Syncronous Motor Does Not Spin At Rated Speed.

Can you give a link to the datasheet for this motor?

Yup Here is the data sheet, it's a PDF

https://myaccount.dropsend.com/storage/download?file_ids[]=2274480

Followed link - had to register to see it - broke one of my rules and registered - got:

"'There was a problem with this page - We're sorry!'

We're currently working hard to fix the problem and DropSend should be working again very soon."

Here ya go I just up loaded it to a different site they this one,

http://www.mediafire.com/?8cjdnlzct4uogye

You don't say what 60Hz motor you are using at the moment. Do they not also do a 50Hz motor? The chinese gearmotor you are trying probably does not have the gear ratios bang on for precisely 1rpm. That sort of precision is often not required, except for clock-motors, so their 1rpm could be approximate, and all the motors you get from them will probably be the same. This type of synchronous motor will only run at its' synchronous speed, so it either runs at the correct speed or not at all. The actual motor speed could be somewhere between 250-600rpm, depending on frequency and the number of poles, but if you want exactly 1rpm they need to get the gear ratios of the gearbox spot on.

Hansen claim to do various voltages, frquencies and speeds. Or Crouzet might be worth trying. Or Rotalink . There used to be a lot of companies making AC synchronous gearmotors, so you should find a lot more if you do a search (unless they've all gone to China!)

You don't say what 60Hz motor you are using at the moment. Do they not also do a 50Hz motor?

The 110 Vac 60 Hz 1 RPM motor is made by Kingston Motors.

And No they do not make a 220 volt 50Hz version.

I wish but nope.

I am practicing my engineering skills by assuming I already know everything and that the details of the customers problems are below me to the level of not being necessary to concern myself with.

Given this example I think I am ready to start charging to least $100 an hour now as an engineering consultant.

(Really I just saw pages of text with pictures of clocks and 50 Hz over 60 Hz with 20% too fast was someplace in there so I jumped to conclusions. I am allowed to have these sort of days too you know.)

If the two clocks ran precisely 20% too fast on 60 Hz, then either motor would be correct for your application. Why not order a bunch of them?

Think it's a case of finding a drop-in replacement (or as near as possible) for the 60Hz job.

The motors would have to be ripped out of one or other clock, and the original design would have to be adapted to accept them (if they were sufficiently powerful - I remember following a link to the OP's site some time back, and at least one of the clocks was damn big!).

Only the standard wall clocks run correct as expected on either 50 or 60 Hz,,

Read the original posting very closely,,

thanks.

On the contrary, you must read my post 8 correctly. You have said twice now that you have two clock motors that run precisely 20% faster on 60 Hz, matching your spreadsheet. Get a bunch more of them, not the original motor that was off.

I Copy''

"So As a better way to compare your motor I ask some of my contacts of stores overseas that want to buy my clock once I'm ready and ask them to send me a actual clock that is made for 220 volt and 50 cycle power.

And two arrive. They look like this.

(Pictures)

First test I did, was to run those two clocks that are made for 220 volt 50 Hz power, but run them on my 220 volt 60 Hz power.

Again I started at 9 AM as before Local and on those two clocks.

They matched the EXCEL sheet PERFECTLY! For Days even they matched perfectly at the expected 20% too fast. So our theory is confirmed. It will run, but just 20% Fast."

It does not say "Motors" It is talking about the two clocks that stores sent to me to use for a control / comparison test. And they are getting very hard to find. People just do not make clocks like this anymore anyway. They all are battery operated and quartz referenced. out of well over 100 people and stores only two were able to find these and they were in basements or yard sales. NONE were for sale in a store.

In addition NONE would be able to run my clock anyway even if I could buy a million of them for 3 cents.

This Crouzet motor may be a bit too pricey - but it looks like it would do the job.

http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=2044800

Would maybe need an adapter plate to mount it and a shaft adaption, and would unfortunately need an external capacitor to set the direction of rotation.

Agree,

They are a bit pricey, But I could just either absorb the extra price, Or the export clocks would just have a different price also. But the main problem with this motor is yeah instead of the ratchet option to make sure it spins the correct way. It uses the cap method, and that is just not cosmetically an option, in addition the extra umm encloseure? to make it voltage safe also would add a lot to the cost.

Thanks.

