Aaa, it works now, on picture above is amp that I attached directly on Supaspin, compact. Temperature of aluminum plate is just lukewarm after fer hours of spinning, about 10 - 15 deg over ambient.
What could be generally interesting is my side-wise mounting of TO220 multiwat chips, learnd that many years ago with LM3886 as I was to lazy and to clumsy to make good boards, like this all odd pins are on top and all even pins on bottom of the pcb. If 2 single layer pcbs are glued together to make one thicker double sided pcb, its thickness is just slightly less than distance between odd and even pins.
Above is sideview, upper board is longer and extended over Supaspin input terminals that are soldered together, making it one board.
What could be generally interesting is my side-wise mounting of TO220 multiwat chips, learnd that many years ago with LM3886 as I was to lazy and to clumsy to make good boards, like this all odd pins are on top and all even pins on bottom of the pcb. If 2 single layer pcbs are glued together to make one thicker double sided pcb, its thickness is just slightly less than distance between odd and even pins.
Above is sideview, upper board is longer and extended over Supaspin input terminals that are soldered together, making it one board.
And finally, this is my draft of PCB, needed just to check sense and drill holes. Of course if anyone wants to do it correct for production, welcome, sure.
Finally installed latest Supaspin firmware from Richard, all works as described, as always. Thanks RICHB!
So, only thing left is to assemble whole thing.
I will still leave possibility to change motor, but only for similar capstan types as I'm curious if slot-less motor will bring any benefit, so far I have only one (No4 from post#50) and is not better than this one.
When time allows I will probably make copy of this board to rest on my bench for further candidates testing.
So, only thing left is to assemble whole thing.
I will still leave possibility to change motor, but only for similar capstan types as I'm curious if slot-less motor will bring any benefit, so far I have only one (No4 from post#50) and is not better than this one.
When time allows I will probably make copy of this board to rest on my bench for further candidates testing.
Had no time to make a proper post last 2 weeks, but I did investigate. While I tough posting my original problem will soon find "silver bullet" solution from community, it turned in deeper and deeper digging where I got a bit addicted. Supaspin by Richard does almost a "silver bullet" for drive, but motor choice is another story....
Further on actual setup from previous few posts, reduced running amplitude of the motor even lower, to the level I can just barely change LP on the fly (motor running) without stopping it, this is my minimum torque requirement; changing LP's on fly.
Speed takes 10-20 min to sit on exact 33.33 (after cold start) and stays there forever. This is also really silent. Guess startup speed being lower is belt slip with not distributed oil. In any case this motor is fit for purpose.
I'm still having opened eyes for decent shaft encoder and software to process signals from it, test LP is not good enough to measure wow and flutter, or to better put it short time speed osculations, of this set up.
And finally from Internet I found some useful explanations about modern BLDC motors relation to low torque ripple at low speed required by what we do.
First and most understandable link not coming from motor manufacturer but from magnet manufacturer:
https://mqitechnology.com/understan...less-dc-motors-with-isotropic-bonded-magnets/
From the same site, bit more in depth:
https://mqitechnology.com/wp-conten...bration-in-pm-motors-sources-and-remedies.pdf
Another site worth looking at:
https://things-in-motion.blogspot.com/2019/01/selecting-best-pole-and-slot.html
With better understanding, looked for sources of ideal motor, here is manufacturer that does almost all that is needed for perfect TT:
https://www.thingap.com/
I'm planning to contact them and I hardly doubt I will get motor from them, they are into space and defense plus lately I learned they supply motors to VPI. Me being diy from Croatia (somewhere in east europe for them) with limited funds and prospects for their business, does not look like a preferable buyer in their eyes, but I'll try.
However is interested in this subject, pls read linked posts above.
Next I will post few several "preliminary" posts on motor choice . Preliminary as this is not the end or final conclusion, and few because our site doesn't easily take big posts at once.
