wheel and gear configuration VS. efficiency

[mattallen37]

linear voltage regulators are actually fairly inefficient. For analog voltage control (or any DC voltage drop), a switching power supply is certainly more efficient. For simple testing purposes though, a linear voltage regulator would be fine (like the Lego 9v train controller, which uses the LM317), and would cost a whole lot less than a switching supply.

Lego axles, though they seem "square", they are actually cylindrical with "corners" "cut" out. There is usually very little friction in a solid, stiff frame.

As far as worm gears, remember that there is also the friction of it pushing on whatever is in front of it (like a technic brick).

[jwiger]

I planned on using a 5-15 meter course, so I don't think the 9v train regulator with wires and all would help the consistency. To get the most direct reading I planned on using the Energy Meter as the power source. It has its own battery (of seemingly small capacity) and a Male PF connector for output.



It is recharged using a Female PF connecter on the back side. And has a Mindstorms connector on the top edge, which I will use to log the data. Using PWM with this set up is simple, I'll simply connect a PF IR receiver to the power out on the Energy Meter, then connect the motor to the IR receiver. Unfortunately I can't think of a way to regulate the voltage in an analogue fashion. There may be ways, but I would prefer to stick to LEGO solutions for the sake of repeatability (that is to say some one else could find the same results).

I planned on giving the Mindstorms brick up to four tasks. Of which it will use its own power, so that it doesn't impact the findings. The tasks are:

1. Use a servo to turn the "power control knob" on the front of the energy meter. This rotary knob has three postitions (-1, 0, +1) and controls the power at the male PF connector on the front similar to the orange switch on any PF AA or AAA battery box.

2. Use a HiTechnic rotation sensor directly attached to an extra drag-along pace wheel: This will be used as an odometer and will cue the Mindstorms brick to end the test run.

3. Log the electrical power used. I believe I can do this two ways: A. attempt to log the power usage in real time. B. Sample the power available right before the run and right after the run, then find the difference. ( I will probably use plan B so that the processor can commit to the odometer.)

4. Operate a HiTechnic IR Link for the PF IR receiver. During the PWM tests the brick will have to operate the control knob on the Energy Meter, switching power to the IR receiver, which its self uses power. Then send an IR signal to the receiver initiating the test run, as well as ending the test at the end of the run. I know that since the IR receiver uses power, that I can run a few side tests to see how much power the receiver uses while idle and while commanding the motor to run, this should be easy enough to figure in to the power used during the test runs and work around it.

What do you guys think? Have I overlooked anything?

[h-g-t]

Re the effects of friction, I replaced a gear train with a simple worm/ gear drive which had a higher ratio but was much easier to turn.

When I rotate it then let go it carries on for a few degrees, whereas the gear train stopped instantly.

[jwiger]

Finally back in Texas! And after taking care of a few things, I've had a few moment to play around with the Energy Meter, Here is what I've learned:

1. The number of Joules displayed is separate from the actual charge on the Energy Meter's battery (Lego Energy Storage).
2. The output port on the Energy Meter is regulated close to 9.0 volts. With switch in the off position it shows about 9.5 volts. While powering the E-Motor with no physical load the output settles at 9.0 volts. I tried loading it down by driving nine PF-M motors unloaded at the same time (load was over 3 watts) the voltage dipped to 7.5 volts but this was right before/as the internal battery died. I will let it fully recharge then try this again, I have more motors and I'm curious what the max load is for the Energy Meter. I will find its limit even if I have to connect all five of my RC Buggy motors
3. When the display shows zero Joules (0J) it stops driving the output. Again the charge level of the internal battery is irrelevant (unless the internal battery dies). If you start putting energy into the Energy meter it restores power to the output port.
4. When the display shows one hundred Joules (100J) it turns the input port off, until it drops down then turns the input port back on until 100J is reached again.

If you have a load that is drawing more power then what is going in you will see the display counting down the Joules until it reaches zero, then the output port will begin cycling on and off as the display measures one Joule (1J) then back to zero. If you have a load that is drawing less power then what is available at the input, the display will count up while driving the load. When the display reaches 100J it begins the cycle as mentioned in step 4 above while continuously driving the load, until the internal battery dies.

I'm not sure what conditions are specifically required to charge the battery on the Energy Meter. I have a few theories but I'll do some more testing on that later. Right now I'm just having fun playing around with it, and getting familiar with what it can do.

