Gun4IR - 4 IR points must be detected - tried 2x Camera Moudles so far

[m1n10n]

Hello,

First Gun4IR build - I am currently hitting an issue after attaching the Camera.
When I go through calibration I'm unable to make it through due to error:

The 4 IR points must be detected in the camera preview:


In the image we can see that there are 3 points detected ... but if I back up further I can see all 4 ... but I still can't do calibration.

I thought at first, this was due to the IR LEDs not being bright enough - but all LEDs were emitting light fine - I then went as far as to bridge one of them in each cluster to make the other two brighter, as I thought maybe the IR LEDs perhaps were not bright enough ... I'm not thinking this is it though (please let me know of your thoughts though).

I've gone through testing here and made a video to share, please take a look - the one strange thing: It's hard to see but: near the end you can see that the tracking of the mouse picked up in the OS, despite jumping around quite a bit, seems to be tracking backwards when moving the gun left to right .... I thought I must have a wire backwards but checked the wiring diagram (pdf and videos) and I seem to have it correct.  I've even taken out my 2nd camera (for my second gun) and wired that one up and my issue is identical.

Thoughts on what I'm likely doing wrong?

[m1n10n]

OK ... so far I figured out:

-The gun tracking backwards was due to the "TOP" logo not being set or the camera being rotated in the wrong orientation - got that all sorted.
-The tracking is still pretty jumpy - going to work on getting more IR light out of these LEDs ... or maybe I guess I need diff LEDs/diff resistor ... setup ... (using 37 ohm 0.5w)
-Looks like I had to have the trigger wired up to make it through this menu (for some reason I though I could use CTRL on keyb or something ...)

[JayBee]

OK ... so far I figured out:

-The gun tracking backwards was due to the "TOP" logo not being set or the camera being rotated in the wrong orientation - got that all sorted.
-The tracking is still pretty jumpy - going to work on getting more IR light out of these LEDs ... or maybe I guess I need diff LEDs/diff resistor ... setup ... (using 37 ohm 0.5w)
-Looks like I had to have the trigger wired up to make it through this menu (for some reason I though I could use CTRL on keyb or something ...)

It looks like your LEDs are awfully weak and low quality, you won't get very far with this model I think.

[RandyT]

I'll second that.  Unless there is an IR blocking filter on the camera being used, those LEDs should be bright and nearly white.

OP: Bridging LEDs is typically a bad idea, depending on the circuit.  But given the low output (unless they are already destroyed) they might still be ok.

What voltage are these running at and how are they wired?  At the typical 1.7 to 1.8v, three of them in series on 5v would need 1ohm or less (really 0, as there is barely enough to light them up). 2 in series is ~75ohm, or twice what you have.  Keep in mind that once you step over the ratings, output will decrease and the LED is immediately damaged irreparably.  Changing things without understanding the circuit and the specs of the parts in it, is a recipe for disaster.

[m1n10n]

I'll second that.  Unless there is an IR blocking filter on the camera being used, those LEDs should be bright and nearly white.

OP: Bridging LEDs is typically a bad idea, depending on the circuit.  But given the low output (unless they are already destroyed) they might still be ok.

What voltage are these running at and how are they wired?  At the typical 1.7 to 1.8v, three of them in series on 5v would need 1ohm or less (really 0, as there is barely enough to light them up). 2 in series is ~75ohm, or twice what you have.  Keep in mind that once you step over the ratings, output will decrease and the LED is immediately damaged irreparably.  Changing things without understanding the circuit and the specs of the parts in it, is a recipe for disaster.

Thanks for all of the inputs so far.

I followed  this video to a "t":



The LEDs are:
CHANZON 5MM IR LED V(F): 1.2-1.5V I(F): 20mA  (940NM)

I have 3x in Series w. a single 0.5w 37r resister for each section ... and the one was bridged in the video because I also thought they were not very bright at all and was trying to isolate the issue to the LEDs/Camera/etc ...
I had a standard 5v phone charger on the end but later found a small AC adapter that was like 6v ... (also messed around w. a variable PSU and cranked it up and just felt the resistors get really hot in that test lol)
... then you can see I bridged on of the leds to try and increase brightness (which it did) and then I went as far as to bridge two of them to give me one single IR LED which was a bit brighter.

