Everything you always wanted to know about the MIMUL motor* (*but were afraid to ask)

[Heikki Repo]

48 minutes ago, Duncan Brown said:

Is there anyone out there who can explain to me why a knob with 5 marked positions (75/50/N/12/8) is attached to a switch that has 6 positions?  I mean I guess it's way easier to buy a 6 position switch than a 5 position switch, but...   As near as I can tell, electrically, it's basically a position under "8" that does the same thing as "8".  Can anyone with one of these motors (that works!) verify that's how it behaves? 

My ACL heavy duty motor from around 1976 (possibly somewhat different compared to yours, but not the latest version) has six speeds, one unmarked after 75 fps: it's the inching speed. The settings are 25/8/12/50/75/inching.

[Duncan Brown]

Just now, Heikki Repo said:

My ACL heavy duty motor from around 1976 (possibly somewhat different compared to yours, but not the latest version) has six speeds, one unmarked after 75 fps: it's the inching speed. The settings are 25/8/12/50/75/inching.

What is an inching speed?  REallllllly slow?  Is that different than just using the knob to do it by hand?  So maybe that's what the sub-8 position is on this motor?

And is that really the order of positions on the knob?  Clearly the designer can attach the wires to the switch in whatever order they want, but in-numerical-order would seem to make more sense.

Duncan

[Heikki Repo]

45 minutes ago, Duncan Brown said:

What is an inching speed?  REallllllly slow?  Is that different than just using the knob to do it by hand?  So maybe that's what the sub-8 position is on this motor?

And is that really the order of positions on the knob?  Clearly the designer can attach the wires to the switch in whatever order they want, but in-numerical-order would seem to make more sense.

Duncan

Yes, really slow speed. Faster than just turning it by hand though ;)

And yes, that's the order on my motor. Doesn't seem logical, except from the point of view that 24/25 is the "normal" and crystal synced whereas other speeds are neither ... ? It's also easy to turn to the normal speed without looking, just by hand if you know which way to turn it.

[Duncan Brown]

27 minutes ago, Heikki Repo said:

Yes, really slow speed. Faster than just turning it by hand though ?

And yes, that's the order on my motor. Doesn't seem logical, except from the point of view that 24/25 is the "normal" and crystal synced whereas other speeds are neither ... ? It's also easy to turn to the normal speed without looking, just by hand if you know which way to turn it.

OK my camera seems to be from 1974, if yours is from 1976 and is NOT the later mirror-parking motor, perhaps they just made a mid-stream improvement by swapping the order of the wires on the switch and remarking the knob.  Easy enough.  If I ever get mine running correctly I'll see if my sub-8 position is actually inching.  I don't think so but again maybe that was another improvement on the later one, to take advantage of the otherwise wasted extra position.

Duncan

[Heikki Repo]

1 hour ago, Duncan Brown said:

OK my camera seems to be from 1974, if yours is from 1976 and is NOT the later mirror-parking motor, perhaps they just made a mid-stream improvement by swapping the order of the wires on the switch and remarking the knob.  Easy enough.  If I ever get mine running correctly I'll see if my sub-8 position is actually inching.  I don't think so but again maybe that was another improvement on the later one, to take advantage of the otherwise wasted extra position.

Duncan

I have mirror parking, but mine is not the ACL2 motor - i.e. the one that has the recessed inching knob.

[Gregg MacPherson]

6 hours ago, Heikki Repo said:

....24/25 is the "normal" and crystal synced whereas other speeds are neither ... ? 

I don't have the electronics knowledge to enjoy Duncan's free fall down the rabbit hole...I can remember how to spell Kirchoff (no, he cant). Great photos D, very trippy...

This issue about which motors are/are not crystal controlled at non sound speeds came up in messages recently. I did notice this in the French ACL 1.5 user manual, which shows the MIVAR motor, with mirror parking icon, exposed inching knob without a protective ring shape around it.

"Motor:
12 volts multiduty type, brushless, Hall effect,
6 crystal controlled speeds 8, 12, 24, 25
50, 75 frames per second. Automatic stop
in viewing position..."

Heikki, was crystal control at all speeds not extant before MIVAR? Was there a definitive source on that point?

