High power RC motor and drive unit production

[CNCAddict]

You can see from the windings that things got pretty hot. The problem is that I always kept close watch of motor temperature and the outside never really went much above 100C (ok for SmCo motors). The windings however were hot enough to melt the solder on the Wye connection. After opening the motor to have a peak around, it seems the windings have poor thermal transfer to the stator. So the delta T between motor temp and winding temps gets out of hand very quickly.

Even so, the motor was still running just fine and never caused any problems until I changed gearing which let the motor rev a tiny bit higher (maybe around 13,000rpm). That's when one of the magnets decided to leave town. The stator is still "ok" and I may build my own rotor to get it working again, but the prototypes I'm finishing up will perform a lot better so there is not much point in bothering with it anymore.

P.S. I'm not knocking the Astro motor. I think under normal use it should be just fine, but I have a habit of running things harder and harder until something eventually blows

 

[Miles]

Thanks for the info,

I'm not planning to run mine at more than 9000 rpm and I don't need more than 2Kw peak, out of it. Disappointing about the heat transfer, though - I was hoping to use rings on the case as a heat sink....

Have you considered doing an inrunner for bikes?

 

[CNCAddict]

For 2kw I wouldn't worry about the temps, I think you'll be just fine. I had this motor geared for over 50mph so it's not really surprising that things went south. I've run 3 different motors on my scooter (Astro, HXT, Hacker) and none of them really worked perfectly. I just want a smooth, reliable, powerful motor for my scooter but so far nothing has done all three at the same time (why I'm building my own).

I can certainly design a standard internal rotor design but it is harder to cool the magnets. In a sealed motor the magnets will get a LOT hotter than on the outrunners or an open inrunner motor. It's really critical to keep the magnets cool to prevent a degradation in motor performance and possibly ruining the magnets. It's all a compromise I guess.

 

[swbluto]

What was specifically wrong with the HXT that you used on your scooter? Was it another pushing-the-limits showdown?

 

[CNCAddict]

The HXT had some severe vibrations and eventually the magnets started to wiggle out of place. I had probably 300miles on the HXT when it actually started to fail. It also lost sync fairly easy during hard acceleration so I had to be gentle on the throttle.

 

[Miles]

It's really critical to keep the magnets cool to prevent a degradation in motor performance and possibly ruining the magnets. It's all a compromise I guess.

I guess that's why AF use SmCo magnets for these. I was thinking one could suck the heat out from the windings before it builds up, but if the transfer to the case isn't that good... Presumably the contact between the laminations and the case is good enough? So, is it the sleeving between the windings and the laminations that is the bottleneck?

I steered away from open motors because I thought there might be problems with magnetic debris getting inside, over time. Am I worrying needlessly about this, do you think?

 

[CNCAddict]

Presumably the contact between the laminations and the case is good enough? I thought there might be problems with magnetic debris getting inside, over time. Am I worrying needlessly about this, do you think?

The stator to case contact seems ok. It's the windings that look like they have minimal contact with the stator slots.

I've never had any problem with debris on the HXT. After opening it up there was just some dust on the inside, but nothing that would affect operation. I suppose it's possible for a small pebble or something to get lodged in the motor and cause problems but it's never happened to me. Even so, if the motor is directly in line with one of the tires then I think it's a good idea to put some sort of mud guard in place to keep things clean.

 

[Miles]

It's the windings that look like they have minimal contact with the stator slots.

Anything to be gained from pumping thermal compound into the voids?

 

[swbluto]

The HXT had some severe vibrations and eventually the magnets started to wiggle out of place. I had probably 300miles on the HXT when it actually started to fail.

So it wiggled out of place like... the glue slowly came undone? Would it have to be reglued to fix it and would this be considered needed long-term maintenance for the HXT motors?

 

[CNCAddict]

Anything to be gained from pumping thermal compound into the voids?

Maybe thermally conductive epoxy would help a lot, but again unless you're only pulling 50A or so this won't be a problem. I personally wouldn't put heatsink paste though, that is designed for heatsink interfaces where the gap is only a few mils thick.

So it wiggled out of place like... the glue slowly came undone? Would it have to be reglued to fix it and would this be considered needed long-term maintenance for the HXT motors?

There is stamped steel magnet retainer in the front of the flux tube which came loose and started rubbing the inside front of the motor casing. I took it all apart and many of the magnets had come loose. I think they use epoxy that is much too brittle. The HXT has a really flimsy flux tube and you need extremely compliant epoxy to be able to "move" with the steel and still remain bonded to the magnets. If you cant find high quality flexible epoxy, then some rubberized CA should work fine. The temperature rating of the adhesive isn't that important since the magnets on outrunners should never get very hot. Even some of the "sumo" glue, or low cost epoxies they have at lowes would probably work better than the glue these motors come with.

 

[etard]

Beautiful motor CNC,
Have you decided on prices and are you able to take orders? If this motor has twice the power of the Astro, does that mean it will produce twice the torque, or take twice the amps? Also, I would like to be able to run it around 50 volts, will it be able to handle this continuously, or could it handle it with a 12 volt booster battery every now n then?

