The farmer and his plow - When I set out to build a record cutter

[farmersplow]

Measuring the cut grooves of the Ortofon test disc:

I use the Ortofon test disc for a few measurements.

First a picture with the microscope (silent cut):
.
.

99_07 ortofon microscope.JPG

.
.
Then the observation and measurement on the PC. The measurement result of the cut width is 95.2µm. This results in a cutting depth of 47.6µm:
Unfortunately, I have chosen a very small font size for the dimensions in this picture. Therefore it is difficult to see.
.
.

99_08 ortofon mess.png

.
.
Here is an image with the microscope (Square Wave duty Cycle 3:7):
.
.

99_09 ortofon microscope.JPG

.
.
Then the observation and measurement on the PC. The measurement result of the cutting width is 93µm. This results in a cutting depth of 46.5µm. The groove spacing is 200µm (127 lpi):
.
.

99_10 ortofon mess.png

.
.
And another picture of the Ortofon disc with the microscope:
.
.

99_11 ortofon microscope.JPG

.
.
Then the observation and measurement on the PC. The measurement result of the cutting width is 87.6µm. This results in a cutting depth of 43.8µm. The groove spacing is 200µm (127 lpi):
.
.

99_12 ortofon mess.png

.
.
To be continued:

Greetings from Austria
Thomas

[zdenek]

In fact, it may be. It seems that you can measure the width without any problem.

[farmersplow]

I think it's high time to report again:

A few months have passed since I last reported. It has been an exhausting time in which a lot has happened and I have also been able to make a lot of progress on the project.

In addition to the thousands of hours of work and development on my lathe, I had to deepen my knowledge of programming microcontrollers, develop new circuit diagrams and PCBs and complete a master's course for mechatronics engineers.

So unfortunately I didn't have much time to write, but now I would like to report on how the work has progressed.

A lot of time has been spent on something that has been on my mind for years. It was about transforming the hundreds of wire connections on my large ‘breadboard’ into a circuit diagram and a PCB. At the same time, other ideas were to be realised or improved.

Here's a souvenir picture of my wire chaos and an idea of my headache:
.
.

MB_01.jpg

.
.
So after a few weeks of developing circuit diagrams and PCB planning, I was able to produce three new circuit boards. Of course the main board (which replaces the tangle of cables), an electrically isolated board for controlling the heating lamps and a new (better) board for controlling the input panel and display:
.
.

MB_02.jpg

.
.
Greetings from Austria
Thomas

[farmersplow]

The electronic control of my lathe:

Once the circuit boards were finished, I just had to fit them with the components:
.
.

MB_03.jpg

.
.
I integrated a variable groove depth control on this main board....
.
.

MB_04.jpg

.
.
...and a groove control.
.
.

MB_05.jpg

.
.
This automatic variable groove control consists of an analogue audio part for signal processing and an adapter board that I planned to be interchangeable for development purposes. Finally, it is controlled by a microcontroller. This allowed me to work on it until I achieved the best result. But I will report on this later.

Greetings from Austria
Thomas

[misjah]

you don't mess around do you?. great work

[farmersplow]

Farmer to mission-control: it's an honour! Thank you very much - and I want to do it right

[farmersplow]

The base plate of my lathe:

In addition to the circuit boards, the base plate also contains part of the internal cabling:
.
.

MB_06.jpg

.
.

MB_07.jpg

.
.
I laid the USB-C cables at the same time to enable subsequent programming (and updating).
The circuit board for controlling the heating lamp was manufactured separately to enable galvanic isolation from the main board and to lay the high-voltage cables in a shielded area. This means that there are no interfering electromagnetic fields in the area of the main board and the audio cables.
.
.

MB_08.jpg

.
.
Greetings from Austria
Thomas

[farmersplow]

The base plate of my lathe:

The underside of the base plate will have a ventilation grille and feet.

However, I still have to change the ventilation grille because the long aluminium bars start to vibrate like the sides of a harp, which could be disturbing in certain frequency ranges.
.
.

MB_09.jpg

.
.

MB_10.jpg

.
.
Greetings from Austria
Thomas

[displacedsnail]

You know, Thomas, every so often I come across the download of that video you made, Cutting in the Dark - such a great looking system! I'm excited to see the final result. Are you beginning to feel the end of the build coming? (But let's be serious, which lathe is ever "done," anyway?)

[farmersplow]

You know, Thomas, every so often I come across the download of that video you made, Cutting in the Dark - such a great looking system! I'm excited to see the final result. Are you beginning to feel the end of the build coming? (But let's be serious, which lathe is ever "done," anyway?)