It might be worth checking on this in case other gearmotors are available. It's been a while, but I'm sure the motors we used were unidirectional and did not require phasing capacitors. I'm pretty sure the SaiaBurgess motors are available as unidirectional.

I checked with Farnell - they have 2 uni-directional versions:

ACW: http://uk.farnell.com/crouzet/82344771/motor-geared-230vac-1rpm/dp/3094200

CW: http://uk.farnell.com/crouzet/82344778/motor-geared-230vac-1rpm/dp/3094250

Even more expensive, but no cap. needed.

I used to use these motors with a previous company, ten or so years ago. To give a rough idea on pricing, we could buy the chinese gearmotors for about $5, and I don't think Crouzet were any more than $10, based on about 10,000/yr. It shouldn't be necessary to pay more than about $15-$20 for a good motor, depending on quantities, which is probably a small percentage of the overall cost of the clock.

Another one worth checking out is: http://www.saia-burgessusa.com/motors/ (Saia was a Swiss manufacturer, I have one of their synchronous motors in my turntable.

The motor the OP is talking about does not run exactly 20% too fast. It runs slightly more than 20% at 60hz. This is where his problem is.

Please reread post 8 correctly; and then read post 16.

Right 16 is correct,

Sorry, the clocks are Clocks not Motors. well yes they have motors in them but they are the reference to compare the questionable motor against. read the other post about using them Not possible

This answer is NOT OFF TOPIC He has it Dead on Correct.

If the gear train includes a 32 tooth gear where a 33 was expected, it will run at 33/32 times expected speed. In your 72 minute text period, that would advance 33 / 32 * 72 minutes, being 74.25 minutes. The display would advance from 9:00 to 10:14:15

If you dismantle the geartrain of one motor you can see precisely what ratios the maker has equipped it with.

Mark Bingham
Relativity PL

I just contacted them to see if they are willing to tell me these deails. numbers of poles and the complete gear train teeth counts and layouts. will see if they are.

Joe

If you are dealing with China direct, they are likely still trying to understand your letter.

Long letters tend to go into the too hard basket.

I suggest you make a test report in table form

"Test of sample 'such and such'"

Use common symbols, terms and numbers.

Give the same format table for specs - put a ±% tolerance on rpm, and on Hz.

Ask for a new sample to specs.

Send a picture of a clock*. 'motor is for clock'

*^{_{(of one you don't care if it gets copied)}}

Bear in mind, they may not have had 50Hz when they assembled and tested it, or even a way to measure rpm accurately. Doesn't mean they can't repeat it, once they get it right.

It seems to be a gear problem.

34point5 is probably right, the Chinese might struggle to understand your letter. It might be better just to tell them you need exactly 1rpm, whilst their gearmotor is approx 1.032rpm. But I disagree with asking them for a gearmotor to spec. If they had the correct gearset for 1rpm at 50Hz they would be using it already, and it would most likely be very difficult to get them to make the correct gearset for you. The output speed of the synchronous motor itself is probably 250rpm (assuming it is a 12-pole motor) so you need a gear ratio of 250:1, whereas they are using about 242:1.

If you insist on using a Chinese gearmotor, you might be be better off going through one of the many US based dealers/distributors that handle Chinese motors (Anaheim Automation is one that comes to mind, but there are many such companies). It might cost a little more, but at least you should get what you want. I suspect the gearmotor accounts for only a small proportion of the overall cost of the clock, so it might be worth paying a little more.

I agree with Tornado, open up the two 220v/50Hz clocks you've received and see what gearmotor they are using (unless their system is quite different). If the speed and interface are suitable, then use them. Otherwise try some reputable 'western' companies such as the aforementioned Hansen, Crouzet or Rotalink.

Looking at your website, it appears you just bolt the gearmotor on the back of the clock, so it should be easy to adapt for different dimensions. In fact there are only two arrangements for these gearmotors which have become pretty standard, the concentric gearbox such as you are currently using, and the ovoid gearbox which has slightly different mounting lug centres and a more offset output shaft. Most European manufacturers favour the latter.

There are many such clocks available in the market with 1.5 volts pencil battery. You dont have to bother about supply voltage and frequency. These batteries last for min. 7/8 months.

"...many such clocks..." I wouldn't say so.

DC is fine for running quartz movements, but small DC motors show a huge variation in speed, and the speed varies under load, and as the battery wears down. You would have to have a very good (expensive) controller to run such a mechanical clock with a DC motor.

