DIY shellac cutting lathe

[LinusN]

Hiya

Joined this forum mainly for this project. I have already introduced myself here, so let's get straight to the topic.

I have an old hand cranker which I really love. Unfortunately, getting hold of good old 78RPM records is not too easy and I also came to think; "wouldn't it be very nice to be able to record whatever I want?".

latetrolls004.jpg

I have googled a lot, and from what I can see no one really seems to do this. That means, making mono shellac records. No micro-grooves or anything else fancy. First I thought I should be able to make my own blanks, but I soon realized this might not be as easy as I initially thought. I found another method which I think will work well; grinding down old records. Finding old, worn out records is not hard at all, and you can buy maybe 2 for a dollar in any thrift shop/flea market.

What about the cutter head then? After some thinking I came up with this:

latetrolls007.jpg

latetrolls008.jpg

The stylus holder is made from aluminum, and weighs no more than about 3 grams. The bearing is the pick-up bearing from an old hard drive. These are extremely high quality, and I like to use them. I actually used one of these as bearing for the vertical tone arm movement in my turntable.
The linkage between the speaker and the cutter head consists of a carbon fiber tube and the tab will be made of thin stainless steel. Since the movement will be very small, I think there is not need for the joint between the tab and the cutter head to be flexible, e.g. I can tighten the screw to get rid of any play between it and the tab.

The speaker is something I bought from Ebay. An Altec Lansing something-something. I think it will do OK.

Yesterday, I started making the stylus holder. Luckily I have access to some handy machinery where I work. I used to operate these before I moved on to the engineering department. Otherwise, I guess I would have no idea how they worked.
The Wire EDM machine is not really as well known as other CNC machines, but it is really quite impressive! It cuts any metal with very good precision.

latetrolls005.jpg

[LinusN]

latetrolls002.jpg

latetrolls006.jpg

This is the result:

DSCF4216.JPG

For moving the cutter head I'll use a small linear drive unit with ball screw and a DC servo that I had laying around.

Since the records are very hard I need to heat them up when cutting. I tried with my hot air rework station and it seems like they soften nicely once they get hot enough. Since I want to get the cutting process quite stable and repeatable, I think I'll use some kind of regulator to keep the temperature right.
I bought one of these from Ebay, and will use a microcontroller to read the temperature from it and control the element of a hot air gun. I think it will work...

By the way; I bought the cutting needle from a guy on Ebay. I think he might be here at the forum to. Later on, maybe I'll try to make something out of tungsten carbide (in the Wire EDM machine), but that will be a much later project. I also really like these kind of projects

[tragwag]

I've had limited success "embossing" and playing back with a bamboo stylus on my victrola.
The hardest part will be creating a playable standard width groove that does not degrade.
As I'm sure you know, shellac records were created from a literal beeswax master.
keep us posted on the progress!!
If it works out, I'll surely buy some shellac 78s from you

[LinusN]

Well, I believe even the "factory made" shellac records did not last very long. I don't really have any experience my own on this, but did not these records wear out much quicker than modern vinyl records?

Maybe I should add that HiFi is not really main focus here. Of course, it's good if it sounds OK but I'm happy as long as I manage to make any kind of record that can be played several times and sounds OK

[LinusN]

Did some more CAD:ing today. I may be able to cut out most of the parts next weekend!

Assembly1.png

Do you think there is a need for angle adjustment of the cutting stylus?
I am also not very sure about what amplitude of the stylus would be suitable.

[markrob]

Did some more CAD:ing today. I may be able to cut out most of the parts next weekend!

Assembly1.png

Do you think there is a need for angle adjustment of the cutting stylus?
I am also not very sure about what amplitude of the stylus would be suitable.

You shouldn't have to worry about the angle too much. You cut will still be lateral. The use of a conical rather than knife edge stylus will have some issues with self erasure of higher frquenies, but that's not a big deal here. You do want to get the 90 degree groove geometry in the ballpark or you might have some issues with tracking on playback.

