Irregular Groove Spacing on DIY Lathe – Need Help Diagnosing the Cause

[Jonaa]

Hi everyone,

I’ve built a DIY record lathe for embossing on PET discs and I’m running into an issue that I just can’t seem to solve. Generally, the lathe is working great and I’m getting some nice sounding records out of it, but there’s a consistent irregularity in the groove spacing that I haven’t been able to eliminate, despite several attempts.

The lathe consists of a linear CNC slide mounted across a turntable, a lead screw powered by a NEMA 17 stepper motor and an aluminum armature that holds the cutterhead (pictures below).

The issue is that the spacing between grooves is not uniform. Under magnification, it becomes clear that, in some places, every other groove is spaced slightly farther apart. This creates a subtle but regular "wobble" pattern in the groove alignment, almost as if the feed rate is fluctuating rhythmically (see picture below).

At first, I suspected mechanical flex in the cutterhead armature assembly. I had originally used some 3D-printed plastic parts, so I replaced all of them with custom aluminum parts to increase stiffness. The entire assembly is now quite rigid, and when I press on it with my hand, there’s only the tiniest amount of flex, and that’s much more force than what is being applied to the armature during cutting. Despite this improvement, the groove-spacing pattern remains completely unchanged.

I then turned to the electronics. The Raspberry Pi Pico I’m using as the controller outputs 3.3V logic, while the stepper driver (a Leadshine DM556) expects 5V–24V signals. I added a proper logic level shifter, powered via the Pico’s 5V VBUS pin, to raise the control signals to 5V. This didn’t make a difference either.

To rule out software timing inaccuracies, I rewrote the control C code to use a hardware timer for generating step pulses instead of relying on sleep_us() delays. The step pulses are now extremely consistent and precisely timed, but once again, the repeating variation in groove spacing is still there.

One thing I’ve been wondering about is whether the issue could be related to how the lathe is physically mounted. Currently, the entire lathe assembly is placed directly on top of the turntable. It’s not mounted to a separate frame with legs or otherwise fixed to the desk; it simply rests over the platter. My thinking has always been that this setup should actually be an advantage — if the turntable happens to move slightly during operation, the lathe would "move with it," reducing the risk of relative motion between the cutterhead and the record surface. If the lathe were mounted rigidly to the desk instead, any movement in the turntable could result in misalignment or uneven embossing. Still, I’m wondering if this lack of structural isolation could somehow be introducing vibrations or micro-movements that affect the groove spacing in a regular pattern. Any thoughts?

If anyone has seen something like this before, or has ideas for what might be causing a repeating fluctuation like this, I’d love to hear your thoughts. This problem has me stumped, and I’d really appreciate any input.

Thanks!

[farmersplow]

hard to say: here are a few things you may not have checked yet:
.) Axial slippage of the plate blank on the disc (because the centre hole is too large)
.) Drive rubber has too large an adhesive joint (a larger diameter)
.) Drive belt, rubber, tension too low
.) the "loop" in the programming of the stepper motor is not even. even if you work with an internal timer, queries in the loop could extend the time, causing some steps to be lost. I don't know at what intervals you send pulses to the stepper, but in my case I only have 3ms between two pulses (with large groove spacing). According to your picture, you have a groove spacing of about 150µm (~ 150 lpi). So you have about 5ms per step. In principle enough time for the processor, unless it has to do something else? I don't know your software.
.) Badly centred wheels of the transmission from stepper motor to drive spindle
.) or something else gg
Thomas

[drdub]

We had the exact same issue on our machines.
We believe it's caused by tolerances in the bearings of the turntable but not entirely sure.

Anyway - what helps is a magnet that pulls the metal of the precision plate to the left side.
That way the grooves are uniform.

See pic attached

[Jonaa]

Thank you a lot for your inputs! I will check all these things and let you know if I manage to find the source of the problem or not.

Have a good one!