The beauty of using AC synchronous motors for clocks (or timers) is the simplicity, and the fact that the speed does not vary as long as the supply frequency does not vary, which I believe is almost always the case.

I have 5 battery driven clocks and one A.C mains driven digital clock.Some times due to fluctuation in mains supply clock starts giving flickering and has to be reset. While battery driven completely stop functioning till battery is replaced. However if you are in need of accurate time then may be syncronous motor is more reliable.

Yes they are made on a license on our patent.

But They are Crap, I have to admit it. And I don't really care they know they are crap, yes they last 7 to 8 months, thats about the life time of the clock itself too. They started originally with an Intermatic Motor as the heart of the little cheap plastic one.

And when they wanted to open to the rest of the world they went the battery option. but even with the 1/2 size ball and shorter arm and 10 times or more of a less in/oz demand those motors die withinn a year the gears get stripped out.

There is a pretty good thriving undermarket of people selling replacement gear parts to repair them with a metal gear final pairing.

But this "Fix" costs twice the original whole clock.

The Plastic served it's purpose. It made the clock available to the masses at a cheap price. But you get what you pay for.

Yes they make them by a license on our patent. But we know they made/ make them as cheaply as possible. They told us once they have less than a buck in costs involved in them.

But we think of the comparison they even said. Theirs is a Quarter Pounder with Cheese, and Mine is a nice sirloin at a 5 star steakhouse.

They are both Beef, but they are not the same.

I had different experience. I have been using these clocks since 8/10 years without any problems. One of them is around 25 years old. I had to only change the batteries. In 2/3 clocks due to dust accumalation I opened the movement mechanism, after cleaning I reassembled the unit it started ticking again. There was hardly any wear on plastic gears. As far as price is concerned you can buy complete movement in India for $2/-. Complete clock is sold @ $5/-. It is really value for money.

Link is http://www.indiamart.com/ajantaassociates/wall-clocks.html

I have read through all of the content on I think, but, apologies if I have missed something. Surely the problem is the difference in the diameter, therefore the circumference of the shaft? A given point on the circumference of a slightly larger shaft will travel through a greater arc. As I say, apologies if you have already taken this into account, but just in case.

Regards

Tom

No, the arm at the back of the clock that lifts each ball is just fastened to the gearmotor shaft, so that one ball is lifted each minute. The shaft diameter makes no difference. Have a look at the video:

http://www.idle-tyme.com/watch.html

Fascinating.

I really enjoyed reading the story on your website - thanks. I am sure I have seen the rolling ball clocks at Paddington Markets in Sydney Australia many years ago - fascinating to watch - maybe it was a copy ?

There have been some excellent suggestions on this site and I would think that if you persist with inquiries you will eventually find a suitable motor but just in case you fail I have a simple mechanical suggestion that may solve the problem.

Synchronous motors have been manufactured internationally by many manufacturers for numerous applications. These include rotisserie motors, micro wave oven turntable drives, record player drives using sleeves or rubber jockey wheels of different sizes for different frequencies, refrigeration defrost timers, dishwasher, dryer and washing machine timers etc. Some of these applications have been supplanted by electronic timers but there are still a few machines which use synchronous motors / timers - they must be very cheap if you can buy a microwave oven for as little as $40.00 retail !

I have just had a look at some old appliance timers and noticed one where the motor mounting position may be altered for frequency. The motor has two mounting holes and screws - it pivots on / at one and may be secured via the other hole and screw into either of two positions marked 50hz and 60hz on the washing machine timer, whereby the timer motor assemblies output gear meshes further or lesser with the timer input gear for adjustable / variable ratio within limits - this simple approach may provide the difference you need although it may require you providing an additional gear and mounting which you would probably prefer to avoid.

Stuart Fox, Bondi Beach, Australia

Fiber (Italy) also does a unidirectional gearmotor. They are probably cheaper than Crouzet. I'm sure they have a US distributor, otherwise you'll find them in Direct Industry.

http://www.fiber.it/sito2010en/default.asp

If you dismantle the problem motor and count the poles, then using the formular detailed on wiki at:-

http://en.wikipedia.org/wiki/Synchronous_motor

The formular is:-

-----------------------------------------------------------------

The "synchronous speed" of a synchronous motor is determined by the following formula:

where v is the speed of the rotor (in rpm), f is the frequency of the AC supply (in Hz) and n is the number of magnetic poles.

