generally if theyre tight or making funny noises theyre fu......finished

they really dont like it if theres areldite in them

Thanks for the advice Terry, but as much as I appreciate it, I probably could have sussed that bit.

I'm looking for a more scientific approach really, Bob? Terry? Whoever it was that made that dark blue car?

JB

A method we've used for bench testing transmission efficiency is to drive an airbrake (a plywood paddle!) and measure the power input to the motor for a given speed at the output, no reason why you couldn't mount an airbrake directly onto the motor shaft to carry out a similar experiment.

Sounds like a cycling turbo-trainer.
Brian

The way we have tested motors is using our battery/motor testing jig which we share with TPM.
http://www.greenpower.beamweb.co.uk/showcase?cmd=view&id=44
What we do is have an old motor wired to our halogen bulb load (two banks of 4 x 12V 50W MR16 bilbs (was 5x bulbs with the Yuassa's)) and use this as a generator. The motor under test is connected to this with a 1:1 chain drive. We have an RPM sensor on the motors (not shown in the picture above). The generator provides a load similar to the car racing at normal power.
We then power the motor and measure its current and RPM under load. We do this for a known good (new) motor first and compare the readings and validate against the Fracmo motor graphs.
Just found some info here:
http://www.greenpower.beamweb.co.uk/files/RotaryRacer/MotorTesting.html

I must admit we haven't done that for a few years (would be good for the new team members to check the one in RR8). As we have data logs from the car racing for each race, we can actually look at the data and compare the motors average power (voltage * current) versus RPM and check from race to race. Although we haven't done the maths for this (and we would have to take into account the speed controllers operation).
Our current motor is about 2years old and was only changed as we had drilled holes in the previous one for cooling and the rules changed. I really think that as long as they are not overheated, they will last for a long long while without much degradation, 5 races a year is only 20 hours of operation per year. I think the motor in RR8 is still not fully run in yet.
With the lower currents now in use, due to the smaller batteries, the heat generated has reduced significantly from about 124W down to 53W assuming the cars gearing is right. This will reduce the internal temperatures. It was noticeable to me walking around the paddocks at various races this year the lack of motor burning smells. It used to be very common ...

Wow... two really useful comments there. Thank you very much Far North and Terry.

I'll see if I can get hold of some MR16 Bilbs :^P

JB

Hi JB
What I'd say is
1) unscrew the brushes & take a look - plenty of 'meat' left? slide smoothly? no massive grooves in the surface that contacts the commutator?
2) bearings - check for endfloat & any run-out on the shaft. With the brushes out it should spin pretty easily with no rattle
3) check rpm, no load on 24V (with the brushes back in of course ;^) It should do 2000 +/- 100 rpm or so. and take about 2A. If it's going faster (say 2300) or taking much more amps it's bazookered (shorted turn or permanent magnet losing strength)
The other likely problem is one (or more) magnets coming unstuck. The symptoms are more subtle, generally just rather lacklustre performance, trouble is you have to take an end off the motor to be sure. You might be able to feel "something wrong" when you twist the shaft by hand. Or you might not...
Brian had magnets fall off after getting on for 3000 miles - probably due to being stored in a container over winter & thus getting damp inside & corrosion over 4 years.
Zebedee had a motor change this year because of ~1mm shaft endfloat & poor performance (the new motor performed EXACTLY the same....;^ ) I think that's after about 2500 miles.
I feel there may be a tendency for teams to blame their motor unfairly - only actual motor problem I've ever had was in Brian at Rockingham in 2009, where 2 of the magnets had fallen off; it still did 107 miles!

PS - apart from the motor we fried big time in max factor in f35 ;^) now edited to correct motor no- load current draw

So, just spinning it up on the car disconnected from any transmission gubbins, with an ammeter and Tachometer attached should give me an idea as to how useful or not this motor may be?

I won't be able to check until the 23rd at the very earliest, but if I post the outcome would you guys be able to give me your opinion? (maybe once you've scoffed some turkey though hey? :^P)

How can I check for a shorted winding?

From what I recall it turns smoothly but does make a bit of a clicky sound. It's been left in a cold wet lock up for about 4yrs.

What's the issue with it spinning faster? isn't that a good thing?

