Poor Man's Feedback Cutter Head

[mischmerz]

I was thinking: What if a cutter head mounted microphone would provide a signal that modulates a LED which illuminates (or not) a photo resistor in the signal path to the amplifier? Would that count as a poor man's dynamic feedback cutter head? Or am I on the wrong path?

Michaela

dynamicled.png

[opcode66]

First issue is that a microphone wont give accurate information about how the inidivdual pistons are moving in a stereo design. It would pick up the movement of both transducer.

Second issue is that a transducer is not a speaker. It has no diaphragm. So, you have to be in contact with the transducer in order to perceive its full range. You dont here the low end of the audio that you play through a cutterhead. That is why all my bench test videos sound the way they do. The head sounds like it is very high end accentric. You dont hear the bass that it is representing unless you touch the torque tube to your ear physically. If you physically connected the microphone element to the transducer you would seriosuly compromise the performance of the transducer. Thereby making feedback irrelevant since it would not work effeiciently and the system would likely sound better without a mic elent attached... If not attached, it won't pick up a huge frequency range. And again, you may as well not do it at all.

The led/photosensor you suggest is what is referred to as a Vactrol. They are notoriously inaccurate. They do not have a smooth curve to them and would therefore be a poor choice for a negative feedback scenario. It is actually maybe 8 or 9 components in total to build a negative feedback system for one channel (using solid state opamps). Not hard and very accurate in comparison to Vactrol attenuation.

FYI, there is no such thing as a dynamic feedback cutterhead. It is one or the other.

Finally, unless you are detailing your work as an inventor or innovator, you should really be posting your cutterhead questions in a different section of the forum. Glad to answer them and happy you are asking. Just pointing out the etiquette.

[mischmerz]

First issue is that a microphone wont give accurate information about how the inidivdual pistons are moving in a stereo design. It would pick up the movement of both transducer.

Well .. it would be for a mono head. Thanks for your answer and your explanations. My underlying thought was the resonance frequency of .. say a Grampian .. that would most likely trigger the microphone / LED combination which in turn would throttle the signal. As to the selected forum: It say's "Experimenters' & Innovators' forum" - I most certainly don't belong into the upper echelon of innovators here, I am mostly focusing on the LoFi end of things and I may even use the wrong terms now and then .. but why should a thought like this not be posted in this forum?

Michaela

[opcode66]

Please don't take my comment the wrong way. The section "Secrets of the Lathe Trolls" is likely best for general questions about gear. Unless you are building this out, this isn't the right section.

Ok, so mono, great. You still won't get proper response without coming in contact with the moving part of the cutterhead that holds the cutting stylus. Feedback only works correctly if you are getting a real idea of the physical movement. Otherwise, feedback can actually do the exact opposite of what you want and will distort signals. If feedback is done incorrectly, you can get self oscillation and blow your drive coils. So, it is either correct or the exact opposite of what you want, or a nightmare. There isn't an in-between.

A contact mic would likely be the best scenario. But, even then... I think that a lot of folks are having a misunderstanding about exactly what Feedback can and can not do. It is not a cure-all. It take s a lot of factors to be correct for feedback to work.

There was a cutterhead made that utilized an RF emitter and receiver. The was a reflector plate on the moving stylus holder. That system was like a localized radar within the cutterhead. It worked great. The alternative to that has been a coil and magnet affair like what is in a phono cart. These are the things that have worked classically. If there was a lofi means to do this, it would have been done and repeated infinitely by now. There just isn't. Even the illustrious Flo (as upper echelon as you can go) has stated that Feedback is not easy to achieve.

You are def not getting there with a Vactrol. Speaking of which, I have messed around with creating extremely accurate Vactrols. But, it is much bigger than a standard vactrol, it doesn't use a typical LED, so it requires a circuit, and the casing is a tube. But, it is waaaaay more accurate than a standard off the shelf component.

[mischmerz]

Please don't take my comment the wrong way. The section "Secrets of the Lathe Trolls" is likely best for general questions about gear. Unless you are building this out, this isn't the right section.

Actually - I am planning to build it. It may not be accurate but I believe it might help me to flatten the curve of the head. Why would it burn the coil out? The signal is set to its value, the resistor will only reduce it and if all else fails, the fuse blows. If it works as intended - even if its horribly inaccurate - it tells me that the principle idea might be correct. I will then replace the photo resistor with some transistorized circuit and try again. Don't get me wrong - I fully understand that this is indeed crude and may not even work - but folks with old(er) mono cutting heads don't have a lot of options. So we explore, we try and well we might fail. After all - if we want to have perfect audio, we would just use DSD streams. But how would that make sense as I am almost only work with vintage music. I am taking great pride in the fact that I am not using anything digital - that I am trying to learn to deal with audio the old fashioned way. What takes me 5 minutes on the computer takes me half a day on the reel-to-reel, equalizer, compressor, de-esser and all the other tools. And after all the work I am ending up with a great MONO record that doesn't have to hide itself behind any other of the great recordings made back in the days. So - I want to squeeze even a bit more out of what I have. And in order to do that, I need to improvise

So - if anyone has additional ideas about how add some form of feedback to a mono cutter head (without tearing it apart), feel free to jump in.