Duncan, is "Hall effect" just referring to the physics used for motor position sensing? Don't distract yourself with a lengthly answer.

Cheers.

[Heikki Repo]

10 minutes ago, Gregg MacPherson said:

This issue about which motors are/are not crystal controlled at non sound speeds came up in messages recently. I did notice this in the French ACL 1.5 user manual, which shows the MIVAR motor, with mirror parking icon, exposed inching knob without a protective ring shape around it.

"Motor:
12 volts multiduty type, brushless, Hall effect,
6 crystal controlled speeds 8, 12, 24, 25
50, 75 frames per second. Automatic stop
in viewing position..."

Heikki, was crystal control at all speeds not extant before MIVAR? Was there a definitive source on that point?

Sorry, most likely I had picked that up somewhere online in the last past few years and when replying, didn't have time to go for the manual to verify. That's the reason for the question mark - I started doubting if that could be true, considering what Duncan had previously written of the electronics. Since I haven't shot high speed footage with my ACL under such conditions where the lack of crystal speed would become a problem, that erroneus idea of them not being crystal speeds had taken root ("surely my ACL can't be so good and modern as I previously thought").

[Gregg MacPherson]

1 hour ago, Gregg MacPherson said:

...was crystal control at all speeds not extant before MIVAR?...

I think Duncan or someone has said that the pre-mirror parking motors, his one being a MIMUL, had crystal only for 24/25 fps. He may have known that or it may have come clear in his digging...which would point to my not properly reading the thread.

[Duncan Brown]

Well the MIMUL motor speeds are all generated ultimately from a crystal oscillator, but the question is whether they all have a feedback system to maintain an utterly precise speed.  I do note that the switch inputs for all the other speeds control one circuit, while the N position of the switch take a completely different path.  (But that may just be because it needs to further determine whether it's 24 or 25fps).

A hall effect sensor is simply a transistor that is controlled by a magnetic field instead of a base current. It is a way to "see" if a magnet is present or not.  I'm guessing in this context it refers to the tachometer function of the motor - the stepper circuits drive the motor, while the hall effect circuits sense the resulting speed.  Just like the motor is driven by magnetic effect, without the need for contact brushes, so too is the sensing done in a contactless manner.  

In scoping around the circuit yesterday I came across a chip not behaving correctly.  Though it could also be the chip it is driving its signals into not behaving correctly.  I think I will have to replace a chip or two before proceeding.

That speed ramping circuit is COOL!  Though that whole daughterboard has the looks of an "oopsie daisy" -  a whole system designed and ready to go and then they find an issue that they have to develop a fix for and wedge in there somehow.  The connections to the logic board are just tacked on to IC pins, they are not given their own connection points like they would be if it had been part of the original design.  I'm imagining what happened is that they had the whole system finished, and it worked perfectly as they switched the knob from one position to the next, everything was great.  Then the first actual user they handed it to spun the knob from 8 to 75 in an instant and the motor locked up, or ran backwards or something,.  Oopise daisy!  So what happens when that circuit is switched on is that the input frequency, which had just been passed through unchanged to the output pin, drops to a much much lower frequency (half or less) then slowly, across a second or two, ramps back up to match the input frequency.. That's more consistent with how stepper motors like to see their speed changed.  I'll try to post a movie of that action on my oscilloscope later in the week when I get back to this project.

Duncan

[Aapo Lettinen]

in brushless motors one needs to use electronics to switch the windings on and off at exactly correct timing and motor position to make the motor run. A brushed motor uses a mechanical commutator to switch the windings on and of at correct timing but the brushless motor does not have any mechanical means to do this switching and thus it needs to be done electrically with some kind of sensor circuit monitoring the exact position of the rotor's magnetic poles at all times.

one way to sense the correct switching positions in a brushless motor is to install hall effect sensors ("magnetic field sensing sensors") inside it which sense when the rotor's permanent manget poles are at the expected position for the control electronics to know when to switch one winding off and another on and at which exact position the rotor is at any given moment. Typically a brushless motor using Hall effect sensors has one sensor for every phase of the motor, so a 3 phase motor typically has 3 hall effect sensors.

