By the time that the box was ready, the PCB was still not ready. Instead, the whole circuit was laying on two breadboards for test. It was just the right time to check if the 70 LEDs per side are enough! So, i run some tests before i make the PCB for the controller. The test was to cut some photo-resist PCB pieces, and try the box for different time and LED power values. Before i present to you the results of the tests, take a look at the following videos:
The control panel
The operation of the box
The test results
All items ready!
So, i had it all prepared: the sodium hydroxide, the PCBs, the artwork printed on transparency, the box itself... I decided (bad choice) to start with very quick exposure time and relatively low exposure power. The first test i did it with the ambitious time of 70'' at 60% LED power. The result was a completely blank PCB... I kept on increasing the time and power gradually. I did this for 6 more times. The more the exposure time and power, the better the result. I have to admit though that the transparent film was not so transparent. I have bought some new transparent films that i intend to use from now on. I believe that the result will be even better with less LED power (or time exposure).
The following picture shows the test-PCB pieces in line. I have mark over each one the exposure time (in seconds) and the LED power (in %):
From the 6th test (560'' @ 100% power) the result was crystal clear. Take a closer look:
But the last test is by far the best result of all. I used to have a Hg 150 Watt lamp for artwork transfer. I also had once a pair of small UV lamps. None of the above can reach the purity and ease of this box. The Hg lamp was very fast in terms of exposure time (about 150-160 seconds), but the results was almost never the same. A few seconds more and the artwork was broken. A few seconds less, and the artwork was not clear. The UV lamps gave a better result, but had a very long exposure time. I am very pleased that this box works so good!
I created 2 panels with 63 LEDs, each led is 1W. Dimensions of one panel are 400x320mm. Each LED is mounted on aluminum heat sink which is attached for 1mm metal plate.
I attached plate in a wooden box I created for these.
I know that, I used 63LEDs on one plate for lightning, each LED is 1W.
I try to ligt few PCBs, but PCBs with ticker lines are ok but with thinner lines they are illuminated a little too much.
I will try to play a little with frosted glass.
I created this box with 1w UV LEDs(140degree), and it doesn't not light very good my PCBs. I created everything by mathematics from page and it should work. So tell me is that frosted glass really needed in front of my LEDs or they are too strong.
Please I really need some help, and sorry on my English.
Hope someone will answer. Ty.
Hi does anyone have the MYINITIALIZATION.INC file and or a working hex. The hex in worklog works fine up to the point it does not switch off the Leds when the timer reaches zero. It goes to 999 and carries on counting down
What kind of UV leds did you use? when I search for UV leds I found a lot of different ones. Could you please post the specifications of the leds? (current draw, luminous intensity, etc) I really want to build this but I dont know what is the correct luminous intensity, if you search you can find from 400mcd to 2000 or 3000mcd. Thanks
Thank you for your response! I am myself also preparing for the new version of my old exposure box made from old scanner, with uv fluo lamps, working with no problem 8-10mils, hardly 7mil. Now I dismantled that construction because I think this can be lowered to say 7mils standard. I want to make it also with different light source ie LEDs.
That specific task of higher quality of production cannot be done easily IMHO only by experimenting, but it has to be supported with practical mathematics/physics. I am in the middle of calculations and believe that something will come up from those.
I have a suggestion: before you decide to fix the distance between LEDs or to fix the distance between the LEDs and PCB, just drop me a note, maybe at that time I shell come up with some special values to suggest them for your (and mine) construction. Of course you can always do it "by hand" but it is easy to check one or two setups before you fix the box permanently. Anyway, keep the good work and go on!
@Fernando Oliveira There has to be command "include" before these files. It is "#include BCDTODECIMAL_3DIGITS.INC". Make sure that the file "BCDTODECIMAL_3DIGITS.INC" is in the same directory awith your project
Thank you for sharing your knowledge.
I am now starting to work with Microchip PIC's. I follow your tutorials and did almost all your examples.
I'm implementing a timer to my Insoladora Leds UV's and i follow your project.
I had some problems when trying to compile the project but have managed to solve, just missing the following:
Error  C: TMP UVBOX DUALPWMTEST.ASM 1104: Symbol not previously defined (BIN8TODECIMAL)
Error  C: TMP UVBOX DUALPWMTEST.ASM 1390: Can not open file (Include File "BCDTODECIMAL_3DIGITS.INC" not found)
Therefore I am noticing missing file "BCDTODECIMAL_3DIGITS.INC" in the project.
Can you help?
@Manolis first, do not install the project into a directory which has your user name. Instead go to d:/root or c:/root.
Then, deactivate the case sensitive option
http://www.pcbheaven.com/picpages/MPLAB-Starting_a_new_project/ [step 10]
and then make sure that all inclusion have at least one space before them! They must not start in col 1...
@Giorgos Lazaridis Hello again George. Well, I did change the setting you mention before I build the .asm file the first time.
