Thread Tools
Apr 27, 2011, 11:20 AM
Registered User
machinate's Avatar
Thread OP
Discussion

Outrunner for rock crawler - Kv and low-speed smoothness, is my theory sound?


I know this isn't entirely appropriate to this subforum, but this is where all the motor gurus hang out and my question probably needs someone with a strong understanding of motor construction to answer.

I'm looking into powering a rock crawler with a small sensorless outunner. It seems this was fairly popular in the rock crawling community a few years back, but ultimately everyone wanted a bit more smoothness at low speed and switched back to brushed motors. However, I've had a thought about this, and I need someone to tell me if my understanding is accurate.

Here's the thing - when the crawler guys started using outrunners, they naturally went for a very low Kv, as that makes sense for the low speed and high torque required in that application. However, it seems to me that you'd actually get better resolution and smoothness with a faster motor, running additional gear reduction to get the same final speed. I figure that smoothness isn't really tied to Kv, instead being mostly related to the number of switching cycles the ESC goes through per motor rotation. Wouldn't this be based on the number of poles and stator teeth? If so, comparing two motors with the same stator and magnet configuration, even if they've been wound for very different Kv, wouldn't they start within the same fraction of a rotation, and have the same angle of rotation per phase? If that's the case, and both motors are geared to the same final output speed, the faster motor should be smoother. It'll switch a larger number of times for the same amount of wheel rotation.

Any thoughts on this?
Sign up now
to remove ads between posts
Apr 27, 2011, 11:37 AM
Silent Majority
Bert's Avatar
You need sensored not sensorless.
Apr 27, 2011, 11:50 AM
Registered User
machinate's Avatar
Thread OP
As noted in my original post, the idea of using sensorless outrunners in crawlers is not new, and at one point a number of national competitors were using them with quite a bit of success. I'm well aware that sensor-based systems offer a lot better low-speed control than similar sensorless ones. That's not what I'm going for, considering how much those sensored setups cost.

Outrunners are dirt cheap these days, and the basic principle has been proven before. I'm looking at a design question, and wondering if making the 'obvious' decision with regards to Kv might actually have been a mistake for those earlier experimenters. It seems to me that the highest Kv available on a given number of teeth and poles would actually be better (provided you then gear the extra speed back out), and I'm hoping someone with more experience in motor design can comment.
Last edited by machinate; Apr 27, 2011 at 11:56 AM.
Apr 27, 2011, 01:22 PM
Registered User
Been there done that...FAIL

The problem with running sensorless is that on start up the ESC needs to see feedback from the motor to properly control it... throttle response and low speed smoothness suffer during this start up period no matter the KV. Also, by increasing KV you loose that wonderful lowend axle/transmission breaking torque

Honestly... even the current sensored setups(Novak/Tekin) I've tried are not all that great when compared to a good hand-wound and properly set up BRUSHED motor(35T being my favorite).

John Rob Holmes is working on producing a sensored outrunner and I think a rockcrawler programable sensored ESC as well that may just solve the start up issues and provide axle wrecking low end torque

http://holmeshobbies.com/product.php...&cat=19&page=1

read about this new motor here:

http://www.rccrawler.com/forum/showt...ored+outrunner
Apr 27, 2011, 01:45 PM
Registered User
machinate's Avatar
Thread OP
Yeah, I've looked into the Holmes stuff and while it looks promising, it also has weight similar to a brushed system (way more power than I need) and the cost of a sensored brushless. For that matter, a hand-wound motor costs more than an outrunner, too.

I suppose I need to clarify my goals a little. I'm not going for competition-level performance, I'm actually looking at making a trail rig and want the lighter weight and non-maintenance as much as anything. It's still a crawler, though, and I do want to be able to creep along and negotiate fairly tough obstacles. That's where this topic comes in - I know people have tried outunners before, and ultimately gone back to brushed motors. Those were always very low-Kv motors, though, which is what makes sense at first glance; crawlers generally use high-turn systems, and for brushed it works great. My view on the matter is that with brushless, just maybe, a faster motor will let you reclaim some of the lost smoothness through lower gearing (which will also return the axle-snapping torque you lost).

Here's an example, to help explain what I'm getting at: Consider two hypothetical motors, both with two poles (so, inrunners, I guess? It's just a simplified explanation). Since they both have two poles, they'll both require the ESC to switch phases three(?) times per rotation. Only let's wind the motors differently, so one is 1000 Kv and the other is 2000 Kv. Now, we'll gear them differently, with the 1000 Kv motor at 30:1 and the 2000 Kv motor at 60:1. They'll both provide exactly the same wheelspeed, and exactly the same wheel torque, for exactly the same power - the fast motor is spinning twice as fast and making half the torque, but the extra 2:1 gearing doubles the torque and halves the speed. Now, at a glance it seems like you're getting the same result, except for one key difference: for each wheel rotation, the slow motor is switching 90 times, while the fast motor is switching 180 times. The way I figure it, that's twice the resolution, resulting in half the twitching. Bump the pole count back up to 14 or whatever's common in smaller outrunners, and it starts to look like the twitching and jerking of the outrunner might take up such a small amount of wheel travel that it might not be noticeable.

