Hi Dr. Roa,

By "patterns of stimulation," I'm assuming temporal patterns (e.g. pulses) through a fiber, not spatial patterning through a window.

Summary : I'd suggest a blue led coupled to a fiber for ChR2 and a green laser with a fast optical shutter for Arch3. LEDs are easier to drive, but a laser is probably required for inhibitory ion pumps.

1) LEDs have nearly nearly linear electrical current to optical power output, so if your study requires highly precise temporal patterns of stimulation, this would be the way to go. However, LEDs emit light in many directions, so it is difficult to focus the light into the fiber. The ability to couple a high percentage of the LED's output into a small diameter fiber, the coupling efficiency, is a key way to differentiate the available products.

Lasers can be coupled at a much higher efficiency. However, cheap diode lasers perform quite poorly when pulsed. A laser rated at 100 mW might only put out 3 mW when pulse for 1 ms. The best way to use a cheap dpss laser is to leave the laser on and then gate the light with a fast optical shutter. However, this approach can only be used for rectangular pulses, not arbitrary temporal patterns.

2.) I'm guessing that miniaturized LEDs help get more light into the end of the fiber. I hope someone else will comment on this. Might be the best solution.

3) If you can get >5 mW out of a 200 micrometer fiber, you'll have more than enough for the ChR2 studies, but inhibitory ion pumps require relatively more light. A green laser with an optical shutter would probably be best.

Additional considerations: Variations using LEDs and cranial windows may suffice if you are working in the cortex.

Hi Dr. Roa,

By "patterns of stimulation," I'm assuming temporal patterns (e.g. pulses) through a fiber, not spatial patterning through a window.

Summary : I'd suggest a blue led coupled to a fiber for ChR2 and a green laser with a fast optical shutter for Arch3. LEDs are easier to drive, but a laser is probably required for inhibitory ion pumps.

1) LEDs have nearly nearly linear electrical current to optical power output, so if your study requires highly precise temporal patterns of stimulation, this would be the way to go. However, LEDs emit light in many directions, so it is difficult to focus the light into the fiber. The ability to couple a high percentage of the LED's output into a small diameter fiber, the coupling efficiency, is a key way to differentiate the available products.

Lasers can be coupled at a much higher efficiency. However, cheap diode lasers perform quite poorly when pulsed. A laser rated at 100 mW might only put out 3 mW when pulse for 1 ms. The best way to use a cheap dpss laser is to leave the laser on and then gate the light with a fast optical shutter. However, this approach can only be used for rectangular pulses, not arbitrary temporal patterns.

2.) I'm guessing that miniaturized LEDs help get more light into the end of the fiber. I hope someone else will comment on this. Might be the best solution.

3) If you can get >5 mW out of a 200 micrometer fiber, you'll have more than enough for the ChR2 studies, but inhibitory ion pumps require relatively more light. A green laser with an optical shutter would probably be best.

Additional considerations: Variations using LEDs and cranial windows may suffice if you are working in the cortex.

Hello Thomas,

thanks a lot for your response.

I have to go deep into the brain so, that makes everything a bit more complex.

As you mention, my feeling is that most of the LED-based systems perform very well for blue light (even miniaturized led get around 8-20 mW at the end of the cannula). However, the problem arrives for green LED. Plexon offers a compact modules (PlexBright LED Modules) that can deliver up to 7.8 mW at the end of a 200 um fiber. However, I'm not sure if this will be enough.

Anybody have experience with this kind of equipment?.

Juan,

You may want to investigate the efficiency of laser coupling for Plexon module. In principle, you can use MM fiber with up to 800um core. Test the system with say, two patchcords of 200 um and 800 um cores and measure the power at the other end. Then you will know if increasing the diameter of the fiber boosts the output power and to what extent.

ThorLabs sells a number of LEDs, including blue and green, that are pre-coupled into multimode fiber. They should do what you need for a reasonable price.

Thanks a lot for your comments.

Hello Jonas,

thanks for this valuable information. I didn't know anything about this modified opsins. The problem it seems this is a very recent technology so, I guess there are not yet cre-dependent adenovirus or transgenic mice expressing this channels, isn't it?