Doing the math ... it seems for 5v in this scenario that 37r would be appropriate ... but I'm sure I've missed something - or these LEDs are just garbage and I need different ones ...

√(.5/37) = 0.19A ... so ... maybe this would have worked for just 1xLED not 3x in Series ... not sure ... (but again even jumping the other two to make this formula work for 1x IR LED it's still not so bright)


(I have yet to test at night - but will do this also - I am thinking it could work "OK" at night possibly ... )


OK ... so ... back to the drawing board on this LEDS setup it seems ...   Feel free to correct my math - I think I'm having trouble calculating the voltage drop in the formula from the IR LED rating...
(Also just thinking of using totally diff IR LEDs)
 
RE: LEDs burning out - I have a pack of 100 - they don't seem to be burning out - but are obvi not very bright either ... when checking on the camera...
NOTE: I didn't think this through and just purchases what was in the video mentioned as I figured it should all work fine if I had the same parts etc ... but ... I'm obvi doing something wrong.  Wonder if it's just due to the tv being bigger than a standard PC monitor and needing a more robust IR setup?

[m1n10n]

I'll second that.  Unless there is an IR blocking filter on the camera being used, those LEDs should be bright and nearly white.

OP: Bridging LEDs is typically a bad idea, depending on the circuit.  But given the low output (unless they are already destroyed) they might still be ok.

What voltage are these running at and how are they wired?  At the typical 1.7 to 1.8v, three of them in series on 5v would need 1ohm or less (really 0, as there is barely enough to light them up). 2 in series is ~75ohm, or twice what you have.  Keep in mind that once you step over the ratings, output will decrease and the LED is immediately damaged irreparably.  Changing things without understanding the circuit and the specs of the parts in it, is a recipe for disaster.

Interesting - yeah ... this is why I started jumping them one at a time and watching the brightness of just one increase ... Seems this is not the best way to do this?

[RandyT]

I'm still not convinced that 940nm is the best wavelength for these, but to each their own on that front.

The video indicated a 27ohm resistor, not 37.  However, 37ohm would have put the per-LED forward voltage at a conservative ~1.4v, so that shouldn't have been terrible, based on the stated specs.   

On the other hand,  it's difficult to know what those are based upon.  Sometimes the Chinese sellers will spec these with a range, with the lowest value being the moment they begin to emit, which is not very useful, but easy to test.  The real question is what the higher value in the range is based upon, sans a proper datasheet with testing charts.  It may be that they are intended to operate at the max of 1.5v, which is what that 27ohm resistor will provide.

Now the issue you have is that you attached different power supplies.  At 6v, with the 37ohm resistor, you fed the LEDs double what they are rated for.  While they may not be completely dead, I am 99.9% convinced that they are severely damaged, and not emitting what they should after that.  If you went higher, then change that to 100

What I would do at this point is rebuild the circuit with new LEDs AND a 27ohm resistor AND use a 5v supply.  All 4 of the modules should be running in parallel off of the supply.  But if for some reason you have the modules in series, that will never work.

If that doesn't work for you, you will likely need to look at some different LEDs. 

[m1n10n]

I'm still not convinced that 940nm is the best wavelength for these, but to each their own on that front.

The video indicated a 27ohm resistor, not 37.  However, 37ohm would have put the per-LED forward voltage at a conservative ~1.4v, so that shouldn't have been terrible, based on the stated specs.   

On the other hand,  it's difficult to know what those are based upon.  Sometimes the Chinese sellers will spec these with a range, with the lowest value being the moment they begin to emit, which is not very useful, but easy to test.  The real question is what the higher value in the range is based upon, sans a proper datasheet with testing charts.  It may be that they are intended to operate at the max of 1.5v, which is what that 27ohm resistor will provide.

Now the issue you have is that you attached different power supplies.  At 6v, with the 37ohm resistor, you fed the LEDs double what they are rated for.  While they may not be completely dead, I am 99.9% convinced that they are severely damaged, and not emitting what they should after that.  If you went higher, then change that to 100

What I would do at this point is rebuild the circuit with new LEDs AND a 27ohm resistor AND use a 5v supply.  All 4 of the modules should be running in parallel off of the supply.  But if for some reason you have the modules in series, that will never work.

If that doesn't work for you, you will likely need to look at some different LEDs.

Sounds good - thanks for the attention here -  hopefully I can go over it all again this weekend and will post findings.