the artifacts/spurious tones/noise limits you on the low end of levels. Not what I expected to find. To add insult to injury, the individual harmonics produced by level changes in complex ways.
So far a lot of it is from poor pcb layout design, grounding, jitter, and over-all build issues.... many which can be over-come with care in build. But the ADC itself isnt that great. I would hate to think this is IT ! This is the best we are going to get with digital consumer products.
Thx-RNMarsh
Last edited:
Everyone, Parasound customers are not expected to add capacitance to 'optimize' their phono cartridges. Most do not have the test equipment to do so, in fact.
The recommended input capacitance for a Shure V15 is 250pf.
Most cartridges, including Grado, could care less, within reason, what the input cap load is. 1 meter of Radio Shack shielded cable is: 100pf. So, the loading is close to optimum and even under the recommended capacitance.
The recommended input capacitance for a Shure V15 is 250pf.
Most cartridges, including Grado, could care less, within reason, what the input cap load is. 1 meter of Radio Shack shielded cable is: 100pf. So, the loading is close to optimum and even under the recommended capacitance.
Here http://www.diyaudio.com/forums/equi...n-audio-range-oscillator-274.html#post3497338 is a comparison (not completely apples to apples but enough so) of a TI PCM4222 and an AK5394A. Neither are cheap but neither are really expensive (and in volume they are not a cost issue on a premium product). I don't have an ADI ADC demo board or device to compare. I think ADI put their money elsewhere. Selling exotic ADC's and DAC's to consumers is low margin. Selling ADC's and DAC's for cell towers will pay very well.
The full scale problem with DAC's goes back to some of the early Philips chips. The low pass anti-aliasing filter causes peaking or greater than FS on the output. The fix (audiophile horror!!!) is to digitally lower the level before the filter so it won't overload. If some current players have problems it usually means someone did not read the datasheet (never happens). However recordings that hit FS are not common in arenas where sensitive listeners are buying. I did get a bunch of demo files from some major labels where digital clipping was present in many of the files, as was a total dynamic range of 6 dB. A 12 bit DAC would have been more than enough for that content (mostly Hip Hop).
That means, if someone changed their interconnecting cable between phono cartridge and preamp, there really could be an audible difference... the freq response would change and the resonance damping would change etc.
Then, audiophiles would describe the sonic characteristic of these various interconnecting cables with low, med. high C and ascribe a sound to them.
-RNM
Last edited:
Well Richard, don't buy a JC-3, buy our new, improved Constellation Pegasus phono preamp. It has more adjustability. Only $19,000. You can afford to be picky.
Seriously, you really don't expect the average audiophile to use both a moving magnet, AND play with cables, do you?
It is professionally impolite for another professional audio designer to criticize another's product design, I might remind you. '-)
Seriously, you really don't expect the average audiophile to use both a moving magnet, AND play with cables, do you?
It is professionally impolite for another professional audio designer to criticize another's product design, I might remind you. '-)
For everyone else: The OPTIMUM way to optimize a moving magnet cartridge would be to have a continuously variable input capacitance AND a proper test record, with some way of measuring the frequency response over a sweep.
I once started the design of a preamp for HK, where I tried to add a variable loading cap, BUT all the small bodied ones, with enough cap range, used Mylar, and that added the 'feared' DA (1979), at that time put forth as important by Richard Marsh, and I took him seriously, so I never put that feature in any product even when it would have looked good as an added feature.
The JC-3 was designed as a practical, no frills, phono stage and we left out many adjustments, except hi-lo gain, and 100 or 47K loading, with about 100pf loading in the box.
I once started the design of a preamp for HK, where I tried to add a variable loading cap, BUT all the small bodied ones, with enough cap range, used Mylar, and that added the 'feared' DA (1979), at that time put forth as important by Richard Marsh, and I took him seriously, so I never put that feature in any product even when it would have looked good as an added feature.
The JC-3 was designed as a practical, no frills, phono stage and we left out many adjustments, except hi-lo gain, and 100 or 47K loading, with about 100pf loading in the box.
Last summer Wavebourn mentioned his friend Alex Zaslavsky's idea to use air variable capacitors for RIAA eq (in high Z valve(?) preamps). Seems like this would be an even better fit for MM/MI input trimming, if desired.
For a terminating/RFI input capacitor, Bob Cordell, in another context, recommended a Zobel with about 75 Ohms in series with the cap, the goal being to terminate the line at RF. No reason this couldn't be done with an air variable input cap if wiring lengths are kept in mind.
One argument against variable input capacitance is that it's just another resonance that *should* be damped down to nothing - which gives a high inductance source the same response as if it were a MC. This is easy to try for DIY'ers - a resistor in the several K Ohm range provides the 75uS pole just fine. A dual pot of the usual volume control size is about right to play with it. Just remove the cap currently giving the 75uS pole.
