DIY wireless sensor bar distance test

[billcsho]

Summary: The pointing distance limit of a wireless sensor bar is around 12 times the distance between the two groups of InfraRed LEDs at the two ends of the sensor bar.

Problem:
I am playing Wii with a projector half of the time. The projected image is usually between 60 and 80 inches (diagonal). I bought a Nyko wireless sensor bar but I was very disappointed by the distance limit when playing Rayman Raving Rabbids. It keeps telling me to get closer to the TV. This really makes it difficult to play Wii on a very large screen at a distance. In both the original sensor bar and the Nyko wireless one, the distance between the left and right IR LED is around 8-9 inches. The fact that the WiiMote can see the IR from a longer distance during calibration (though not much longer) made me suspect the problem being the spatial resolution of the two groups of IR by the WiiMote.

Solution:
I got a 20 pack IR LEDs from Dealextreme for $2.90 shipped. For each side, I connected 6 of them and a green LED in serial and put in a 1.75"x1.75"x1.75" plastic box. I don't think you need that many IR LEDs but this will ensure a stronger IR signal to be detected at a longer distance and a total of 7 LEDs eliminated the need of resistor when it is connected to a 9V DC power adapter (~1.2V drop per LED). I made 2 of this LED lamp boxes (sensor cubes) and place them at various distance. First, I tried the sensitivity calibration. Surprisingly, I found the signal strength from these "wireless sensor cubes" to be just on par with the the original sensor bar even they have 3 times the number of LED. It gives around 22 feet of detection limit (before it starts blinking in maximum calibration setting). Nevertheless, through a digital camera viewer, the 6-LED sensor cubes showed a stronger intensity than the Nyko wireless sensor bar. I think the LED I got has a shorter emission wavelength. Even the intensity is higher, it was not detected by the WiiMote with higher sensitivity. More LED may be needed if you need to play Wii at a distance longer than 20-25 ft from the sensor. As for the pointing sensitivity test (spatial resolution of the 2 IR LED groups), I found placing the 2 sensor cubes farther apart gives me longer functional pointing distance. There is roughly a factor of 12 (i.e. placing them 10" apart and the WiiMote pointer works up to around 10 ft). With the original sensor bar, the resolution of the 2 IR groups limits the gaming distance in Rayman Raving Rabbids to less than 8 ft. It was very annoying keep seeing that "Get Closer to Your TV" message blinking from time to time. With the homemake sensor cubes placed 16" apart, I can play the same game up to 18ft away. I've tested up to 24" and I can play up to 22-25ft where it is limited by the IR signal intensity.

More test:
One last thing is to test the optimal ratio in the distance between the 2 IR LED groups and the screen size. This is critical for shooting game using one of those gun adapter that you want to line up your WiiMote with the onscreen pointer. I found if the sensor cubes are placed around 1/4 of the width of the screen (not diagonally measured), then the WiiMote points almost right at the pointer on screen from side to side. I tested it with the shooting game in WiiPlay. In other words, the original sensor bar is best for a 37"-42" TV. For an 84" projection screen, one should place the sensor cubes (IR LED lamps) at ~16" apart. The actual point angle would depends on how far away are you sitting/standing from the screen. Using 2 separated IR LED sources allows you fine tune the pointing angle to improve gaming experience.

Final words:
I also tested the IR LED lamps with a regular 9V battery, but the signal was too weak. Therefore, I have to use a DC adapter and I put a 3.5mm port on the box for that. I have tried using a retractable wire from a USB cable thinking that would be neat to have adjustable distance between the 2 lamps and use only 1 DC adapter. It turned out that the wires inside the retractable cord are too thin (high resistance) and the coating made it hard to solder. I ended up using an audio cable from a desktop speaker to connect the sensor cubes. You may want to get slightly thinner cable to connect the 2 boxes. My cable is too heavy that keep dragging the box and change the orientation of the boxes.
:aureola:

 

[iceblade]

bilcho thank you very informative thread. I'm thinking of using 3 IR LEDS for
IR each source with two AA batteries. Which type IR LEDs and which type of resistor do you think I should use?

 

[Wiinter]

I was actually designing a sensor bar (mainly for my PC) with a focus on adjustability--a slide for adjustable distance between both sources, a left-right swivel for both lights and a separate up-down swivel for both, presumably allowing me to get perfect placement and then mark the settings so I can easily move it between different locations. Yours isn't quite as feature rich, but it's a lot easier and still has a lot going for it. You should have this up on its own website, or least link to this thread in your sig so others can find it.