PS ; one trivial problem, but suggestions are welcome:
Struggle to find right glue to seal- laminate copper layer inside of the housing and plinth. Last time I used old fashion solvent rubber glue (smelly sticky yellowish stuff like we used to repair tubes on bicycle tires) It lasted for decade, but not good enough. One problem is that these surfaces are huge and don't allow solvent to come out or humidity - air to come in. Second problem is one side is metal, other MDF or stone (one surface is barrier, second surface is porous). Liquid epoxy would fix this but is mess to work with.
Further on actual setup from previous few posts, reduced running amplitude of the motor even lower, to the level I can just barely change LP on the fly (motor running) without stopping it, this is my minimum torque requirement; changing LP's on fly.
Speed takes 10-20 min to sit on exact 33.33 (after cold start) and stays there forever. This is also really silent. Guess startup speed being lower is belt slip with not distributed oil. In any case this motor is fit for purpose.
I'm still having opened eyes for decent shaft encoder and software to process signals from it, test LP is not good enough to measure wow and flutter, or to better put it short time speed osculations, of this set up.
And finally from Internet I found some useful explanations about modern BLDC motors relation to low torque ripple at low speed required by what we do.
First and most understandable link not coming from motor manufacturer but from magnet manufacturer:
https://mqitechnology.com/understan...less-dc-motors-with-isotropic-bonded-magnets/
From the same site, bit more in depth:
https://mqitechnology.com/wp-conten...bration-in-pm-motors-sources-and-remedies.pdf
Another site worth looking at:
https://things-in-motion.blogspot.com/2019/01/selecting-best-pole-and-slot.html
With better understanding, looked for sources of ideal motor, here is manufacturer that does almost all that is needed for perfect TT:
https://www.thingap.com/
I'm planning to contact them and I hardly doubt I will get motor from them, they are into space and defense plus lately I learned they supply motors to VPI. Me being diy from Croatia (somewhere in east europe for them) with limited funds and prospects for their business, does not look like a preferable buyer in their eyes, but I'll try.
However is interested in this subject, pls read linked posts above.
Next I will post few several "preliminary" posts on motor choice . Preliminary as this is not the end or final conclusion, and few because our site doesn't easily take big posts at once.
PS ; one trivial problem, but suggestions are welcome:
Struggle to find right glue to seal- laminate copper layer inside of the housing and plinth. Last time I used old fashion solvent rubber glue (smelly sticky yellowish stuff like we used to repair tubes on bicycle tires) It lasted for decade, but not good enough. One problem is that these surfaces are huge and don't allow solvent to come out or humidity - air to come in. Second problem is one side is metal, other MDF or stone (one surface is barrier, second surface is porous). Liquid epoxy would fix this but is mess to work with.
Two things to clarify, and are not too obvious. Pls feel free to dissagree.
1. 3-phase BLDC rotational speed vs commutation power suppls frequency:
apperently it doesnt matter how many windings-slots-poles-teeth stator has. Speed of BLDC motor is dependent only on number of magnetic poles pairs of the rotor. More stator windings etc will result in less torque ripple and that is beneficial, but does not affect speed.
To calculate speed (RPM) vs frequency in traditional formula rpm=60(seconds) x frequency/Pp(number of pole pairs) , for Pp use number of magnets in rotor devided by 2.
Example; If rotor has 8 attraction points as seen by small piece of iron, that is 4 (8/2) pole pairs and the speed of this motor wil be 750rpm @50hz.
2. cogging (torque) vs (torque) ripple
this seems to be mixed often:
- Cogging torque (or just cogging) is static, not powered, attraction of magnets in rotor to iron teeth in stator. It can be felt by finger rotating motor and ripple is felt. This is not necessarily bad (it is clearly observable with Airpax - Premotec etc motors as used in Linn, Rega etc) as it is static observation.
When stator teeth are energized by coils and motor turns, picture changes completely.
- Torque ripple is change in motor torque between poles in rotation, this is what counts. Cogging is affecting torque ripple to some account (still not known to which extend by me) but it is not at all eliminated in slottles (ironless, cogging -free) motors that have no iron core (just coils in air) and theoretically can't have cogging.
