LEGO® Power Functions Infrared Receivers
Power Functions infrared receiver exist in three version. The initial one, had a firmware bug in Single Pin mode. This bug was corrected in second version, identified by a short blink on the green LED when power is connected. Finally, a third version with a more powerful motor driver was introduced in 2012 in Rock Crawler set (9398).
The new driver chip used is a Texas Instruments DRV8833, while previous versions (hereafter collectively called "V1") use an On Semiconductor/Sanyo LB1836M. While many capabilities can be deduced from these datasheets (eg. nominal current per channel is 1.5A for DRV8833 vs. a mere 400mA for LB1836M), nothing beats some experiments to test the limits. So I couldn't resist doing an extensive comparison test...
The setup is similar to the one I use to compare motors. The IR receiver is powered by a laboratory power supply, and one output is directed to the load (motor or power resistors) through Mindsensors CurrentMeter and VoltMeter (thanks to Mindsensors for these nifty devices!). Torque is again applied by weight lifting, but this time I used a hook and pulley to get longer test runs and more real life conditions. A HiTechnic IR-link is used to control the PF receiver by the NXT. Most tests were performed at 9V (nominal voltage of alkaline batteries) and 7.2V (nominal voltage of NiMH rechargeable cells).
No load consumption
This is the intrinsic power consumption of the receiver.
Not the most important parameter since it's pretty low compared
to motor current, but for the sake of completeness...
V2 consumption is constant, while V1 depends on active function and number of active outputs.
In this test, we measure current and voltage that can deliver the IR receiver at various load levels. To get constant load, I used power resistors, instead of motor. Motor drivers have integrated current limitation, and also thermal protection (they shut down if they are getting too hot). Using short pulses, driver chip doesn't warm up, and thermal protection doesn't trip. More precisely, outputs are activated, and half a second later, we measure current and voltage, then shut down outputs, see NXC program here.
This time, we load receivers and let them run for one minute. At high load, temperature in motor driver raises. When there is a risk to go above safe value, the driver shuts down its outputs and will restore them only after it has cooled down enough. If the load is still there, this cycle repeats, resulting in a "hiccup" behaviour.
- V1 receiver
- V2 receiver
Power dissipated in driver depends little on receiver supply voltage, as we saw in pulse tests. So thermal shutdown occurs at roughly the same current when using 7.2V supply.
Real world testing...
In order to make a more realistic test, I built a simple vehicle. I used a power hungry 5292 RC buggy motor to stress the IR receivers. The video below shows the compared behaviour of V1 and V2 receivers.
As you can see V2 receiver performs a lot better than V1... but only if the battery pack is up to the task! Regular Power Functions AA battery pack simply can't provide enough current to allow V2 reach its full potential. Each channel of V2 receiver can provide 1.5A continuous, totalling 3A. Internal resettable fuse of AA battery pack is a Bourns MF-MSMF075 rated only 750mA. Looking at the datasheet, you see that at ambient temperature, it trips in 20 seconds at 1.5A and 0.8 seconds at 3A!!!
That said, I can't recommend shorting the protection as I did in the video! In case of problem (eg. cable short) this could result in deterioration of battery box switch or cable contacts - and even burning/fire! Replacing the fuse with a higher rating one (eg. a MF-MSMF150) would probably be a good compromise between safety and performance.
Curves showing performance of various motors and motor pairs driven by PF receivers are available here.