Tamiya 'toy' motor testing

A place to talk about anything that does not fit in one of the above categories.

Tamiya 'toy' motor testing

Postby nexisnet » Wed Jan 09, 2008 5:10 pm

A lot of people, myself included, are interested in just how high you can push the voltage on the motors in the small Tamiya gearboxes. I would be happy to halve the operating life of my gearbox if it means being able to run the motors faster, especially if I can run them straight from a greater-than-5V battery pack, so long as that half-life is still longer than I plan on using the motor. Lets say 20 hours of continuous operation would make me happy, and 50 would make me very happy.

Unfortunately, since these are small 'toy' motors, the manufacturer really doesn't offer any cycle-life numbers, even when operating the motor at the recommended 3V or less. I happen to have a bunch of Mabuchi FA-130s around, those are the motors that the 70168 Double Gearbox, 70167 Single Gearbox (4-Speed), 70097 Twin-Motor Gearbox, 70093 3-Speed Crank-Axle Gearbox, 70110 4-Speed Crank-Axle Gearbox, and 70103 Universal Gearbox are based on. Recently I've been doing a little testing with these motors, basically I've been running them into the ground...in a controlled, pseudo-scientific way of course:

Image

For light loading and downgearing, I'm fitting the test motors into the Solarbotics GM9 gearbox (143:1 reduction). The gearbox output shaft is turning a piece of card-stock which interrupts a photo-gate once per revolution. It looks like the gearbox is just sitting there, but I super-glued it to a chip socket so it presses in place on the breadboard. The photo-gate interruptions are counted by an ATTiny2313, which drives an LED display (readout in RPM) and transmits the speed readings over a serial connection to a laptop. The motors will all be driven in the same direction at a constant voltage, straight from a regulated DC power supply, working only against the internal resistance of the gearbox.

The first thing I noticed is that running the motor way over the recommended voltage makes it even more loud and annoying than these gearboxes normally are, which might be a consideration for some robots. To keep from driving myself crazy, I'm conducting these experiments with the breadboard on a piece of foam inside a (room temperature) minifridge (it REALLY helps!).

Image

The second thing I noticed about overpowering these motors was that they generate some horrendous electronic interference. The motor is powered from an isolated output of my power supply, and I added capacitors across the motor leads, but the USB-Serial adapter kept flaking out whenever I was running the motor (the LED would flicker and my laptop would stop receiving data). I even tried surrounding the adapter in grounded foil, but I eventually had to cut it out of the setup. Instead I ran the TTL serial TX line from the ATTiny through a logic inverter and into my laptop's serial port. Neither the inverter chip nor the AVR seem affected by the interference, but still watch out (and use those capacitors!).

So far the data is less than encouraging (at least for us fans of 7.4V LiIon packs). I started out at 7.2V, the nominal voltage of a 6 cell NiMH or NiCd pack, and just under the nominal voltage of a 2 cell Lithium Ion or Lithium Polymer pack (my favorites). The motor was dead after one hour and 44 minutes. Not temporarily overheated, not slowing down, dead dead. When I took it out of the gearbox the shaft still spun freely, but it wouldn't turn under power, even with no load, and here's why:

Image

In the picture above, the bottom motor was never used at all (sacrificed in the name of science!). The brushes and commutator (the brush-contact barrel) are pristine, shiny, and well-lubricated. The middle motor is one I had used for some amount of time, but only ran at or below 3V. There's some grit in it, but the commutator is still smooth and the brushes are in fine shape. The top motor is the one I ran at 7.2V. There are groves maybe 20 thousandths of an inch deep on the commutator, and the brushes are seriously damaged, by which I mean that one brush is just plain gone! There is also copper dust spattered all over the inside of the casing. Less than two hours earlier this was a brand new motor, just like the bottom one in the picture. Wow.

I just finished testing another motor at 6V, which did a little better, lasting for five hours and 28 minutes. It now looks the same inside as the 7.2V motor. It's even the same brush that's missing (the one I had connected to power, if that makes a difference), which leads me to believe that the damage is caused by arcing, and more severe in one direction. If you run your motor in reverse some of the time it may last longer, but both brushes are damaged so it probably won't double the life. I don't think the ware is caused by friction directly, since the 6V motor lasted for almost three times the number of actual rotations of the 7.2V motor. Anyway, here's the performance data I have so far:

Image

I have a 5V test running right now, and I've got four more motors to burn after that. I'll keep posting results as I get them, which I can only hope will take a while!