I would think +/-.003-.005" would be more than enough. 3 grams might be a bit much in terms of mass (you also have to add the driver moving mass too). It may be ok since you are not planning to cut any higher than about 5Khz I assume.

It looks like you are not planning to cut given the conical shape of the stylus. The plan is to heat the platter to soften and emboss the groove. Is this correct?

You may need to add some stiffness to your system. Right now, the main resonance its going to be pretty low since the restoring force is provided by the driver surround. The system is constant amplitude below (compliance controlled) the main resonance and transitions to falling off at -12db/oct (mass controlled) above this point. You may also need some damping to control the resonance.

IIRC, acoustical era recordings have no EQ added and are basically constant velocity within the recorded bandwidth. So you will have to do some EQ to your head to get the right result.

Mark

[LinusN]

I guess frequencies above 5kHz will not be interesting. I will also have the option to slow everything down. I could go half speed or even slower if that would be a benefit.
The actual arm has a mass of 3 grams, but since it pivots around a point the effective mass will be lower than that. I don't think the mass will be an issue in this application, but I can't say for sure

You're right. Embossing it is, and I plan to heat everything up. Since I want to keep the temperature stable I plan to use an IR thermometer so control the heating automatically.

EQ will probably be a must. No problem! I could even record a few sweeps without EQ, play it back and adjust. And I guess I could fiddle around some with both stiffness and damping.

Regarding stylus; I have not made any decisions about what stylus to use. I have two different types, and will probably in the future make more models.
I do believe though, that cutting would be a bit tricky in such hard material. Maybe that is also possible if applying some heat, as long as the material does not get to elastic

Thanks for your input!

[markrob]

I guess frequencies above 5kHz will not be interesting. I will also have the option to slow everything down. I could go half speed or even slower if that would be a benefit.
The actual arm has a mass of 3 grams, but since it pivots around a point the effective mass will be lower than that. I don't think the mass will be an issue in this application, but I can't say for sure

You're right. Embossing it is, and I plan to heat everything up. Since I want to keep the temperature stable I plan to use an IR thermometer so control the heating automatically.

EQ will probably be a must. No problem! I could even record a few sweeps without EQ, play it back and adjust. And I guess I could fiddle around some with both stiffness and damping.

Regarding stylus; I have not made any decisions about what stylus to use. I have two different types, and will probably in the future make more models.
I do believe though, that cutting would be a bit tricky in such hard material. Maybe that is also possible if applying some heat, as long as the material does not get to elastic

Thanks for your input!

You have the equivalent of a impedance transformer given the fact that large speaker travel results in smaller stylus movement. So the speaker will less mechanical load than the stylus point. That might be a good thing.

I'm sure you'll have no problem playing with all of the mechanical and electrical parameters to get this working well. You are not shooting for Hi-Fi performance in any event. I'll be interested to see how this all works out.

Mark

[THEVICTROLAGUY]

it should be noted that original 78 discs have pumice added to the mixture to wear the stylus as you play the disc, this prevents the stylus from wearing the groove. one needle, one play. the playback stylus must be changed every time you play another disc, or the same disc for that matter. this is going to create a great deal of noise in your recording that is strictly mechanical in nature, a literal abrasive that is added to the disc itself.

when doing recordings on cylinders i encounter some really noisy surfaces from mold deposits in the wax. no matter how smoothly i shave a cylinder, if there is mold or any contamination to the surface of the cylinder you get a noisy recording. i don't know how you will get a mirror smooth finish on an original disc, and even if you manage to "shave" a disc you are going to have a tremendous amount of surface noise inherent in the disc itself, this cannot be removed, you will simply have to record on top of it. in many instances the surface noise is as loud as the recorded information, sometimes louder. just by " shaving " the disc your are going to induce a great deal of additional surface noise. these discs are as brittle as glass, and cannot be flexed at all. i have snapped original discs wiping them with a rag while holding them by the edge.

in order to make a high quality recording for playback on an acoustic phonograph you will have to start with a noiseless blank disc that is soft enough to record on, yet hard enough to washstand the weight of the acoustic reproducer as well as being able to stand up physically to a needle.

it will be difficult to get any original victor or columbia disc quiet enough to make a quality recording and the abrasive in the disc itself is designed to quickly wear steel needles, your cutting stylus is going to have to be something other than steel.

