A prototype Lavet stepper motor

[Tony JeffreeParticipant @tonyjeffree56510]

[Tony JeffreeParticipant @tonyjeffree56510]

See Youtube video here:

The motor is being driven in parallel with the motion work for my Free Pendulum clock. This was a feasibility exercise with the aim of building motion work for the clock rather than using the existing re-purposed Quartz clock movement. Obvious problem is that it is very noisy, which is mostly a consequence of using a small but very strong permanent magnet; I may have to find a source of smaller/weaker ones.

[duncan webster 1Participant @duncanwebster1]

Excellent, are you going to write it up? The video seems to show it doing 3/4 turn followed by 1/4 turn, is this an optical illusion? I have a Gents silent slave with a Lavet motor, but it does 1/4 turn per pulse, which helps with the gearing

[Tony JeffreeParticipant @tonyjeffree56510]

Thanks! Yes, it's an optical illusion – the blob of white tac is actually slipping on the shaft because the motor's acceleration from rest is quite high. I will be writing it up in due course, but it is early days!

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Tony Jeffree on 24/03/2021 20:03:03:

Thanks! Yes, it's an optical illusion – the blob of white tac is actually slipping on the shaft because the motor's acceleration from rest is quite high. I will be writing it up in due course, but it is early days!

.

That’s a relief … I thought my eyes were going funny again

The write-up will be most welcome.

MichaelG.

[Tony JeffreeParticipant @tonyjeffree56510]

Fooled me to start with too!

[Michael GilliganParticipant @michaelgilligan61133]

Whilst not wishing to steal your thunder, Tony …

For anyone who is unaware of the general design of Marius Lavet's motor :

This is worth watching : **LINK**

MichaelG.

.

Edit: __ See also : https://en.wikipedia.org/wiki/Lavet-type_stepping_motor

Edited By Michael Gilligan on 24/03/2021 22:18:25

[Michael GilliganParticipant @michaelgilligan61133]

PostScript

The collection of drawings in the patent is wonderful … but sadly let-down by the scanning

**LINK**

https://worldwide.espacenet.com/patent/search?q=pn%3DFR823395A

… The hunt is on for a better copy.

MichaelG.

[Tony JeffreeParticipant @tonyjeffree56510]

I've removed the white blob and marked the rotor with red/black pen to make the movement more obvious.

[Tony JeffreeParticipant @tonyjeffree56510]

Hard to see from the vid, but the necessary asymmetry in the magnetic circuit was achieved by making the two pole pieces moveable – they are offset slightly so the rotor's rest position is not straight across. The pole pieces started life as a 3/8" X 1/8" steel strip with an 8mm hole in it, held in slots in the ends of the armature side pieces. They are each offset by about 1/2mm to get the asymmetry. The shaft is pivoted in tiny ball races, but that is probably overkill.

[Tony JeffreeParticipant @tonyjeffree56510]

A nice feature of these motors is that you don't have to worry too much about end float in the rotor shaft – the PM in the rotor centres itself on the pole pieces and holds the rotor in place with significant force.

[John HaineParticipant @johnhaine32865]

Excellent Tony, I have been thinking about the same thing, you beat me to it! Do you use a short "bar" magnet in a brass (or ali) cylinder?

I've also been trying to find out how those clocks with apparently continuous motion work. I believe that they use the identical type of motor but drive them with trick pulse trains, that's the next challenge, should make them a lot quieter.

[John HaineParticipant @johnhaine32865]

I get magnets from these people:

**LINK**

Good range of small rod magnets, also have various materials and diametrically magnetised rods.

[Tony JeffreeParticipant @tonyjeffree56510]

Posted by John Haine on 25/03/2021 15:15:21:

Excellent Tony, I have been thinking about the same thing, you beat me to it! Do you use a short "bar" magnet in a brass (or ali) cylinder?

I've also been trying to find out how those clocks with apparently continuous motion work. I believe that they use the identical type of motor but drive them with trick pulse trains, that's the next challenge, should make them a lot quieter.

Thanks John!

It is a short cylindrical (but diametrically magnetised) magnet sandwiched between two brass cylinders of the same dimensions, with axial 1.5mm holes to take the steel shafts. Not the easiest way to go and a pig to align, even badly – but I have found some donut shaped magnets that will do the trick and will be easier to use. I've sourced some from Ebay, some from Amazon; the donut shaped ones are these:

LINK

Supplied by Magnet Experts – haven't been delivered yet though!

[Tony JeffreeParticipant @tonyjeffree56510]

Posted by John Haine on 25/03/2021 15:15:21:

Excellent Tony, I have been thinking about the same thing, you beat me to it! Do you use a short "bar" magnet in a brass (or ali) cylinder?