[zdenek]

Tilt the whole machine with the turntable a few degrees to the right according to your photo [2-3 degrees of angle or more] this should eliminate the play in the turntable bearing. If this does not help, there is no noticeable improvement in the gap between the grooves, then check if the drive between the motor and the pulley is in the axis, but I doubt it is such a big problem. There may also be a problem with the turntable motor, it will not be noticeable when playing vinyl records, but when cutting all the inaccuracies of the motor will be visible. Although the belt drive is not very precise in terms of the precision of the turntable platter, it works well in recorders and lathes, the tensioned belt eliminates the play in the turntable bearing. The machine you are making looks massive, rigid, but the turntable as a drive reference is a very unreliable drive reference source in recorders. Finally, try my patent, tilt the device in its entirety to the right by a few degrees, I am curious about the effect myself, let me know.

[Jonaa]

Hello again!

It turned out that my turntable was the source of the problem.

I have been using a modern (and quite cheap) direct-drive turntable, with the assumption it would be ideal due to its robust construction and smooth platter. Today I decided to swap it out for an old belt-(I think)driven turntable from the 1960s—one that actually feels far less sturdy, with noticeable platter play when nudged by hand.
To my surprise, this older, less precise turntable produced beautifully even grooves, completely eliminating the waviness I was struggling with (see pictures below).

I tried tilting the direct-rive turntable and lathe setup slightly to the right, as per zdenek's instructions. Unfortunately, that didn’t solve the problem so I need to have another turntable for my setup. I don't want to use the old 60's one since I can't fine-tune the speed of it, which is necessary to compensate for the weight from the cutterhead against the disc surface. Any suggestions on what kind of turntable to use? I’ve always been under the impression that direct-drive turntables were the way to go due to their power and torque, but based on my results, I’m not so sure anymore. I can’t afford a high-end Technics DJ deck like many here seem to use, but I’d really appreciate recommendations for something reasonably stable, quiet, and—most importantly—won’t introduce inconsistent groove spacing again!

To the left: silent groove with the old turntable from the 60's, to the right: silent groove with the modern direct-drive turntable. The last picture shows the set up with the old turntable.

[Jonaa]

Anyway - what helps is a magnet that pulls the metal of the precision plate to the left side.
That way the grooves are uniform.

I wanted to try this as well but apparently, the platter on my direct-drive turntable isn't magnetic. Thank you anyway for the tip!

[jjwharris]

I've had the exact same problem - haven't quite figured it out. I have 4 lathes of the same design - 3 are fine, 1 is not...I'm also using a nema17 for the carriage motor.

I've checked everything mechanically, swapped out steppers and drivers, reuploaded code from the working machines....My next step was checking whether it was a power supply issue.

Seeing this, I might take a closer look at the turntable - I have another lathe of another design with the same issue - both belt driven, one is driven with a servo and the other a nema 17.

[themagician]

I had this issue prior to upgrading to a better lead screw, but I can see you already have a decent one in place..?

Mechanically, (ideally) the record lathe carriage should move laterally consistently, however the groove pattern suggests there is some (small sine wave) oscillation in that lateral movement, either adding (and subtracting) to the lateral movement, or by means of a vertical oscillation - or both! (elliptical / circular 2D oscillation).

How about taking a time lapse video of the carriage running from min to max, as these oscillation are sometimes easier to see at an alternative frame rate?
This would at least confirm/deny this theory.

[Jonaa]

I had this issue prior to upgrading to a better lead screw, but I can see you already have a decent one in place..?

Mechanically, (ideally) the record lathe carriage should move laterally consistently, however the groove pattern suggests there is some (small sine wave) oscillation in that lateral movement, either adding (and subtracting) to the lateral movement, or by means of a vertical oscillation - or both! (elliptical / circular 2D oscillation).

How about taking a time lapse video of the carriage running from min to max, as these oscillation are sometimes easier to see at an alternative frame rate?
This would at least confirm/deny this theory.