[LinusN]

You have the equivalent of a impedance transformer given the fact that large speaker travel results in smaller stylus movement. So the speaker will less mechanical load than the stylus point. That might be a good thing.

I'm sure you'll have no problem playing with all of the mechanical and electrical parameters to get this working well. You are not shooting for Hi-Fi performance in any event. I'll be interested to see how this all works out.
Mark

Yep, that was kind of the idea. To be able to use more power for the lateral needle movement, keeping a small amplitude. I have not made any calculation on it, so it could either work or not work at all. Just a test. I suspected that I might need a bit more stability than when embossing vinyl so that's how the idea came up.

[LinusN]

it should be noted that original 78 discs have pumice added to the mixture to wear the stylus as you play the disc, this prevents the stylus from wearing the groove. one needle, one play. the playback stylus must be changed every time you play another disc, or the same disc for that matter. this is going to create a great deal of noise in your recording that is strictly mechanical in nature, a literal abrasive that is added to the disc itself.

when doing recordings on cylinders i encounter some really noisy surfaces from mold deposits in the wax. no matter how smoothly i shave a cylinder, if there is mold or any contamination to the surface of the cylinder you get a noisy recording. i don't know how you will get a mirror smooth finish on an original disc, and even if you manage to "shave" a disc you are going to have a tremendous amount of surface noise inherent in the disc itself, this cannot be removed, you will simply have to record on top of it. in many instances the surface noise is as loud as the recorded information, sometimes louder. just by " shaving " the disc your are going to induce a great deal of additional surface noise. these discs are as brittle as glass, and cannot be flexed at all. i have snapped original discs wiping them with a rag while holding them by the edge.

in order to make a high quality recording for playback on an acoustic phonograph you will have to start with a noiseless blank disc that is soft enough to record on, yet hard enough to washstand the weight of the acoustic reproducer as well as being able to stand up physically to a needle.

it will be difficult to get any original victor or columbia disc quiet enough to make a quality recording and the abrasive in the disc itself is designed to quickly wear steel needles, your cutting stylus is going to have to be something other than steel.

Grinding old discs down was no problem at all, but I do not think I can get even close to the original surface finish. I may have missed something, but does surface roughness really have such a bit impact on the quality of the recording? I mean, the information is not recorded on the surface. I am not questioning you. I'm the guy with no experience at all here, so I have made a lot of assumptions.
Have to add that I should have always used the word "embossing" instead of "cutting", since embossing is what I plan to do.

Regarding the stylus; The plan was to make a stylus out of tungsten carbide later on.

[markrob]

Hi,

If you were cutting, the surface finish might be an issue as it would affect the cut depth. Are you planning to control embossed depth using an advance ball or via a floating damped suspension?

Not sure about how your approach of heating the surface until you are near the glass transition temperature will work as compared with cutting. I suspect the stylus will move the grit out of the way as it plows through the surface. If you were trying to cut, it would be like hitting a rock (not good) It will interesting to see if the recorded surface noise using your proposed method is worse than a standard pressed 78 shellac disk.

How are you going to grind the disc? Using a surface grinder?

Mark

[LinusN]

My first plan was to use fixed depth but then I realized that many people got good results with floating suspension. I believe the surface roughness would have a bigger impact on sound quality if using floating suspension, but I will give it a try.

Regarding heating; I just made some quick tests, but to me it seemed like the material softened really nice at a specific temperature. If I can find an optimal temperature and keep it stable during the entire process (not just a heat lamp, but an IR thermometer and a regulator) maybe it will work. Cutting seems like out of the question to me, both because of the hardness but also the abrasive properties of the material.