In short, it is torque ripple that is measure of merit for TT drive, not cogging.
1. 3-phase BLDC rotational speed vs commutation power suppls frequency:
apperently it doesnt matter how many windings-slots-poles-teeth stator has. Speed of BLDC motor is dependent only on number of magnetic poles pairs of the rotor. More stator windings etc will result in less torque ripple and that is beneficial, but does not affect speed.
To calculate speed (RPM) vs frequency in traditional formula rpm=60(seconds) x frequency/Pp(number of pole pairs) , for Pp use number of magnets in rotor devided by 2.
Example; If rotor has 8 attraction points as seen by small piece of iron, that is 4 (8/2) pole pairs and the speed of this motor wil be 750rpm @50hz.
2. cogging (torque) vs (torque) ripple
this seems to be mixed often:
- Cogging torque (or just cogging) is static, not powered, attraction of magnets in rotor to iron teeth in stator. It can be felt by finger rotating motor and ripple is felt. This is not necessarily bad (it is clearly observable with Airpax - Premotec etc motors as used in Linn, Rega etc) as it is static observation.
When stator teeth are energized by coils and motor turns, picture changes completely.
- Torque ripple is change in motor torque between poles in rotation, this is what counts. Cogging is affecting torque ripple to some account (still not known to which extend by me) but it is not at all eliminated in slottles (ironless, cogging -free) motors that have no iron core (just coils in air) and theoretically can't have cogging.
In short, it is torque ripple that is measure of merit for TT drive, not cogging.
"preliminary" guide to motor choice, #1.
Most of motors I have chosen are for now days cheaply available video tape machines, VHS, beta, digital beta and DVCAM, even others were considered.
A, bearings:
When motor is removed from unit, check if motor has ball bearings or bushings. Ball bearings are horribly noisy, bushings are what is needed. Unfortunately many better motors have ball bearings as video machines cared not about noise but about precision where ball bearings are better.
If ball bearing motor has other good factors (see coming posts) it can be considered to replace rollers with bushings.
B, motor stator configuration
Here is picture of 3 typical stators:
On top: very bad: 9 slots, no dummy teeth, no skewing (VCR drumhead)
Middle: very good: 24 slots, max number of dummy teeth which gap almost equals slot gap (so far favorite VCR capstan)
Bottom: probably the best , no iron core (slot-less, iron-less, no cogging) (Capstan motor from 50k Euro DVCAM recorder).
I still need to understand benefits of iron-less motor like this with few coils (as space doesn't allow for more air coils) vs cheap motor in middle that has iron cores (and slots) but many of them with dummy teeth to add....
Most of motors I have chosen are for now days cheaply available video tape machines, VHS, beta, digital beta and DVCAM, even others were considered.
A, bearings:
When motor is removed from unit, check if motor has ball bearings or bushings. Ball bearings are horribly noisy, bushings are what is needed. Unfortunately many better motors have ball bearings as video machines cared not about noise but about precision where ball bearings are better.
If ball bearing motor has other good factors (see coming posts) it can be considered to replace rollers with bushings.
B, motor stator configuration
Here is picture of 3 typical stators:
On top: very bad: 9 slots, no dummy teeth, no skewing (VCR drumhead)
Middle: very good: 24 slots, max number of dummy teeth which gap almost equals slot gap (so far favorite VCR capstan)
Bottom: probably the best , no iron core (slot-less, iron-less, no cogging) (Capstan motor from 50k Euro DVCAM recorder).
I still need to understand benefits of iron-less motor like this with few coils (as space doesn't allow for more air coils) vs cheap motor in middle that has iron cores (and slots) but many of them with dummy teeth to add....
"preliminary" guide to motor choice, #2.
How to access UVW 3 phase input.
If we want to move any further a simple way to aces input posts for stator UVW windings is necessary.
Each UVW winding input in common BLDC will be ending at 2 mosfet bridge drivers, wire connecting to one source and one drain. Some lower power motors have mosfet drive embedded in a chip (here chip diagram will help to determine UVW posts, see post #153).
