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

[farmersplow]

Feedback Amplifier - Part 1

I wasn't able to post anything last month. But not because I didn't do anything. On the contrary, I worked so much on my project that there was no time for the presentation. And because I also want to do my contribution properly, I first had to find some time to prepare this chapter. Because just throwing in a few pictures without any explanation didn't seem good to me. So I found some time - between one and four in the night, when some people are partying and most people are asleep.

Now let's talk about the Feedback Cutterhead and Feedback Amplifier. I prefer to call it the ‘Feedback Compensation Amplifier’ because this part is supposed to compensate for the resonances of the drive of the driver coils. Ideally, it should compensate for the main resonances that occur in the range from 500Hz to 2kHz (depending on the cutterhead) in order to achieve the flattest possible output curve.
In addition (if the amplifier works well), it should also compensate for the second head resonance, which can be found at around 10kHz. The amplifier can also take over another task: the use of the IRIAA filter, which reduce the low frequencies for cutting and boosts the high frequencies.

So I had spent months trying to develop a circuit for it. There were many attempts that did not lead to a satisfactory result. I do have some knowledge of electronic circuits, but this is where my limits became apparent.

About two and a half years ago (Jan. 2022 https://www.lathetrolls.com/viewtopic.php?f=15&t=9578#p61050 ) I rebuilt the circuit of the Caruso amplifier, with 50% success. One channel had not worked as it should, which had repeatedly led to the destruction of my first cutting heads. The smoke threads rose up in delicate spirals and there was a smell of burnt resin in the air. It confirmed my theory that there must be a lot of smoke in all the electronic parts (tightly sealed). As long as the smoke stays in there, the part works. But as soon as the smoke comes out, the part is broken.

About a year and a half ago, I visited Flo (the master craftsman) in his workshop in Switzerland. There were lathes, electronic measuring devices and audio equipment everywhere. There were circuit boards here and cutting heads from all decades there. It smelled of machined metal and soldered circuit boards. Each box contained a jewel from the past that the master craftsman had brought into the present day. But also current, new and fresh things like the coffee he offered me. I love coffee and the machine he used to make it was in keeping with tradition! All in all, it was a special experience for me and I humbly tried to soak up the knowledge from decades of experience. As a parting gift, Flo gave me two circuit boards of the Flokason-Caruso feedback amplifier. Thank you Flo!

After consulting Flo again, I decided to use this circuit diagram from the Caruso board - simply because it works. With his permission, I can also present my MOD. The circuit diagram corresponds to that of the Caruso board (Flo had thankfully published it) - I have redrawn the circuit diagram because I want to make my own circuit boards in the future. Not because Flo's boards are bad, but simply because they don't fit into my case in terms of space. I also need to make some changes to the resistors and capacitors for my head to customise the board.

So first of all, here is the circuit diagram of my feedback compensation amplifier (based on the Caruso board).
.
.

A01 FP9_FB-comp-amp 1_3.png

.
.
There are a few things to consider:
R24/C25 regulate the lowpass at the line-in (to filter out high frequencies)
R7 regulates the gain of the feedback coil (more or less gain is required depending on the coil/magnet)
C33 adjusts the forward compensation for the first resonance
R33 adjusts the bass forward correction at low end. (~ 30dB)
C32 adjusts the forward compensation for the second resonance
C37/R37 is responsible for feedback activation
R39 adjusts the gain of the feedback (higher resistance = higher FB)

Here is the component list:
.
.

A02 FP9_FB_Comp_Amp_1.3 parts.pdf

.
.
In one of my first posts, I wrote about the recommended fuses with protective circuit. I found these in connection with the Caruso board.
https://www.lathetrolls.com/viewtopic.php?f=15&t=9578&start=20#p61061
.
.

A03 FP9_FB-comp-amp 1_3.png

.
.
Later I realised that this RC circuit causes a strong bass drop and therefore left it out again (I only kept the fuses):
https://www.lathetrolls.com/viewtopic.php?f=15&t=9578&start=40#p61137

Now, however, the IRIAA characteristic and the structure of the feedback compensation amplifier require this part of the bass cut at least partially outside the amplifier board.
This means that the amplifier board is only complete with the external RC circuit.
So I installed it again. Directly at the socket input of the amplifier in my device. (Power amplifier input to the fuse, then RC and then on to the cutterhead with a cable).

How is everything connected together?

