NXT battery test

[jwiger]

Unfortunately my stuff is still somewhere between Germany and Texas. But please go ahead and test yours out. When i get a chance I will do the same with mine. I have a fairly simple multi-meter that can measure current, I'm just not sure if it's rated high enough for these motors. But I will try. perhaps there is some kind of shunt we can use...

[kvols]

As soon as I have a few hours, I'll make some initial measurements. I think I can borrow 9 or even 12 motors, set the brick to run on a stabilized power supply, and then measure the motor power consumption under different conditions. The goal should be to find a way to generate a well-defined amp. load. I'll keep you posted. Thanks again for your interrest!

[jwiger]

My job in the military has me dealing with portable radios a lot, and we have a bunch of rechargable batteries that I wanted to test. A generator mechanic at my shop suggested using light bulbs as a load, since the wattage on them is actually very consistent. and rigging a timer would be fairly easy. but then again we want to test what batteries are best for the NXT. I think we should look at what types of loads a typical NXT brick will see, most will only have 1-3 servos. I don't think testing with a large gang of motors would reflect pertinent data.

[kvols]

I agree very much, that light bulbs, or other resistors would probably be more precise that motors. Using the motors does have a bunch of disadvantages (from precision load, regardless of the wear on the motor, the test will be noisy, and the test will cause some wear on the motors.

Regardless, I think it would be very interesting to be able to:

• - Create a quite precise "standard" test, that everyone with a MXT Mindstorms set could perform, without any other special LEGO or other hardware.
  - Make many tests, perhaps of many different battery types, or test a specific battery type many time, adding better precision.
  - write the software test, figuring out how to create a well-defined load given the restrictions.

I have seen a few battery tests on the net, but there is always one more battery I'd like to compare to the tested batteries, but I would have to depend on the test being performed on that specific battery, because I cannot reproduce the exact same test, and compare the results.

There are also other factors to think about:

1) Usage pattern
you already mentions mimicking a typical LEGO usage pattern, but maybe there a more patterns: Very low load (no motors) to very high drain (all motors running on full throttle or so), and something in between....
2) Battery load characteristics.
How does the battery behave under load: Does its voltage fall drastically under load, or not at all?
3) Battery voltage characteristics.
We might not only be interested in the number of mAh, but also in which voltages the power is distributed. (Will a robot run fast for a short period of time, then slow for a long time, or will it run fast, and the suddenly run out of juice?)

I hope that an easy to do battery test could create enough tests to describe a lot of batteries, given the size of the community...

I'll keep you posted!

[jwiger]

I know this is an old thread but I'm still interested. Good news is I got my Lego Energy Meter, bad news is the military deployed me until the summer...

Have you made any progress finding amp draw on the servo motors? Looking at the literature for the Energy Meter it seems it can basically "absorb" any voltage between 0.1-9.9 volts to charge up. Then can dump 100 Joules at 9 volts. I'm interested in seeing how many times I can get 100J out of a single alkaline AA battery, maybe repeat the test with 3-4 AAs to obtain an average. Then repeat the process with AA's of different chemistry.

I think doing this with any of the NXT servos will be largely based on averages; due to age, and health of everybody's servos.