Michaela

[opcode66]

It is all about the frequency of the audio and the latency of your transfer. Those factors can cause phase shift. Which will cause amplification in a negative feedback system. Which, with high frequencies, can damage your coils or blow. That's why.

Now you know what if feels like to be me. You know how big the target is painted on my back....

Good luck. And I mean that sincerely.

[mischmerz]

Now you know what if feels like to be me. You know how big the target is painted on my back....

Hahaha .. well .. thanks. I will buffer the led with a small capacitor and a bleed resistor so that it won't actually pulse a frequency. Only some average volume. I could even include some form of a simple low pass filter to get rid of anything way above the resonance frequency. This should take care of any negative feedback.

Michaela

[opcode66]

[sunkingrecords]

I don't think the LED, and even less the LDR, are resposive at audio frequencies. You will most likely get an envelope of the audio intensity, but not the audio itself.

Kind regards

Miguel

[opcode66]

[mischmerz]

I don't think the LED, and even less the LDR, are resposive at audio frequencies. You will most likely get an envelope of the audio intensity, but not the audio itself.
l

LEDs are sure enough responsive to audio frequencies. There are quite a few experimental setups transferring audio via LED. The LDR might indeed be to slow. But we'll see.

mm.

[opcode66]

There is a nice little bit about it here.

https://books.google.com/books?id=-5UPyE6dcWgC&pg=PA76&lpg=PA76&dq=audio+via+led+ldr&source=bl&ots=wwGlsrDBkU&sig=5DScoVdEiOvd9-vxcaI5zRTR5bA&hl=en&sa=X&ved=0CDsQ6AEwB2oVChMIquDem_GdyAIVVC6ICh1mNAtS#v=onepage&q=audio%20via%20led%20ldr&f=false

High frequencies would be an issue. Inaccuracy of the components is an issue. Temperature fluctuations, yes again, an issue.

I want to see a response curve of the LED/LDR combo before saying it is viable.

That being siad, I've seen these as well http://diyaudioprojects.com/Solid/DIY-Lightspeed-Passive-Attenuator/

You'll notice there is no actual data given about the performance. Just personal opinion. No response plots. I wonder why?

[mischmerz]

You'll notice there is no actual data given about the performance. Just personal opinion. No response plots. I wonder why?

There's no doubt in my mind that the proposed solution is rudimentary at best. The only question to be answered is: Is it better than no feedback? And this I will try to find out. So - maybe somebody has an idea on how to improve the circuit?

Michaela

[markrob]

Hi,

It will be a great learning experience for you. As you have it defined, it will not work. But, you need to find this out on your own. You'll need to gain a better understanding of feedback systems in order to be successful. What improvements are you looking to achieve by using feedback?

Mark

[mischmerz]

As you have it defined, it will not work. [...] What improvements are you looking to achieve by using feedback?

Why wouldn't it work? The idea is to be able to pickup resonance from the head that will (what I think) be distinguishable by increased volume picked up by the microphone and to use the signal to reduce the incoming volume accordingly. I am currently not concerned with quality issues as those can be addressed with filters and fast circuits in the signal path, I am currently only trying to get some form of rudimentary volume control based on the audio signal as received by the microphone. If fed into and analyzed by a computer the microphone signal might even be used to detect distortions caused by too much volume, bad styli or other problems.

Michaela

[markrob]

Hi,

Just cut some pink or white noise (depending you your choice of response analyzer) cut this, playback, and characterize the open loop response of your head. Develop an EQ to flatten the response as needed. Season to taste.

I guess what you are trying to describe is a slow acting compressor/limiter. I think that would be a very iffy way to approach the problem. Give it a go and let's see how it works out.

One obvious problem is that the acoustical output you hear while cutting is not that well related to what is being cut (as Todd mentioned). Are you planning to use this with a standard moving iron head? If so, the system resonance is already pretty well damped out by design. If you are looking to use this with a DIY Mono moving coil design, then you will have to deal with a primary system resonance, but also the fact that the open loop response is not flat vs. stylus velocity. So even if your comp/limiter detects and knocks down the primary resonance, it still will not address the overall open loop response. In addition, the slow response of your level detector will allow initial transients to happen before the gain can be reduced. You will also have to somehow differentiate between peaks due to the resonance and peaks due to program material. I guess you could add in a bandpass filter to listen only in the resonance range, but at that point, you might as well EQ it out up front.

Mark

[mischmerz]

I guess what you are trying to describe is a slow acting compressor/limiter. I think that would be a very iffy way to approach the problem. Give it a go and let's see how it works out.

Thanks Mark. As the microphone is attached to the cutter head it will listen to everything happening in there. With some smart analysis software one would be able to recognize not only resonance but all sorts of possible issues. As the setup is described in my first post, it would indeed act as a limiter - with the idea (and careful adjusting) that resonance would add a distinguishable spike that would trigger a limiter response. I understand and agree with your observations - especially the timing problem if the resonance would just be a short blib. However - if the the signal induces a longer resonance response from the head, the limiter would be able to catch and limit it. I ordered a few SL-32SR2 optocouplers and will soon be able to see the results.

Michaela