KNOP lathe DIY build

[Gus]

Hi
Very nice improvement of your cutterhead, did you use carbon fibre as pushrods?
Keep it up!

[grooveguy]

VERY encouraging results, do keep us posted. Sounds fantastic.

[KNOP]

To Gus: Yes that is with the carbon fibre pushrods. They weight 0.3 grams instead of 1 gram for metal pushrods.

I have some days leave from work this week, I will install a new stylus and try to get the noise floor as low as possible, also try out the heating. On my current stylus the heating wire has come off.

[KNOP]

Hello.

I've changed the stylus for a new one, took me most part of the day getting everything dialed in again. Needed to re-EQ the cutterhead, made some improvements on the cutterhead connection to the suspension mechanism, reajusted the counterweight and so on.

Noise floor is much better, recorded at 45RPM 140LPI this time. The wow issues are much less pronounced.
https://clyp.it/xvrlkgvh

[grooveguy]

Boy, that sounds magnificent to me; if I can achieve results like that I will be happy as a clam. Remarkable stereo separation. In your closeup of the head, it appears that the 'torque tube' (or whatever you call the member that connects the stylus holder to the back of the head) goes straight back; that is, there is no obligatory 20-degree angle that Noel Keywood says is necessary in his paper. Any comment?

Question: You speak of a counterweight. Are you using a counterweight, exclusively, to support the head and set the depth of cut, or do you have a spring of any kind in use for that? Also, can you divulge: 1) distance from the stylus to the vertical pivot point of the head, and 2) height of the pivot point above the disc surface? (Don't tell if you'll have to kill me afterward, but I'm collecting this sort of info.)

Keep it up!

[KNOP]

In your closeup of the head, it appears that the 'torque tube' (or whatever you call the member that connects the stylus holder to the back of the head) goes straight back; that is, there is no obligatory 20-degree angle that Noel Keywood says is necessary in his paper. Any comment?

Well this is because I didn't know about this angle until recently. I'll have to do some research in this matter.

Question: You speak of a counterweight. Are you using a counterweight, exclusively, to support the head and set the depth of cut, or do you have a spring of any kind in use for that?

Only a counterweight at the moment. I made provisions for using a spring, but have not experimented with that option. The problem I have with the weight is that the wires on top of the cutterhead and the suction tube have quite an influence in the weight of the downforce, and its quite touchy to dial in.

Also, can you divulge: 1) distance from the stylus to the vertical pivot point of the head

The stylus to the V spring is 22mm (The V spring is the vertical pivot point I guess?)

, and 2) height of the pivot point above the disc surface?

About 10mm

[grooveguy]

Okay, thanks for that info. As far as the pivot point goes, there seems to be quite a range among commercial machines, and even a wider one among homebrew ones. Mechanical engineers will argue this physics, but all these recorders seem to work just fine. I was looking for some magic 'golden mean,' but perhaps it's not as terribly important as the purists make out.

I thought I had posted the Keywood paper here somewhere, but can't find it. If you need it, it's here:
https://www.dropbox.com/s/140rmbnaulsl2xa/Vertical%20Tracking%20Angle.pdf?dl=0

An old Fairchild lathe I had years ago used only a counterweight, and I recall the same problem maintaining a constant depth of cut because of the vacuum hose and wires. I think I clamped those to the carriage near the pivot point, so at least the head was no longer sensitive to changes in the dressing of the longer run to the chip jar and amplifier. A spring allows adjustment on-the-fly, of course, but it does raise the resonant frequency of the vertical component, and if the 'Q' of the system is high, can lead to nasty vertical oscillation. I guess part of the trick is keeping the length of the spring as long as you can so that the stylus pressure doesn't change much as the vertical position of the head changes. I don't know about these mechanical analogies of equivalent electrical circuits.

But the bottom line is, of course, that you are getting splendid results. Good job and good luck!

[sifis1983]

Hi,

Congratulations! That sounds excellent to me! And thank you for your great information...

Some queries:

What about the "new stylus" ? What kind of stylus you use now and what before?
Why you "re-EQ" your cutterhead? Is all about the resonances of your design? or you just compensate some frequencies to "flat" response, after iRIAA?
Is this audio example cut in PVC?

Regards,
Sifis

[KNOP]

Hey,

The stylii are both from myshank, grampian type. The old one had a broken tip. The eq is for flattening the resonances and irregularities of the cutterhead, I do not use iriaa, as it is compensated directly via the fabfilter plugin. Its all cut on blanks from myshank as well. I had to redo it because of different weight on the stylus.

Took me and a friend about two days to get the weight and suction working reliable and consistant. Made about 20 records, ranging from meh to pretty awesome.

Added gmonobass in the signal chain, also JB multiband compressor for hf limiting, and a lfp and hpf at the end. The hf limiting goes a long way reducing distortions in the high frequencies.

Still looking for decent (free) vst plugins for de essing and stereo expansion.

[KNOP]

Did a little ghetto upgrade to the lathe. No time to clean up the build this week.

Added a belt and a foam vibration damping pad to the motor. A bit crude, but works. At lead in and out speeds there is now only a faint low end rumble instead of a whine. Still not perfect but at least makes the lead in and outs usable. At regular song speeds the motor noise is undetectable.

IMG_20170329_211427.jpg

[KNOP]

Ok forget about yesterdays post. Found a way better solution. Now the motor housing is fully decoupled from the lathe on a pole next to the lathe, and replaced the belt by rubber elastics(really dont know the word in English). Now there is no vibration noticable at all on the lathe.