Thanks for your input, themagician!
I’m fairly confident at this point that the issue wasn’t related to the lead screw. I actually tested with a different one (with a tighter pitch) earlier, and it didn’t change the irregular groove spacing. What really seemed to make the difference was switching to another turntable — with that one the problem disappeared. That makes me think the uneven speed of the turntables I was using before (or possibly their platter bearings) was the real cause.
I’m planning to upgrade to a Technics SL-1200/1210 MK2 soon, and once I’ve tested it I’ll update this thread with the results. Hopefully that deck will have the stability I need.

[Jonaa]

Hi folks!

I just wanted to give you all an update after my recent experiment.

This past week I was able to borrow a Technics SL-1200 MK2 from a friend, hoping it could finally be the endgame solution to the groove spacing problem. To my disappointment, it turned out to have the same problem as the other direct-drive turntables I’ve tried (see picture 2 below). So, unfortunately, I won’t be buying an SL-1200 anytime soon.
Instead, I’m moving forward with Plan B, which was suggested by drdub earlier in this thread: placing a magnet next to the precision plate to reduce the play in the platter bearing. The issue is that I don’t own a precision plate, and the platter on the cheaper turntable I wanted to use for the test is made of aluminum, so it’s not magnetic. I don't want to spend like 300 USD on a "real" precision plate just to test my theory, so my plan is to make one myself out of a steel disc and see how that works. I haven’t built it yet, but once I do, I’ll run some tests and report back here with the results.
If that works, I might reconsider the SL-1200 in the future. But if not, I guess I have to build my own deck with a heavy and rigid platter, driven by belt from a stepper motor or something similar. That will be a bigger project, but I’m ready to go there if necessary.

Here's a picture of the current setup with the SL-1200 MK2.

tempImagejTJ9xX.gif

Picture 1

And here is a picture of some low volume grooves embossed into a PETG blank with the SL-1200 MK2. You can clearly see the long-short-long-short pattern we're all familiar with.

S20250920_002.jpg

Picture 2

And just for comparison, here's a picture of some louder grooves made with the old Garrard deck I talked about before. If ignoring the overcut to the right, I think that's quite a good representation of how it should look if everything works correctly.

S20250913_003.jpg

Picture 3

[PLD]

It seems to me that wear (or perhaps manufacturing tolerances) in the bearing/spindle somehow causes the platter runout to change on every other rotation, it's very obvious when using a dial indicator, and I can only assume it's something to do with the magnets and coils of the direct drive motor units. One of my SP10s exhibits the problem. For what it's worth, I recently examined a disc made using a fairly new 1200GR turntable as the platter, and the groove was 100% accurately spaced, although I can't comment on speed stability.

[Thelatheofus]

I'm having a similar issue, but only on one side of the record. I don't have a vacuum platter, so I usually get records that are far from flat, and it seems to me like those issues only appear on the "bent" side of the record.
I'm using a stanton st-150 turntable, my theory is that it has a bearing issue aswell, and that when the needle goes on the higher part of the "bent" record it applies more wiehgt to the turntable platter, and thus makes the bearing problem appear.
I never noticed this issue before changing my pulley for a u-shaped one and an o-ring as a belt, but the way those irregular spacing are placed on the record make me think that the belt and pulley are not responsible for this in my case. I changed the turntable at the same time as the belt and pulley so I'll go back to my Numark and see if it fixes the issue.

[farmersplow]

I'm having a similar issue, but only on one side of the record. I don't have a vacuum platter, so I usually get records that are far from flat, and it seems to me like those issues only appear on the "bent" side of the record.
I'm using a stanton st-150 turntable, my theory is that it has a bearing issue aswell, and that when the needle goes on the higher part of the "bent" record it applies more wiehgt to the turntable platter, and thus makes the bearing problem appear.
I never noticed this issue before changing my pulley for a u-shaped one and an o-ring as a belt, but the way those irregular spacing are placed on the record make me think that the belt and pulley are not responsible for this in my case. I changed the turntable at the same time as the belt and pulley so I'll go back to my Numark and see if it fixes the issue.

