Joachim, I'm sorry, once again, if I insulted you in some way. What I call this is CROSSTALK, something that is usually not permitted in 12 step programs, for good reasons. Crosstalk is: 'The trouble with you is', 'I like my way better',etc.
I started to describe an IMAGINARY 'People's Preamp' a SIMPLE design, yet elegant, in its way, with minimal features, easy to understand and build, BUT well engineered and executed with the 'right' parts, not necessarily the newest or most exotic parts. This is what I hoped to do, and to get feedback from you and others in the process. I did not foresee a 'design contest' or 'Let's add some marketing features' etc. Just an engineering EXERCISE, just like in a classroom situation. Perhaps the exercise can be completed, still, it is simple enough.
I started to describe an IMAGINARY 'People's Preamp' a SIMPLE design, yet elegant, in its way, with minimal features, easy to understand and build, BUT well engineered and executed with the 'right' parts, not necessarily the newest or most exotic parts. This is what I hoped to do, and to get feedback from you and others in the process. I did not foresee a 'design contest' or 'Let's add some marketing features' etc. Just an engineering EXERCISE, just like in a classroom situation. Perhaps the exercise can be completed, still, it is simple enough.
The way I test a turntable, is to remove the rubber mat, (if possible) and DING the metal turntable with my fingernail. If it goes ding! That is bad. If it goes dunk, it is OK. It is a simple test, but revealing.
At price. The current price for a cd/dvd player which sounds identical to my reference unit is $20 for a one disc player. Four years ago it was $40 and $78 for a five disc changer. Twenty years ago it was $200. Twenty-two years ago one almost as good cost $750 list, $600 actual retail. That's what I call progress. Players made prior to that were unacceptable because of poor sound quality.
Wow! I can NOW retire in peace, as what needs to be done, has been done. The wisdom of the past should lie in a library somewhere, and we should all concentrate on the latest 'widget' that needs to be optimized. Robots, anyone? ;-)
Early cd players sounded awful. Their d/a converters made strings sound steely. That was almost eliminated by 1989 with the 20 bit 8x Denon. By 1991 it had been solved by the 1 bit 8x IMO. The 192 khz 8x Toshiba chip sounded indentical in 2007. That was good enough for me. Whatever problems are blamed on cds, they have nothing to do with the technology, it's how the technology is used.
Having lived on the infield of WTFM's transmitter (within six blocks) for years and across the street from a ham radio operator, I'm well aware of the problems of induced RF noise. You basically have to build a Faraday cage around high gain elements including the cartridge, tone arm wiring, and low level preamp stages to have any chance to shield it. Mu metal for cartridges works best. I shield my phono cables the way the manufacturers of high quality signal cables do, with a drain wire and foil. It costs virtually nothing to do. And it works. Multiple foil/100% braids have not been necessary in my experience although the best manufactured signal wires use them. Some like thick Ethernet coax were quad shielded.
I hate the pops and clicks of vinyl records. I also don't like not having a remote control for them. Maybe I should have bought an Accutrac.
CDs combined with close miking of pianos create some sounds and problems I rarely heard on vinyl or in live performances. Spurious sounds from reverberation within the harps, strings on unstruck notes which vibrate sympathetically, the damper release/pedal action, clicking of fingernails on the keys, and even occasional humming by the pianist (Billy Taylor is one.) There are other spurious sounds recordings now pick up.
"P.S.: everyone is able to calculate necessary slew rate from highest audio sine frequency and wave amplitude. But it is not enough at all."
I've heard that mantra from the time I was a kid when University claimed its Sphericon supertweeter extended to 40 khz, when Harman Kardon sold tube amplifiers with FRs to 70 khz (A500 and A700) and sold state amps to 1 mhz (A1000) and Audio Fidelity lauded their "Frey Stereophonic Curtain of Sound" that extended to 25 khz. But it flies in the face of well accepted Fourier analysis. On what evidence do you base your statement?
I think Salas has covered that quite well. Simple and you are free to match and select exotic parts if you wish. If built carefully with a good supply I would defy you to do the "no peeking" test.
So you won't have a turntable with a ding in it. I don't blame you. I've heard of dunkin' donuts but not dunkin turntables.
I've met some pretty strange people on the internet. Back Several years ago on another site I ran into this guy, I'll call him Mr. X, who had quite a following. But he was a bit of a snob. It was said that he wouldn't exchange postings with anyone who was not a degreed electrical engineer. Then one day Mr. X made a statement which proved conclusively that he didn't have the slightest idea what the mathematics of electrical engineering was all about. When a real electrical engineer corrected him, Mr. X did the strangest thing, he ascribed his own elementary blunder to the engineer and the correction of it to himself. The engineer later took particular delight referencing Mr. X's postings showing not only his blunders but his bizarre need to always be right even when he wasn't. It turned out that Mr. X wasn't an electrial engineer himself, in fact it wasn't even an engineer of anything. He admitted he'd merely take a few courses in electrical design. His designs had more to do with tinkering than engineering. He used a trial and error method until he found somethng he liked. He never had a design goal or objective despite claiming he had very expensive, sensitive, and sophisticated test equipment. He said he couldn't measure what there was to differentiate some circuits he liked and others that he didn't. Eventually the engineer got completely bored with the site and decided to leave with a bang.
Just to finish my 'lecture' on how to make a simple, effective preamp for a Shure moving magnet cartridge of today. Well, the Shure best audiophile cartridge for sale today from Shure has an inductance of 650mH and a series resistance of 1.5K ohms. NOW, does quibbling about 200 ohms and 600 ohms seem rather trivial? I would hope so.
