Piezo Feedback

[Babooino]

Since no question can be stupid enough...
I'm trying to understand the feedback in a head.
Would,a piezo disc work instead of the secondary winding?
Makes sense?

[EmAtChapterV]

I've asked about this as regards a Fairchild 541A head. (Between the damping material and the armature.) The consensus seems to be that something with such low current would need its leads to be very well shielded to avoid inductance from the drive coil. It would need to be a very small piezo element too, about 10 x 10 millimeters.

[Babooino]

I've asked about this as regards a Fairchild 541A head. (Between the damping material and the armature.) The consensus seems to be that something with such low current would need its leads to be very well shielded to avoid inductance from the drive coil. It would need to be a very small piezo element too, about 10 x 10 millimeters.

Why so small?
Is there a link where i can get more info on this?

[DjMikiRia]

I have built my own stereo lathe, can someone tell me the need for feedback please? I am open to anything that can get a better sound with designs.

[opcode66]

feedback can be used to counteract physical resonance. every cutterhead as well as every physical item has a resonant frequency. in a cutterhead the resonant frequency is the are of the frequency spectrum where response will not be flat (i.e. it will be exaggerated). Feedback can be used to determine what the real world reaction to the drive voltage is. and, when necessary it is used to attenuate the drive voltage in order to maintain flat response.

[dmz007]

Feedback is a common principle of many engineering control systems. Basically you use feedback to compare the actual output (ie cutterhead motion) to desired input (the audio signal). If there is a difference, then a controller (like a feedback amp from Flozki) can be used to drive this difference to zero.


Here's a link: http://en.wikipedia.org/wiki/Control_system.

[Babooino]

How to use feedback in a diy head?

[opcode66]

Without a second set of coils in your cutterhead you can't use feedback.

One set of coils are the drive coils. Any cutterhead including homebrews have one or two drive coils (mono or stereo head).

With an additional coil per channel you can then get feedback. Literally, the drive coil gets voltage and starts moving in the presence of the fixed pole magnet. The feedback coil is attached to the same moving part as the drive coil. The feedback coil has a very small voltage induced within it by being moved in the presence of a fixed pole magnet (just like in a phono cartidge).

So, the feedback coil will create a small voltage used to check the actual response of the cutterhead given the drive voltage. At points of physical resonance, the drive voltage can be lowered to assure flat response. This involves a preamp circuit where the feedback voltage is negatively summed with the drive when necessary to avoid resonance.

[Babooino]

Without a second set of coils in your cutterhead you can't use feedback.

One set of coils are the drive coils. Any cutterhead including homebrews have one or two drive coils (mono or stereo head).

With an additional coil per channel you can then get feedback. Literally, the drive coil gets voltage and starts moving in the presence of the fixed pole magnet. The feedback coil is attached to the same moving part as the drive coil. The feedback coil has a very small voltage induced within it by being moved in the presence of a fixed pole magnet (just like in a phono cartidge).

So, the feedback coil will create a small voltage used to check the actual response of the cutterhead given the drive voltage. At points of physical resonance, the drive voltage can be lowered to assure flat response. This involves a preamp circuit where the feedback voltage is negatively summed with the drive when necessary to avoid resonance.

Great explanation!
You mean, the feedback signal is phase inverted and suumed to the original program, thus cancelling the resonant peaks? Is that it?

[opcode66]

Yes, and the amount is controlled by a potentiometer. Therefore you can tune in a certain amount of feedback via calibration. On a Neumann rack we do -9db of feedback.

[Babooino]

Yes, and the amount is controlled by a potentiometer. Therefore you can tune in a certain amount of feedback via calibration. On a Neumann rack we do -9db of feedback.

Could possibly a phono cartridge be used for that?

[opcode66]

No. You need realtime feedback. Using a phono cartridge for playback some distance form the cutterhead would not provide realtime feedback.

If you are asking if a cartridge could be disassembled and parts used for this purpose, also probably a no. All you need to do is add another coil to a coil based cutterhead and monitor the induced voltage from the secondary coil.

[tubefan]

yes, potentially a phono cartridge could be mounted so that its cantilever could be glued to some position along the torque tube of a cutterhead, and the displacement used to provide the feedback. or two mono heads could be used, or just the magnetic circuits and coils could be salvaged...

[opcode66]

Well, that's true. But, it would then be physically impede the free movement of the coilform.... I think that would alter the sound cut to groove.

I originally thought that they were asking if a playback tonearm playing the just cut groove could be used for feedback.

[Babooino]

Cool , has someone experimented something like that ?
I'm still not convinced why a piezo disc glued to the torque tube could not be used as feedback...
I use a piezo disc in my diy plate reverb with very good results.