-----------------------------------------------------------------

That way you can first find out how fast the motor turns at 50hz (or 60!) and then count the teeth on each gear to tell you what the resulting shaft speed will be........

I hope it helps!

Thanks,

I started to do this, but got to the point of WOW the first three gears the teeth I need better eyes for! they are tiny is an understatement!

The magnets on the motor itself are stronger than I expected. you can turn the motor but it really wants to settle in with the magnets..

You need to make a scale for say 10 teeth (for each different sized cog) on a piece of paper.

Mark the first tooth, using the scale count how many tens there are, add the rest......

A magnifying glass that is held in a stand, or used like spectacles could be a further help too!!

OK,

Here we go gang, I opened that motor, and below is the gear train.

Now I'm not sure how many "Poles" the actual "Motor" is. But when you turn it you can feel the magnetic poles trying to seat it in certain locations. one complete rev of the motor has 10 of these ummm holds? anyway where it wants to rest there are ten of them.

Now I don't know for sure if my math is correct or not, but with that gear train did I calculate correctly tht the geartrain is a 580.740740740....... to 1 ratio?

Heres the geartrain,

Unless my math is wrong it's the freaking geartrain!!!!

120 X 50 Hz = 6000

6000 divided by 10 poles = 600

600 divided by Gear train ratio, 580.740740740 = 1.03316 lookey there

take that times 60 seconds = 61.989795918446415556018426550454

almost 2 seconds FAST!!!

I always knew that the chinese wrenches were junk, but can this be this simple? Am I doing that corretly?

Joe

My calc matches yours, and it explains the ~2 min/hr overspeed. They really need an aggregate 600:1 reduction.

You probably want something like this - [guessing at the stacking physicals]

But teeth can be any number to suit centers, but must yield the nominated ratio

14 x 7 = 2:1

27 x 9 = 3:1

28 x 7 = 4:1

40 x 8 = 5:1

50 x 10 = 5:1

Total 600:1

Yes, it is that simple, and I think we established many posts back (see #2 and #20) that the gear ratios were wrong. The chinese manufacturer might not even consider it a problem, since for the vast majority of applications for this gearmotor it wouldn't matter if the speed is 3% out. You could ask them to supply a gearmotor with the correct gears for precisely 1rpm, but I wouldn't get my hopes up. So what do you do next?

Just sent the report to them to see what they say, I love their test results say it was accurate to their coronograph at 1 rpm.

Boy I hope they dont need that for sailing the ocean.

Maybe their local grid runs chronically slow at about 48.7 Hz! (Pun intended.)

Andy,

Can you tell me please where the 120 part of the formula comes from?

Joe

I just looked it up, my first theory was wrong, completely!! So I am correcting this post........sorry I cannot give you a fully correct answer...I simply don't know, but guessing it is something to do with minutes x 2, but that is only a guess.

Both half cycles actually cause movement of the motor due to pole changing magnetic polarity.....

Maybe you can find a proper explanation here:-

http://www.brighthub.com/engineering/electrical/articles/42183.aspx

and here:-

http://en.wikipedia.org/wiki/Synchronous_motor

Might be off topic, but have you looked into "stepper motors"?

You may be able to get a motor with the right step number (and thus not need a controller) for the offshore frequency systems.

Maybe there "coronograph" is also synchronised to their local electrical mains frequency so they have a perfect match to an incorrect baseline.

Yes I thought about steppers as a replacement for all the places.

But unless I was looking in the wrong place no matter what I tried it was extremely cost prohibitive.

But a good thought.

Joe

You still need a controller to drive the stepper motor, you cannot drive it from just the mains sine wave.

But it could be a PIC with a crystal controlled frequency for accuracy (with a trimmer capacitor for fine tuning of the time), a few transistors and a small program......if the mains was used to set the timing it could be made as accurate as a synchronous motor, which is as accurate as the mains......

It would be even simpler if the stepper motor had multiples of 60 as a step.....angle would need to be 6°, 3°, 1.5° or 0.75°.....which are values I personally have not seen that often........like never! But they might still be available somewhere.....

There are two ways to address this problem (assuming for the moment that only minutes are needed, not seconds, but the method is similar for both!):-

1) Attach gears to give the 6° per step(s). This would allow almost any cheap easily available motor to be used. Cheap small motors are generally 48 steps or 7.5° per step.

or

2) Make up a "correction" table that specifies the nearest and best position to select for each minute using either a 1.8° motor or better for best accuracy..