1: Audio signal source (usually PC to sound card, or tape or other line source)
This signal can be EQ-processed, but does not require an IRIAA EQ!
2: LINE-IN Feedback Compensation Amplifier
3: LINE-OUT Feedback Compensation Amplifier
4: This LINE-OUT is routed to the Power Amplifier input
5: Then it goes on to the fuse and the RC circuit (bass damping)
6: From there the cable goes directly to the DRIVE-COIL of the cutterhead

A: The movement of the drive rod (to which the feedback coil is attached) induces a corresponding voltage (FB signal) due to the movement in the magnetic field. +/- is passed on by cable from the cutterhead.
B: The FB voltage +/- is connected to the FEEDBACK-COIL-IN of the Feedback Compensation Amplifier. The line-in is ‘compared’ with the FB-in. Resonances (i.e. unwanted increases in volume) are then compensated for.
C: A commercially available dB meter is connected to the MONITOR connection. It is used to calibrate the stereo compensation correctly. I will show how this works in practice in one of the following reports.

Here is a diagram showing the basic setup as described:
.
.

A04 FP9_FB-comp-amp 1_3.jpg

.
.
Greetings from Austria
Thomas

[farmersplow]

Feedback Amplifier - Part 2

Based on the circuit diagram, I then made circuit boards that fit my case exactly:
.
.

B01 FP9_FP-comp-amp 1_3 PCB.jpg

.
.
I hope everything works as planned and I don't disgrace Flo!

Greetings from Austria
Thomas

[spinnertownblanks]

This is absolute gold and somehow you keep making the thread I find most informative and interesting get even better! Endless thanks for all of your work.

[farmersplow]

This is absolute gold and somehow you keep making the thread I find most informative and interesting get even better! Endless thanks for all of your work.

Thank you very much! I endeavour to

[farmersplow]

Feedback Amplifier - Part 3

Now that the components have finally arrived, I can start assembling the circuit boards.
First the small and flat parts (diodes and resistors). Then the sockets for the ICs and the small capacitors. Then the larger capacitors (tantalum and ElKo's) and finally the transistors, potentiometers and connectors.
.
.

C01 FP9_FP-comp-amp 1_3 PCB.jpg

[farmersplow]

And because it's so much fun, here we go again (because it's stereo)!
.
.

C02 FP9_FP-comp-amp 1_3 PCB.jpg

.
.

C03 FP9_FP-comp-amp 1_3 PCB.jpg

.
.

C04 FP9_FP-comp-amp 1_3 PCB.jpg

.
.
Whatever you think, I think it looks great. If it works as it looks, then all is well.

Greetings from Austria
Thomas

[Iron]

Although I don't have the knowledge to fully understand what is going on here, I just wanted to thank you for sharing such an amazing work with us!!

[farmersplow]

Although I don't have the knowledge to fully understand what is going on here, I just wanted to thank you for sharing such an amazing work with us!!

Ha Ha Ha! It doesn't matter that you don't understand everything! I don't understand everything either! But it gets better and better with time. I've read a lot of posts over the last few years and at the beginning I didn't understand anything either. Now that I've spent thousands of hours working on the project, it's much easier to understand. Thank you for your contribution and I'm glad you like it.

[zdenek]

nice board, but not for the head with 1" coils, I'm waiting for super sound effects... unless it's supposed to be a bass head...and without a sopran:)

[farmersplow]

nice board, but not for the head with 1" coils, I'm waiting for super sound effects... unless it's supposed to be a bass head...and without a sopran:)

Perhaps we should agree that we have different goals

[farmersplow]

Feedback Amplifier - Part 4

These circuit boards should ultimately fit into the housing of my lathe. In the following picture you can clearly see that there is not much space in this area. However, I didn't want to hide it somewhere in the device, as this would make it more difficult to carry out a proper calibration.

Rendering image of the installation location of the FB-Comp.Amplifier. I have hidden the side panel:
.
.

D01 FP9_FP-comp-amp 1_3 PCB.JPG

.
.
The calibration is then carried out from the rear. The IRIAA EQ should be deactivated for the calibration process.
Rendering image of the rear calibration area
.
.

D02 FP9_FP-comp-amp 1_3 PCB.JPG

.
.
And so I'm getting closer and closer to my goal - or so it seems.

Greetings from Austria
Thomas

[zdenek]

Feedback Amplifier - Part 4

These circuit boards should ultimately fit into the housing of my lathe. In the following picture you can clearly see that there is not much space in this area. However, I didn't want to hide it somewhere in the device, as this would make it more difficult to carry out a proper calibration.