As far as the E-Motor it has a few unique characteristics that I didn't really expect. For one it's larger than I expected, physically it's between the size of the M-Motor and the XL-Motor. Philo broke out the weights for each and if you study pictures it's obvious that the size is between the two PF motors. Still seeing it in person, it's larger than I expected. Secondly it has very low start up inertia and current draw with no load. I've connected it to a PF-M motor and when the M motor is slowly turned by hand it easily drives the E motor. When the E motor is turned by hand it is very hard to get the M motor to start turning. I haven't tried any gear train to increase speed, but cranking by hand with a direct drive is a challenge to see life from the M motor. Additionally connecting the E motor to an XL motor and hand turning the XL motor cause the E motor to turn very easily (excepting of course how much force it takes to turn the XL motor by hand on its own).

More to follow...

[inxt-generation]

Very interesting so far! Can't wait to see what else you find out.

I will find its limit even if I have to connect all five of my RC Buggy motors

Woah. You have five Buggy Motors?!?!?! Most people are lucky just to have one. *sigh* I only have one. I would be happy even with two, let alone five. All the BL prices are way too much, and they rarely appear on EBay. *sigh* So is the way of a 13-year old where the best parts went out of production before he even heard of MindStorms.

[jwiger]

Yeah I got two of them with a buggy set. Then I was lucky enough to be stationed in Germany for four years. I found the third in the motor accessory set, sitting randomly on a shelf in a German toy store. The fourth and fifth were also motor accessory sets in a small town just South of Billund in Denmark, I was lucky to spend a week up there doing training with the Danish military, and just happened to find them.

[inxt-generation]

Yeah I got two of them with a buggy set. Then I was lucky enough to be stationed in Germany for four years. I found the third in the motor accessory set, sitting randomly on a shelf in a German toy store. The fourth and fifth were also motor accessory sets in a small town just South of Billund in Denmark, I was lucky to spend a week up there doing training with the Danish military, and just happened to find them.

Well, then. I guess being sent everywhere by the Military does have it's pluses! (No offence, I've got nothing against the Military. I just wouldn't want to be sent who-knows-where for who-knows-how-long all the time. Not exactly me tasse de thé.)

[jwiger]

I may have just thought of a way to control motor speed through variable voltage, I have an old Radio Controlled truck with an electronic speed control. I think it's built for 7.2-8.4 volts, but with a rather high amperage. It connects to the radio receiver as a servo.

Matt, et al: Is there a easy way for me to throttle this speed control with stock NXT hardware (I suppose I could hack a cable) and NXT-G software?

[mattallen37]

You want to use the NXT to control the H-bridge of an old RC vehicle? If it's a cheapo RC car with the RF receiver and motor controllers all on one board, you would need to isolate the H-bridge from the rest of the RF receiver and such, and wire directly to the control pins of the H-bridge. I haven't ever done it, but I suppose it should be possible. If it's a real RC car (servo motor for steering, separate ESC, etc. like a traxxas), then you should be able to control it with any servo controller.

[jwiger]

Matt: It is a real one individual components and all, and built from a kit (Tamiya Super Blackfoot with Futaba 2ch control). I think you indirectly answered my question. Can you confirm that there is no simple way to throttle the ESC from the NXT brick with out a third party servo controller.

Also for those following my experiment I just learned something about the Energy Meter that may complicate things for me when I attempt to use the PF IR Receiver with it. I had planned to use the IR Receiver inline between the output port on the Energy Meter and the motor. However the IR Receiver derives power from the outer pins on the PF connector, which are not switched by the orange rotary knob. In other words as soon as the IR receiver is connected and the Energy Meter is powered up it begins drawing power (~150mW). This will make capturing the energy used a little more complicated, but it should still be do-able. In my favor I have also learned that if the orange knob is turned to +1 or -1 it will send power as soon as the Energy Meter is powered up, I think my strategy will be to use a servo from the NXT to press the green power button. Or just skip the IR receiver all together and stay to my original concept of just comparing gears, wheels, and also motors.

As far as the input on the Energy Meter, it only seems draw power from the C1 and C2 pins, not the GND and +9V outer pins, this means you can use a PF extension cable with any older 9v motor as a generator and charge the Energy Meter. However, connecting to a PF IR receiver output that is not in a drive state will not charge it. Once you command the IR receiver to a drive state it begins charging.