[SOLVED] suitable lamp for uv pcb curing machine

hi all,
i think this is the best place to ask this question

i am now thinking to upgrade myself from toner transfer to photoresistive method for my future pcb design so i can handle smd components.

i bought every chemical and a piece of photoresistive film from china market. but the main challenge is to design my uv exposing machine. i cannot buy one so i have to build my own. i did some research on the net and saw a couple projects.

i have a 3w uv led with driver but no heatsink, can i use this small light for small pcbs, if not can you advise me any halogen lamp (not flourasent) for this purpose?

thanks

adnan_merter said:

hi all,
i think this is the best place to ask this question

i am now thinking to upgrade myself from toner transfer to photoresistive method for my future pcb design so i can handle smd components.

i bought every chemical and a piece of photoresistive film from china market. but the main challenge is to design my uv exposing machine. i cannot buy one so i have to build my own. i did some research on the net and saw a couple projects.

i have a 3w uv led with driver but no heatsink, can i use this small light for small pcbs, if not can you advise me any halogen lamp (not flourasent) for this purpose?

thanks

Click to expand...

Yes you can use the UV LED with a good heat sink. An easy way to get an UV light source is to use the sterilization fluorescent tubes used in hospitals. Also there are small UV discharge tubes used in old-style UV lights to irradiate children (caution,use special goggles to protect your eyes).

jiripolivka said:

Yes you can use the UV LED with a good heat sink. An easy way to get an UV light source is to use the sterilization fluorescent tubes used in hospitals. Also there are small UV discharge tubes used in old-style UV lights to irradiate children (caution,use special goggles to protect your eyes).

Click to expand...

Sterilization UV is short wavelength, usually 254 nm (i.e., the extremely intense mercury vapor line). Not only is that not necessary, it probably won't work very well because the glass plate used to hold the transparency to the PCB and maybe even the transparency itself will be opaque to it. It is UVC, and one should avoid unnecessary exposure to it.

@adnan_merter :
You say nothing about the pre-sensitized PCB's you have. Did the manufacturer give you any information? Are they positive acting or negative acting? Are they DNQ-sensitized Novolak. The vast majority of pre-sensitized plates are positive-acting, DNQ sensitized. Those photoresists work from 360 nm to over 430 nm. Using a shorter wavelength provides no practical advantage, assuming the intensities are the same.

Here is a reference that describes them: https://en.wikipedia.org/wiki/Photoresist The I,H, and G lines of mercury are approximately 365 nm, 405 nm, and 436 nm, respectively.

Assuming you have the most common type of pre-sensitized boards, you can use any light that emits at 390 nm or so. That includes ordinary incandescent lamps; although, heat from such lamps may need to be taken into consideration. Ordinary white fluorescent lamp bulbs work fine and are what I use.

There is a plethora of different fluorescent lamps on the market with slightly different phosphor for different lighting effects. You can check the emission spectrum for the ones available to you. I use an ordinary Philips F15T8 bulb. Type B will work. I happen to have type BL, which emits in the 366 nm region plus plenty of visible light. I would not recommend it over the type B lamp. It works and was available to me. The phosphor in both lamps appears white in natural light. You do not want a lamp that is clear (no phosphor) nor one that appears dark blue. As a side note, fluorescent lamps are noted for causing bleaching of colors on fabrics. photographs, paintings, and such. That effect is from the UV they emit. This Wikipedia shows spectra of several fluorescent lamps: https://en.wikipedia.org/wiki/Fluorescent_lamps

John

You say nothing about the pre-sensitized PCB's you have. Did the manufacturer give you any information? Are they positive acting or negative acting? Are they DNQ-sensitized Novolak. The vast majority of pre-sensitized plates are positive-acting, DNQ sensitized. Those photoresists work from 360 nm to over 430 nm. Using a shorter wavelength provides no practical advantage, assuming the intensities are the same.

Click to expand...

i am not using pre-sensitized pcb i am using an ordinary pcb and i am covering it with photoresistive dry film, and no idea about the wavelength neccesity. but the seller gives some information about exposing this film. it can be cured with normal economical white energy saving lamps within 30 minutes and, i think the best wavelenth is between 390-405 nm wihch is the same as uv light. and also the seller says it is negative dry film, so the light basically strengthen the film.

there are plenty of light types but i couldnt decide. i dont want to spend so much money for this machine so the reasonable one is uv led.

If it is a film that is laminated to the board, it may be a negative-acting resist. That is something you need to confirm. The developers for many negative-acting resists (e.g., the DuPont products) contain an organic solvent in a mixture. Positive-acting resists often use just an inorganic (NaOH, KOH, sodium meta-silicate) or organic (e.g., a tetraalkylammonium hydroxide) base for development.

As for cost, a household fluorescent bulb is quite easy to use and cheap in the USA. UV led's may take more of your time to put together and get working. But that decision won't really affect your final product. I agree that 390 to 405 nm is the wavelength range you should try to use. Basically, that was the whole point of my post.

John

A mercury fluorescent UV tube is most effective with moving film under the light using a slow print feed mechanism. THe tube has no phosphor for converting to daylight.

Halogen generates too much heat and will shrink the Mylar.

SunnySkyguy said:

A mercury fluorescent UV tube is most effective with moving film under the light using a slow print feed mechanism. THe tube has no phosphor for converting to daylight.

Click to expand...

I don't see how you can conclude that:

1) Mercury fluorescent lamps with no phosphor produce 85 to 90% of their light at 254 nm. An additional 5 to 10% is at 185 nm, which is usually ignored unless one is doing vacuum spectroscopy or under nitrogen or other transparent gases. (http://en.wikipedia.org/wiki/Ultraviolet ) Thus, only a very small percentage of the emission of a low-pressure mercury lamp is at wavelengths greater than 254 nm.

2) You mention using Mylar as the carrier for the pattern. Mylar has transmission properties similar to window glass ( **broken link removed** ). It simply does not transmit wavelengths less than 314 nm. Window glass cuts off at 350 nm. Thus, using Mylar film held down with window glass will block all of the 254 nm emission from the mercury lamp. Even if one omits the hold-down glass, the Mylar will "look" opaque at 254 nm. No pattern will be discernible.

That would be neither effective nor efficient. As mentioned above, it is far better to use a phosphor coated (household) fluorescent lamp. I hope that is the direction the OP has decided to take.

John