IMG-20170330-WA0002.jpeg

[grooveguy]

Splendid! It looks as though the belt pulleys are about the same diameter; that is, no speed reduction is afforded there. Is that some sort of gear box between the driven pulley and the lead screw? I am doing work along this same line and wonder what actual speed the stepper motor is running when you record at a specific l.p.i., such as 180 or so for microgroove? I'll bet that the flexible rubber belt help isolate not only the noise from the motor, but tends to either smooth-out the steps or at least take the higher-order components out that might be audible as simple lateral modulation of the groove at the stepper rate. Good job!

[KNOP]

Dear Grooveguy,

It is indeed a 1:1 transmission, there was none needed since the original version was tuned to have the lowest vibrations at 100-200lpi. The green thing is no gearbox but a stepper motor stripped of its stator. It serves as an axle with double bearing now, saves me time and money to build an extended axle.

The following times are for a 1mm pitch leadscrew, using a 200 step per revolution stepper motor with 1/16th microstepping (so 3200 steps per revolution). Did the calculations by hand, could be off a little bit. The values are the time needed between on and off signals to the stepper driver.

#define R100 1111 //time between on and off signal a in uS, for stepper motor 200 steps per rev, 16 microstepping with 256 microstep interpolation, 33RPM
#define R120 1333
#define R140 1555
#define R160 1777
#define R180 2000
#define R200 2222

#define F100 823 //Same at 45RPM
#define F120 987
#define F140 1152
#define F160 1317
#define F180 1481
#define F200 1646

digitalWrite(STEP_PIN, LOW);
delayMicroseconds(LatheSpeed);

digitalWrite(STEP_PIN, HIGH);
delayMicroseconds(LatheSpeed);

[grooveguy]

Thanks for that; it all looks really good. I shied away from ball bearings on my lead screw because ball bearings are generally considered probable sources of noise in disc recording applications. Most of the older overhead mechanisms had cones drilled in the ends of the lead screw, and you put either a single, captive ball or a point into that cone as a bearing and to hold the lead screw steady along its axis.

The machinist I'm using had no experience with that sort of support, plus whatever fixed parts bear into the ends of the screw have to be adjustable, yet another machining process. So I went for bronze bushings, which serve both as a bearing for the screw to turn in and a thrust bearing to keep it in one place along the axis. BUT... (and there's always one of those) that means that the bearings have to be in very good alignment with respect to one another or the screw will bind. This is a very possible problem in my design.

So now I'm thinking along the lines of pillow blocks, which other Trolls have used to support their lead screws, and hope that the ball bearings will be good enough not to impart any 'rumble' along the axis of the screw as it turns. My screw is a massive 3/4 inches/19mm in diameter with 10 turns to the inch, or a pitch of 2.54mm, so it's turning pretty slowly. I do use a 4:1 rubber O-ring belt reduction, both to keep the stepper in a happy range and to help low-pass-filter the microsteps from the driven pulley.

Keep up the clever work and keep us posted, please.

[KNOP]

at Grooveguy:

Yes I've noticed that indeed on the older overhead style lathes they use a kind of ball and recess at the end of the feedscrews. I've considered this, but found it too expensive a part to make. I've still not got the noise floor down to the level I would like, so that may be masking some motor noises. Probably chipped my needle again, after a stupid swarf mishap. As of now I'm nothearing the bearings make any noise.

On the noise issue: Seems the rubber belt did not work out, tried a couple of other sizes and stuff, but while at lead in and out speeds it works great and silent, at normal recording speeds it makes speed irregular. So back to using the GT2 belt, but with just enough tension not skip. This translates less vibrations. I also made an external transformer mod, to take care of that annoying 100Hz peak. Seems like an engineer thought it was a good idea to put the mains transformer right next to the platter motor in the Hanpin 5500's. Since the local electronics stores are closed today I just mounted the transfo to wall, will get an enclosure asap.

IMG_20170401_193015.jpg

Results are good: the 100Hz peak is gone. Makes for a more transparent and defined low frequency.

[grooveguy]

Well done, KNOP! With the rubber belt you may have experienced a sort-of "stick/slip," where the lead screw might actually stop turning because of a change in load, and then break loose and be pushed quickly ahead for a short period by the elasticity of the belt. A good point and something for me to keep in mind. I do plan to use a pretty stiff O-ring however. Thanks.

[KNOP]

Hello everyone.

I've installed a nema 14 size motor instead of a nema 17. I cannot hear any difference in noise translated to the record. I did notice that the quality of the stepper bearings varies by quite a lot. I've got a Lin engineering stepper motor, which has good bearings, and a wantai stepper motor + a no name chinese stepper, both of which make some bearing noises.

I've made some recordings in various genres to get a feeling what translates good to vinyl and what not. The suction hose was completely stuffed with swarf. Thought that the white in the tube was talc powder, but it was the swarf. I cleaned the tubing and cut about 10 records straight without the suction acting up. There is a lot of needle noise however, I will install a fresh stylus soon: hoping I won't break it again.

https://clyp.it/3w1gizxt

[grooveguy]

Very nice results! Judging from the crisp and open top-end, your stylus must not be in as bad shape as you make out. I do hear the surface noise at the beginning, where the level is low, but it sounds to me more like irregularities in the polycarbonate than a stylus problem, which is a more constant noise. I'll bet if you cut a real lacquer blank it would be virtually noiseless. Good work and an extraordinarily good cutterhead.

[Gus]

BRAVO Wim!

[record cutter 76]

hi knop whill you share the cutter head 3d printer file white me i would like to test it you have done a verry good job and it sounds reely good
hope that you can healp me white my request