Sometimes is easy and sometimes almost impossible to reach them.
Here is example of 2 seemingly identical stators:
Both have mosfets inside of coils (bigger chips) and hall sensors inside of other 3 coils(small chips).
But! Motor on the left has drive connected on wire that ends inside coil. Since rotor comes on top of this it is almost impossible to take a wire out of the middle of the coil and connect to new drive or scope. There is possibility to drill small holes through metal PCB and drive wires under the assy, but I did not try that yet.
These terminals we need are also connected to 3 electrolytic caps you see on outside for filtering (on left motor utmost 3 caps, C2, C3 and C5). This is possibility to hook up but traces on PCB are very thin and questionable if sufficient for long term operation.
Motor on the right however has drive connected at outside of the coil, see white and yellow wire, and missing one near letter V. This makes it much easier to take UVW terminals out to be connected.
In any case , it is needed to take wires out where UVW 3 phase is driving windings, if that is not possible in sane manner, I would wait and continue with another motor.
How to access UVW 3 phase input.
If we want to move any further a simple way to aces input posts for stator UVW windings is necessary.
Each UVW winding input in common BLDC will be ending at 2 mosfet bridge drivers, wire connecting to one source and one drain. Some lower power motors have mosfet drive embedded in a chip (here chip diagram will help to determine UVW posts, see post #153).
Sometimes is easy and sometimes almost impossible to reach them.
Here is example of 2 seemingly identical stators:
Both have mosfets inside of coils (bigger chips) and hall sensors inside of other 3 coils(small chips).
But! Motor on the left has drive connected on wire that ends inside coil. Since rotor comes on top of this it is almost impossible to take a wire out of the middle of the coil and connect to new drive or scope. There is possibility to drill small holes through metal PCB and drive wires under the assy, but I did not try that yet.
These terminals we need are also connected to 3 electrolytic caps you see on outside for filtering (on left motor utmost 3 caps, C2, C3 and C5). This is possibility to hook up but traces on PCB are very thin and questionable if sufficient for long term operation.
Motor on the right however has drive connected at outside of the coil, see white and yellow wire, and missing one near letter V. This makes it much easier to take UVW terminals out to be connected.
In any case , it is needed to take wires out where UVW 3 phase is driving windings, if that is not possible in sane manner, I would wait and continue with another motor.
"preliminary" guide to motor choice, #3.
Once 3 correct UVW wires are connected and available, several things can be tested:
- Finally, most important, connect 2 wires (out of 3) to scope and rotate motor by hand (samples of some motors EMF on scope are in post #130):
a) If the shape of signal (back EMF) is close to sinewave; that motor is going to run good with sinewave drive .
b) If back EMF is away from sine, one can try to run it with original drive chip (+hall sensors + mosfet bridges) and check how well this drive reassembles motor EMF (see again post #153).
However driving it with original chips (hall + chip + mosfets) corrected by DC voltage input will make close to impossible to maintain the speed without feedback. And feedback with belt drive is no-go as there is no enough time reserve for feedback. With DD it is different but I'm not there with this post, its about belt drive so far.
Pls also pay attention @Pyramid post #140 and others in this thread, those are good posts describing feedback benefits for DD,
I think for belt drive this will be more difficult and we need stable rotation w/o feedback
Once 3 correct UVW wires are connected and available, several things can be tested:
- check resistance and record it for later (must be same in all 3 combinations)
- connect small 1.5 V battery to 2 wires and probe each stator tooth with piece of small iron. This will indicate which tooth's are energized per phase. I always got every third tooth energized, this is not necessarily like this as sometimes 1 phase is connected to few teeth in row.
- put the rotor back on and energize one coil with battery as above, when rotating the rotor there will be some attraction points to overcome. Number of these attraction points is number of pole pairs to be used in formula rpm=60(seconds) x frequency/Pp(number of pole pairs)
- Finally, most important, connect 2 wires (out of 3) to scope and rotate motor by hand (samples of some motors EMF on scope are in post #130):
a) If the shape of signal (back EMF) is close to sinewave; that motor is going to run good with sinewave drive .
b) If back EMF is away from sine, one can try to run it with original drive chip (+hall sensors + mosfet bridges) and check how well this drive reassembles motor EMF (see again post #153).
