Ohhhhhhhhhhhhhhhh !!!
I am honoured by your reply, GREAT John Curl, my teacher !
Raymond
The power supply is OK, in principle, except that better damping can be achieved by putting a 100 ohm resistor across each common mode choke coil, rather than any extra resistance put in series. However, the second part of the power supply is missing. The inter-buffers. They are very important to isolate the AC-DC power supply
Yes, that's what I'm talking about. Some people may get the idea that what's been shown is just fine. The juicy part is missing.
Since, I am now working a great deal on phono circuits, one aspect of phono reproduction not covered very well, is phono EQ, or what is commonly called RIAA Equalization. As my Blowtorch preamplifer has phono EQ included and a few other units, as well, we might discuss it.
As we have learned from previous inputs, MC cartridge loading does little to change the gross characteristics of the phono cartridge. Therefore, in order to get the response of the cartridge to be relatively flat with frequency when playing back records that have been made with different disc cutting systems, we have to apply a standard EQ called the RIAA EQ CURVE, and then elect or decline to modify the out-of-band components of the phono cartridge output. This can be controversial. Just wait, and see. ;-)
As we have learned from previous inputs, MC cartridge loading does little to change the gross characteristics of the phono cartridge. Therefore, in order to get the response of the cartridge to be relatively flat with frequency when playing back records that have been made with different disc cutting systems, we have to apply a standard EQ called the RIAA EQ CURVE, and then elect or decline to modify the out-of-band components of the phono cartridge output. This can be controversial. Just wait, and see. ;-)
Accurate phono eq is something I have worked toward for a long time. Small errors (+/- .1 dB) can have a disproportional effect on the perceived sound. The out of band treatment has a major effect as well, both low frequency and high frequency. There are a lot of preconceived notions about correct treatment that don't hold up to scrutiny. Both low frequency poles and high frequency poles that don't actually exist in the chain, or are not at the specific frequencies they are claimed to be at.
Should you provide compensation for the cartridge response, and how? And adjustment for left/right balance in the cartridge? Getting one correctly/optimally aligned is also difficult.
Cutting systems are not perfect, neither are the tape machines upstream. However their gross errors do not seem to eliminate the sensitivity to errors in the playback eq. (I know this seems contradictory.) In terms of feedback, the no feedback obsession is at odds with the up to 40 dB of feedback around the cutterhead itself. It would be interesting to look at the delay around that loop. Cutter head repair service for Neumann stereo cutterheads!!
The new interest in additional eq options for mono playback is interesting, and there are a lot of options to support.
Should you provide compensation for the cartridge response, and how? And adjustment for left/right balance in the cartridge? Getting one correctly/optimally aligned is also difficult.
Cutting systems are not perfect, neither are the tape machines upstream. However their gross errors do not seem to eliminate the sensitivity to errors in the playback eq. (I know this seems contradictory.) In terms of feedback, the no feedback obsession is at odds with the up to 40 dB of feedback around the cutterhead itself. It would be interesting to look at the delay around that loop. Cutter head repair service for Neumann stereo cutterheads!!
The new interest in additional eq options for mono playback is interesting, and there are a lot of options to support.
Do you have any proof to support this extraordinary claim? Simple physics shows that 0.1dB at 10ft. distance means moving the listener by 1/8" off axis. Unless you are using some sort of head jig and a GEB member as subject, this is unlikely to be heard or reproduced experimentally.
I agree that it seems a little beyond something that can be supported. I don't have a formal test to support it, just years of annoying feedback from customers, and my own experiences. However an ABX test can be skewed by a .1 dB difference. The frequency range with the small difference is broad, it may have a corresponding time/phase difference etc. The easy way to address it is to hold the eq to very close tolerances.
Another outrageous claim- some of the sonic differences attributed to circuit design could be just response differences.
Another outrageous claim- some of the sonic differences attributed to circuit design could be just response differences.
Balancing the channels to 0.1dB is already hard to believe it's audible. But matching the RIAA to 0.1dB is insane. Fortunately, it's not that hard (or expensive) to do. It would catch a few more customers
Last edited:
I agree with Demian that small changes in the RIAA make a difference. However, even bigger changes in frequency response happen with individual types of MC cartridge, and even high frequency response from the beginning to the end of the record. Most serious designers use 1% caps, usually matched to even better tolerance, for stereo reproduction and .5%-1% resistors, also matched in pairs. I can only speak for serious designers, experienced in the field, who make their living in competition with each other. We DO need a standard to compare with each other properly. We would not want Dr. Lipshitz to criticize our efforts with a poorly designed or sloppily made RIAA network.
Demian,
I completely agree with your statement that 0.1dB channel difference is audible in a RIAA network.
For 30 years I've held that truth and always match the RIAA caps and Rs to better than 4 digits to achieve this.
To the doubters: The stereo image achieved is superior to that obtained from a looser matching of parts. Why? Because if the fundimental of an instrument is placed (say) at 20 degrees from center, but it's harmonics spread out over an area due to a non perfectly matched eq curve, then the image will be diffuse, to a greater or lessor degree.
Try it, really match the RIAA parts, correctly align the cartridge, and hear the image really tighten up, and if it's near perfect, the image will 'snap" into focus!
Regards, Allen
I completely agree with your statement that 0.1dB channel difference is audible in a RIAA network.
For 30 years I've held that truth and always match the RIAA caps and Rs to better than 4 digits to achieve this.
To the doubters: The stereo image achieved is superior to that obtained from a looser matching of parts. Why? Because if the fundimental of an instrument is placed (say) at 20 degrees from center, but it's harmonics spread out over an area due to a non perfectly matched eq curve, then the image will be diffuse, to a greater or lessor degree.
Try it, really match the RIAA parts, correctly align the cartridge, and hear the image really tighten up, and if it's near perfect, the image will 'snap" into focus!
Regards, Allen