Robotic venus flytrap

Projects General Guidance
1

Any suggestions for making something that's sensitive enough to detect a fly landing on it?

I want to make a mechanical/robotic venus fly trap. I figure the trap will be quite easy, the tricky bit will be designing a trigger.

A fly is probably too small and light for any pressure or weight sensor, so I was thinking of maybe a photo diode trigger. Take a photo-diode with a light shining on it, if a fly lands on it the light will be blocked, the diode would be in shade and the trap triggered.

Option two might be to etch a pair of interlocking tracks onto a 1 inch square PCB, if the fly lands on the PCB and 'shorts out' the two tracks, then the resistance should (hopefully!) change enough to be measured - although I'll admit I'm not sure what the resistance is of a fly!

Any better/alternative ideas?

Thanks

2

Capacitive sensing might work. Or the camera from an optical mouse

3

Capacitive sensing might work

Good idea. I was going to reply that it was too complicated to impliment - then I found this article...Arduino Playground - HomePage

Or the camera from an optical mouse

I think this is a similar idea to a photodiode and a light source.

Thanks

4

if you have an LDR, the fly landing on it would have a drastic change of resistance.

if someone walked by, they would cast a much lower value.
put your hand over, large change....

probably need to continuously adjust for background light.

5

Option two might be to etch a pair of interlocking tracks onto a 1 inch square PCB,

Skip the detection bit and mechanical trap, and just put a couple of kV on adjacent tracks :slight_smile:

6

AWOL:
Skip the detection bit and mechanical trap, and just put a couple of kV on adjacent tracks :slight_smile:

Maybe he wants to capture the fly alive...
Don't now much about the resistance of a fly but i could imagine that the model is not that dissimilar from the human body. Rtotal = Rskin + Rinternal where Rskin exhibits highly non linear characteristics and breaks down at a certain voltage (for human R internal is about 1200Ohms and Rskin breaks down at about 500V). Of course the actual values for fly will be quite different (lower?) but the model could apply. If this is correct the voltage needed to detect a fly should be some 10's V. Experimenting is certainly needed.

7

instead of a normal LED and photo-diode, try it with one of those blue "commercial-fly-electrocution-devices" lights.
It's a fly magnet!

8

C-F-K:
instead of a normal LED and photo-diode, try it with one of those blue "commercial-fly-electrocution-devices" lights.
It's a fly magnet!

Dont know if it has to be UV or if a blue led will have a similar effect. If it has you could use two of them. One as a emitter and the other one as a Light detector

9

It would need to be a UV.