Multiple motor & controller approach?

[Delinquent]

I've been an avid RC'er for some time now, I build all my own motors for my electric aircraft, anything from a few watts using a stator from a CDRom drive up to kw's using the stators from power drills.

Multiple motors has been utilised in the RC scene on occasions, sometimes with more success than others, usually with a ring of motors around a central gear.

The cheaper motors on the market tend to have poor quality bearings, but I've yet to find one I couldn't replace bearings with good quality alternatives. Reliability wise, RC folks tend to get quite upset when motors and controllers are unreliable - having a very expensive helicopter fall out the sky tends to put you off a brand for life, so in general they aren't too bad, that said there is a good reason why some cost £50 and some cost £500, same as with controllers. The one positive with controllers is that problems found with the cheaper ones in RC applications won't be an issue to land based vehicles - interference caused by a noisy controller can bring your aircraft down in seconds, whereas you won't even notice it ambling along a road! Component failure on some of the cheaper ones is higher (I've had a fair few burn fets out while operating well within design parameters) however again, if you are using multiple motors in a vehicle the most likely result is you will be heading home just on reduced power.

Cooling is a distinct issue - the rating on a lot of these motors assumes it's got a 1ft+ fan in front of it creating a shed load of ram air. One motor I have built will put 4kg of thrust out from a 14 inch prop all day long, but if you take the prop off and put a load swing on it you get about 5 minutes before it overheats almost to destruction (I did in fact destroy the first incarnation during bench testing - melting the windings to the stator...). That would be my biggest concern.

It's an approach I've been considering on my own project, utilising 2 motors per front wheel. I have since found a possible avenue that will allow me to build a 300mm "scaled up" version of the smaller motors I build which would be preferable purely in terms of simplicity, but if that fails then I'll be going back to the multi-motor approach I think!

Something I forgot to mention... using multiple motors there have sometimes been issues getting them all to "fire up" properly - sometimes one will get stuck - not had it happen myself so I'll see if I can dig up the info on that.

The other major concern with one of the commercially available motors would be transmitting the power through the shaft - for some reason I just can't see those 8mm shafts every taking the strain.

[qdos]

Hi Del I'm sitting on the sidelines here following this thread with keen interest. I have to say it's good to have some input from someone who uses these motors. At a recent show we did there was a youong 14 year old lad who had oodles of knowledge in batteries, motors and controlers who chatted with us for quite some time. There is certainly a lot to be learnt from this field. Thank you for posting

[Jeremy]

Thanks that, I'm a relative newcomer to the RC motor scene - I bought my first outrunner to fly a scale model of a new design of full size aeroplane only a years or so ago.

The reliability issue comes down to only two main things with the motor; keeping it cool and making sure it has decent bearings. Both seem straightforward to resolve, I think. Replacing the bearings with the very best quality ones will be cheap and easy. Cooling can be sorted by adding a high pressure cooling fan and some well-designed ducting. Coupled with a stator temperature sensor it seems easy enough to protect the motor from over-heating.

The shaft size issue made me instinctively think that it would be a problem. I was surprised when I did the stress calcs to see just how low the maximum stress induced by the motor max torque was. The surface stress at the end of that 8mm shaft, with three motors coupled together, was less than 1% of the yield stress for silver steel (which seems the most readily available shafting). The bigger issue was torsional deflection, but this was minute also. Some torsional deflection could be useful, as coupled with a pair of rotary encoders it would be possible to turn it into a direct torque measurement system.

The controller start up and synch issue, plus the paucity of available RC controllers for voltages over about 26 volts presents a bit of a challenge. I've been giving this some thought though, and think I may have a way to get around this.

There are quite a few good brushless controllers around now that will work at voltages of 36 to 80 volts os so. Some are fairly easy to modify to deliver currents up around the 100 to 150 amp mark without too much trouble. The snag is that these are being manufactured for electric bikes and scooters, and inevitably use Hall position sensors. One other snag is that quite a few of these controllers are designed for use with hub motors, and won't handle the high frequencies needed to drive a motors like these. I've spotted one design that will though, and it's pretty cheap at around £75.

This means adding Hall sensors to the motor, but that shouldn't present a problem, as they are cheap and easy to fit. I would need to add an external trigger magnet, as these motors seem to have an unusual magnet count. Sensor type brushless controllers need to see either 60 degree or 120 degree Hall outputs, which doesn't jive well with 14 magnets.


Jeremy

[Delinquent]

Hi Del I'm sitting on the sidelines here following this thread with keen interest. I have to say it's good to have some input from someone who uses these motors. At a recent show we did there was a youong 14 year old lad who had oodles of knowledge in batteries, motors and controlers who chatted with us for quite some time. There is certainly a lot to be learnt from this field. Thank you for posting

It's nice to find a thread in here I may be of some assistance on, rather than just a pain with lots of questions. incidentally, the vehicle I'm looking at using them on is something I believe you are aware of - Stuarts trike (well, based on it anyway...)

Thanks that, I'm a relative newcomer to the RC motor scene - I bought my first outrunner to fly a scale model of a new design of full size aeroplane only a years or so ago.

The reliability issue comes down to only two main things with the motor; keeping it cool and making sure it has decent bearings. Both seem straightforward to resolve, I think. Replacing the bearings with the very best quality ones will be cheap and easy. Cooling can be sorted by adding a high pressure cooling fan and some well-designed ducting. Coupled with a stator temperature sensor it seems easy enough to protect the motor from over-heating.