Again, I'm not looking for recommendations. I'm looking for comments on a specific theory.
Apr 27, 2011, 03:29 PM
Scratch builder
Yes, your theory is correct. Gearing the crap out of a high RPM outrunner will give more "resolution" as you say. There is a limit to the switching frequency govered by the pole count and the ESC used. From what I remember reading, DLRK wound motors switch 7 times per revolution. These are quite common.

Is it practical to devise a coupling that has some rotational slop built into it? This would give the motor some wiggle room.

It comes down to how many extra parts are needed to accomodate the outrunner and is it worth it as opposed to a simple brushed set up.
Last edited by KenSt; Apr 29, 2011 at 11:14 AM.
Apr 29, 2011, 07:04 PM
Registered User
How about a clutch?
Apr 29, 2011, 11:37 PM
Registered User
machinate's Avatar
Thread OP
Some kind of mechanical solution would be interesting, but I think very difficult to pull off. Using a slipper clutch would limit the total power you can use, and using a centrifugal clutch would still give you poor fine control at low speeds. A flexible coupling like Ken suggests (a guibo) might be able to help the motor start in certain situations, but you'll still have a minimum steady speed based on the lowest RPM the controller can maintain without switching back to the jerky startup sequence. That's what I'm trying to target with this idea - minimizing that lowest speed.

The best-looking motor I've found so far is a Scorpion SII-2212-1850. Within the size and power levels I'm looking at, it's the fastest motor I can find with lots of teeth and poles. Unfortunately, I'm not looking to spend that much money on an experiment, but cheaper motors don't usually specify pole counts. I'm probably going to try this motor from HobbyPartz and hope it doesn't show up with a really low pole count. I figure if my theory actually holds true, even if it's got 6 teeth and 8 poles (generally not seen at this size until 3000 Kv or so) it'll still be slightly better than the really low-Kv motors everyone else seems to have tried. From what little I can see of the stator in the picture, I'd say there's a fair chance it's what I'm looking for.
Apr 30, 2011, 12:15 AM
Registered User
How about a fluid coupling? I know there are probably none available in model sizes, but it shouldn't be too difficult to make one yourself. Just an idea.
May 01, 2011, 12:08 AM
Dieselized User
gkamysz's Avatar
What makes a sensor setup expensive? Nothing difficult about adding sensors to a motor.

Greg
May 01, 2011, 01:09 AM
Wake up, feel pulse, be happy!
Piece's Avatar
How about this motor? 12 poles, 14 magnets (I can assure you of that; I've got two)
May 01, 2011, 01:13 AM
Wake up, feel pulse, be happy!
Piece's Avatar
You could also try rewinding the one in my link above. It's got 6 turns stock; if you went down to maybe 4 turns (or less, perhaps) the Kv would be ... very high. Probably not usable/practical for a plane at that point, but it might work for your crawler.
May 01, 2011, 08:56 AM
Registered User
machinate's Avatar
Thread OP
TP16: I've looked at that motor. It's good to know it's the right configuration, but I'm pretty sure the shaft is too short. I would have to find a replacement before I could use it. Is the shaft reversible?

Rewinding something is certainly worth considering if all else fails, but I'm not quite that ambitious. 2200 Kv is about what I'm going for, anyway; and faster and I'll run out of gearing.

Gkamysz: It's true that adding sensors to a motor shouldn't be particularly expensive, but even if you got a homebrew setup working well, commercial sensored controllers are still pricey.
May 01, 2011, 02:51 PM
Wake up, feel pulse, be happy!
Piece's Avatar
The shaft is indeed reversible. It's held with two setscrews and a ton of friction; you might need a vise to get it moving.

If you need a longer shaft, just get a piece of 1/8" stainless steel rod (cheap; 3x 12" pieces are like $1.19 at my hobby shop) and cut a longer shaft to replace the short one. That's what I use to replace bent shafts and it works great.
May 06, 2011, 08:26 PM
Registered User
The ebike guys are all over this stuff. They are adding 3 hall effect sensors to RC brushless motors just above the outside of the case) and sensing the outrunner magnets

If you are sensored, then you can deliver much higher currents (more torque) at very low RPM and also maintain very low RPM. You can't do this on sensorless because at low RPM then controller has to guess what is happening.

State of the art is using sensored inputs at low RPM, and then switching to sensorless drive once RPM (and back EMF) gets high enough..

Check out this video:
Turnigy Aerodrive Custom Brushless ESC Sensored Drive (1 min 27 sec)


And

C80100 motor modified with external hall sensors.wmv (0 min 28 sec)


These low RPMs are quite impressive. Not needed for airplanes, but on a climber it'd be a big help


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Looking for Small Low KV outrunner for XF5U power Grejen Power Systems 23 Aug 14, 2014 11:57 AM
For Sale Nice Clodbuster rock crawler for sale $250 shipped!! Stampede720 Cars - Trucks and Parts (FS/W) 3 Apr 27, 2011 11:35 AM
Sold losi mini rock crawler redrob Cars - Trucks and Parts (FS/W) 4 Mar 27, 2011 06:19 PM
Sold Losi Micro Rock Crawler helifrek Cars - Cars and Parts (FS/W) 1 Mar 15, 2011 04:48 PM
My 1/35 4x4 Rock Crawler aka "Rock Biter" killer_d76 Micro Cars 10 Jan 12, 2010 08:57 PM