However driving it with original chips (hall + chip + mosfets) corrected by DC voltage input will make close to impossible to maintain the speed without feedback. And feedback with belt drive is no-go as there is no enough time reserve for feedback. With DD it is different but I'm not there with this post, its about belt drive so far.
Pls also pay attention @Pyramid post #140 and others in this thread, those are good posts describing feedback benefits for DD,
I think for belt drive this will be more difficult and we need stable rotation w/o feedback
"preliminary" guide to motor choice, #4.
Rotor magnet configuration.
I got myself very useful tool; piece of green magnetic film. It shows magnetic fields without throwing iron dust at them and making difficult to clean the mess. It seems to be available at amazon for 9.99$ per sheet.
This is the picture of rotors corresponding to 2 slot-less motors (stators shown 2 posts above):
Obviously patterns follow trapezoidal shapes of stator coils. I still need to decide to make bushings for one of these and get rid of racers in order to make progress meaningful.
Next picture is of preferred Phillips capstan with 24 slots, dummy slots and running very well
It has just normal straight magnets.
Finally picture of Sony drumhead motor that I discarded in post #179 for noise and luck of speed stability.. (must admit this motor was run with feedback, a lot of it, in original machine. so it was not made for no feedback run)
This one has skewed magnets which is nice tool against torque ripple.
Rotor magnet configuration.
I got myself very useful tool; piece of green magnetic film. It shows magnetic fields without throwing iron dust at them and making difficult to clean the mess. It seems to be available at amazon for 9.99$ per sheet.
This is the picture of rotors corresponding to 2 slot-less motors (stators shown 2 posts above):
Obviously patterns follow trapezoidal shapes of stator coils. I still need to decide to make bushings for one of these and get rid of racers in order to make progress meaningful.
Next picture is of preferred Phillips capstan with 24 slots, dummy slots and running very well
It has just normal straight magnets.
Finally picture of Sony drumhead motor that I discarded in post #179 for noise and luck of speed stability.. (must admit this motor was run with feedback, a lot of it, in original machine. so it was not made for no feedback run)
This one has skewed magnets which is nice tool against torque ripple.
As it seems, different motors follow different strategies to defeat torque ripple, but none uses all, except might be ThinGap company products I mentioned earlier
For me as so far Phillips VCR capstan with many stator poles , dummy slots and good number of rotor magnets seems best so far, but I need to go deeper. Slotless motors have just few coils and few rotor poles but achieve good results with feedback . ..than again feedback is no option in belt drive....... please shut me for this if this is wrong... in my world I cant find enough time to allow stable PID feedback between platter and motor distanced by belt and 2 pulleys....
For me as so far Phillips VCR capstan with many stator poles , dummy slots and good number of rotor magnets seems best so far, but I need to go deeper. Slotless motors have just few coils and few rotor poles but achieve good results with feedback . ..than again feedback is no option in belt drive....... please shut me for this if this is wrong... in my world I cant find enough time to allow stable PID feedback between platter and motor distanced by belt and 2 pulleys....
Finally I managed to finish rebuild of almost complete TT. Don't thing it is best in the world, but at least in my opinion it is one of the prettiest, at least for me who like understated retro design loaded with contemporary features:
It is still on my work bench. Need to clean and adjust tonearm, and play with preamp again. Also still looking for shaft encoder to properly measure W&F.
Also I envisioned cover, made from commonly available poly-carbonate sheet that I bent with help of steel pipes and heat gun. Cover is still in design phase, but will come in soon:
And back side:
To the right is original Amphenol 7 pin connector that I needed as it was feeding mains power + 3phase 160V in my original set up. I kept it (even it is overkill, some DIN connector would do here) just to avoid drilling new holes. Now it feeds 12V from external HDD PS (12V - 1.5A) which allows me not to connect safety ground from this side. On the pieces of cable (as I have many pins ) I connected Supaspin Menu button (not needed daily) and 2 diagnostics: 2 wires to measure system current via 1 ohm resistor and secondary tach sensor for which I need to develop precise 360 point strobe.