Bearings agreed, it's often the first thing I do with a purchased motor (which due to the price I'm doing more and more now - it's just not worth the time building them!) Cooling wise, ducting and fanning is certainly going to aid, the only thing I'd say is test thoroughly! My ducted fan jet motor got hot enough to cause me fair concern despite running well within spec, and I still wonder if it helped it on its way to an early demise (anything to avoid facing the fact it might have been my dumb thumbs...)

The shaft size issue made me instinctively think that it would be a problem. I was surprised when I did the stress calcs to see just how low the maximum stress induced by the motor max torque was. The surface stress at the end of that 8mm shaft, with three motors coupled together, was less than 1% of the yield stress for silver steel (which seems the most readily available shafting). The bigger issue was torsional deflection, but this was minute also. Some torsional deflection could be useful, as coupled with a pair of rotary encoders it would be possible to turn it into a direct torque measurement system.

That's good to know. Some of the cheaper motors have quite soft shafts, I'd certainly question the quality of the silver steel used, but much like the bearings it's not something that you can't swap out quite easily with many motors.

The controller start up and synch issue, plus the paucity of available RC controllers for voltages over about 26 volts presents a bit of a challenge. I've been giving this some thought though, and think I may have a way to get around this.

There are quite a few good brushless controllers around now that will work at voltages of 36 to 80 volts os so. Some are fairly easy to modify to deliver currents up around the 100 to 150 amp mark without too much trouble. The snag is that these are being manufactured for electric bikes and scooters, and inevitably use Hall position sensors. One other snag is that quite a few of these controllers are designed for use with hub motors, and won't handle the high frequencies needed to drive a motors like these. I've spotted one design that will though, and it's pretty cheap at around £75.

This means adding Hall sensors to the motor, but that shouldn't present a problem, as they are cheap and easy to fit. I would need to add an external trigger magnet, as these motors seem to have an unusual magnet count. Sensor type brushless controllers need to see either 60 degree or 120 degree Hall outputs, which doesn't jive well with 14 magnets.


Jeremy

I'll have to ask my flying mates for some links - there are more and more high voltage / high current RC controllers coming out which would overcome the hall sensor / magnet issue. Not as cheap as the lower ones admittedly, but still in a price range that could make it worth while.

[qdos]

Excellent stuff Del. Yep I'm aware of the Etrike I have the plans right here and I'm currently doing an article on it for Plugged In having driven it at the Electrathon. I too see this set up as a distinct possibility with this vehicle and hopefully we can give it a try. I'm definately up for trying it out

[Jeremy]

I too am interested in a very light, aerodynamically clean, reverse trike, this motor (if it works) could be a contender as a power unit.

Years ago I contemplated building a trike from the front of an old Formula Ford, with a motorcycle engine and rear wheel grafted on to the back. I looked far and wide for the front half of an old FF at a reasonable price, but, quite understandably, the front is the bit that's least likely to survive any prang, so proved difficult to buy.

I've found a 200A Kelly BLDC controller that has regen, reverse, works up to 50V and costs about £150, which might be a contender. It's a bit big and heavy though - I'd prefer something as light and compact as possible - so I'll keep looking.

Jeremy

[qdos]

LOL You work on the motor system and I'm onto the body we've got the chassis and running gear sorted

[Delinquent]

qdos, without wanting to go too far OT, what sort of overall KV'age would you foresee for a multi-motor setup in that trike? Sorting the motors out is the one thing stopping me buying the plans at the moment (well, that and spending £50 before I really have to ) I'm considering doing away with the chassis altogether and building the body as a composite "pod" - though at this rate the only thing I'll be taking from the plans is the front wishbone setup

Back to speed controllers - the kelly are what I've been looking at with my single large brushless DC idea - just goes back slightly to the hall sensor issue you mentioned earlier Jeremy. I must remember to take that into account if I get these stupidly large stators organized.

In the RC world my favoured ESC is the Castle Creations Range. They tend to work faultlessly and have superb service (including replacing one for me FOC despite me informing them I'd been a bit dense and overloading it!) $160 will get you 2kw@50v, $200 will get you 4kw@50v. With the state of the dollar even with import duties it can be cheaper to buy direct from the US than a UK supplier!

Obviously that would leave you lacking in Regen. Reverse I've got around before with these by simply disarming the power up sequence and adding a crossover switch between two of the phase wires. I did hear something about them removing the ability to disarm the power up sequence (because it can quite literally be lethal if you plug in on high stick with a RC helicopter!) but I'd have to confirm that.

[Delinquent]

oops - forgot the link to the CC ones...

http://www.castlecreations.com/products ... eries.html

extensively programmable either by beep-control or through PC link.

[Jeremy]

Thanks for the info, Del. Have you seen the electric recumbent powered by a Plettenburg Terminator? Here's a link: http://www.recumbents.com/WISIL/shumaker/default.htm

Matt started out with an AXI, then upped the power with the Plettenburg; the details of this are on the second page of his site. Matt seems to be a great fan of the Castle Creations controllers, too. He's mentioned that CC are in the process of developing some products aimed at the electric vehicle market - I think they are due to launch later this year.

If it's any help with regard to the Kv issue, my RD50 has a 50kV motor, running at a nominal 48V (more like 52V), via a 3.7:1 reduction to a rear wheel that has a rolling diameter of around 24". The controller is set to limit the motor current to around 150A, which is more than enough for pretty good performance on a bike that only weighs about 90kg, plus an 85kg rider.

I am planning on using a reduction drive from whatever multi-motor set-up I end up with that will have an output shaft speed of about 2500rpm. This should be a reasonable speed to couple to most vehicle wheel with a simple chain or belt drive.

Jeremy