Automotive Starter Motors as Drive Motors

Technical Electrical
1

Okay… first of all, this is not for an FRC application.

A colleague of mine has suggested using recycled automotive starter motors as drive motors in an application not entirely different from how we would use CIMs on our FRC robots. In fact, my suggestion was that we use CIMs… but my colleague liked the “recycled/maker/DIY/low-cost” vibe of the starters. I thought I’d seek out some more information before getting too fixated on CIMs… I really like CIMs, but want to keep an open mind on this.

Here’s what I’m thinking… I’d appreciate advice from anyone with practical experience using starter motors this way.

It appears that many late-model starter motors are permanent magnet DC motors with a planetary gear drive reduction. Sounds promising, especially the internal gear reduction. Even 3:1 helps.

They tend to be more powerful than CIM’s (one guestimate is that a starter motor will be about 1/50th the power of the engine it is starting) and thus draw higher currents. This is not only likely to cause a problem with overheating, but also for motor control circuitry, batteries and wiring. I expect most starter motors to pull over 200A on start up and 100A under load. I may be able to dial that back with a motor controller and put a soft limit on current draw.

Older, series and shunt wound motors can be converted to forward/reverse operation, or even for use with a speed control. Should be able to do the same, but even more easily, for a permanent magnet one.

Mounting the starter motor and putting a useful gear on the end of it will be a bit of a pain relative to a CIM, but we do have access to a full machine shop including CNC lathes, mills and a waterjet. (Yeah, sucks to be me…:slight_smile: )

So it seems like there aren’t any insurmountable problems with going with recycled starter motors… but I’m worried that something weird is going to pop up… something like having them timed so that they don’t run well in reverse or… well, I don’t know what.

I know starter motors work great with car batteries in on/off applications for relatively short bursts, like bar stool racers, but has anyone used them with speed controllers in forward/reverse applications? Any reason why they *wouldn’t *work? (Edit: and alternatively… any success stories where they do?)

Thanks for your thoughts!

Jason

2

I don’t have any personal experience with them, but I’ve heard of starter motors being used in combat robot applications more than a couple times, generally for drive.

3

Starter motors and winch motors will provide much more power than a cim. Finding a speed controller that can feed them may be more of a problem.

4

Jason,
Besides the intermittent design feature of these motors, they are intended to produce a lot of power depending on the motor they are used with. A small four cylinder engine in a warm climate will produce a much different load than a full size V8 in Calgary. I would expect a significant bias on direction but that should be easily overcome with software. Many starter motors will also contain a temperature cutout to prevent damage under hard starting conditions.
In general I would expect a larger motor to easily run several hundred amps stall current. The battery recommended for my Odyssey V6 has a cold cranking spec of 800 amps at 75 AH, so I would not be surprised to find the starter in 300-400 amp range.

5

They’ve got great power for their size, but that comes at the cost of durability.

I would make sure whatever ones you wind up using have replaceable brushes a and can be taken apart and cleaned without great hassle.

6

Intertesting. I recently bought one of the modern permanent magnet starter motors for my 55 chevy, the motor I got was originally used on a 1999 chevy truck. It cost well over $100. Previously I had been using the older (1960s-70s) style starter motor, which can be purchased for less than half as much.

If you have an application that requires a motor that is significantly more powerful than a CIM, then I can see how it might be worth looking into. But you didn’t tell us what it’s for…and the $28 CIM is a real bargain.

7

Back in the day, when I raced ministocks on a track in southern Georgia, I learned that the starters we used required 24V rather than the 12V the CIMs use. Keep that in mind.

8

A typical starter motor can deliver 1.5 to 2 kW of power, briefly. That is quite a bit more than a CIM. At those power levels, cooling is an issue. A very reduced load (still tons more than a CIM) would reduce that a lot.

Starter motors are not designed to run backwards, the brushes are optimized fir one direction only. They are also not designed to run for long periods of time, maybe a lifetime of a few hundred hours.

But they ARE cheap.

You will want to disable the solenoid that extends the pinion. An easy interface to the rest of the world is a sleeve that fits over the pinion gear teeth, with an output shaft to fit what you want. Make it on a lathe to ensure it is concentric.

9

Thanks, all, for the advice. I didn’t want to say too much about the application as we haven’t announced the details to our students yet, but, on the other hand… if I’ve got them surfing CD often enough to catch this, then I’m happy.

A bit of background… after 13 years teaching high school “shop” classes, I took a job at the British Columbia Institute of Technology, teaching shop teachers on the technical aspects of what they will be teaching. Our students come in with a variety of backgrounds in trades and technologies, and at least one year of general post-secondary studies. They spend two years with us studying woodwork, metalwork, mechanics, drafting/design, electronics (that’s me), and a few other support courses. I speak from experience when I say that the program is pretty much as fun as it sounds… I was a student in it 15 years ago (and my Dad was a student in it about 40 years ago, we’ve got more than a few second-generation shop teachers in BC!). After spending two years with us, the students spend a year at the University of British Columbia, where they do their practicum and complete their B.Ed.

As a capstone course in our program we do a major design project called the “Vehicle in a Box”. The challenge is announced just before Christmas Break, and the competition takes place at the end of May. The students are doing other courses at the same time, and evening access to the shops is very limited, and they work in groups of three or four, but aside from that, it isn’t entirely dissimilar from FRC… well, with the exception that the students who are driving the machines are actually in the machines. The vehicles have to dissassemble to fit in a fixed-size storage box, and the competition begins by unpacking and assembling the vehicles, with bonus points for the best time.

We’ve done various challenges and used various power sources in the past, ranging from human power to 1hp “electrathon” style motors and, most recently, brushless hub motors from electric bicycles. Normally the challenges are outdoors, but this year we are taking things indoors and playing a game somewhat similar to Breakaway (the soccer game, but without the bumps) but using a giant earth ball like the one from the “around the track” racing game of 2008.

We usually get decent media coverage and try to ensure a good design/build challenge for our students while coming up with an entertaining, media friendly presentation to help promote our program.

So given that the mass of the vehicles has to include a driver, I’m expecting vehicle masses in the range of 200-300lbs. Two CIMs would have to be geared pretty low in order to provide decent maneuverability, but I don’t think I can justify going to four motors/controllers per vehicle. I’d rather have the students look into building a shifting gearbox if they wanted top speed, or we could double up the batteries to run at 24v. Yeah, the motors might not be rated for it, but they’ll take it for as long as we need them to. The extra “kick” from starter motors would be nice, but it comes with the added “cost” of having to build high-current controllers and connections, as well as some massive battery drain… not to mention characterizing the performance of essentially a random collection of motors pulled from a scrapyard.

We are working on developing the final details, including the motor specifications over the next week or two. We know we can use CIMs as a fallback, if the starters don’t work, but I’m trying to work out right now just how much “hassle factor” is going to be involved. There will definitely be some… and we want the students to learn about motors… but we also want their learning to be in a positive sense of “what works well”.

So that’s what we’ve got planned. If any of my students are reading this, I hereby offer up a BEvERage bribe to keep quiet about this until the game is fully announced. :slight_smile:

Thanks,

Jason

P.S. Yep, you can do that in post-secondary. The average age of students in our program is 28.

10

Jason,
For what it’s worth it shouldn’t be too hard to get a dozen identical starter motors from a medium-sized scrap yard. Pick a brand of car that is pretty common and then harvest the starters from those that have roughly the same size engine. Car manufacturers don’t change their starter designs too often.