Help to build a UV box for PCB

[mvs sarma]

In your opinions what is the best UV tube from Philips that can fit for this pcb job?

did you see an article on PCB exposure box using UV LEDs in elektor magazine?
that would be cool and efficient I feel.as far as crystal glass , you may try a recovered glass from any flatbed scanner

 

[jpanhalt]

In your opinions what is the best UV tube from Philips that can fit for this pcb job?

Your current lamp, code TL-D, appears to be a cool white color code for lighting purposes. For PCB exposure you want a BL color code, not BLB. BL has a phosphor to cause emission at around 370nm (UVA). BLB is basically a germicidal lamp with an emission of 254nm (i.e., the main low-pressure mercury line).

I use a F15T8/BL that is an 18" bulb. If you need a 24" bulb, the number would probably be something like F20T12/350BL or F20T12/BL. I am not completely sure of the difference between 350BL and BL emission color codes. I suspect the 350BL has a different phosphor and shifts the emission to a slightly shorter wavelength, namely near 350nm instead of 370nm. I suspect either type will work, but you will need to calibrate your time for each.

Here is a Wikipedia link on fluorescent bulbs. The color is discussed near the end.

John

 

[Boncuk]

did you see an article on PCB exposure box using UV LEDs in elektor magazine?
that would be cool and efficient I feel.as far as crystal glass , you may try a recovered glass from any flatbed scanner

Hi mvs sarma,

the problem with the LEDs used in Elektor is a narrow angle of emission. To minimize losses the distance between LEDs and object must be very close. (Radiation wavelength contains a small amout of 400nm) To cover an area without gaps (for equal exposure of the board) the LEDs must be arranged very closely. It takes as many as 240 LEDs to cover DIN-A4 size.

The latest price is €0.38 per piece, which amounts to a total of €91.20. Further the LEDs require a constant current source being split to groups not to exceed the supply voltage into dangerous areas. Last not least the LED arrangement requires active cooling.

Hans

 

[mvs sarma]

Hi mvs sarma,

the problem with the LEDs used in Elektor is a narrow angle of emission. To minimize losses the distance between LEDs and object must be very close. (Radiation wavelength contains a small amout of 400nm) To cover an area without gaps (for equal exposure of the board) the LEDs must be arranged very closely. It takes as many as 240 LEDs to cover DIN-A4 size.

The latest price is €0.38 per piece, which amounts to a total of €91.20. Further the LEDs require a constant current source being split to groups not to exceed the supply voltage into dangerous areas. Last not least the LED arrangement requires active cooling.

Hans

I feel constant current is a simple issue with LM317 or a similar device. at 25mA , perhaps cooling is not involved. Ofcourse we can have a 12V dc fan.
the distance between the pcb and lamps point is to be set properly. Infact i was proposing to make one Once I get Cramaline positive resist spary. But blocked as Franel India unit was having issues with getting hazadus (inflammable ) by flight.

 

[picbits]

Hi mvs sarma,

the problem with the LEDs used in Elektor is a narrow angle of emission. To minimize losses the distance between LEDs and object must be very close. (Radiation wavelength contains a small amout of 400nm) To cover an area without gaps (for equal exposure of the board) the LEDs must be arranged very closely. It takes as many as 240 LEDs to cover DIN-A4 size.

Unless you mount them on an old scanner head and slowly scan the head along the PCB. This way you can only expose the area you are after, switch of say half the LEDs if you're making a narrower PCB etc.

 

[mvs sarma]

Unless you mount them on an old scanner head and slowly scan the head along the PCB. This way you can only expose the area you are after, switch of say half the LEDs if you're making a narrower PCB etc.

Here is a link for one such design using an old scanner box, ofcourse one has to make arrangement for reducing the speed of lamp strip motion.
http://hackedgadgets.com/2007/02/24/uv-led-pcb-exposure-scanner/

 

[Hero999]

Hi Hero999,

try looking into the beam of an UV-radiating LED of Seoul semiconductors for a split of a second. It radiates 70mW at 340nm solely at an angle of 20degrees!

To make it safe order an escourt dog before you do that. A Seoul semi engineer required one month of intense medical care to regain his vision after he ignored the safety precautions.

Hans

Black light fluorescent tubes throw the radiation in all directions, not in one focussed beam that can easilly ruin the retina. Filtered blacklights are probably the most dangerous because the Wood's glass blocks most of the visible light which causes the iris to contract protecting the retina. The insect killer tubes used in PCB developers give off plenty of visible blue and violet light which causes the iris to contract, thereby protecting the retina from any harmful rays.