Some brushless motor driving circuits can sense the rotor position from the induced back emf current of the windings without using any kind of separate sensor inside the motor itself. This kind of "sensorless brushless motor" is mechanically simpler, a bit smaller and cheaper and easier to make but it needs a lot more complicated driving circuit.

 

It is possible to extract the motor's running speed from the same hall sensor signals used for driving the motor but that is just a "extra benefit" of the arrangement and the main reason the hall sensors are in the motor is to enable the switching to make it run in the first place

[Aapo Lettinen]

brushless motors have the main drawback of needing a extremely complicated driving circuit compared to brushed motors and it makes designing new control electronics for brushless motors pretty difficult unless one happens to find a suitable off the shelf circuit just by pure luck which even happens to fit inside the camera and which happens to be possible to control with the actual crystal sync electronics. often the driving circuit needs to be customised in many ways and off the shelf does not work in the application.

that is the reason why I have tried to avoid brushless designs whenever I can, they are just a huge pita to work with by my opinion and only worth it if making dozens of exactly similar devices

[Duncan Brown]

I got video of the speed ramping effect as seen by the scope, I'll try to post a link to that soon.

I took out the chip that was being driven by the seemingly misbehaving chip.  Still misbehaving.  (A counter, not having the correct frequencies coming out of its various divided-frequency pins.)  So I took out the misbehaving counter chip.  I have some of those coming, probably get here next week.  I'll put in the counter chip and see if its output then look correct.  If so I'll put back in the chip it's driving and see if things still all look more correct.  If so I might try temporarily reassembling the motor and seeing if it's working better.

The ACL cycles through one complete frame in precisely one revolution of the motor.  Now the question is, how many steps per revolution of this stepper motor?  This stepper doesn't have strong enough magnets for me to "feel" the steps when spinning it slowly by hand.  I'll obviously eventually figure out the answer by mapping out all the logic and doing the math, but it would be nice to know the answer from the motor side, as a sanity check.  I'm not ready to try to pull that motor from the casing and look for a model number.  SO many wires going to it through holes in the case.

Duncan

[Duncan Brown]

Here's the speed ramping circuit in action on a scope:

 

At rest it passes through the main 48KHz input signal to the output.  (My scope is on the output.)  When a 5V signal is applied to one of the connections to this board (as is done by the switch being in 50 or 75 positions in the assembled motor), the output signal frequency becomes a fraction of the input signal, quickly coming back to about half the frequency, where it pauses for a moment...before smoothly ramping back to the original frequency.

I took a closer look at the board.  There is a trace that takes the 48KHz signal from the basic crystal/counter circuitry and sends it to the dividers and logic that develop the signals for the stepper motor.  This trace is cut on the original circuit board.  The 48KHz signal instead gets sent over to this kludge board, along with 5V and ground.  Then another wire sends it back again to where it was originally going as an input to the dividers and logic.  So this is definitely a fix for problems that were found after the original design was "completed" but before this motor was sold.  This logic is all automatic so that by giving it a lower input frequency, it sends a commensurately lower frequency to the stepper windings.  This is a way to bring the stepper up to high speed a little more slowly, rather than the pulses to it instantly increasing in speed....something steppers don't generally like.

Duncan

[Duncan Brown]

On 4/20/2022 at 10:13 PM, Duncan Brown said:

I took out the chip that was being driven by the seemingly misbehaving chip.  Still misbehaving.  (A counter, not having the correct frequencies coming out of its various divided-frequency pins.)  So I took out the misbehaving counter chip.  I have some of those coming, probably get here next week.  I'll put in the counter chip and see if its output then look correct.  If so I'll put back in the chip it's driving and see if things still all look more correct.  If so I might try temporarily reassembling the motor and seeing if it's working better.

OK, the new counter chips arrived.  Put one in, and gloriously perfect precision square waves at all the right frequencies coming out of it.  Put back in the old logic gate chip it was driving... everything all screwed up again.  Pulled that chip out again and the counter outputs look perfect again.  So it looks like BOTH of the chips I thought could be "the bad chip" are... "the bad chips".  I'll have a new one of that second chip here next week and will try again.  One chip dead after 50 years could be random.  Two connected chips dead is more worrisome.  Someone was in here static zapping things by touching them, or fed more than 5V in on the 5V line, or something, possibly.  We'll see where this adventure leads next!