I've also been trying to find out how those clocks with apparently continuous motion work. I believe that they use the identical type of motor but drive them with trick pulse trains, that's the next challenge, should make them a lot quieter.

Thanks John!

It is a short cylindrical (but diametrically magnetised) magnet sandwiched between two brass cylinders of the same dimensions, with axial 1.5mm holes to take the steel shafts. Not the easiest way to go and a pig to align, even badly – but I have found some donut shaped magnets that will do the trick and will be easier to use. I've sourced some from Ebay, some from Amazon; the donut shaped ones are these:

LINK

Supplied by Magnet Experts – haven't been delivered yet though!

Yes, getting them to work quietly is an issue – even the tiny ones in the quartz movements generate a significant "tick". With the relatively large magnet I am using the tick is very loud indeed, especially when attached to a sounding board. The smaller ring magnets should be better from that point of view.

I'm wondering what happens if you simply drive the Lavet motor using a Sine wave…? Might be worth a go.

[Tony JeffreeParticipant @tonyjeffree56510]

Of course, I had to try it…I have a crude sig gen that will give a ~1 Hz sine wave and a cheap Class D amp, so this is the motor driven from the output of the amp:

LINK

Not very smooth, some of which may be down to the quality of the signal of course, but interesting!

Edited By Tony Jeffree on 25/03/2021 16:52:33

[duncan webster 1Participant @duncanwebster1]

Not easy to see without taking it apart but I suspect the gents has an axial magnet with steel pole pieces. This could be achieved using a number of small magnets rather than one polo mint. Can also give more than one pair of poles. A non magnetic axle through the middle would avoid short circuiting the field 

Edited By duncan webster on 25/03/2021 16:55:36

[Tony JeffreeParticipant @tonyjeffree56510]

Managed to delete one of the earlier videos – here it is again:

LINK

[Tony JeffreeParticipant @tonyjeffree56510]

Posted by duncan webster on 25/03/2021 16:53:33:

Not easy to see without taking it apart but I suspect the gents has an axial magnet with steel pole pieces. This could be achieved using a number of small magnets rather than one polo mint. Can also give more than one pair of poles. A non magnetic axle through the middle would avoid short circuiting the field

Edited By duncan webster on 25/03/2021 16:55:36

More poles would certainly help with smooth running of course.

[John HaineParticipant @johnhaine32865]

Nice to see it works on a sine wave too. I thought the Lavet motor was limited to 2 poles? You would need a more complex stator structure with more rotor poles. From what I can glean the "continuous motion" quartz clocks have a conventional motor but a clever drive waveform with multiple pulses of varying width – I'll try to dig out what little documentation I could find (mainly a patent).

[Tony JeffreeParticipant @tonyjeffree56510]

Looking at the patent drawings that Duncan posted, it seems that he built variants with more than 2 poles, but I haven't (yet) got my head round how that works in detail.

The continuous Quartz clocks could well be using a fairly conventional 2-phase bipolar stepper – if you apply sine waves (one phase-shifted) to the two phases you get a close approximation to continuous rotation. The tricky thing then (for a pendulum clock) is generating the sine waves at the exact frequency of the pendulum.

Edited By Tony Jeffree on 26/03/2021 10:04:47

[John HaineParticipant @johnhaine32865]

I'm pretty certain that they do use the standard Lavet pattern, as these are extremely cheap to make and standardised. This patent is the best I've been able to find that describes how they are driven, also this describes some current measurements that show a pattern of current pulses. The idea seems to be to give the coil a small kick or a few to get it moving and just nudge it over the dynamic equilibrium point, then pulse it in the opposite direction to slow its acceleration towards static equilibrium, then finally stop it briefly at that point so it doesn't overshoot.

You could use a standard stepper, using a processor to do the microstepping so it looks like it's doing equal seconds increments, but of course it takes more current.

[duncan webster 1Participant @duncanwebster1]

Here's a couple of pictures of the Gent's, definitely 4 pole and 1/4 rev per pulse. The white gear is driven by a worm on the motor shaft. It is very quiet

[Tony JeffreeParticipant @tonyjeffree56510]

Interesting! I suppose current consumption is an issue for a clock that you expect to run from an AA cell, but as I am using a 12V "wall wart" as the power source, using a conventional stepper motor with microstepping would definitely be an option.

[Tony JeffreeParticipant @tonyjeffree56510]

<duplicate post>

Edited By Tony Jeffree on 26/03/2021 11:34:59

[Author]

Posts