As an EE, I'm curious to know more about this process. From my quick searches, I did read that it takes more power for the lathes to etch the stamping masters than the lower frequencies and that the coils on lathes heat up quite a bit. Please direct me to some links that I could read up on. Is it possible for records to contain harmonics beyond 25kHz other than quad CD4 albums?Frequency range of harmonics found on records
Hello everyone. I just want to get this out there that I have next to no clues when it comes to the process of producing records. I have a rudimentary understanding that lathes are used to cut the grooves into the stamping masters which in turn gets used to press vinyl records. Thats as far as I got in this process and even there, I'm thinking I'm skating on thin ice. As an EE, I'm curious to know more about this process. From my quick searches, I did read that it takes more power for the lathes to etch the stamping masters than the lower frequencies and that the coils on lathes heat up quite a bit. Please direct me to some links that I could read up on. Is it possible for records to contain harmonics beyond 25kHz other than quad CD4 albums?
As an EE, I'm curious to know more about this process. From my quick searches, I did read that it takes more power for the lathes to etch the stamping masters than the lower frequencies and that the coils on lathes heat up quite a bit. Please direct me to some links that I could read up on. Is it possible for records to contain harmonics beyond 25kHz other than quad CD4 albums?
Re: Frequency range of harmonics found in vinyl.
Hi,
To start, I'd get ahold of these publications (see links below). I think they are still in print. Required reading if you are into the technology.
What are you looking to accomplish and why do you want to discount CD4? That they exist, in fact, proves that you can cut and playback high frequencies.
BTW, no etching involved here. You cut a "V" shaped groove into a lacquer blank. The stamper is an electroformed negative of the lacquer.
https://www.aes.org/publications/anthologies/downloads/jaes_disk-anthology-1.pdf
https://www.aes.org/publications/anthologies/downloads/jaes_disk-anthology-2.pdf
Mark
To start, I'd get ahold of these publications (see links below). I think they are still in print. Required reading if you are into the technology.
What are you looking to accomplish and why do you want to discount CD4? That they exist, in fact, proves that you can cut and playback high frequencies.
BTW, no etching involved here. You cut a "V" shaped groove into a lacquer blank. The stamper is an electroformed negative of the lacquer.
https://www.aes.org/publications/anthologies/downloads/jaes_disk-anthology-1.pdf
https://www.aes.org/publications/anthologies/downloads/jaes_disk-anthology-2.pdf
Mark
Re: Frequency range of harmonics found in vinyl.
CD4 was cut at half speed IIRC, which does of course nothing to help with the geometric limits but does significantly reduce the current required in the cutting head coils.
There are three limits on the amplitude you can cut up there:
One geometric imposed by the back angle on the cutting sapphire (the radial velocity must never exceed the tangential velocity), which imposes a limit dropping at 6dB/octave on the input program.
A second due to the assumed osculating circle of the replay stylus, which is why CD4 needed a line contact stylus.
The third is heat, the output of the IRIAA card is essentially a velocity signal, and rises at 6dB/octave from about 2kHz,the coil current however produces a force, so the required coil current rises with frequency (for constant velocity) also by 6dB/octave... Thus across the region where heating is a headache power rises at 12dB/octave from about 2kHz, winding up some 40dB (10,000 times) higher to cut the same program level at 20k as it is at 1k.
It is fortunate that the other limits stop you doing that, but 500W for a peak was not unheard of, and peak loudness wars saw 1kW cutting amps being deployed (And lots of blown out heads even with helium injection).
Finally, the issue that confuses some folks, the cutting heads are generally feedback designs and the feedback gain (obviously) falls with frequency, by the time you hit 20k there is not that much left, and all the various non linearities are becoming somewhat more prominent, so you can expect to see a mess of low level intermod products on replay up there simply because the loop is running out of puff.
The classic Neumann lathe electronics has a two pole LPF at 50kHz right before the cutting EQ and also has an opamp in the equaliser that runs out of GBP at about 100khz or so, in addition to a LC LPF in the HF limiter card that seems (Yay for spice!) to be something like an inverse Chebyshev LPF going over at 50kHz or so.
You can cut up there, you just cannot cut much level up there and expect the intermod to disappoint.
There are three limits on the amplitude you can cut up there:
One geometric imposed by the back angle on the cutting sapphire (the radial velocity must never exceed the tangential velocity), which imposes a limit dropping at 6dB/octave on the input program.
A second due to the assumed osculating circle of the replay stylus, which is why CD4 needed a line contact stylus.
The third is heat, the output of the IRIAA card is essentially a velocity signal, and rises at 6dB/octave from about 2kHz,the coil current however produces a force, so the required coil current rises with frequency (for constant velocity) also by 6dB/octave... Thus across the region where heating is a headache power rises at 12dB/octave from about 2kHz, winding up some 40dB (10,000 times) higher to cut the same program level at 20k as it is at 1k.
It is fortunate that the other limits stop you doing that, but 500W for a peak was not unheard of, and peak loudness wars saw 1kW cutting amps being deployed (And lots of blown out heads even with helium injection).
Finally, the issue that confuses some folks, the cutting heads are generally feedback designs and the feedback gain (obviously) falls with frequency, by the time you hit 20k there is not that much left, and all the various non linearities are becoming somewhat more prominent, so you can expect to see a mess of low level intermod products on replay up there simply because the loop is running out of puff.
The classic Neumann lathe electronics has a two pole LPF at 50kHz right before the cutting EQ and also has an opamp in the equaliser that runs out of GBP at about 100khz or so, in addition to a LC LPF in the HF limiter card that seems (Yay for spice!) to be something like an inverse Chebyshev LPF going over at 50kHz or so.
You can cut up there, you just cannot cut much level up there and expect the intermod to disappoint.