Questions:
Did you use 880nm or 940nm IR? What viewing angle did they have? What radiant intensity?

Where'd you get the 3.5mm port you put in the cube to plug your DC converter into (I was planning on stripping the end of the DC converter and using one of the universal plugs from Radio shack)?

Did you angle any of the LEDs for a wider viewing angle (the more off to the side you are, the less strength the IR signal is going to appear to have), or are you counting on just directing both cubes to point straight at you? Do you think it would be effective at multiplayer where you and the other guy are a few paces apart?

Since you've got 12 leds that can pull 1.2 volts each, and only one DC brick (right?), you could have something closer to 14 volts and get a little more intensity, right? Did you chose th 9 volt for a particular design reason, or was it just what you had?

Finally, the only effect moving the sources away/towards each other should have is to make the source appear closer/more distant to/from the wii remote. It's helpful if you can't move it any further away for reasons of space or signal strength--but given this, it's important to know how far away you were standing when you did your reticule calibration test. Your 1/4 screen distance measurement shouldn't lead to identical calibration at all distances, should it? If you get a good formula perfected, Wiili.org could benefit from your experimentation.

 

[billcsho]

Questions:
Did you use 880nm or 940nm IR? What viewing angle did they have? What radiant intensity?

It does not specify on the site, but I think it is only 850nm. Read my report for more details. I have the LED protuded out and you don't worry about viewing angle at such distance. I tried at least 90 degree (+/- 45 from the center)

Questions:
Where'd you get the 3.5mm port you put in the cube to plug your DC converter into (I was planning on stripping the end of the DC converter and using one of the universal plugs from Radio shack)?

Did you angle any of the LEDs for a wider viewing angle (the more off to the side you are, the less strength the IR signal is going to appear to have), or are you counting on just directing both cubes to point straight at you? Do you think it would be effective at multiplayer where you and the other guy are a few paces apart?

I pulled the 3.5mm port out from an old desktop speaker. It is mono but even a stereo 3.5mm plug would also work. Of course you can get it from RS.
I think I answer the other question already. I played fishing and tank with my daughters sitting 6-8 ft apart at around 6-8 ft from screen without problem.

Questions:
Since you've got 12 leds that can pull 1.2 volts each, and only one DC brick (right?), you could have something closer to 14 volts and get a little more intensity, right? Did you chose th 9 volt for a particular design reason, or was it just what you had?

Every LED takes 1.2-1.5V. It is not linear when you daisy chain several LEDs together as they are not simple resistor, but you care much if each LED in the same group has slightly different intensity. My original thought was to use a 9V battery and avoid the use of resistor. So putting 7 LEDs togethers sounds perfect (rechargeable 9V batteries are actualy 8.4V). The boxes from the 2 sides are connected in parallel (not in serial) so I only need to provide a 9V DC power for the whole thing.

Questions:
Finally, the only effect moving the sources away/towards each other should have is to make the source appear closer/more distant to/from the wii remote. It's helpful if you can't move it any further away for reasons of space or signal strength--but given this, it's important to know how far away you were standing when you did your reticule calibration test. Your 1/4 screen distance measurement shouldn't lead to identical calibration at all distances, should it? If you get a good formula perfected, Wiili.org could benefit from your experimentation.

Well, that is what I exactly said at the end of the report. When I did the pointing calibration initially, I was around 8ft away. Then when I try it again at around 12 ft, the pointing position changes as expected. So one need to calibrate the pointer at their position. Since I don't have a Zapper-like adapter to really aim at the pointer, 1/4 screen size distance measurement is only a rough estimate. But the basic idea is one may adjust the distance between the two IR signal to calibrate the pointing position on screen according to their screen size. I did not mention the Y-axis calibration, but I assume this should be proportional to the x-axis and it can be easily adjusted by the position of the IR signal.

 

[billcsho]

bilcho thank you very informative thread. I'm thinking of using 3 IR LEDS for
IR each source with two AA batteries. Which type IR LEDs and which type of resistor do you think I should use?