For example, from 12 o'clock, the first minute 12:01 or 6° from 12 o'clock, would be at 3 x 1.8° = 5.4° (0.6° less than optimum, nobody would notice).

12:02 should be 12°, would be 12.6°. 1.8° x 7, also un-noticeable.

Using 0.9° motors it would be even more accurate angle wise. For example 12:01 would be 6.3° (7 x 0.9°). Only 0.3° too much.

12:02 would be 11.7° (13 x 0.9°), only 0.3° less than the correct angle.

Further thoughts:-

If micro-stepping was used, the accuracy could be even higher, even with very cheap 7.5° motors that you can buy for a few Dollars each........microstepping is very easy to implement with a PIC.....

Achieving 1/2, 1/4, 1/8 or 1/16th steps is really, really simple......

Thats a great way to go with stepper motors.....easy and accurate.....plus there are steppers for any power needed, just pick your motor......and the drivers to match.....nice little project too, very interesting.

Reading up on the web will supply everything you need to achieve this. If someone was paying me, I would do it for you........

"For example, from 12 o'clock, the first minute 12:01 or 6° from 12 o'clock, would be at 3 x 1.8° = 5.4°"

What are you going on about, Andy?

The requirement is for a motor that has an output shaft which does one full rotation per minute. Did you read the OP, or follow the link to Idle-Tyme?

Jon, I was theorising on a possible stepper motor solution. Sadly, steppers using clock type steps 6°, 3° or even 1.5° are not easy to find, but 7.5° stepper motors are everywhere very cheap and easy to interface.....

I was of the opinion you knew your way round steppers......

Do you need any further infos from me......?

Yes, stepper motors or even brushless dc motors could do a good job, but are very expensive compared to AC synchronous, especially since you also have to consider the power supply and motor controller. We know that the AC synchronous gearmotor is a good, simple, effective, low-cost solution, and we know that it works, so why bother looking at other technologies, just look for other suppliers.

If the chinese don't know they have the wrong gear ratios then you'd best steer clear of them and their incompetence. As we've seen, the calculations are very simple, and they should know what gears they are using. And you can't blame it on the local supply, a motor manufacturer should be able to test their motors on any supply. The last few companies I've worked for (pump manufacturers) certainly could.

I have a suspicion that they could just be cobbling together gears that they have available from other gearmotors that give a result they deem 'close enough' so they don't have to manufacture the correct ones. Since you've taken the trouble to dismantle their motor and count the gear teeth, you should show this to them, along with the calculations. Then ask them if they can supply a true 1rpm 50Hz gearmotor. If they say yes, get a sample. If they say no, look elsewhere.

Good post. GA.

I don't think that a stepper solution would be that expensive, as I mentioned in my previous post the really cheap motors (48 steps 7.5° per step) are available for a few dollars, the PIC would cost far less than $10 and be programmable in Basic if it was bought from Rev-Ed or their suppliers. All the software needed is fre from here:-

http://www.rev-ed.co.uk/picaxe/

Synchronous motors without gears are around a dollar here, for example:- 600RPM 50Hz., with no gearbox..

A complete controller, stepper driver and motor could be built for around $20-$30 in quantities of 25 or more.....maybe even less in a third world country.....

Make the clock a bit more expensive......

An accurate synchronous motor would be cheaper though I feel.....if you can find one.

The stepper solution would work in any country with any frequency, even on DC if the mains was not used for timing......

I sent this to them last night,, and below this is their reply,

Here is what I sent to them,

Sara, & Sam,,,

I have found out where the problem is. And who ever is doing the testing is Lying to you all. And maybe needs to be fired? Sorry to be so brunt about this, but who ever designed the motor and is telling you all that this is indeed a 1 rpm motor is totally wrong.

Since what they were telling you was not matching what we were seeing here, I had to break it down and dissect a motor to see what is really going on in this thing. And it became very clear as to why this is NOT a 1rpm motor. Yes it's close. But it is not and here is how this was discovered buy a dozen engineers and myself today.

The final out put speed of this "Gear Motor" of course all depends on how fast the actual motor is spinning AND the gearbox.