-Adam
User avatar
nexisnet
 
Posts: 1216
Joined: Mon Mar 21, 2005 9:29 pm
Location: Boston, Massachusetts

Postby jan » Wed Jan 09, 2008 6:41 pm

Wow, that's awesome! I look forward to your other test results.

We did a lot of tests with the motor noise when we were working on the low-voltage orangutan, and I think those brushes are bouncing a lot. Looking at the current, we saw several instances where the current dropped to zero per phase, which nicer motors didn't do at all. With the motors running from 3.6V batteries, we'd induce many-volt spikes in small loops of wire many feet away!

- Jan
User avatar
jan
Pololu Employee
 
Posts: 2176
Joined: Sun Dec 28, 2003 7:12 pm
Location: Las Vegas, NV

Getting better

Postby nexisnet » Thu Jan 10, 2008 7:54 am

Yikes! Nothing produces interference quite like a cheap DC brushed motor. I remember when I was little my parents had a coffee grinder that would totally destroy the image on a TV 20 feet away.

The five volt motor made it through the night, and died after 15 hours and 31 minutes, getting in a respectable 240,232 rotations. Same mode of failure as before, but this time it's the other brush that's missing (although I'm not 100% sure that I hooked up power in the same direction as the first two). There seems to be some sort of logarithmic trend going on:

Image
(Yay PNG chart)

I just threw on another motor at 4V.

-Adam
User avatar
nexisnet
 
Posts: 1216
Joined: Mon Mar 21, 2005 9:29 pm
Location: Boston, Massachusetts

Oh well

Postby nexisnet » Fri Jan 11, 2008 7:08 pm

The vaguely-logarithmic trend of motor life tripling for each volt less may be over.

Something went catastrophically wrong with the first motor I tested at 4V, its speed started slowly dropping at around eight hours, and it died right around 14 hours (an hour and a half LESS than the 5V motor). I tested another motor at 4V, which has a failure curve that looks much more like the previous tests, but it only lasted about 17 and a half hours, just two hours longer than the 5V motor.

The stats so far:
Image

My guess is that at higher voltages, arcing is the primary source of brush and commutator damage, but at these lower voltages and longer lifetimes heat and friction are limiting the time and cycle life as well. The motors all look the same inside in the end, but at lower voltages the motors also seem to be venting smoke, which builds up as soot on the windings and motor cases. It may be that this happens all the time, but at higher voltages there is less time for the soot to build up.

I've got a 3V motor running now, and I'd like to think it will last the weekend, but...

-Adam
User avatar
nexisnet
 
Posts: 1216
Joined: Mon Mar 21, 2005 9:29 pm
Location: Boston, Massachusetts

The results are in!

Postby nexisnet » Wed Jan 16, 2008 10:30 am

Well the testing is all done, and the results speak for themselves.

The Results:
Image

There is a pronounced operational lifetime shortening as you increase the voltage (no surprise). Also, I had trouble with a couple of individual motors. Again, the first motor I tested at 4V (purple) died an early, but long and drawn out death, so there may have been some defect in it to begin with. The first motor I tried at 3V (dark blue) kept stopping short, but would mysteriously come back to life (and full speed) hours later, when no one had so much as opened the mini-fridge. I took it apart and found that the brushes were in fine shape, but the grooves had already started to form in the commutator and were full of soot. So, if you're having trouble with a motor stopping for no good reason, a blast of compressed air or some non-conducting lubricant (i.e. white teflon grease) through the vent in the plastic cap could make all the difference.

Speaking of difference, how does all this compare to the motor's performance at 3V (the recommended peak voltage)? Wow:

Image

Seriously! Stepping up to 4V from 3V cut the operational life down to less than 1/4 of nominal! Another way to look at the relationships between the good motors normalized to 3V, since these relationships should (ideally) hold for any gearing/loading (I defined the end of the "useful lifetime" as when the output speed drops past half of the overall average speed):

Image

There is a vaguely logarithmic relationship after all, and besides, every technical presentation needs at least one chart with a log-scale axis:

Image

The Verdict:
If you want more speed, you can have it, but will pay dearly for it in decreased operational life. That isn't to say don't do it, keep in mind that most hobby robots are run in short bursts infrequently, and you can replace the motors (Tower Hobbies has them here for $2.39).