That's great! Jonas. I see floxed AAV C1V1 on Deisseroth list but iC1C2 is not available yet.

http://genetherapy.unc.edu/services.htm

Let's wait also for the AAV ChloC. This will be a very interesting tool.

Jonas, congrats on the ChloC paper. Very exciting work.

Thanks Jonas for the comments and for the great job in building new tools for the optogenetics. It will be very valuable for many scientists from many different fields!.

After listen several comments I think I will be conservative and I will go for a laser-based system just in case I need additional power (no provided by LEDs).

However, now I'm facing a different problem. As far as I know, laser diodes are better than DPSS because the low price and higher stability at high frequencies. However, I could not find reasonably cheap laser diodes for blue (470 nm) and green/yellow (560-590) lights. That means that I will have to use a DPSS laser attached to a shutter?.

Regards,

Juan

The wavelengths around 470-530 for direct diodes are fairly new, and as a result they are quite expensive. DPSS lasers are a bit cheaper, and do have the advantage of a better spatial mode, which (in theory) means they should couple better into fiber. The drawback you have already considered is you can't directly modulate them, so you will need an external shutter. Depending on the speed and whether you need on/off or full analog power control, you can get away with anything from a high speed mechanical shutter to a polarizing liquid crystal shutter to a Pockels cell at the high end.

Hi Jonas, Doric sells dual channel laser diode at 3500 EURO. The problem is wavelengths for that system are 515 and 415 nm which is a problem if you need blue and green/yellow lights sources. The other options that I found go around 5000-8000 EURO each wavelength (I don't know why but yellow ones are far more expensive). Now I'm looking at Laserglow and they have a DPSS laser with two colors (473 nm & 589 nm) around 10500 € for 100 mW. The problem is I will need a shutter, as Craig mentioned. 

So, I waiting some quotes but I'm still not sure about what is better. I guess the final cost of each setup will decide!. 

Thanks for your comments. 

Have you looked at Thorlab's multi channel source? It can have anywhere from two to four different outputs, although they come out of independent ports. You configure it with what you want at purchase. They also can make combiners to get the outputs all coming out of the same fiber. I'm just not sure it would have enough power for your application but it may work for you, and it is internally modulated.

http://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=3800

Thanks Craig for your suggestion. I'm not quite sure whether the power from Thorlab's multi channel source will be enough. In any case, they have nothing around 560-590 nm. 

Thanks Jonas, 

but this only can deliver one wavelength at a time, right?. I would need 2 of those to deliver one wavelength in one nucleus and a different one in another, isn't it?

I already asked for this kind of setup. The problem is the power :

http://www.thorlabs.de/newgrouppage9.cfm?objectgroup_id=5206

Only 0.5-0.8 mW for green-yellow using 200 um fiber. So, maybe not enough for ArchT or halorhodopsin. Also, if you want to use 2 wavelengths at the same time coupling both to a rotary joint (for in vivo experiments) like the Doric FRJ 2x2 FC-FC, the light at the end of the cannula will be even lower (aprox 20% lower).

This is why I discarded LEDs. Is the cheaper option but maybe not very useful for some applications.

The standard UM-System might help. This System is very simple:

https://www.researchgate.net/publication/44665186_3D-reconstruction_of_blood_vessels_by_ultramicroscopy

https://www.researchgate.net/publication/220514305_Ultramicroscopy_-_a_novel_light_sheet_based_imaging_technique_created_by_various_research_disciplines

Article 3D-reconstruction of blood vessels by ultramicroscopy

Article Ultramicroscopy - A novel light sheet based imaging techniqu...

 I agree that regular LED and microLED solution do not provide satisfactory output power. 

In our lab we are using 125um diameter, dispersive or straight cannula (Doric Lens) which should output >0.5mW (usually ~2mW, for both blue and green) for good behavior results and recently we found a good alternative. We thought that only crystal laser were could output >10mW at the end of the patch cord resulting in ~2mW at the end of the cannula, but this company makes LED based laser that can output up to 40mW (patch cord, I have reached up to 15mW output from a dispersive type cannula) when correctly calibrated: cnilaser.com (made in china).

It seems that those laser have a very fast response to TTL drive and provide really nice power for in vivo applications. Hope it helps.

Well, I would like to know how you have solved your set up. Cheers!