Ha ha! You're right, it never seems to get really finished! My dear wife always asks me the same question: “Do you have the feeling that it will be finished soon?” After four years, I can say yes, I have that feeling. I've been saying that for two years, but now it's quite strong - the feeling. I think I'm on the verge of finishing.

[farmersplow]

The marriage.

In car manufacturing, we called it a marriage when the lower part of the chassis was joined to the engine and the body.

In my case, it is the connection of the base plate, including the entire electronic control system, with the lathe.

I had to build a temporary lift for this because the lathe is quite heavy and can't be lifted easily.
.
.

MB_11.jpg

.
.
But this made it quite easy to screw the base plate in place.

In the end, all the electronics are now installed and the many wires that had been causing me headaches for so long are finally off the table.

As already mentioned, I have laid the USB-C cables to enable subsequent programming. I first attached the connection sockets for three microcontrollers to the front. This gives me easy access for updates or programme improvements.
.
.

MB_12.jpg

.
.

MB_13.jpg

.
.
Greetings from Austria
Thomas

[zdenek]

will there be a lathe for sale?

when will we hear the recordings because that is the most important thing, said a German poet: the method is not important, the important thing is that it is effective, so when?

[farmersplow]

will there be a lathe for sale?

when will we hear the recordings because that is the most important thing, said a German poet: the method is not important, the important thing is that it is effective, so when?

While we are on the subject of proverbs, I would like to refer to one that has had a profound meaning for over 2500 years:
"The way is the goal". It emphasizes that the experiences, efforts, struggles and developments on the way to the goal are just as important as reaching the goal itself.
The recordings will come soon, just be patient. I wish for patience too, and as soon as possible!

[farmersplow]

Automatic, variable groove spacing control - Part 1:

Now that the new motherboard has been installed, I would like to carry out some initial tests. My greatest curiosity concerns the midi control and automatic, variable groove spacing control. I will come to the midi control later and would like to satisfy my curiosity about the groove control first.

About the preliminary work:
I spent several months with the calculation and the software. The programming caused me a lot of headaches. The reason for this is the short time the processor has to read the audio signal, calculate the new spacing and implement the changes. The audio hardware also had to be prepared for this.

What is the aim of variable groove spacing control?

As a rule, the drive spindle is moved at a constant speed using a stepper motor or BLDC motor. This creates a constant forward movement and a uniform spiral is cut or pressed into the record blank.

The system does not care whether a silent groove is cut or with high volume and strong bass. The distance always remains the same. To prevent the recorded signal from cutting into the groove of the previous round from one revolution to the next (overcut), the distance must be large enough to accommodate the recording volume and the bass (both of which require a lot of space due to the wide back and forth movement of the cutting stylus). For example, a groove spacing of 180µm (~141 lines/inch) must be set. You cannot record louder than this, otherwise the groove will collide.

I have created a few sample images here to explain this:

PIC. 1: fixed distance with relatively quiet recording.
In the picture you can see that the signal is louder at first, then it gets quieter and then louder again.
There is a lot of space (wasted land) between the grooves:
.
.

PIC_1.jpg

.
.
PIC. 2: fixed distance with medium-loud recording.
The same signal is recorded only slightly louder. There is still space (wasted land) between the grooves:
.
.

PIC_2.jpg

.
.
PIC. 3: fixed distance with relatively loud recording.
It is the same signal. There is hardly any space (land) between the grooves in the loud areas. The recording should not be louder here. However, space is wasted in the quiet areas:
.
.

PIC_3.jpg

.
.
PIC. 4: Fixed distance with recording too loud.
It is the same signal. There is an overlap between the grooves in the loud areas! The needle would jump during playback and the recording is destroyed. Nevertheless, space is wasted in the quiet areas:
.
.

PIC_4.jpg

.
.
To be continued:

Greetings from Austria
Thomas

[farmersplow]

Automatic, variable groove spacing control - Part 2:

PIC. 5: variable distance with relatively loud recording.
It is the same signal. There is still (little) space between the grooves (land) in the loud areas. This space (land) between the grooves is always constant and just as narrow in the quiet areas as in the loud areas. No space is wasted:
.
.

PIC_5.jpg

.
.
Depending on the piece of music, this can save up to 80% (usually around 30%-50%) of space.

Advantages:
Anyone who thinks that this is only there to give you more recording time is mistaken. You can also record louder with the same recording time or a combination of both advantages.
Another advantage is a drastic reduction in background noise when recording quietly. As the feed rate of the spindle is very low with silent cuts, the speed of the motor of the drive spindle is much lower and therefore also its noise and transmission into the recording. Even if these are already quiet, they will be many times quieter.

I will explain what to look out for in a variable control system in the next article.

