Modifications to allow alternate LEDs?

[John Ball]

Hello all,

I'm a member of a vision research lab where we are scheming how to use the FED3 to test color discrimination. I'm only just now stepping in to this task, so there are probably a lot of things I don't understand yet. My apologies in advance.

One problem that we have is that we would like to have more control over the wavelength of light stimuli used for learning tasks. If I understand correctly, the main array of LEDs is this neopixel strip: https://www.adafruit.com/product/1426#technical-details, which in turn uses these LEDs: http://www.world-semi.com/Certifications/WS2812B.html. The datasheet on those LEDs states the emitter wavelengths are Red: 620-625nm, Green:522-525nm, Blue: 465-467nm. If I'm wrong on these facts, please let me know.

We are interested in testing stimuli using shorter wavelength light (closer to the long end of the UV spectrum, ~400nm). There two possible paths that I can envision for this:

1) Replace the LEDs with alternate emitters at the wavelengths we want. I haven't been able to find anything like the neopixel that includes UV LEDs, but I did find this: https://www.ledtechnologies.co.uk/proflex-rgb-uv-led-strip-395nm-24v-ip20-black-pcb.html, so it seems that RGB+UV LEDs exist, at least.

2) Use an external device for stimulation and somehow synchronize it with the FED3 for behavioral testing. I would have to source or build my own LED stimulator in this case and somehow control it from the FED3 unit.

Does anyone have any insight on these questions? I'm not opposed to trying to modify the design, but I'm trying to avoid getting too far into the weeds by staying close to the original.

Thanks in advance for any help you can give,

John Ball

[Lex Kravitz]

Hi John!  Thanks for the question, and for doing your research on this!  You are correct in the parts being used, the "Neopixel" is Adafruit's branding of the WS2812B or SK6812 LEDs (these are very similar LEDs, Adafruit uses one or the other in their different Neopixel products).  For FED3, we are using the SK6812RGBW-BW "Blue White" variant, data sheet here.  Here is the spectra, including the white LEDs - we are "BW":

Unfortunately there is very little emission at 400nm for the "BW" white LED.  However, if you can use 425nm, one solution could be to put a UV bandpass filter in front of the LED strip and use the white LED to generate UV?  This LED is very bright, 10% emission would still be bright enough for them to see.  That could be a simple solution.  It would also be possible to remove and replace the LEDs with the "WS" variant to get more UV (or maybe get OEPS to swap this LED when they assemble them). 

Your second option is also do-able.  FED3 can send 3V TTL pulses from its output port which can control an external stimulator.  I'm not sure what kind of stimulation device you are using but there are plenty of TTL-controlled LED light sources, hopefully you don't have to build your own.  This is how we do optogenetic stimulation with FED3.

Finally, if you just need a small UV light source a cheap mod would be to just attach a small UV LED directly to the output port.  This port can source enough current to drive a small LED (<40mA) on its own.  So any small 3V LED can be connected to it and illuminated, with the pattern of illumination controlled by FED3. Something like this will work.

Hope this helps! -Lex

[John Ball]

Hi Lex,

I'm glad I checked the group forum, somehow this message didn't come to my inbox.

Regardless -- thanks for the discussion, this is both thorough and helpful. Building on your bandpass suggestion, the "warm sunlight" version of the LED (WS, seems like Adafruit calls them "warm white") seems to have good emission near 400nm...what would it take to make a FED3 using that strip instead?  We have a pre-made kit (not the fully assembled unit) from open ephys, I'm assuming if we built it ourselves it would be a simple matter of buying the alternate component.

However, my real problem is that we are interested in testing discrimination between green and UV light. I'm concerned that if we used filters to create LEDs that could ONLY emit in the UV, then I'm not sure we'd be able to test whether mice see differences in color or whether they are learning based solely on which emitters are lit at any given time.

I looked again at the LED strip I linked above. I think I was wrong, it contains combined RGB emitters as well as separate UV LEDs, not a single unit that combines both. A shame, that would be a perfect unit for what I need.

Do you have any recommendations for external TTL-controllable LED sources? We don't have a stimulator at this time and any ideas would be welcome. I'm more familiar with LED sources for microscopes, which are much too high-powered and expensive for this application.

Thanks again for your help!

John

[Lex Kravitz]

Hi John,

Yes GMail sometimes flags these forum posts and I find them in my Spam folder! 

If you're experienced with surface mount soldering you could swap the LEDs from an assembled FED3 board.  However if you aren't experienced with this I would not recommend trying - the surface mount parts aren't made to be swapped in and out so it's challenging. I think the best option there would be to contact OEPS and see if they can switch this part for you in assembly, or assemble your own PCB. Assembling your own also might not be a good option unless you're experienced with surface mount soldering but it can be done.  There is a lab here at my University that recently assembled 20 PCBs on their own.  In either case, during assembly it would be a trivial thing to swap the part.

But you bring up another issue that's more of a big deal - there might not be a single LED module that emits green and UV from the same location (or very close locations).  If they have to be in the same physical location you may have to make something that combines two LEDs under a diffuser or something.  I would guess this would likely end up being an externally controlled device, and not the Neopixel strips on FED3.  Getting FED3 to communicate with an external LED to tell it whether to turn on green or UV would require a bit of tweaking, depending on how it's controlled. I can help you with that if you can find a lightsource you want to use. 

In terms of external LEDs, it depends on what you are trying to do.  We often connect FED3 to Plexon LEDs for optogenetic stimulation.  These LEDs take a TTL input, work well, and it's the brand my lab bought.  I know ThorLABS and Doric sell very similar products.  These are fiber coupled and not super cheap though.  If you just need a visual stimulus you can do it with a simple cheap LED like this:

Or if you have the newest OEPS boards with the 3.5mm jack it can be a little cleaner looking:

Hope this helps! -Lex

[John Ball]

Hi Lex,

Thanks again, this is all helpful. I have a pretty clear view of my options now -- I just get to think about how to implement them.

(Also -- I've done just enough surface mount soldering to know that I can, but that I'd like to avoid it if possible...)

Best,

John