Try using 2 AA with 2 LEDs or 4AA with 4 LEDs (in serial). Others has posted the use of 4AA with 2 LEDs with a resistor. You'll need to calculate the amount of current going through to determine the resistance required to drop the voltage from the battery source to the level required (~1.2-1.5V per LED). This may depend on the specification of the LEDs you get.

 

[rrg]

I saw Radio Shack has the 940nm. Do you think that will help the distance?
They are more expensive.:sick: One for $1.79.
But if fewer are needed for the better results it might be worth it.
I looked around and can't seem to find what the Wii bar uses.
What's your take?

edit:
I just found that the latest speculation is that they are 940nm.

 

[billcsho]

If the emission intensity is the same, then the 940nm should have a much higher efficiency to be detected by the WiiMote. The OEM sensor bar has 5 IR LED on each side arranged in o ooo o pattern. The Nyko wireless has one big and 2 small IR-LEDs on each side in ooo. The middle one of the 3 appeared brighter due to the radial arrangement and the middle one is pointing forward. Mine uses 6 of the shorter wavelengths ones (emitting red light) and the signal is comparable to the OEM and the wireless one on WiiMote calibration. If you get a higher intensity longer wavelength one, you should be able to use less LED. The picture below has the Nyko, my home made one, and the OEM one (one side of each shown).

 

[kenn]

Quick question:

Is there a technical limit, within reason of course, in the distance apart? I thought it would be interesting to put half an IR emitter on the left and half on the right side of my TV halfway from top to bottom. This would put the wii pointer very close to it's actual pointing location on screen rather than an "offset".

 

[Wiinter]

Quick question:

Is there a technical limit, within reason of course, in the distance apart? I thought it would be interesting to put half an IR emitter on the left and half on the right side of my TV halfway from top to bottom. This would put the wii pointer very close to it's actual pointing location on screen rather than an "offset".

There's a relative limit. If you're too close, your wiimote won't be able to see both sources, and if too far that the IR signal would be too weak to detect.

 

[billcsho]

There are a couple problem with putting the IR on the 2 sides of the TV. First, in the Setting menu, you only have the options to put it either above or below TV. Placing it in the mid level is better for the sensing, but you don't have the way to calibrate it. Second, the detection range of the IR sensing is between 2x to 12x the distance between the two IR. Practically, you need to be at least 3x the distance between the two IR sides in order to use on screen pointer. When you are at 2x the distance, the 2 IRs would be at the edge of the calibration screen. You need to leave some room to allow the 2 points to move around. If your TV is 30" wide, you need to be 8-10 ft away. On the other hand, if the IR signal has a detection limit of ~25ft (depending on the intensity), there is no reason to place the 2 IR groups greater than 25" apart.

 

[Wiinter]

There are a couple problem with putting the IR on the 2 sides of the TV. First, in the Setting menu, you only have the options to put it either above or below TV.

If you're reason for putting it there in the first place is that the "calibration" in Wii settings doesn't achieve accuracy, then you calibrate by moving the IR sources. That's not a problem, it's a solution.

I tell my Wii that the sensor bar is underneath the TV, and I put it on top. Partly because I love pulling one over on Revie, and partly because that just plays more comfortably to me.

 

[billcsho]

It is better to put the sensor bar closer to your arm level that the IR image would be placed in the center of the calibration image when you are holding the WiiMote the way you like, disregarding it being on the top or bottom of the screen. That would allow the detection of maximal movement.

 

[kenn]

Ok, thx for the feedback. I think I'll just leave it as-is. It seems to work fine. I have noticed, though, that when the batteries in the Wiimote are "fresh" everything seems to be crisper, sharper, more accurate, etc. than when they are about to get a "warning" message from the console. When the batteries are 2/3 gone, I get the "move closer" message from wii sports occasionally but with fresh batteries, I'm good to go. I just bought a Nyko charger base with rechargeable batteries so I'm hoping now that I'll always be "fresh" since I really don't ever play for more than a couple of hours at a time.

 

[billcsho]

Personally, I think those WiiMote chargers are just waste of money. The battery capacity is actually lower than my NiMH AA batteries and there are several reports of killing the WiiMote. Well, at least it is not the Nyko one, so you should feel better.

 

[Wiinter]

I buy my energizer rechargables, 2500mah. I have to take them out to recharge them, but the battery life is decent.

I worry about whether Nyko et al were using a circuit to prevent overcharging. That would suck if they didn't