Optical sensors for commutation -Working

Bob from Astro actually built a 3215 with one of these encoders on it about 18 months ago. He never got back to me about what board he used, but this one is exactly what he was describing.

I wish I was more electronically inclined. I would immediately buy one and begin testing.

Matt
 
mdd0127 said:
flathill said:
http://www.amtencoder.com/Product/AMT303

These look really cool. I looked at them a while back but had a concern. There is only one thing that makes me question their reliability for our application. It appears that it's using a compression fit on the plastic adapter that interfaces the shaft. I didn't see what material was specified for the adapter but it would have to be some kind of sticky urethane or something to work

Nope check out keyless taperlock shaft couplings. Some can hold hundreds of foot pounds of torque with non sticky metal on metal. They actually allow you to use a smaller shaft because the keyway is what causes the shaft to fail in many cases. I agree the adapter looks a little flimsy but I doubt it will slip if well designed.
 
I did some time as an industrial maintenance tech and am more familiar with taperlocks than I want to be. One slipped and welded itself to a 3 inch diameter shaft. I tried every trick in the book to get it off but ended up slicing it off little bits at a time with an oxy torch and a bunch of wet rags to reduce the HAZ. After a few hours of that, then some grinding, then sanding, I was able to save the shaft and attached 6 foot blower. I got yelled at the whole time to not hurt the $100K shaft and blower, then got in trouble later that day for taking most of the day to surgically remove that thing. $12/hr x 6 hrs = $72 versus $100K!

The little plastic insert in the picture looks pretty iffy to me. I'm ordering one soon so we'll see.

I can't wait to order some optical units from my inventive friend across the pond though! :wink:
 
Warning: electrical noob posting...(I apologize in advance)

I finally remembered to dig out the DIY axial-flux motor document sent to me by Axel from Sweden. I wanted to look up the IR/optical sender/sensor part numbers that he used, so I could add that info to this thread as an option for those who wished to experiment. He didn't list specific part numbers, but I DID stumble across some info in the instructions that may prove to be useful.

I know from past research that optical sensoring has been easily and cheaply used to monitor the shaft speeds of turbochargers that spin over 50,000 RPMs, so opticals have more than enough speed capability to do anything that we could ever want. Axel has included the instructions to solder together a DIY simple ESC, and he makes a point of saying that since the op signal can be sensed so rapidly, that he chose to PWM the op-signal for throttling the motor.

I assume there is some benefit to doing it this way as opposed to PWM-ing the controller in the place that it normally occurs? Could this possibly be an as-of-yet unused benefit of the Op-sensor method?
 
Even if he is PWMing the optical signal for some reason, in order to actually control the voltage to the motor (and thus power and speed) it still requires PWMing the phases of the motor. (well, you could do it with linear voltage control but there would be a huge amount of wasted power at the controller)
 
Okay 8) Now I'm not a lecky-techie, but I've always thought that if you had 2 opto sensors sensing a spinning disk with 1 slot (aka an encoding disk), and you feed the on/off signals to a processor, you would always know the direction of rotation and you could put any advance / retard of timing you like, when and where you like (based on rpm / load). No precise placement of anything, everything configurable. And Burtie's the man, isn't he!

Of course the drawback would be the cost of the processor that is powerful (fast) enough to do all that.

BTW, I sure am glad you guys type s-l-o-w-l-y enabling people like me to grasp some of the concepts (or not :D )

Cheers,
GT
 
Thanks Bill, it was interesting, and I learnt something.

But with the can encoded, that encoding has to be positioned very accurately in relation to the optos, yes?

What I was suggesting is that provided we know where the can is at one point in time, then no matter what the physical relationship was between that point and the optos, it could be aligned with the optos (as near as) perfectly by the software in the processor, which then provides the "hall" signals to the controller. And you only need one opto to do it. The second opto is to provide clockwise / counter-clockwise sensing.

Jeez, I hope that makes sense :shock: :lol:
 
gtadmin said:
Thanks Bill, it was interesting, and I learnt something.

But with the can encoded, that encoding has to be positioned very accurately in relation to the optos, yes?

What I was suggesting is that provided we know where the can is at one point in time, then no matter what the physical relationship was between that point and the optos, it could be aligned with the optos (as near as) perfectly by the software in the processor, which then provides the "hall" signals to the controller. And you only need one opto to do it. The second opto is to provide clockwise / counter-clockwise sensing.

Jeez, I hope that makes sense :shock: :lol:

Do you mean interpolating between trigger points? If so, then yes. Something like an Arduino is more than fast enough to pretty accurate timing between trigger points.



--
Bill
 
Tiverion said:
...Do you mean interpolating between trigger points? If so, then yes. Something like an Arduino is more than fast enough to pretty accurate timing between trigger points.

