Negative feedback coil for the Edikow machine.

[Viewmaster]

In my pursuit to record NBTV onto Edison wax cylinders (my Edikow machine) I am building a high powered, ferrofluid cooled and highly damped disc cutter.
This will also incorporate a feedback coil, whose tiny signals will be amplified. (This coil and its magnet is mounted near to the cutting stylus to ensure that it truly reflects the stylus movement.)

Then using my double beam oscilloscope I will compare the main input NBTV signal with the feedback signal, to ensure that they are 180 degree out of phase with each other.
The feedback signal will then be mixed with the main NBTV signal, (180 deg out of phase with it) at the main amp input stage, to attain the negative feedback.

I am not very well up in electronics (mechanics is my line) and wonder if anyone has any comments on this approach to attaining and using negative feedback from a coil?
Albert.

[Steve Anderson]

...This will also incorporate a feedback coil, whose tiny signals will be amplified. The feedback signal will then be mixed with the main NBTV signal, (180 deg out of phase with it) at the main amp input stage, to attain the negative feedback. ....any comments on this approach to attaining and using negative feedback from a coil? Albert.

Albert, this is tricky and not easy even for those who manufactured disc cutting lathes. I would be hesitant to try it on a home-brew machine. One of the major problems is the delay time between the input to the cutter head and the output from the feedback, such that at higher frequencies with the introduced phase-shift it simply goes into oscillation.

Philips tried this with bass drivers in loudspeaker systems years ago, it wasn't that succesful and doesn't seemed to have caught on, and this was for below 500Hz.

Not quite sure why you're considering this, but if you're getting distortion of low frequencies this could be due to over-excursion of the cutter head. Reducing the signal level during cutting will increase the high-frequency noise.

If you're not using it now I suggest pre-emphasis during cutting, and de-emphasis during playback. A simplified version of the RIAA characteristic as used on records since the 50s should do fine for NBTV.

Steve A.

[Stephen]

I think that Albert is trying to overcome the low natural resonance frequency of a high power cutter head due to its moving mass. The commercial cutter heads since the 1950s have had this natural resonance in the 1-2 kHz range. Contrast this to the low power amateur and professional cutter heads of the 1940s and earlier that have natural resonances in the 6-15 kHz range. There was an incentive to minimise moving mass to raise the resonance as high as possible because the low power of available valve cutting amplifiers made any degree of negative feedback impractical.

Steve, I think that the RIAA curve, being approximately 6 db per octave above about 2 kHz and resulting in an approximate 20 db boost compared to 2 kHz at 20 kHz and a 40 db boost compared to 20 Hz, may be impractical for NBTV signals. Audio signals have relatively low harmonic content amplitude at higher frequencies, whereas NBTV signals may have relatively high fundamental frequency content at high frequencies that may overload the cutter head.

For NBTV, a better equalisation curve might have a 250 Hz turnover frequency with 6 db per octave attenuation and no high frequency pre-emphasis. This would be similar to the equalisation of early electrical grammophone records.

[Steve Anderson]

Steve....Audio signals have relatively low harmonic content amplitude at higher frequencies, whereas NBTV signals may have relatively high fundamental frequency content at high frequencies that may overload the cutter head.

For NBTV, a better equalisation curve might have a 250 Hz turnover frequency with 6 db per octave attenuation and no high frequency pre-emphasis.

Stephen, all agreed. I did say a simplified version of the RIAA curve(s), I guess a lot for NBTV would have to be determined by experiment. But without feedback (pun intended) from Albert, I'm not sure why he's contemplating this arrangement of cutter head feedback.

Steve A.

[Viewmaster]

Thanks to both of you for replies.
Yes I am going to have a modified trial equilisation, both for cutting and playback.
I have obtained ghostly looking images off wax but nothing I can show which would make any sense.
I do like trying neg feedback to increase freq response and help iron out resonances. If it doesn't work out I just disconnect the coil!

Also I appreciate very much the phase shift problem and cutters do burn out due to this oscillation at the top end.. I may also try a low pass filter set at say 20KC to reduce the unwanted HF. Some pro cutters do this.
Maybe a fast blow fuse in cutter circuit would stop any fireworks!

I found with my old cutter designs that to get 10kc recorded then I had to
cut the LF end in order to turn up the wick! I decided that what was needed was a big power input into the cutter but severly damped to prevent excess excursion of the stylus. "Power with restraint!"

I now have a better pickup too. Previous were piezo, now I have a Shure moving magnetic PU with better freq response.

Another line of attack is to record at half speed, again many pros did this way back.
Thanks again for your thoughts.
Albert.

[Stephen]

Albert, if by "ghostly looking images" you mean that the images comprise mainly outlines without fullness, especially in large areas, there may be low frequency attenuation somewhere in the recording/reproducing chain. I wonder if it might be due to the low frequency response characteristic of the pickup. It is much easier to get a good low frequency response from a cutter head than a pickup.

[Viewmaster]

Albert, if by "ghostly looking images" you mean that the images comprise mainly outlines without fullness, especially in large areas, there may be low frequency attenuation somewhere in the recording/reproducing chain. I wonder if it might be due to the low frequency response characteristic of the pickup. It is much easier to get a good low frequency response from a cutter head than a pickup.

Yes, Stephen, but other problems were too much varyation in speed and vibration too on my model lathe I originally used so the image was never stable. I could actually hear the wow and wobble on the NBTV output from the recorded wax!

I wasted much time on this lathe but now have an old Ediphone dictation machine run by a DC motor I installed.
This cuts out speed variations (I hope!) and should get me going when the new high powered cutter is finished. I am mounting it on the carriage so that I can alter the angle of attack of the cutting stylus to the wax cylinder (The rake angle I believe it is called in disc cutting parlance).

With a grooove spacing of 6 thou the depth of cut is only 2 thou on nil cutter input (hill and dale as per original Edison cylinders) so there is little room for error in all this....
....Gee, why don't I stick to high quality, simple to record and play CDs?

Albert.

[Stephen]

....Gee, why don't I stick to high quality, simple to record and play CDs?

Albert.

I think that the answer is the romance of cutting recordings. I used to have a great deal of fun cutting records.