[opcode66]

Yes it could. But, it would have to be small so as to add negligible weight/mass. The way the bobbins work with the torque tube, weight and mass are very critical. Especially when considering physical resonance. You wouldn't want to mess up an already working system by replacing a part that would electrically work but would physically throw things out of balance.

The other point that was made by EmAtChapterV is that the piezo element creates such a small voltage. If the copper wires attached to the piezo element were not electromagnetically shielded then the drive coils might induce a higher net voltage than what the piezo element itself was generating. In this case, the feedback signal would be less than useful.

[markrob]

Cool , has someone experimented something like that ?
I'm still not convinced why a piezo disc glued to the torque tube could not be used as feedback...
I use a piezo disc in my diy plate reverb with very good results.

I've done some experimentation using a ceramic stereo pickup coupled to my DIY head for feedback. It is indeed real-time and does work. But there are issues. At first glance, it would look to b e good idea. They have very high output voltage as compared with a magnetic pickup. But, the downside is that they are also very high impedance devices, so you have to take care to avoid pickup from the drive coils. If the cross talk from the head drive is greater than the feedback signal, you are out of luck. A coil based feedback approach also has to deal with this too. If you couple the piezo such that it defects as a cantiliever (a stereo pickup would work this way), it is a amplitude sensing device. If you mount piezo disks to the torque tube they would function as an accelerometer. Closing the loop as a position servo is a much easier than based on acceleration. But I'm sure it could be done with some thought. The effect of the element on the open loop response of the system would have to be taken into account (add moving mass and/or stiffness). Also, you would have to make sure that the sensing is really responding to the stylus motion at high frequencies (no decoupling). If you have some experience with feedback systems, you should be able to cobble something up an and make some open loop measurements before closing the loop.

Mark

[opcode66]

My comment about real-time need not reiterated. I thought he was suggesting to use a pickup. Like on a tonearm. You know to play the groove as it was being cut. He wasn't very specific at first and I got the wrong impression. And, that would certainly not be in real-time.

[Babooino]

Cool , has someone experimented something like that ?
I'm still not convinced why a piezo disc glued to the torque tube could not be used as feedback...
I use a piezo disc in my diy plate reverb with very good results.

I've done some experimentation using a ceramic stereo pickup coupled to my DIY head for feedback. It is indeed real-time and does work. But there are issues. At first glance, it would look to b e good idea. They have very high output voltage as compared with a magnetic pickup. But, the downside is that they are also very high impedance devices, so you have to take care to avoid pickup from the drive coils. If the cross talk from the head drive is greater than the feedback signal, you are out of luck. A coil based feedback approach also has to deal with this too. If you couple the piezo such that it defects as a cantiliever (a stereo pickup would work this way), it is a amplitude sensing device. If you mount piezo disks to the torque tube they would function as an accelerometer. Closing the loop as a position servo is a much easier than based on acceleration. But I'm sure it could be done with some thought. The effect of the element on the open loop response of the system would have to be taken into account (add moving mass and/or stiffness). Also, you would have to make sure that the sensing is really responding to the stylus motion at high frequencies (no decoupling). If you have some experience with feedback systems, you should be able to cobble something up an and make some open loop measurements before closing the loop.

Mark

no experience whatsoever...

But i'm willing to try to add some sort of feedback to a stereo head i plan to build.
I've been using piezo discs for a lot of things at things at my studio. I dont thing that thing would add any representative weight to a head as it is VERY light.

So, i if i:
1-get small piezo discs (have a stash of 20mm)
2-use some sort of electromagnet shielding on the leads (coaxial?)
3-couple it to the torque tube (or drive to styllus holder links?)
4-load it to a hi-z preamp
5-phase invert
6-sum it to to program (using a potentiometer) before the cutting amp

If i use a paper cone design, i should put a piezo in each one, right?

Would that improve anything in the sound? What am i missing?

Also, i have a friend who owns a metal industry. He is gonna machine it for me in his cnc lathe as long as i give him a 3d project (igss).
If there is someone interested in send me a 3d file with a good design, in exchange for a machined unit. Please pm me.

[EmAtChapterV]

It's important to note before you do this that if the signal is out of phase at a nominal test frequency (say 1 kHz), it's not necessarily going to stay that way once you approach and exceed the secondary resonance frequency. At some point the moving mass (coils, torque tube, et cetera) is going to stop acting like one cohesive unit and start acting elastic, with phase changes between what's being fed into the coils and how they're actually moving. It will be important to run frequency sweeps with an X/Y oscilloscope trace on the open-loop feedback coils versus drive coils, to find out at what point this starts to happen, and design a passive low-pass filter to put on it. Otherwise it's going to oscillate itself to destruction.