Pololu Beacons Going crazy

Hello
I’ve recently bought a pair of Pololu IR beacons and I noticed that they often go crazy like this :
http://www.youtube.com/watch?v=lws4N8kiIQk <== this is a video of one of the beacons, supplied with a 9v battery as shown in this video
==>http://www.youtube.com/watch?v=v1lVDEkx5io
Is this normal?

When I try to get them to work together, one sometimes gives the opposite direction from where the other actually is, as you’re gonna see on the video below :
==> http://www.youtube.com/watch?v=–u3UEidmCw

I’m not saying they don’t work at all, I succeeded in writing a pretty good program and the robot follows the beacon correctly, but I don’t want for there to be any hiccups the day of the competition I’m preparing.

I would like for it to work exactly like another one I saw on youtube => Here is the link :
http://www.youtube.com/watch?v=8_2L0kTIe0A Do you see the huge difference between this beacons and mine?

Do you have any thoughts?

Thank you

Hello.

I cannot see your third video (looks like it’s private), but the first one makes me think you have a power problem, with the voltage dropping as the beacon starts transmitting. A 9V battery is not necessarily the best power supply for these, but it should work better than that. Do you have a different power supply available? What is the battery voltage doing when the beacon is connected?

- Jan

I noticed the battery voltage dropped to 8,70 V, but as you can see on the third video (I changed the video status to public), the other beacon on the robot is supplied with a 12V battery along with the motors. But I don’t understand why there would be a power problem since the datasheet says 6v minimum?

Your voltage is probably falling lower than the 6V but not showing up on your meter. Can you look at the power input with an oscilloscope? If you need to stick with the 9V battery, another capacitor (100uF or bigger) right at the board should help.

For the other problem, you can add some optical shielding between the sensors so that only the one facing the other beacon can see the IR. The sensors work much better from the front side, but they can still detect through the backs of their packages, so if one happens to be more sensitive than the other, it might detect more even when it’s facing a different direction.

- Jan

Hello,

I have searched all IR Beacon related posts and the above closely matches the issue I face with my IR beacon pair. I have one IR transceiver attached to an Arduino Mega micro-controller that reads the digital outputs and serially prints logic HIGH or LOW (1 or 0) for each corresponding direction. The Arduino is powered by an external power supply (8AAA NiMH battery pack) and so is this transceiver (powered through the Vin pin on the arduino). The other transceiver (not connected to the arduino) is powered separately by another 8AAA battery pack. I am also seeing very sensitive readings from the beacons. I suspect the problem is due to the power source not being regulated,
which brings me to my question -

What would be a reliable source of power for the beacons? I am aware of the input voltage range (6-16V). Would one of the Pololu voltage regulators be useful?

Also,

Could someone shed more light on optical shielding? What material can be used? I understand that the idea is to prevent the beacons seeing any other direction but from the front.

Thank you.

-Suj

Your power supply should be fine. Just 9V alkaline batteries in particular are not good for high-current applications, which the beacon sort of is.

Cardboard usually works well for the shielding. Be careful with some plastics and tapes, which might look opaque but can be transparent to IR.

- Jan

Hi,
What I don’t understand, is why the other guy’s pololu beacons on the video: http://www.youtube.com/watch?v=8_2L0kTIe0A

are working perfectly fine? And without any optical shielding?

The environment might not be the same. It looks like you have some IR distance sensors on your robot; were they on when you did your test? Do you have any other IR devices on during your test? You are also on the close end of the intended range of these units; does the performance get better if you have the beacons a bit farther apart? If you have them relatively close together, all of the sensors are flooded and it’s difficult to determine where the signal is from when they can all see it all the time.

- Jan

It’s true that I noticed a big difference when I put the beacon/base far from the robot. The problem happens only at close range, the lighting is average (60 or 75 W lamps), and as for the IRs, they’re Sharp GP2D12 and since they weren’t facing the beacons I don’t think the problem would come from them. My only worry is that when the robot is at a few meters from its arrival point(where the beacon/base is positioned), it would get wrong readings, and take a few seconds to right its trajectory (seconds which are very valuable in the competition).

Hello.

That’s where the optical shielding might help out. Have you tried it to see if it makes close-proximity detection more reliable?

- Ben