Duncan

[Duncan Brown]

My chips are missing in transit.  Working with the seller to find them or send new ones.  In case you were waiting with bated breath for the next installment of this saga...

Duncan

[steven jackson]

I have the latest version heavy duty motor that will only go slow and too fast with the amber light staying lit all the while.  I was in contact with an Eclair tech in France and he told me the motor speed control is aided by two very small incandescent lamps which act as a tachometer, that, like normal light bulbs, burn out, and my motor displays the symptoms of that being the issue.  The bulbs are still available but are covered in epoxy somewhere within the motor...I don't know much more than that.  I'm sending him my motor for repair and hopefully that will fix it up.  I'm wondering if your motor had this issue and if, when dismantling the motor, you noticed these small lamps??  According to the tech, the verification is simple, the repair less so.

[steven jackson]

oh, and he referred to the bulbs as "position detectors"

[Aapo Lettinen]

11 minutes ago, steven jackson said:

I have the latest version heavy duty motor that will only go slow and too fast with the amber light staying lit all the while.  I was in contact with an Eclair tech in France and he told me the motor speed control is aided by two very small incandescent lamps which act as a tachometer, that, like normal light bulbs, burn out, and my motor displays the symptoms of that being the issue.  The bulbs are still available but are covered in epoxy somewhere within the motor...I don't know much more than that.  I'm sending him my motor for repair and hopefully that will fix it up.  I'm wondering if your motor had this issue and if, when dismantling the motor, you noticed these small lamps??  According to the tech, the verification is simple, the repair less so.

it sounds weird that they would be incandescent as IR leds are much better for that use and should have been available back then when the motors were made as even earlier motors use leds for the purpose. optical tachometers are common in crystal sync motors because it is easier to get more slots ("steps per revolution") in them than if using a magnetic hall sensor and magnetic encoder("tachometer") disc and it may be easier to get a clean signal out of an optical tachometer than magnetic depending on what kind of hall sensors and preamp are used. Optical sensors are often used for shutter parking too depending on the motor even when a magnetic sensor would work for that application just as well.

anyway, I am pretty sure that if they are incandescent bulbs they could be relatively easily replaced with LEDs by the techician as one can get leds in all different sizes and shapes to fit almost any application and they can be made much smaller than incandescent having still substantial light output. probably leds would be still need to be manually casted to epoxy to form a new light assembly for the application but it would be similar with any other light source so to me would sound like a good replacement. Haven't repaired any original ACL motors though so your tech knows better

[steven jackson]

2 minutes ago, aapo lettinen said:

it sounds weird that they would be incandescent as IR leds are much better for that use and should have been available back then when the motors were made as even earlier motors use leds for the purpose. optical tachometers are common in crystal sync motors because it is easier to get more slots ("steps per revolution") in them than if using a magnetic hall sensor and magnetic encoder("tachometer") disc and it may be easier to get a clean signal out of an optical tachometer than magnetic depending on what kind of hall sensors and preamp are used. Optical sensors are often used for shutter parking too depending on the motor even when a magnetic sensor would work for that application just as well.

anyway, I am pretty sure that if they are incandescent bulbs they could be relatively easily replaced with LEDs by the techician as one can get leds in all different sizes and shapes to fit almost any application and they can be made much smaller than incandescent having still substantial light output. probably leds would be still need to be manually casted to epoxy to form a new light assembly for the application but it would be similar with any other light source so to me would sound like a good replacement. Haven't repaired any original ACL motors though so your tech knows better

He was very specific in calling them incandescent and comparing them to household lightbulbs, so I reckon that's what he is really describing.  I'm just hoping this will cure the problem.  

[Aapo Lettinen]

1 minute ago, steven jackson said:

He was very specific in calling them incandescent and comparing them to household lightbulbs, so I reckon that's what he is really describing.  I'm just hoping this will cure the problem.  

if the motor itself is older design than the crystal sync electronics then it would make more sense to use incandescent in the design as it would be part of the original motor drive.  but the motor drive is supposed to be brushless which is newer technology and thus I can't see any reason why one would use incandescent inside the motor drive as incandescent is not very long-life and additionally does not tolerate mechanical stress very well.