To the left is my preamp "drawer" that was shown before, with all ever needed connectors.
It is still on my work bench. Need to clean and adjust tonearm, and play with preamp again. Also still looking for shaft encoder to properly measure W&F.
Also I envisioned cover, made from commonly available poly-carbonate sheet that I bent with help of steel pipes and heat gun. Cover is still in design phase, but will come in soon:
And back side:
To the right is original Amphenol 7 pin connector that I needed as it was feeding mains power + 3phase 160V in my original set up. I kept it (even it is overkill, some DIN connector would do here) just to avoid drilling new holes. Now it feeds 12V from external HDD PS (12V - 1.5A) which allows me not to connect safety ground from this side. On the pieces of cable (as I have many pins ) I connected Supaspin Menu button (not needed daily) and 2 diagnostics: 2 wires to measure system current via 1 ohm resistor and secondary tach sensor for which I need to develop precise 360 point strobe.
To the left is my preamp "drawer" that was shown before, with all ever needed connectors.
And few pics from inside of the build:
Housing is again layered in copper. I spend some time to figure how to effectively laminate it and decided on epoxy + staples. Not nicely nailed staples I removed, and nicely nailed ones I left. This is method for building wooden yachts by west system, how to pressurize not flat surfaces so that glue sticks. It is not so neat but it is inside, and not visible, function matters here.
All separate pieces of copper are electrically connected by pieces of mesh shield from some cable and soldered with 60W iron to the copper. With metal enclosure this would be no issue, but building wooden enclosure that should behave like metal is bit demanding.
Rubber pads you see all over are not for some additional damping, but just to adjust height of the plinth and motor to where I liked it. I left bigger piece of rubber under preamp board and that is just to avoid accidental short circuit.
On left back, you see input connector followed by Supaspin with attached power amp. Tested this for several hours, copper laying is more than enough of heat sink for power amp.
Motor is left front, in silicone damped housing on its own, I showed earlier in the thread.
In the middle is connector for tach sensors mounted in the plinth.
On right - back is still unpopulated preamp board.
Detail of ground connecting copper pieces:
And this is plinth from underneath:
Cooper layer on plinth I glued with double sided glue tape, much easier here as it is flat.
Bearing is also grounded heavily. I checked assembled unit and all metal parts are on same potential. Even spindle is at ground through ball trust bearing.
This is more to remove static potential from LP than for shielding, even it does have RFI shield purpose.
You can see cable connector from plinth , that is for 2 tach sensors.
Housing is again layered in copper. I spend some time to figure how to effectively laminate it and decided on epoxy + staples. Not nicely nailed staples I removed, and nicely nailed ones I left. This is method for building wooden yachts by west system, how to pressurize not flat surfaces so that glue sticks. It is not so neat but it is inside, and not visible, function matters here.
All separate pieces of copper are electrically connected by pieces of mesh shield from some cable and soldered with 60W iron to the copper. With metal enclosure this would be no issue, but building wooden enclosure that should behave like metal is bit demanding.
Rubber pads you see all over are not for some additional damping, but just to adjust height of the plinth and motor to where I liked it. I left bigger piece of rubber under preamp board and that is just to avoid accidental short circuit.
On left back, you see input connector followed by Supaspin with attached power amp. Tested this for several hours, copper laying is more than enough of heat sink for power amp.
Motor is left front, in silicone damped housing on its own, I showed earlier in the thread.
In the middle is connector for tach sensors mounted in the plinth.
On right - back is still unpopulated preamp board.
Detail of ground connecting copper pieces:
And this is plinth from underneath:
Cooper layer on plinth I glued with double sided glue tape, much easier here as it is flat.
Bearing is also grounded heavily. I checked assembled unit and all metal parts are on same potential. Even spindle is at ground through ball trust bearing.