I think a stable temperature would be a great benefit to all you guys out there cutting in different kinds of plastic!? I thought, maybe I could come up with something that is useful for more than me. But maybe there is already a good economical solution to this, or maybe the temperature is not very critical!?

I used a surface grinder. I think the hardness of the material is actually a benefit when it comes to grinding.

[LinusN]

Hiya

Did some measurements on the surface today.
The surface roughness of the blank area of an original disc was 0,05 Ra (very low indeed!), 0,89 Rz.

I used a surface grinder to remove the old tracks, and finished with a 500 grain paper. Measured the surface roughness and got 0,34 Ra, 3,4 Rz.

hardness.jpg

Please note that I did not do any polishing. I think the surface quality can be increase very much with little effort if one uses the right tools.

Yesterday I used my temperature controlled hot-air station to get an approximate figure of what temperature would be suitable for the embossing. Around 100 deg Celsius seemed ok. I took one of Jerry Tate's needles, held it at a 20 deg angle and scratched the surface a bit with it. The only thing I can say is that it does at least seem possible to be able do do what I intend to

A factory made record:

3.jpg

A ground down record with some test "scratches". The track in the middle looks quite OK.

1.jpg

[LinusN]

After working with some other projects (incl this) I have gained some more time for this.
I have assembled a simple heating unit consisting of a modified heat gun (brought out separate leads for the fan to be able to control the heating elements separately), a solid state relay, an IR temperature sensor and an Arduino running a PID loop keeping the temperature of the disc stable.

One day I happened to pass the old CNC mill at work and got a great idea; i could use that one instead of building a new spindle and linear transport for the cutting head. At least for the initial tests. I guess later on I'll come up with a neater solution. Anyway, I put everything up and did some tests.
It turned out that about 105 deg C seems to be a good temperature for embossing.
This is about it at the moment.

To make more tests, I have to prepare more records, and complete the suspension for the cutting head.

[tragwag]

this is next level stuff. congrats!!

[Angus McCarthy]

A lot of people said this essentially couldn't be done. Bravo.

[LinusN]

It's not done yet, and may end up a complete fiasco, but as I say; "failure is always an option". You always gain something from it

[megrimm]

Hello,

Cool thread! I was wondering what happened with all these experiments? Status update?

I just picked up about 1000 shellac records for $5 (the place just wanted them gone!) and I took some of the broken ones, crushed them up in a drywall bucket with a sledgehammer, and tried melting them in a crock pot (on high). Maybe a crock pot does not get hot enough but the material did become elastic/gooey but not really liquid enough to be moldable. As a side note, it also seeped some awful noxious fumes in the form of black smoke when removing the lid, something 1930's era toxic I'm assuming.

Questions:

Can I get some kind of biscuit that has material constancy and can be molded/pressed using reclaimed/recycled shellac 78's?
What heating method might be best for semi-liquefaction?
How were shellac records originally heated/pressed? I have only found useful info for vinyl but not the material formula, heating temp/method, etc for shellac.

There must be some reference material out there on the original process but I have yet to find (or even know where to look) something useful to move experimentation further. Maybe someone knows?

Thanks!
mark

[LinusN]

Hi megrimm

I have so many projects going on in parallel, so this particular one is stalled at the moment.
You can see in my earlier posts that I tried to do some embossing. It kind of worked, but I have to make a new platter. The platter I used was something I found in the scrap bin. It is not solid, but has spokes just like a wheel. When the record reached the temperature which made it really soft it sunk in between the spokes and got crooked.
I need to make a solid platter and make some new tries but at the moment, this project is higher in priority: http://svarthaletracing.se

I have pretty limited knowledge in this area. The knowledge I have is gathered during the last year when dealing with this project. I guess there are other people that can give you better answers for your questions, but I can try sharing the little I know.

Every manufacturer had its own "secret" recipe for making the record. They are made of shellac, but with other material mixed into it. There is also an abrasive mixed into the material. This makes the needle/stylus wear out faster than the record, i.e. the record lasts longer.