I found absolute proof of why it's running fast, and it's irrefutable.

matter of fact using math to calculate it. It shows to no doubt whatsoever that this motor will indeed run faster than 1 rpm and has no chance of doing anything but that.

I will explain below.

First lets determine how fast the actual motor is spinning.

This is simple. All we need to do is see how many poles the motor is made with, and insert that into the formula below.

Now lets calculate the speed of the motor itself, to do that we use this formula, The "synchronous speed" of a synchronous motor is determined by the following formula:
120 * f / n = rpm
f is the frequency of the AC supply (in Hz) and n is the number of magnetic poles.

OK dissecting the motor is it learned this is a ten pole motor. So we have the numbers needed ten poles and a power frequency of 50 Hz.

So lets run the numbers. 120 Times 50 Hz = 6000 divided by the 10 poles = 600 RPM.

600 is logical, makes total sense. a easy ratios to make for minutes, seconds, hours even, so this is a good starting point, Now lets see the total reduction ratio of the gear box.

Output Speed Calculation, Going Backwards From Output Shaft To The Motor Itself.

I see this table gets messed up in the formatting here at CR-4, But it's a table from a spreadsheet that is broken down like my graph paper gear train, but just numbers,

Driven Teeth
Driver Teeth Ratio Final Output Ratio.

Output Shaft 24 8 3 to 1 3 to 1

2nd stage 28 10 2.8 to 1 8.4 to 1

3rd Stage 28 9 3.111111 to 1 26.133333 to 1

4th Stage 40 9 4.444444 to 1 116.148146 to 1

5th Stage 50 10 5 to 1 580.740733 to 1

I think you can see that already we have a problem if the final output speed is to be 1 rpm, the final reduction ratio should have been 600 to 1 not 580.740733 to 1 600 rpm motor divided by the reduction ratio of 580.740733 to 1 = 1.0331632790 RPM. HMMMM? That rpm in the terms of seconds = 61.989796744634408139578526860454 seconds! Does this look familiar? 2 seconds too fast per minute! Did they install the wrong gearbox on this motor possibly? The 600 Rpm of the electrical part of the motor sounds like it's a good choice. But the gear box is wrong to make a 1 rpm output.

Joe

And they replied with this morning,,,

Dear Joe,

Good day. Thank you for having provided so much infomation.You are cool and professional indeed.

I've discussed it with other engineers.We'll reply to you ASAP.

We'll have to see what happens. I'll Post updates as they happen.

But thank you All for all the help. This group is just plain fantastic!

Joe

Cool!

By the way, just noticed, 10 pole is not unusual, but 12 and 24 pole are more common. If your Kingston 60Hz motor is 12 pole, the motor speed will also be 600rpm (ie it will have the geartrain your chinese motor ought to have).

While we all wait for their reply, here some sideline info that makes ya go HMMMM?

OK as stated we sold a license on our patent to the company to make the cheap plastic version. And they started with a cheap motor made by intermatic. It was a 110 volt 60Hz motor as we have been using just smaller, weaker, and less rugged.

But then they also wanted to go for the foriegn markets and ran into a similar problem. So they went into a battery powered version. But it is NOT like you would think. Steppers etc. Nope.

Their arm rests in one location till a ball is needed and then quickly it scoops one up and deposits it where needed the complete rotation is only like 5 seconds or so then it stops and waits.

They added a ticking second display so there is still something to watch, a video of the motion can be seen here, check it out first before going on in this message,

http://www.youtube.com/watch?v=Mcyb3Ypys9o

In the cheap plastic clock to get around the problem of different parts of the worlds having different power and freq of power they made a battery operated version it ran on like 4 "D" sized batteries.

OK, Now it's pretty clear how the arm moves it a simple high speed DC motor and a small gear train as seen here,

And the Motor and Gear Train,

But what amazes me is the board that controlls it,

I see 6 resistors,
1 diode,
2 caps
and what 4 transistors,

How did they from a dc source of 4 "D" sized batteries ever keep accurate time with something this simple!

And it did keep accurate time.

Joe

I suspect that there's a DC 'quartz' type movement up the top - the thing that drives the minute wheel and actually does the timekeeping - and the circuit in your photo (which must if I'm right have an input from the quartz movement) just runs the lift motor. How else would the minute wheel move?

I may have to buy one just to see how they do it.

Great link, thanks.

It appears the the ball for 1 O'Clock is glued in place and never moves!!!! Nor does it ever need to move either!!!