--
Bill
Yes. The uC would calculate all trigger points once the initial offset was configured in software. Of course, I would then need to understand what needs triggering, when it needs triggering, and in what order it needs triggering, something I haven't figured out yet (or more truthfully, haven't bothered to figure out yet). I think that if Burtie (for instance), who has a far greater understanding of these things, were to use this tactic, why reinvent the wheel? :)

Cheers,
GT
 
Very encouraged by the amount of interest in this field.

Thanks for all the comments and suggestions. Some interesting products also brought to light (pun intended) :D


I have had a go at 'productionising' this stuff to fit most of our favorite RC motors:

The result is 2 PCBs;
One of which is designed to read a linear encoder fixed around the can of an outrunner. (Linear version)
The other is designed to fit inside the rear cover of an Astro 32xx and read an encoder stuck on to a disc. (Rotary version)

The rotary version can also be used externally, reading an encoder stuck to a small hub that could be attached to any motor shaft between about 5 and 18mm diameter. :p

Here is the rotary version fitted to an Astro 3220...

SDC12830.JPG

SDC12824.JPG

View attachment 3

http://www.youtube.com/watch?v=JcoR4yhM9B0
[youtube]JcoR4yhM9B0[/youtube]



The same Rotary version can be used externally, here on a Turnigy 80mm outrunner...

SDC12835.JPG

http://www.youtube.com/watch?v=jnpl893hnFE
[youtube]jnpl893hnFE[/youtube]


Finally the linear version...

SDC12863.JPG

SDC12860.JPG

Edit to add link to linear sensor test.
http://www.youtube.com/watch?v=6yB6Rh_bAzU

[youtube]6yB6Rh_bAzU[/youtube]

For further info /blatent plug: http://endless-sphere.com/forums/viewtopic.php?f=31&t=31961
Burtie
 
Game-changer.

Great work, Burtie. :D
 
I don't remember if it was already answered yet... But will the sun or bright light mess with the optical sensors? Are we going to need some sort of shield for them?

I can't wait to buy a few of them and make my mountainboard sensored 4wd!!! It will be a hill climbing beast with sensors!!!! It already climbs really well, just not from a start below like 2-3 mph...
 
Count me in for a couple.... one to screw up and one to use:) on the Astro that I bought from Recumpence a couple of years ago. The Halls that I fitted didn't work as well as I wanted them to.

Thank you,
Snath
 
HumboldRC, -Yep,
Water, Mud, Bright light -will all mess with the sensors. The external versions will need to be protected from these.


Thanks for the positive comments folks. I hope to make some of this stuff available soon to play with, just got to buy some parts, assemble a few pcbs and get some instructions written down.
:)
 
I have one quick question. Have you thought about internal motor temps with regard to the selection of components?

If so, I need at least two for the astros.

I'm so glad you're making this stuff! The TA is amazing!
 
The sensors are rated to work to at least 212F.

For the internal sensors, I guess the thermal situation would be similar to typical hall sensors mounted inside a hub motor. Or hall sensors glued into the slots of an outrunner.

The adhesive holding the encoder disc to the rotor also needs to withstand the internal motor temperture. I have yet to temperature test any of this prototype stuff, and am open to suggestions :wink:

For the real hard-core motor melting, stator cooking addicts, -they may be best off using the external sensors.

EDIT update:
Some temperature testing of the sensors inside an Astro here: http://endless-sphere.com/forums/viewtopic.php?f=2&t=28966&start=60#p455368
 
its like music to my ears hearing that british accent saying internal optical sensors.....

one thing for the external turnigy....maybe the disc could just bolt to the endcap of the rotor with CS screws and have the "pattern" (whats the real name for this) in a sticker form to add to the disc once the right rotation is found, actually how do you find this>??? atleast i gather you wouldnt have to run the change the hall wire till it works routine, you would just move the encoder disc a position around till it did.

Excellent work burtie, I have just the project brewing for this, which has a 80-100 130kv internally mounted into a frame so no sun light will see it. Will work wonders for the sensors picking up the signal but I wonder how hot the motor is going to get.

Friggin legend!
 
IF these are readily available, I also will be interested in using it for my kits, it will keep me from having to make up hall sensor boards for each kit! PLEASE put me on the list of "contact when ready to sell"!!
 
Thanks guys!

@Rodgah,
I did wonder about using those screw holes on the end of the outrunner to attach a disc. But like you pointed out, it would then not be possible to rotate the disc relative to the shaft to tune the timing.

Getting the right hall sequence is still best done by swapping the sensor wires until you get it spinning. Then you can move the hub relative to the shaft to fine tune the timing. When you are happy, just nip up the grubscrew, and it is done!


@Whiplash. The stuff should be available pretty soon. I will be sure to put you on the list :)


I managed to test the linear optical sensor board today, here is how it went:

http://www.youtube.com/watch?v=6yB6Rh_bAzU


[youtube]6yB6Rh_bAzU[/youtube]


Do you like the way it all got colour-co-ordinated ? ... completely by accident 8)


Burtie
 

Attachments

  • SDC12888.JPG
    SDC12888.JPG
    62.2 KB · Views: 3,585
Back
Top