Adding sensors to a Astro 3210 motor

The magnet ring is an aluminum disc with pockets for the magnets and a hole in the center and there's an aluminum spacer behind it. The plastic case is three layers of polycarbonate with slots milled for the hall wires. I used magnet wire to further insulate it and cut a chunk of circuit board out of a computer power supply that got some dremel action. It all gets epoxied together. I'd be happy to send you the parts for one if you want to play with one! I'll be messing with your amazing TA for the first time tomorrow!

I also have a really neat piece of news that I'll have to post more about in a new thread tomorrow. :shock:
 
Most excellent solution!

Seems fit for using a polycarb waterjet for fabbing the parts in sheet form and laminating them together with magnets in place could make production less time consuming but I'm in envy :)

-Mike

PS: PM Sent.
 
Hi Mike and any others that might have the same question,

I've had some crazy stuff going on lately and am planning on refining that design and offering encoders for a bunch of different motors. The first one works great but I have some ideas for improvements.

I don't know how many people have the ability to drill and tap the end of the shaft so one of the improvements will be related to that. I believe Astro's can be ordered with a little extra shaft sticking out of the back and it's easy to re-shaft out-runners so the in the new design, the mag ring will clamp on to the shaft. I'd also like to incorporate a bearing to keep everything aligned, and go to three hold down slots.

If anyone is wanting encoders for their motors, unless it's a 3220, turnigy 80-100, or mini colossus, I'll need at least a good cad drawing of the motor to design from, even better would be to have the motor.

We're trying to get moved and set back up to be financially independent and free from investors again. After getting our hopes built up and smashed so many times, it's time to finance the project ourselves. I'm planning on selling little parts like encoders and custom pcgt pulleys, and doing e-bike conversions to continue to fund the prototype bike build.

With all of that said, I hope to be able to make some stuff available within the next month or so and will post up an appropriate thread in the for sale section.




On another note, I calibrated my CA and the real no load current at full RPM is 1.34 A. This seems pretty close to me and no amount of adjustment will get it any lower. I still haven't had time to get the dyno together or get Burtie's TA going but will very soon!
 
mdd0127 said:
On another note, I calibrated my CA and the real no load current at full RPM is 1.34 A. This seems pretty close to me and no amount of adjustment will get it any lower. I still haven't had time to get the dyno together or get Burtie's TA going but will very soon!

Wow, if that's really right it would be fantastic. My BMC motor sucks nearly 5A at no load.

My guess is your adjustment range is way more than you would need as a very small change in hall location will result in a large timing change. It might be nice if there was a mechanism to do fine adjustment of the angle. A worm gear might be one approach, or simply some slots on the outside edge that you can engage a flat blade screwdriver into and something fixed to pry against. Old ignition breaker points on some cars had something like this.
 
I calibrated the CA with my ammeter inline, adjusting the shunt value until the meter matched the CA to two decimal places so I think it's right on. If I pin the throttle all the way, the controller makes a strange buzzing sound, the motor speeds up to 8400 rpm and the current briefly spikes to 1.8A, then settles back down to 1.32-1.38. I'm not sure what's going on there and probably won't ever be. 7500 rpm on 45 volts is plenty for my application. :wink:

The timing adjustment is very easy to get spot on. I just held the throttle at a fixed position and slowly rotated the assembly watching the CA for the lowest amps. With the mounting screws just slightly loose, it's just like adjusting a distributor on an old car.

That first one tested the theory. The new designs have alignment bearings and are designed to clamp on to the shaft instead of bolting to the end.
 
mdd,

That's cool - the 3220 and 3210 have the same case I belive (maybe longer) but same bolt pattern and shaft diameter?

Put me down for one please - I have a working and tuned solution but it looks like rat scat and I will be showing off the latest incarnations alot!

I'd buy a copy (without the additions) if you would make it available, PM me for contact info / payment, etc)

-Mike
 
Just linking this thread to an optical implementation.

While the external hall sensors with a separate magnetic rotor appear to work well, I could never get an internal hall sensor setup to work satisfactorily with an Astro motor.

After some experimentation fitting halls inside the Astro, I found problems caused by stator flux interfering with the sensors, causing bad commutation, even at very low power levels.
In an attempt to find a good internal solution, I tried this optical implementation instead:

http://endless-sphere.com/forums/viewtopic.php?f=2&t=28966&start=30#p434772
 
Back
Top