Greetings from Austria
Thomas

[farmersplow]

Automatic, variable groove spacing control - Part 3:

What to look out for with variable control:
If you look closely at the PIC_5 image, you can see that not only is the distance (land) before the groove constant, but also after the groove. This sounds logical and simple at first, but for this to be possible, the distance must already be increased (or decreased) before the recording signal is available!
So you need a PRE-SIGNAL!
.
.

PIC_5.jpg

.
.
In addition, you also need to remember which signal has already been played in order to be able to set the spacing for the next round.
With a long-playing record that rotates at 33.33 rpm, it is necessary to store the deflection at least half a revolution before and after the signal. One revolution requires 1.8 seconds. This means that a lead time of 0.9 seconds is required for the PRE-SIGNAL (half a revolution).
A shorter lead time entails the risk of having to move a greater distance too quickly. This rapid movement would then already produce a sound.
A longer lead time reduces the advantage of the variable groove control.


How do you get a PRE-SIGNAL?
In the days of Neumann and co, the PRE-SIGNAL was obtained with a special tape recorder. These devices are very rare and expensive and, as far as I know, are still the only way to obtain a pre-signal in analog form.
The tape recorder was modified in such a way that it was equipped with two pickups. In addition, the tape between the pickups was extended by pulleys to ensure a 0.9 second gap.
Here is an example: Studer A80 with pre-listening function:
.
.

Studer A80 double.jpg

.
.
Since almost everyone today (with a few exceptions) makes the recording signals available digitally via a DAW, there are also other ways to get a pre-signal.

In addition to the DAW, an audio interface that is capable of providing two full stereo outputs is also required. Unfortunately, I discovered that although my MOTU M4 has two stereo outputs, they are not fully-fledged. This means that not both can be controlled via the DAW!!! - A bad buy.
But never mind. I bought a Solid State Logic SSL 2+ and it does what I want to do.

To be continued:

Greetings from Austria
Thomas

[farmersplow]

Automatic, variable groove spacing control - Part 4:

Practical implementation:

I recorded the same music signal twice into my DAW. Once for the PRE-SIGNAL, which is linked to output B.
This signal is not processed. Neither with IRIAA nor with any other EQ. It is only brought to max. 0dB.
The actual recording signal is at output A and is (if necessary) tweaked with a little EQ.
I send the recording signal (output A) to the DAW with a delay of 0.9 seconds (see picture):
.
.

105_02 0.9 seconds delay 02.jpg

.
.
On the audio interface, the cables are connected to stereo output channels A & B.
.
.

105_03.jpg

.
.

105_04.jpg

.
.
I connect the other ends to my lathe:
.
.

105_05.png

.
.
To be continued:

Greetings from Austria
Thomas

[farmersplow]

Automatic, variable groove spacing control - Part 5:

Before I start the first test run with a short sequence, I would like to explain the principle of the hardware and software behind it. In principle, the PRE-AUDIO-SIGNAL is processed analog. I have developed an audio board that amplifies the signal to a value of 0 to 4.5V (max 5V). It is important that the signals that require a lot of space (low frequencies) are amplified relatively strongly in addition to the loud signals, whereas quiet signals and high frequencies are hardly amplified at all.
These signals are then smoothed using an RC circuit and read in and stored by the micro controller. A large number of measuring points are stored in a data register in a time range of 1.8 seconds (0.9s pre and 0.9s post). The time required to perform the next step on the stepper motor is then calculated from this register data using an algorithm.
In one of my previous posts I talked about headaches. This came from the fact that I had a big timer problem because I only had 2.8ms between two steps under extreme conditions (45rpm and 200µm groove spacing) and all the calculations had to be carried out within this time. In my first attempts, however, I needed almost 5ms for this. But after several weeks and changing the sequences, I managed it in 1.9ms - that was close! I love these challenges.

So here is the first test with a music sequence. In the video you can hear the music from the PRE-SIGNAL (from the loudspeaker) and you can clearly see the change in speed of the drive belt to match the volume (and bass).
Everything works well so far. (In the video, the drive belt seems to jerk. But this is only due to the compression of the video from 80MB to 8MB. In reality it runs smoothly).
.
.

107.jpg

VIDEO:

107_video.mp4

.
.
The tests at 45 rpm also worked perfectly.

Greetings from Austria
Thomas

[misjah]

great job....fyi. it's easier to just route the main signal into another channel and apply delay there digitally, then there is no need for playing back 2x2 channels and when you change the main signal level the pre-view will follow.

[evildrome]

Wow. I thought the speed changes would be very slight and difficult to discern but no. There's really quite a difference in speeds.