JB

Yes, its a good simple test, but may be a bit off due to bearings being run in and manufacturing differences.
Two reasons I know of why it may be spinning faster, off load, than expected are:
1. The bearing and brushes have run in and so there is less holding it back (or warmer so the grease is thinner) . A good thing.
2. A motor will spin faster and faster until the back EMF (electro motive force, (generated voltage)) equals the voltage applied (24V in our case) minus a bit for the energy overcoming friction. The back EMF is is proportional to the speed x the magnetic flux. The magnetic flux is proportional to the strength of the permanent magnets and the number of turns of wire on the armature. So if the magnets are weak or there are shorted turns then the motor will spin faster until its back EMF equals the 24V applied. This is bad (especially the shorted winding as this will consume energy and produce heat).
http://www.animations.physics.unsw.edu.au/jw/electricmotors.html
We have found that a good, run-in, free running motor (after a year of racing) does about RPM: 2180 @ 2.0A from our rough tests. When this motor was nearly new it did 2120 RPM @ 2.21 Amps (different to your 5 Amps Bob ?? (Have to check the meter we use at GP sometime ...))
We have found it difficult to determine if there is a shorted winding, hence our motor tests under load, the power input levels for the same output (efficiency) show this up well (I think ...).
The motors do sound a bit "clicky" due to the brushes on the commutator, especially if you turn it backwards to its normal running direction.

Argh - sorry - I was just going by (increasingly dodgy) memory with the 5A & I obviously got it wrong.... I remember it being a nice "round" figure ;^) I'm not in a position to check any more (school locked up for the season....)
Also remember, 2 average batteries will probably give more or less than 24V...... it's pretty important to do the test at an exact voltage!
I'm trying to think of a way to measure rpm accurately if you don't have a calibrated tachometer - and I really can't think of anything sensible - any bright ideas out there?

A strobe (and a dap of tippex) give accurate speed readings - don't know how much they cost though.
Brian

We have a pretty standard brake dynamometer setup - a heavy steel drum on the shaft with a fabric/leather belt to brake it. When we're back I'll do some more digging into it and get a picture. I know it doesn't give a power output but we're using it to provide load so we can test the control software.

To measure speed we use an IR optical sensor and a disc of laser printed OHP film, the lines on it calibrated to give an easily loggable reading. Simples and incredibly lightweight, and seems to work well while driving too.

Must admit that we haven't benchmarked anything, been too busy getting the car running this year!

Hi Terry, i am currently making a similar test rig to the one you have which will help us evaluate a NuVinci CVT as well as testing our motors and was wondering if your light bulbs are connected in series or parallel? Does this matter?
Thanks, Andy

For RPM measurement, our way was like PeterF's although we went simpler by attaching a strip of cardboard from the back or an A4 pad to the motor shaft (about 8cm x 4cm with a hole in one end for the motor shaft and then used an IR LED and sensor either side of its path fed into our spare PIC based car computer's speedo input. Could use a bike speedo with a pulse divider IC before it as well instead of the computer possibly (unless the speedo can handle 250Km/Hour ish :) )...

AndyP, we have two separate banks of 4 x MR16 bulbs (for 17A ish, 5 x bulbs for 23A ish). Each bank has the bulbs wired in parallel for 12V operation. When testing an individual 12V battery we use one of these banks. If we test two batteries, 24V, or drive them from a motor as a generator then we connect these two banks in series for 24V (12V + 12V) operation. You could also use BobC's resistor based load system documented in one of these forums.
By the way soldering to the bulb pins is difficult. We used a small "chock block" terminal connector strip to do this and remember each bank of 4 MR16's will produce 200W of heat ...

Thanks very much Terry, that was the bit i was finding it difficult to get my head around. I think some other students in the team are working on Bob's system to test the batteries but for now i think i will stick with the bulbs as it seems simpler. Do the pins have to be soldered? Or can i just buy some normal bulb holders like you get in table lamps or similar and just wire them together?

Hi Andy - the MR16 has two pins protruding from the rear and usually have dedicated fittings that are housed in a ceramic cup; what Terry suggests is to use 'choc block' (terminals that have two screws to make contact and are heavily insulated - available in long lengths that you cut to the quantity needed) onto the two pins and take wires from those fittings to wherever you need.
Mr T

Oh right, that makes more sense now, thanks