Now, what about voltage noise in the IC itself? Is 8nV/rt Hz too much? Would 5nV/rt Hz be really that much better? Let us add the noise contributions up and see.
We now have: 600 ohms (feedback resistor, worst case), + 1500 ohms from the cartridge making a total of 2100 ohms, or 1700 ohms in Joachim's original case to 4000 ohms that would be the equivalent noise of the 4134 that I first selected. So the total noise would be the equivalent of about 6000 ohms (+100, -300 ohms) depending on the RIAA network scaling. In other words, Joachim and I were debating whether a 0.3dB difference in noise was enough of an inprovement, when the IC got loaded with 3 times the output load to get it. Well, is it worth it? Maybe the IC deserved a lighter load, just to keep it happy. The second problem, that we can only hope to minimize is the need for roughly equal impedance loading on both the + and - inputs of the fet IC op amp. Everybody interested, please read both the detailed spec sheet of the OPA 4134 and the OPA 1644 to understand why this is important. Well, 600 does NOT match 1600, but it is a better fit than 200. See my point? Or am I going too fast for you?
Now, what about the OPA 1644? Well, it looks interesting. Is it commonly available, what packages? Darn, would I have to go to surface mount exclusively. What is its advantage? Somewhat lower noise, the equivalent of a 1.6K resistor. Now what is the difference in noise with RIAA scaling? Well, it is 1.6 + 1.7, or 1.6 + 2.1
These total 3.4 and 3.7 respectively. Take the ratio, square root, and convert to dB.
We lose 0.75dB in S/N with 1/3 loading of the output stage. Do you see the tradeoff? The caps get smaller and cheaper too!
Now I did this 'on the fly' at my keyboard. I could have made a mistake somewhere. I will gratefully accept any correction from you eagle-eyed critics out there, but the point that I want to make, has been presented. Enough said.
Now, what about voltage noise in the IC itself? Is 8nV/rt Hz too much? Would 5nV/rt Hz be really that much better? Let us add the noise contributions up and see.
We now have: 600 ohms (feedback resistor, worst case), + 1500 ohms from the cartridge making a total of 2100 ohms, or 1700 ohms in Joachim's original case to 4000 ohms that would be the equivalent noise of the 4134 that I first selected. So the total noise would be the equivalent of about 6000 ohms (+100, -300 ohms) depending on the RIAA network scaling. In other words, Joachim and I were debating whether a 0.3dB difference in noise was enough of an inprovement, when the IC got loaded with 3 times the output load to get it. Well, is it worth it? Maybe the IC deserved a lighter load, just to keep it happy. The second problem, that we can only hope to minimize is the need for roughly equal impedance loading on both the + and - inputs of the fet IC op amp. Everybody interested, please read both the detailed spec sheet of the OPA 4134 and the OPA 1644 to understand why this is important. Well, 600 does NOT match 1600, but it is a better fit than 200. See my point? Or am I going too fast for you?
Now, what about the OPA 1644? Well, it looks interesting. Is it commonly available, what packages? Darn, would I have to go to surface mount exclusively. What is its advantage? Somewhat lower noise, the equivalent of a 1.6K resistor. Now what is the difference in noise with RIAA scaling? Well, it is 1.6 + 1.7, or 1.6 + 2.1
These total 3.4 and 3.7 respectively. Take the ratio, square root, and convert to dB.
We lose 0.75dB in S/N with 1/3 loading of the output stage. Do you see the tradeoff? The caps get smaller and cheaper too!
Now I did this 'on the fly' at my keyboard. I could have made a mistake somewhere. I will gratefully accept any correction from you eagle-eyed critics out there, but the point that I want to make, has been presented. Enough said.
Last edited:
The TD-150 is an OK turntable, it is certainly better than the Marantz SL15 that I have in storage. I bought a TD-150 first, in 1967, then another in 1977. Still, have the second one, with a Shreve Rabco arm on it, in my lab.
"Just to finish my 'lecture' on how to make a simple, effective preamp for a Shure moving magnet cartridge of today. Well, the Shure best audiophile cartridge for sale today from Shure has an inductance of 650mH and a series resistance of 1.5K ohms."
Excellent point. That the optimal design of the phono preamp stage must take into account the variables of the electrical characteristics of the source impedence and be able to be a correct match to it. Since when taking the cartridge, tone arm wiring, and interconnect wiring into account the number of permutations of these parameters can get large, it will take a lot of research or some other clever method to cope with it. But I'm certain that your $40,000 version will solve this problem handily.
The LCR circuit model gives the predicted result. Ironically it is exactly the same ordinary second order differential equation that is Newton's second law of motion which predicts the mechanical resonance model. When solved for forced oscillation that result will tell you exactly where the mechanical resonances will occur and how large they will be. Correlation between the model and actual experience is excellent and has been demonstrated for nearly a hundred years for the electrical example, far longer for the mechanical example.
47K is the world standard to which the cartridge is designed around. Only load capacitance is to be optimized. Get up to speed, please, PMA. You will confuse people with your nit-picking.
The Shure cartridge is a 4th order low pass filter. Please come to understand this, before confusing the issue further.
I just did not get it that it was a kind of lecture. Yes, your solution is more economical and may do the job even better. I just wanted to comunicate that i have had good experience with the OPA164... series and i will try to avoid in the future to give the impression that i know things audio better then you. That was not my intention. Sorry.
To continue, the Shure cartridge and many like it, use both a mechanical resonance along with the electrical resonance (47K, 650mH, and 200pf) for example to make the flattest response when added to the relatively undamped mechanical stylus mass, contact surface area and vinyl springiness. A VERY FLAT frequency response to 20KHz and sometimes beyond, is possible with this approach.
Last edited:
- Status
- Not open for further replies.