Experimental question detecting light scattering with photodiode

[ihggin]

Hi, so I am trying to build a cell counting apparatus using light scattering. The setup is a (green) laser shining through a clear bottle full of (Dicty) cells suspended in a stirred solution. The idea is to put a photodetector at some angle by the bottle; a greater cell density should result in more scattering (or “flashes”) or light, which can be counted. However, I am running into a number of problems early on.

I am using a UDT-500D “photop” (which I guess is a photodiode with a built-in op amp). The photop is outputting to an oscilloscope. The first problem is that even when there is no light (for example, covering the sensing area with black paper while the lights are off), the scope still shows a signal, with greater amplitude and smaller frequency than a white light signal. I am hoping to see close to zero output signal in the dark, with spikes corresponding to the scattering. Any ideas as to what is going on / a solution? (I’m new to photodiodes and fairly inexperienced with oscilloscopes.) I was thinking a solution might be to hook up a high-pass filter; any suggestions as to what frequency would drown out the dark signal yet still detect scattering?

Another, maybe more serious, problem is that a control bottle filled with just water also shows significant amounts of scattering (to the naked eye). Do you think that this approach to counting cells is inherently flawed? Could there be another method to accomplish this goal?

Thanks, and please let me know if you need any more details / specifics.

 

[anorlunda]

Yes it is flawed. I think the cells are too small to resolve that way.

The better approach is to make a digital picture of a horizontal slice of the fluid, then use image processing software to recognize and count the cells.