Rendering image of the installation location of the FB-Comp.Amplifier. I have hidden the side panel:
.
.
D01 FP9_FP-comp-amp 1_3 PCB.JPG
.
.
The calibration is then carried out from the rear. The IRIAA EQ should be deactivated for the calibration process.
Rendering image of the rear calibration area
.
.
D02 FP9_FP-comp-amp 1_3 PCB.JPG
.
.
And so I'm getting closer and closer to my goal - or so it seems.

Greetings from Austria
Thomas

nice, of course: the end justifies the means"

[Thelatheofus]

Amazing work, thanks again for sharing all this!

[farmersplow]

Amazing work, thanks again for sharing all this!

Thx

[farmersplow]

Feedback Amplifier - Part 5

With a temporary holder for the circuit boards, it's time for the installation.

I use shielded cables and channel separation for all audio signals (one cable per channel). The cables are each two-pole with shielding. After all, I don't want any hum.

The following two pictures clearly show why I developed my own customised PCBs. Unfortunately, the Caruso board didn't fit and the plug-in connections on the Flokason Caruso board were also omitted because I wire directly in the lathe.

.
.

E01 FP9_FP-comp-amp 1_3 PCB.jpg

.
.

E02 FP9_FP-comp-amp 1_3 PCB.jpg

.
.

E03 FP9_FP-comp-amp 1_3 PCB Adjust Video Pix.jpg

.
.
But now everything fits very well and is perfectly integrated. It's almost a shame that you won't be able to see it all when a housing is put over it at some point.

Now I'm getting ready to calibrate the FB control unit! I will create a few videos for the next few posts.

Greetings from Austria
Thomas

[farmersplow]

Feedback Amplifier - Part 6

Calibration Feedback Compensation Amplifier

First of all, I have to make an important prediction! I now know why I always burnt out one channel of the Cutterhead years ago! Because even with this calibration, the fuse always blew on one channel. Fortunately, these coils are very stable and so they didn't suffer any damage. The reason is simple and straightforward if you know it!

As in previous reports, I have already explained why the polarity of a driver coil in a stereo head must be reversed (side cut instead of depth cut). In my meticulous and precise adherence to all connection instructions, I have also done this. However, I only did it with the driver coil, not with the feedback coil! Experienced specialists already know what's coming next. The FB coil sends an inverted signal to the FB amp. The amp tries to compensate for this and ‘corrects’ it. This gives the driver coil even more juice and the FB amp corrects and amplifies with everything it has! - Everyone knows the scenario of feedback and what happens in the loudspeakers.

No matter! The fault has been rectified and I used the oscilloscope to check again whether the driver signal and FB coil signal are in harmony.

I use a frequency of 1kHz for the calibration. (Or even better, the resonance frequency of the cutterhead). Why don't I use a white noise? Because there is too great a risk that something will go wrong with the amplification and the calibration will take some time. I don't want to chase 16kHz (or more) through my cutterhead in the long term. It works just as well with 1kHz.

IMPORTANT NOTE: Ideally, a test signal without IRIAA and without EQ should be cut with the cutterhead beforehand to find out the main resonance. Play this signal on the turntable and view it with spectral analysis. If the main resonance is 1350Hz, then you should (ideally) make the settings with this frequency.
I show here in general how the setting is made for a cutting head with a main resonance of 1kHz.
My electronics also differ from the original Caruso board because the direction of rotation of the potentiometers is reversed. On the Caruso board, the three potentiometers have to be turned in the other direction beforehand and this also applies to the settings.
Finally important: I show an ‘attenuation’ of -27dB! This value differs for each type of cutting head and can be -15dB or even -42dB! It depends on the characteristic curve of the cutting head (and which attenuation it requires).

Flo (inventor of the Caruso Board) wrote a lot about this in a post many years ago:
https://www.lathetrolls.com/viewtopic.php?f=15&t=5250&p=38871&hilit=caruso+head#p38871

[farmersplow]

SETTING in the PRACTICE:

I switch the IRIAA to OFF (so that no high frequencies are amplified). In my case at 1kHz it would not matter whether the switch is on or off.

Here is the arrangement of the components:
Below the control knobs with the IRIAA switch
Above the DB meter
.
.

F01 FP9_FP-comp-amp 1_3 PCB Adjust Video Pix.jpg

.
.
BASIC SETTING:

Before a signal is sent, I set the amplifier to the basic setting:
Connect the dB meter to the dB MONITOR input (both channels)
FB-dB MONITOR Pot. --- turn fully anti-clockwise (both channels)
FB-COIL MODULATION --- turn fully clockwise (both channels)
LINE-IN MODULATION --- turn fully clockwise (both channels)
IRIAA-OFF

See video:
.
.

