To be precise, I'm trying to implement an adaptive condenser-aperture in a brightfield microscope. It's remote and I don't want to have it mechanical moving all the time when changing the objective lens. I also want to address different grayvalue-patterns which wouldn't be possible with mechanical apertures.
Right now I'm using a 2.2' 320x240 RGB TFT (ILI9341) with an Arduino a as a driver. The results are quiet ok. I've placed the Screen in the condenser aperture plane and centred the aperture in discrete steps to find the best fit. The entire pupil (NA=0.9) has 160 px in Diameter. When looking through the Bertrand-lens (telescope which images the pupil), I can see the RGB-pixels. I'm addressing a B/W-Pattern (Oblique, Obscuration, etc.), which follows in the desired colours.
So my question: What could be the effect using a RGB-LCD instead of a pure BW-LCD?
Some thoughts:
- diffracted light due to pixelated structure (also wavelength dependend)
- passing “white” only means, each RGP pixel is in “on-state”; therefore spectrum has gaps
- extinction/contrast might be decreased due to lower fill-factor of RGB-LCDs
An alternative could be to get an old projector and use one of the LCDs as a screen. The pixel size is way smaller and results into a better source shape estimate (compared to mechanical apertures), but the "high-voltage" part in the lab is to dangerous. Using an old Smartphone with "Retina" screen could be another choice, but again, it's RGB. Professional SLMs are too expensive.
Do you have any advices? Thanks a lot!
Hello Benedict,
If i understand well, you are modulating intensity in your system, and not phase. If that is the case, you can obtain binary monochromatic patterns through a relatively cheap digital micromirror device, like this:
http://nl.mouser.com/ProductDetail/Texas-Instruments/DLPLIGHTCRAFTER/?qs=sGAEpiMZZMtr3BLCQWqPFTtjT4BhY7X6IrM1pKtGiSM%3d
it's basically an array of really small mirrors that you can turn either left or right at very high frequencies; with a couple of lenses it would be quite easy to conjugate the mirror plane to the back aperture of your condenser, with any magnification you want. The one i linked is the most cheap "plug and play" device you can get, and it comes with three decent leds in red, green and blue. It is very easy to take apart and use with your own light source.
Just a couple of warnings:
- use with lasers is a bit of a mess, the grid like structure of the micromirror generates an array of dots on the far field...very pretty to look at, but not at all practical in an optical setup.
-the modulation is binary, you can obtain a grayscale, but it is made by just switching the mirrors faster than the human eye can perceive, so if you are trying to use interferometric principles you have to stuck to completely white or completely black pixels.
Hope this helps
Hey Paolo,
thank you for your reply! I didn't even know, that there is also a cheap version of the DLP-evaluation board. Until now, I was always trying to find broken projectors or cheap pico-projectors on ebay to get my SLM-projects running. Controling one of those devices is not nice, due to internal processing of the input data..
But in my case, I can't use this great solution. Too bad. I have to use a transmissive Display with high contrast ratio and high resolution.
My idea:
- Using an old iphone 4 (~40 € used version)
- extend LCD-Phone connection with this cable:
- Remove backlaight and LCD reflector
- Connect iphone to PC via USB
- Use it as 2nd monitor with an app (there are a few, working quiet well)
The result in another experiment was quiet good. Too bad, that there are no LCD-cables for android devices. This OS is better for programming ;)
Using the Psychtoolbox in Matlab one could push the figures as fullscreen on to the iphone.
But still, I keep the question in my mind: What effect has the RGB instead of pure Monochrome? The LCD lays in objects fourierplane. So there will be a modification of the wavelength in its spectra. There will be a wavelength depending attenuation for example..
Any ideas? Thank you!
http://www.ebay.com/itm/iPhone-4-GSM-CDMA-4S-LCD-Display-Digitizier-Lens-Touch-Test-Tester-Flex-Cable-/121025869898
Hello Benedict,
too bad you can not use a DMD, but it is nice to look for an alternative, especially considering the market monopoly of TI on the DMD technology. I'll try to answer your question, with a couple of considerations:
- Consider the low efficiency of an LCD. Since it is a brightfield microscope, i guess you are using a standard white lamp. Even if your panel was "ideal", due to color pixels and polarization you will lose around an order of magnitude of light intensity from your lamp. This light will be converted into heat, so be careful if you are using a powerful microscope transillumination lamp, keep the temperature of the LCD under control. Also, remember that even the "dark" pixel of an LCD is never perfectly dark.
- From a physical point of view, excluding the light efficiency, i see no particular problems in the RGB panel, if you are not using it with lasers. The internal structure of the pixel (which varies in geometry from panel to panel), will, by definition, be the tinyest detail in your structured illumination. According to the Fourier transform relation between back aperture and focal plane, this means it will mostly influence the edge of the field of view of your illumination, which is generally outside of the field of view of the objective anyway. The center of the field of view, which you will be seeing from the objective lens, will only be modulated by lower spatial frequencies, which should be more or less equal for the three colors.
- Consider the fact that LCDs are made for the human eye, which is very sensitive in the green part of the spectra, and less so in red and blue. In general, good screens compensate for this effect, so expect your transmission spectra to have a dip in the green region, compared to the blue and red one.
- From a technical point of view, beware of the usb connection! you will be sending 960x640, 32bit pixels to the screen per each frame. This amounts to 2.5 Mb of data, without considering overhead. considering the bandwith of usb 2.0 is only 15 Mb per second, any external monitor software will probably apply some compression. Since compression methods are Fourier based, this will have weird effects when you put your screen in the back aperture of your lens. Make sure you have lossless image compression to avoid this problem.
- Also, consider a small, high resolution hdmi display, like the ones made for the raspberry pi (this one, for example: https://www.adafruit.com/products/2232). It will not have the high resolution of the iphone, but it would be at least comparable, and much easier to disassemble, setup and use.
Good luck with your project!
- Badges
- Science topic