I was referring to fused quartz that shows a UV cutoff <200nm. I don't think the scanner platten is made of that, as opposed to polished glass that will pass UVA to an extent, but not UVB or UVC. Quartz would be prohibitively expensive for a scanner and of little value, as the bulb is coated to block short wavelength UV. Fisher Scientific has quartz 6" X 6" for about $230. The scanner is 9X11.

Modern rolled window glass does not cause enough optical distortion to matter on the PCBs I make.

John

Black light tubes are made of the same kind of glass that most windows and visible fluroscent tubes are made from. This makes them safet because it blocks and damaging UVC radiation not absorbed by the phosphor and allows through the safer UVA rays.

Edit: Some plastics such as polypropylene and polyethylene, also pass UVA, even UVB, but one must be careful as UV blocking agents are often added to increase UV stability. Data on UV transmission of specific plastics is hard to find. I've tried to get that information for the films I use to make my transparencies, but the manufacturers just will not give it to you.

That's perfectly true, some plastics block UV rays and some pass them. However, I doubt that a plastic that normally passes UV would be doped with a blocking agent to protect the plastic because if it passes UV then it won't be damaged by it. I only think they would add UV filters if the radiation either damages the plasic or bleaches any added dye.

 

[jpanhalt]

@Hero999
Passing UV is not an all or nothing proposition. Of course, what you say is true if there were absolutely no absorption. However, what I said:

Some plastics such as polypropylene and polyethylene, also pass UVA, even UVB, but one must be careful as UV blocking agents are often added to increase UV stability.

did not mean to imply an optical density of 0 (zero) at all wavelengths or even at relatively long wavelengths. The light that may be absorbed can still damage it. That is particularly true when the plastic is exposed to sunlight or other broad spectrum sources (e.g, germicidal lamps).

Unlike glass, plastics are more easily damaged by absorbed light and hence often have UV protectants added to them. As you point out, the lens in our eyes also passes UV. However, it is also damaged by UV, as it is not completely transparent to it.

Another reason UV blockers are often added to plastics is to improve their properties. There are many applications in which UV transparency is not desirable. For example, it's not desirable for safety glasses, window coverings, and transparency films used in art work. I was not aware of the latter, but became aware of it when I discovered that a new formulation of the transparency film I used to make PCBs would not work. The "improvement" was to add a UV blocker, which apparently enhances some artistic effects. The brand, FWIW, is called Pictorico. I continue to use the premium OHP film. The improved film (the one that won't work) is "ultra-premium."

John

 

[mvs sarma]

I have read that the UV has to be around 400nm for proper exposure of pre-sensitized copper clad boards. Thus we need to select the lamp/LED accordingly, perhaps.

 

[jpanhalt]

Thus we need to select the lamp/LED accordingly, perhaps.

That's why the BL, not BLB lamp is recommended.

Judging from the photochemistry involved, going to a slightly shorter wavelength would still effect the change, but the very short wavelength of the BLB bulbs (namely 254nm) won't be passed by the typical materials used (window glass and plastic transparencies) and might cause secondary and undesirable reactions, if one used materials that would pass it. As I mentioned, a 350nm lamp would probably also work.

Some of the presensitized plates are orangish and others are greenish. It is possible that the greenish ones might respond to an even longer wavelength of light. I have never read the product manual for the green version.

John

 

[Hero999]

That's why the BL, not BLB lamp is recommended.

Judging from the photochemistry involved, going to a slightly shorter wavelength would still effect the change, but the very short wavelength of the BLB bulbs (namely 254nm) won't be passed by the typical materials used (window glass and plastic transparencies) and might cause secondary and undesirable reactions, if one used materials that would pass it.

That's not true.

The UV spectrum of a BLB is nearly identical to a BL, both emit strongly in the 350nm region. The only difference is a BLB is made of Wood's glass which is deep violet/blue in colour and absorbs most of the visible radiation.

A BL is probablty better because the shorter visible wavelengths help to expose the film.

Germicidal and rock hunting tubes should not be used because because they emit 254nm which is both dangerous and useless as it's absorbed by ordinary glass and most plastics. I recall a project over on Electronics Lab that incorrectly uses germicidal tubes and is a perfectly good project providing you subsitute them with BL of insect killer tubes.

 

[jpanhalt]

My error. I confused the BLB lamp with germicidal (254nm) lamps. Everything I said about BLB should be read in that context. I agree, germicidal lamps should not be used, even if you have access to quartz equipment. John