Since the problem has not yet been resolved, I took another look at the design. From a technical point of view, the overall design is unstable. I cannot see exactly how stable the overhead construction is mounted on the turntable (this could be a possible cause if it is not fixed). What looks unstable to me, however, is that the turntable stands on its (wobbly) feet. In addition, the drive motor for the spindle is located next to the device. When the turntable rotates, it causes small vibrations. If the turntable vibrates slightly towards the floor (table), the tension on the belt drive between the stepper motor and the spindle changes. Theoretically, this can lead to irregular movements of the spindle and thus to a change in the groove spacing. This occurs at the same frequency as the turntable speed. We are talking about deviations of around 30µm.
I would try the following. Is there a difference between 33 and 45 rpm? Does it change when the turntable is fixed and the attachment is fixed?
I would rule out the platter bearing as the cause of the error, as the distances are too regular.

[Jonaa]

For what it's worth, I recently examined a disc made using a fairly new 1200GR turntable as the platter, and the groove was 100% accurately spaced, although I can't comment on speed stability.

That's interesting. The SL-1200 MK2 I tested was about 25-30 years old and had been used a lot in clubs and stuff, so it wouldn't be strange if the bearing on that deck was a bit too worn. Maybe I should just try a newer deck like the 1200 GR or an MK2 with a recently replaced platter bearing someday.

I'm having a similar issue, but only on one side of the record. I don't have a vacuum platter, so I usually get records that are far from flat, and it seems to me like those issues only appear on the "bent" side of the record.
I'm using a stanton st-150 turntable, my theory is that it has a bearing issue aswell, and that when the needle goes on the higher part of the "bent" record it applies more wiehgt to the turntable platter, and thus makes the bearing problem appear.
I never noticed this issue before changing my pulley for a u-shaped one and an o-ring as a belt, but the way those irregular spacing are placed on the record make me think that the belt and pulley are not responsible for this in my case. I changed the turntable at the same time as the belt and pulley so I'll go back to my Numark and see if it fixes the issue.

Yes, the irregular spacing does (most of the time I think) appear on one side of the record for me as well. What you're saying about the bent discs making the problem in the bearing appear is what I think too. Please let us know if changing back to the Numark solved the problem for you.

From a technical point of view, the overall design is unstable. I cannot see exactly how stable the overhead construction is mounted on the turntable (this could be a possible cause if it is not fixed).

Thanks for your input! The feet of the overhead are not fixed to the turntable with screws or bolts. I use a similar solution to that of the MC-02 lathe, with slidable feet with notches at the bottom, which enables the overhead construction to be clamped very tightly to the turntable. However, although I think the whole system feels very stiff and robust, I of course won't rule out the possibility of the problem being linked to the overhead construction and how it's mounted to the turntable. It might very well be the case that my design isn't stable enough. What speaks against that suggestion though is the fact that the system produces nearly perfectly spaced grooves when it's being used with a non-direct-drive turntable. I will therefore try my idea with the precision plate and magnets first to completely rule out the bearing being the cause of the problem before considering changing the overhead design.

What looks unstable to me, however, is that the turntable stands on its (wobbly) feet. In addition, the drive motor for the spindle is located next to the device. When the turntable rotates, it causes small vibrations. If the turntable vibrates slightly towards the floor (table), the tension on the belt drive between the stepper motor and the spindle changes. Theoretically, this can lead to irregular movements of the spindle and thus to a change in the groove spacing. This occurs at the same frequency as the turntable speed. We are talking about deviations of around 30µm.

That's a good point. However, I had the groove spacing problem even when the stepper motor was attached directly to the lead screw and linear rail, without any pulleys or belts in-between.

Thank you all for your valuable inputs, I appreciate it a lot! I will continue experimenting with this and keep you posted.

[tragwag]

it may work under the more slight cutting force (15-20g), but for embossing I've found I need something much beefier than technics 1200 or even stanton.
it's gotta be the Numark TTX, or Technics SP10 + SP15
it's always worth the extra cost in my opinion