Cordell's LA article gives a more elaborate version but it's not really necessary just to try the idea. It's plenty quiet the simple way.
All good fortune,
Chris
For a terminating/RFI input capacitor, Bob Cordell, in another context, recommended a Zobel with about 75 Ohms in series with the cap, the goal being to terminate the line at RF. No reason this couldn't be done with an air variable input cap if wiring lengths are kept in mind.
One argument against variable input capacitance is that it's just another resonance that *should* be damped down to nothing - which gives a high inductance source the same response as if it were a MC. This is easy to try for DIY'ers - a resistor in the several K Ohm range provides the 75uS pole just fine. A dual pot of the usual volume control size is about right to play with it. Just remove the cap currently giving the 75uS pole.
Cordell's LA article gives a more elaborate version but it's not really necessary just to try the idea. It's plenty quiet the simple way.
All good fortune,
Chris
The question of whether or not phono cartridges otherwise identical, but with either MC or MM/MI generators, would perform differently with frequency response differences, etc. removed, hasn't been approached yet. Back in vinyl days there was lots of loose talk about generator linearity and such, but how much of that was just talk? I get more sceptical the older I get.
Thanks,
Chris
Thanks,
Chris
Oh, John;
Its not a critism of your or anyone else's design. High-End Audiophiles do change phono cables/interconnects and hear differences. This applies to any preamp. They think it is something about the cable which they are hearing.... Then the war starts on the other side - cables are perfect - they have flat freq response and measure it that way as poof there can be no such thing as a cable sound. But of course, it does affect the sound, just not for the magic they think is in the cable itself. rather, for the reasons explained here -- cartridge C loading and its affects on freq response, damping and distortion with the input stage., etc. There, happy now.
Thx-RNMarsh
Last edited:
OTOH, you could call the rest of us stupid, dumm, amateurs and pitch fork wielding villagers with impunity and still be polite, right?
jan
Absolutely true. John's claim that most MM's "don't care" about capacitive loading is ludicrous, especially coming just a day or two after he did some rough calculations to show exactly the opposite.
It may come as a great shock to sneering fashion designers, but people who buy their equipment are actually capable of reading spec sheets and doing basic arithmetic. It does not take a major investment in test equipment to look at a recommended cartridge load, subtracting the cable capacitance, and setting the preamp capacitance for the difference. A committed designer realizes this.
I'll leave it to others to explain why a 797 is not an optimal choice for MM.
Of course yes; that's not an insult, it's a simple statement of fact. Well, except for the pitchfork, I don't have one of those anymore, but John never mentioned pitchforks anyway. Just speaking for myself of course, but that's why I hang out here in the first place - to try and learn something.OTOH, you could call the rest of us stupid, dumm, amateurs and pitch fork wielding villagers with impunity and still be polite, right?
jan
Professional audio designers must be very delicate flowers, then, if they regard criticism as impolite. The only time that fair criticism would be improper is if the critic had commercial motives (e.g. he makes a rival design) but did not disclose this. Or should we expect audio designers to behave like some other 'honourable' professions and close ranks on the plebs who actually pay their wages by buying their products/services?john curl said:
I found it to work very well as the first stage in a two stage MM preamp with passive RIAA network in between (cobbled one together recently to test new PS idea). Even at 1KOhm source resistance, opamp current noise contribution is still negligeable.
The highish and temperature dependant input bias current is a downside, so it is not a device that might be optimal for consumer applications, but for DIY it is great imo. Or am I missing something?
"Not optimal" doesn't mean "doesn't work," it just means that there are better choices. The issue is the current noise. Fortunately, the rising output of a velocity sensor with frequency counteracts the rising induced noise from the cartridge inductance, but if you can start with low in, you can end up with better noise performance.
Using the classic rule of thumb (Ropt ~ vn/in), the optimal source resistance at 1kHz is about 1k (edit: per jcx's comment, closer to 500R at 100Hz, but the inductive reactance is lower as well), somewhat smaller than the Rdc plus XL of many MM cartridges. Not a huge deal, especially when considering the noise floor of the cartridge itself, but when the preamp is selling for big bucks, one might expect attention to detail.
Using the classic rule of thumb (Ropt ~ vn/in), the optimal source resistance at 1kHz is about 1k (edit: per jcx's comment, closer to 500R at 100Hz, but the inductive reactance is lower as well), somewhat smaller than the Rdc plus XL of many MM cartridges. Not a huge deal, especially when considering the noise floor of the cartridge itself, but when the preamp is selling for big bucks, one might expect attention to detail.
- Status
- Not open for further replies.