F02 FP9_FP-comp-amp 1_3 PCB Adjust Video.mp4

[farmersplow]

SETTING LEFT CHANNEL

Play a 1kHz signal (Line In connection of the FB-AMP.).
Power Amplifier (left channel only) slightly above the centre position:

LINE-IN MODULATION --- turn slowly anti-clockwise (left channel)

• After two or three turns, the signal should be audible.

• Continue turning slowly until it becomes loud (but not so loud that the coil melts!)

• A display should already be visible on the dB meter. Ideally, this should be set to a value between -3dB and -15dB

• If the value on the dB meter is above 0dB or below -18dB when the signal is ‘loud’, you should stop and readjust the dB meter (both channels). You can then start again.

Read off the displayed START VALUE:

• In my case, I have set the volume to a displayed value of -15dB.

• My target setting is an ‘attenuation’ of -27dB. I do this in three steps (-9dB; -9dB; -9dB). In between, I always increase the line-in modulation. If your cutting head requires less attenuation (e.g.: -18dB) then only do the process twice. If your cutting head requires more attenuation (e.g.: -33dB) then do the process once more (-9dB; -9dB; -9dB; -6dB).

FB-COIL-MODULATION --- turn slowly anti-clockwise (left channel)

• The signal becomes quieter and the DB metre decreases. Set to a value that is -9dB lower than the START VALUE

• In my case, this is a display value of -24dB (-15dB -9dB = -24dB).

• Now I have an attenuation of -9 dB

LINE-IN MODULATION --- turn slowly anti-clockwise (left channel)

• The signal becomes louder again.

• Turn until the START VALUE is reached again (in my case -15dB).

FB-COIL-MODULATION --- turn slowly anti-clockwise (left channel)

• The signal becomes quieter again and the DB metre decreases. Set to a value that is -9dB lower than the START VALUE

• In my case, this is a display value of -24dB (-15dB -9dB = -24dB).

• I have now achieved an attenuation of two times -9 dB, i.e. -18 dB

LINE-IN MODULATION --- turn slowly anti-clockwise (left channel)

• The signal becomes louder again.

• Turn until the START VALUE is reached again (in my case -15dB).

FB-COIL-MODULATION --- turn slowly anti-clockwise (left channel)

• The signal becomes quieter again and the DB metre drops. Set to a value that is -9dB lower than the START VALUE

• In my case, this is a display value of -24dB (-15dB -9dB = -24dB).

• I have now achieved an attenuation of three times -9 dB, i.e. -27 dB

LINE-IN MODULATION --- turn slowly anti-clockwise (left channel)

• The signal becomes louder again.

• Turn until the START VALUE is reached again (in my case -15dB).

• This ensures that no loss of overall volume is caused by the attenuation.

Before continuing, turn the power amplifier (left channel) down again!

If the signal is distorted during attenuation, turn it back a little (one turn) and switch it off. The limits of the attenuation options have been reached at this point!

Video about the setting:

F03 FP9_FP-comp-amp 1_3 PCB Adjust Video.mp4

[/size][/b]

[farmersplow]

SETTING RIGHT CHANNEL

Play a 1kHz signal (Line In connection of the FB-AMP.).
Power Amplifier (right channel only) slightly above the centre position (same position as before for the left channel):

LINE-IN MODULATION --- turn slowly anti-clockwise (left channel)

• After two or three turns, the signal should be audible.

• Continue turning slowly until it becomes loud (but not so loud that the coil melts!)

• A display should already be visible on the dB meter. Ideally, this should be set to a value between -3dB and -15dB

• If the value on the dB meter is above 0dB or below -18dB when the signal is ‘loud’, you should stop and readjust the dB meter (both channels). You can then start again.

Read off the displayed START VALUE:

• The same starting value should be set as for the left channel

• Here too, my aim is to set an ‘attenuation’ of -27dB. Again, I do this in three steps (-9dB; -9dB; -9dB). In between, I always increase the line-in modulation.

• Now everything is the same as for the left channel

When everything is set, turn the power amplifier back again!

If the signal is distorted during attenuation, turn it back a little (one turn) and switch it off. The limits of the attenuation options have been reached at this point!

Video for setting the right channel:

F04 FP9_FP-comp-amp 1_3 PCB Adjust Video.mp4

[farmersplow]

CONTROL BOTH CHANNELS SIMULTANEOUSLY

Play a 1kHz signal (Line In connection of the FB-AMP.).
Power Amplifier (both channels) approximately centre position (as before for the left channel):

The display on the dB meter should be the same.

Video to check both channels for equality:

F05 FP9_FP-comp-amp 1_3 PCB Adjust Video.mp4