EDIT: Pololu now carries replacement motors for the Tamiya gearboxes, as well as some nice higher-voltage motors you could swap out for the stock ones here.

It looks like somewhere between 4 and 5 volts gives you the best payoff, but beyond that you're really doing some damage. Speaking of damage, I have a nice little motor graveyard going on my desk:

Image

-Adam
Last edited by nexisnet on Tue Sep 29, 2009 1:50 pm, edited 1 time in total.
User avatar
nexisnet
 
Posts: 1216
Joined: Mon Mar 21, 2005 9:29 pm
Location: Boston, Massachusetts

Postby mprobo » Wed Jan 16, 2008 10:41 am

Wow! :shock: Thank you for doing this! :-D

That's some great info to have.
mprobo
 
Posts: 18
Joined: Tue Apr 04, 2006 10:01 am
Location: USA

Postby Ben » Thu Jan 17, 2008 12:39 pm

Thank you for your excellent work on this, Adam! This is extremely useful information.

- Ben
User avatar
Ben
Pololu Employee
 
Posts: 3440
Joined: Mon Aug 28, 2006 1:05 pm
Location: Las Vegas, NV

Postby shimniok » Mon Jan 21, 2008 8:55 pm

Many thanks!! I have been asking myself this very question over the last few days in designing my 1st competition bot. Extremely helpful experiments!! -michael
Michael Shimniok
http://www.bot-thoughts.com/
User avatar
shimniok
 
Posts: 42
Joined: Mon Jan 21, 2008 8:14 pm
Location: Centennial, CO

Re: Tamiya 'toy' motor testing

Postby slurp » Thu Aug 07, 2008 5:20 am

This sort of info is great - Thanks! :-D

I guess the novice needs to remember that motor controllers have voltage losses, the battery voltage is not always what is seen at the motor.... the technology used will have an effect on the voltage experianced.

I've not got to the experiments but I have been thinking about similar tests with regard to load vs. current. Steady state is one thing but shock loads really push the peaks up - I'm not sure how I'm going to measure/capture those.

regards,
colin
slurp
 
Posts: 15
Joined: Thu Aug 07, 2008 5:07 am
Location: York, UK.

Re: Tamiya 'toy' motor testing

Postby rva1945 » Wed Jan 07, 2009 3:53 pm

Hi:

I installed the Tamiya double gearbox in an Academy 1/25 german tank (jagdpanther) converted to RC by me. The power comes from a 6volts pack, an AM radio system and two electronic speed controllers. I know I must run the motors at 3V max most of the time (just small 6V bursts).

At first it was impossible to RC it because of the sparking in the motors. I soldered 3 104 (100nF) capacitors (one connecting both terminals and the other two between each terminal and the motor casing) and the problem was gone.

I wonder whether the sparking inside the motors is gone too, so I guess they will last longer. I will not make the lab test, just run the tank.

Regards
Robert
rva1945
 
Posts: 16
Joined: Mon Jan 05, 2009 5:24 am

Re: Tamiya 'toy' motor testing

Postby nexisnet » Wed Jan 07, 2009 6:11 pm

I look forward to hearing how long your motors hold out Robert, particularly since they're in use in an actual robot with varying power and shaft loading. I did have 0.1uF (aka 100nF) capacitors across the leads of the motors in the original test, but not capacitors between the leads and the motor case.

-Adam
User avatar
nexisnet
 
Posts: 1216
Joined: Mon Mar 21, 2005 9:29 pm
Location: Boston, Massachusetts

Re: Tamiya 'toy' motor testing

Postby rva1945 » Fri Jan 09, 2009 4:34 am

Please visit my topic

Motors replacement for Tamiya double gearbox

Regards
Robert
rva1945
 
Posts: 16
Joined: Mon Jan 05, 2009 5:24 am


Return to General Robotics Discussion

Who is online

Users browsing this forum: No registered users and 3 guests