UV LED BOX - problem with the radiation pattern

[Pedro Duarte]

Hello guys,

based on a few projects out there I decided to make my own UV LED exposure box with an old scanner I had.
I removed the original controling board and replaced it with a costum in order to make use of the scanner buttons, status LEDS and LCD. See pictures:

Scanner - **broken link removed**
Custom board - **broken link removed**
Final product (idle) - **broken link removed**
Final product (working) - **broken link removed**

It turned out quite nice but when I expose my project boards I notice the following pattern: It turns out that the cone of radiation that comes out of the LEDs is not wide enough to overlap the cone of radiation of the adjoining LED. Even if i let it expose a little longer, develop for longer and etch them for longer... see pictures:

current radiation pattern - **broken link removed**
Board after etching - **broken link removed**

I though I might push my LEDs a little bit further away form the glass in order to give the cone enough space to broaden and overlap the other radiation cones, however I am both out of room (the scanner is quite thin) and I am afraid that it wont help that much based on this projection:

expected radiation pattern - **broken link removed**

There seems to be always some bits of the boards that will get twice the exposure of the other bits. And that is not good...

I though that maybe what could help me out would be some sort of radiation diffusion layer, maybe some plastic layer that could refract the radiation in so many different ways that when you look at it you cant really see the LEDS, just their light. Like the ones we use in the kitchen fluorescent lamps:

plastic diffusion layer - **broken link removed**

Do you have any ideias to help me out?

Thanks!
Pedro

 

[Pedro Duarte]

Trying to solve this I bought a PVC mate layer: **broken link removed**
Then i covered the interior of the scanner with a layer of the stuff: **broken link removed**
When the machine is on we can still see the leds, but there is a strong blur that is was I was going for: **broken link removed**

I then exposed another bit of the PCB for 3 minutes, before I was doing 2.5 min, since there was a radiation filter in the middle.
The final result is still not perfect... but much better then the previous: **broken link removed**
Tomorrow I'll add another layer of the stuff to see if results improve again.

But please, if you have any suggestions, please let me know. I'm quite new to this so you're experience is really welcome.

Pedro

 

[ccurtis]

I think you have the right idea using an optical diffuser. It then only becomes a matter of the amount of diffusion and the losses through the diffuser. For UV, the fineness of the irregularities in the surface of the diffuser that create scatter needs to be greater than for greater wavelengths and the material determines the amount of loss. Acrylic or glass passes UV better than some other materials. From what I can tell, the rough side of the diffuser should be away from the source of the light, though I don't know if that is always true.

Edmund Scientific sells optical diffusers for UV, but they are not cheap.

 

[bance]

I had a similar problem, I just sanded the leds with 400 grit wet and dry paper, works well....

HTH Steve

 

[Pedro Duarte]

Thank you very much for your input.

I tried the approach of placing 2 layers of mate PVC under the glass of the scanner. The results were really good.

Board after developing: **broken link removed**
Board after etching: **broken link removed**

I think I'll keep it this way but if I come across the same problem again with thiner traces I'll give the wet and dry paper a try!

Thank you all once again!