This is more to remove static potential from LP than for shielding, even it does have RFI shield purpose.
You can see cable connector from plinth , that is for 2 tach sensors.
One take on bearing oil. I tried again with common synthetic engine oil. Did not work nicely.
Returned back to motocross drive chain oil: it is rather thick oil but it has solvent in it that soon evaporates. My bearing design is sub optimal in a way that there is no easy predictable oil flow, both bronze bushing and SS shaft are flat over whole lenght, without oil groove. (bushing honed and shaft polished).
To overcome initial design shortcoming, I fill the bushing with motocross oil just to cover trust ball. This ensures that oil will flow up when shaft is inserted , unavoidably leaking some oil on top of the plinth. There I have small reservoir to collect that surplus oil and prevent it leak further.
And here is view of top of the plinth, around bushing you can see my messy mistake done when originally building this. It is described in post #3 when I needed to hammer out half sealed bushing out of the plinth.
This damage doesn't affect plinth performance, but left over depression is becoming reservoir for surplus bearing oil.
Here you can see flat mounted two CNY70 optical sensors for tach.
Returned back to motocross drive chain oil: it is rather thick oil but it has solvent in it that soon evaporates. My bearing design is sub optimal in a way that there is no easy predictable oil flow, both bronze bushing and SS shaft are flat over whole lenght, without oil groove. (bushing honed and shaft polished).
To overcome initial design shortcoming, I fill the bushing with motocross oil just to cover trust ball. This ensures that oil will flow up when shaft is inserted , unavoidably leaking some oil on top of the plinth. There I have small reservoir to collect that surplus oil and prevent it leak further.
And here is view of top of the plinth, around bushing you can see my messy mistake done when originally building this. It is described in post #3 when I needed to hammer out half sealed bushing out of the plinth.
This damage doesn't affect plinth performance, but left over depression is becoming reservoir for surplus bearing oil.
Here you can see flat mounted two CNY70 optical sensors for tach.
Further comment on limestone from Brač island. I forgot how soft and homogenus this material is. Needed to drill some more holes in it, it works with steel bit and with wood bit, so easy and soft. I also extended arm mounting hole with just a wood router. This is bit harder than working in MDF, might be like Corean, but with much much higher specific mass and density.
Dirt accumulated over years I just sanded down with 240 grade sand paper, than resealed with stone impregnation. Very easy.
Finally, have no enough words to praise @richb for Supaspin project, this made it all super easy possible!
Dirt accumulated over years I just sanded down with 240 grade sand paper, than resealed with stone impregnation. Very easy.
Finally, have no enough words to praise @richb for Supaspin project, this made it all super easy possible!
One more thought on mettalizing otherwise non conductive surfaces. I obviously used a lot of copper sheet, but now seems new type of conductive paints are avaliable, particularly made for RFI shield. Once I will try some, if that is any good, it would reduce labor dramatically.
A question, brainstorming quest about home brew shaft encoder:
If I would make very precise strobe disc with 360 points and precisely fixed it underneath platter, resulting signal would be very useful to measure speed oscillation live. Software still needs to be developed that would with high clock rate measure time between pulses.
But first things first: anyone with suggestion how to make really precise 360 strobe disc underneath platter?
- above is shown secondary cny70 sensor installed in plinth and looking very closely at perfectly flat white stone platter from underneath
- in post #160 I tested this sensor and figured that stable resolution can be with 1.5 mm black strip, plus white strip it's together 3mm for one pulse. Given diameter of platter than allows for 360 pulses per revolution, or 200hz @33.33 rpm.
If I would make very precise strobe disc with 360 points and precisely fixed it underneath platter, resulting signal would be very useful to measure speed oscillation live. Software still needs to be developed that would with high clock rate measure time between pulses.
But first things first: anyone with suggestion how to make really precise 360 strobe disc underneath platter?
I did some trials, printers don't really like angled lines, always got bit of zig zag effect under magnifier... You think it's worth putting effort?
Than I also need to decide paper quality as it's crutial
Than I also need to decide paper quality as it's crutial