Just trying to understand why incandescent would have been better in the use than using leds as leds are far superior in that kind of use and just need a small resistor to work in the same application.

Opening one of the actual motor drives and posting images of the insides would be interesting because have not found any pics of the insides of the actual motor drive, just of the control boards like on this thread

[steven jackson]

2 minutes ago, aapo lettinen said:

if the motor itself is older design than the crystal sync electronics then it would make more sense to use incandescent in the design as it would be part of the original motor drive.  but the motor drive is supposed to be brushless which is newer technology and thus I can't see any reason why one would use incandescent inside the motor drive as incandescent is not very long-life and additionally does not tolerate mechanical stress very well.

Just trying to understand why incandescent would have been better in the use than using leds as leds are far superior in that kind of use and just need a small resistor to work in the same application.

Opening one of the actual motor drives and posting images of the insides would be interesting because have not found any pics of the insides of the actual motor drive, just of the control boards like on this thread

Maybe Duncan Brown will do us that service and possibly be able to repair his motor!

[Aapo Lettinen]

hope you get the motor working with the lamp repairs. It looks like my 16-speed ACL crystal motor did not raise that much interest and have to phase down the project after getting the presold ones finished so trying to repair the original motors is probably the most practical way to keep the ACLs working overall

[Aapo Lettinen]

1 minute ago, steven jackson said:

Maybe Duncan Brown will do us that service and possibly be able to repair his motor!

yes hope he gets the motor working, would be great to get certain that the original motors can be repaired and kept in reliable condition to be used in the future. I have heard lots of reports about the original motors falling apart recently and it is definitely a problem affecting many filmmakers. Making new motors is possible as I have demonstrated with my 16-speed crystal motor but a new motor is inevitably either a bit bulky OR very expensive. The additional issue with making completely new ones is that they need to be made in large-ish batches to keep the prices tolerable and it is difficult to sell that many motors at once which effectively prevents developing any new technology because it cannot be funded reliably

[steven jackson]

Just now, aapo lettinen said:

yes hope he gets the motor working, would be great to get certain that the original motors can be repaired and kept in reliable condition to be used in the future. I have heard lots of reports about the original motors falling apart recently and it is definitely a problem affecting many filmmakers. Making new motors is possible as I have demonstrated with my 16-speed crystal motor but a new motor is inevitably either a bit bulky OR very expensive. The additional issue with making completely new ones is that they need to be made in large-ish batches to keep the prices tolerable and it is difficult to sell that many motors at once which effectively prevents developing any new technology because it cannot be funded reliably

This tech came up with his own replacement motor a few years back (10 years ago?), using the original motor block, but discontinued it for lack of interest.

[Aapo Lettinen]

2 minutes ago, steven jackson said:

This tech came up with his own replacement motor a few years back (10 years ago?), using the original motor block, but discontinued it for lack of interest.

yes that sounds like the typical story, one uses years developing a replacement and then one or two are sold in the end and the product discontinued right after that for creating so much financial losses. Making a new motor really necessitates that something like from 8 to 15 pieces are sold at once so that one could get even a little bit for the work put on the project (something like 5 bucks an hour tops, probably more like 2 bucks an hour or so).

If making a completely new motor now it should be preordered and paid in full beforehand to be possible to make and then a large enough batch, say, at least from 10 to 20 made at once with the funds depending on how expensive it is to manufacture. So some people would need to commit to upgrading their camera even if the original motor would still work fine. As far as I have seen, it is extremely uncommon that anyone would like to do that even if the replacement motor would be much better than the original one and thus there would not be enough orders to fund the project even if single customers would like to order some motors here and there. One cannot wait for 15 years with the expensive-to-manufacture-and-paid-with-borrowed-money motors for them to be purchased one by one if ever so it is pretty evident that the best option would be just to try to repair the original motors until they totally give up and have to be thrown away, then just stop shooting with the camera body completely or wait for 10 years to pool up enough non-working cameras so that it would be practical to make a new crystal motor for them which, at that point, could finally be pre paid completely and thus be possible to develop in the first place