Thanks, I look foward to seeing one in Matt's gearbox.

 

[recumpence]

Hey Guys,

Glad to see all the posts!

CNC,

Sorry I never called you back. I get busy right before and after a holiday. I will call you soon. :wink:

I am thrilled to see the motor parts. This should be a fun winter mounting everything.

Dave is cutting drive unit secondary chain idler assemblies today (you will see when they are done). After these are run, he will be making the universal drive mount which is basically a split clamp like the drive pieces that just comes to a blunt end with some threaded holes in the sides. That mount can be bolted or welded to a bike frame. Also, there are tapped holes in the sides of the secondary stage link for mounting without the mount I mentioned. The last thing is to go ahead and get some freewheel adaptors and 1/2 inch to 5/8 inch tube to shim up the shaft.

I am working hard, but taking my time.

Matt

 

[CNCAddict]

Beautiful motor CNC,
Have you decided on prices and are you able to take orders? If this motor has twice the power of the Astro, does that mean it will produce twice the torque, or take twice the amps? Also, I would like to be able to run it around 50 volts, will it be able to handle this continuously, or could it handle it with a 12 volt booster battery every now n then?

Thanks, I look foward to seeing one in Matt's gearbox.

Haha, no orders just yet. The soonest the production motor could be ready would be sometime in February but that is maybe too optimistic. I need to see what part of this motor may need to be redesigned before doing a production run of 100 or so units.

I can wind the X-era 6050 for 300V is anyone wants that. If wound for an identical Kv as the Astro then this motor will produce over 2x the torque at over 2x the current draw. I'm going to test the motor above 15,000rpm to see how it holds up. I'll be sure to post some pretty pictures and video of some scooting...just hope my A123's and Spin 200 ESC hold up

 

[GGoodrum]

The drive mount sounds perfect, but will this be able to be used with a single-stage setup as well, or just the two-stage version? Also, for my single-stage version, which is using the chain drive, will the idler assembly work with the motor plate?

-- Gary

 

[recumpence]

CNC,

Cool...... 8)

Gary,

Yes the mount with work single or dual stage. Also, yes, the idler will work for your application. However, you probably won't need it bacause the motor will slide fore and aft in the mount for shain tension adjustment.

Matt

 

[dimitrib90]

Haha, no orders just yet. The soonest the production motor could be ready would be sometime in February but that is maybe too optimistic. I need to see what part of this motor may need to be redesigned before doing a production run of 100 or so units.

Do you have an idea of what the price could be?

 

[Miles]

Do you have an idea of what the price could be?

The only down side is the fact that technology like this does not come cheap. David's motor will probably come in at roughly $400. My dual stage drive for that much power will be about $300 (roughly) and the SHV 200 required to run it at it's max capacity will cost $750. However, a single stage reduction will be between $150 and $200 (depending on layout), a more basic motor would be $200 to $300, and a HV110 is $250. So, you can get away with spending less for RC stuff. But, we want to see just how far we can take this ultra light weight, extremely powerful setup.
Matt

 

[dimitrib90]

Thanks. Definitely not cheap. Looks like Im going to have to save my christmas money.

 

[Miles]

Compared to Plettenberg prices it is

 

[recumpence]

Cost is relative.

CNC's smaller motor (the 40mm motor) is $190 and a VERY nice motor. Oops, I just realized, I never posted pics! I will today.

Anyway, it looks like the 60 series motor will be twice the price of the 40, but with 3 to 4 times the power! Not a bad tradeoff.

Remember, if you want to stick with 2kw or so, RC stuff is really reasonable to get into. Mounting it and making it work is the problem. That is why I am making my drive system.

Matt

 

[recumpence]

Mitch,

Thanks for the contact. I spoke with Jason at length. It looks like I will be ordering 60 adaptors (30 in aluminum and 30 in steel).

So, who wants to draw up this adaptor for me in CAD to email Jason?

Matt

 

[ericslai]

Matt,

I will volunteer for this CAD work. Just let me know what I need to do. I use Pro/E daily and have been doing design work for over 20 years.

You can PM me if it is too much detail for this thread.

Erics.

 

[Miles]

Matt,
I will volunteer for this CAD work.

Time to learn MCAD, Matt..... I don't think it would take you long to get up to speed... :wink:

 

[recumpence]

No kidding!

Does anyone have one of those FW adaptors they can measure up? I could order one, but this would save time. Jason said it would be 3 weeks minimum, maybe more, to make these. So, I would love to get them ASAP.

Hey, Mitch and Gary want a series drive (chainring to motor with one chain, then motor to cassette with a second chain). This will make the shaft overhang from the reduction bearing kind of far. I am concerned about shaft flex. Also, the pedal sprocket on the output shaft should be the outer most sprocket and the motor drive sprocket to the rear cassette should be the inner. This would reduce the load on the shaft considerably.

Thoughts?

Matt

 

[Miles]

No kidding!
...Thoughts?

I'm never sure what you guys mean by this.....

In UK, we say "I couldn't care less". In US, you say "I could care less", with apparently the same intended meaning