The very first error I get is about a missing .inc file and I think it has to do something with all the others, as you can see in the txt file I posted. Here's the exact error:
C:DOCUMENTS AND SETTINGSUSERDESKTOPNEW FOLDERNEW FOLDERDUALPWMTEST.ASM 97 : Cannot open file (Include File "MYINITILIZATION.INC" not found)
And then I get a lot of "Symbol not previously defined (ZERO)" and "Missing argument(s)" errors which I believe have something to do with the missing .inc file.
Towards the end of the build log I get another "Include file not found" error, the file being: "BCDTODECIMAL_3DIGITS.INC" this time.
I am really sorry to bother you again but we are very close to completing the project and we seem stuck at this point where the timer and the power meter don't work like they're supposed to.
We also tried flashing the PIC with the latest .hex you posted but still no luck!
So, I really don't know what would be a way to fix this. Maybe send us the missing include files and try building the initial .asm file again? I suppose the best solution would be to send us the final .hex after you read it from the completed UV box. This would be the "ultimate criterion" for our circuit as well since if it still doesn't work, then there will be something wrong in the hardware and not the software.
I understand that opening the box and reading the .hex might be too much trouble but if you find the time to do it, we would certainly appreciate it! Of course any other workaround you suggest is more than welcome!
@Giorgos Lazaridis Hello again George and thank you for your response!
Well I tried following the procedure according to the link you posted but I seem to be getting too many errors and some missing .inc files.
Here's a link of the output I get when I try to build the project:
(if you can't see the link, please tell me)
Would it be too much of a trouble for you to post the FINAL firmware version? I mean the one you used in the video presentation? Either in assembly or (preferably, to be honest) as a .hex file?
@Manolis hmmmm that is so strange indeed. Is it possible that you download the code instead and recompile it? Maybe the hex is corrupted or something.
Download MPLAB and follow the instructions here:
After all those months, our build is almost over but we seem to be facing a problem with the values for the timer and the power.
From the video presentation, I can see that you have no problem going up to 999 seconds (ok. 995 to be precise!) and the power stops (I assume) at 100%. In our design though, the timer stops at a value of 500 and the power is able to go OVER 100% which is quite unreasonable.
So is it possible that you could give us any hints as to why this happens? Could it be a hex thing or something else? We are using the hex provided here of course and so far we haven't modified it at all!
I used a self adhesive foil diffuser used to create a frost effect on glass or by people that make back illuminated signs and logos.
It is not expensive and you can find it easily.
like George said, after you determine the led to led distance based on the
led's angle of illumination you can place two accordingly spaced leds on a cardboard and use a piece of difuser on glass or acrylic to experiment with the led-diffuser and diffuser -PCB(use a white paper) distance.
Try to find the shortest possible leds to pcb distance (greater ligth intensity) that at the same time allows for the ligth from the two leds to blend uniformly.
1. Sure it is relevant. First of all the exposure time has to do with the intensity. Also, the spacing of the LEDs has to do with the angle of the LED beam. BTW: Cheap LEDs mean cheap results...
2. Yeah, trial and error is the way. Too close and the diffuser does not work very well, too far, and the light intensity is decreased.
Hi, congratulations again from the great project, one of the more well-made and studied, IMHO.
I'm looking forward realizing it, and the shadow-spots thing did really worry me, but it sure did relax myself seeing that someone thought at it as well.. most of the UV LED exposure box out there don't even bother about it.
However, I've got a couple (maybe many more) questions that I would ask you, eveyone who has made this project and encountered good results and problems.
1. is the type of LED really relevant?
As far as I can see, the problem of the radiant angle can be corrected making some calculations and finding the right distance between LEDs and PCB; instead, how much does the intensity of LEDs (mcd) affect the design? only in the exposure time?
There are really many kinds of UV LEDs out there, from cheap ones (ca. 6 dollars x100 or less) to much more expensive ones.. knowing that i can make a good design even with cheaper ones would be a great news!
2. how do you calculate the distance between the LEDs and the diffuser? I understood how to get the LEDs-PCB distance, but how do we manage distances using a diffuser? Only via try and repeat?
3. @Peter : which kind of diffuser paper did you use, and which one would you suggest?
3-a. Instead of using the diffuser paper (maybe it's quite expensive), a DIY frosted acrylic transparent glass would do as well?
Many thanks to everyone who takes some time to answer this long questions.
I have used the light diffusing film and i found it works just fine.
I used one uv led a small piece of diffusing film on a glass and a piece of white paper as a screen to determine the correct led to film and film to pcb distances.
I saw this light diffusing film, and remembered Georgios' efforts with sand blasting acrylic, so for anyone making this project, this film may be worth a look. I haven't used it myself, but will place a link here, if anyone uses it or has any experience with it, please post a comment.
Hello Giorgos, greeting from Italy.My compliment for the well done project. I'm starting with Pic/Avr electronic project my self and I will like to build your UV Box and Timer. Where I'm not proficient is the Pic programming, will you tell me please the fuse setting (if any) because I will program the Pic after the board is assembled.
@Manolis hmmmm prices have gone really high... yes this is the one i talk about, this is the one i use by the way. and i bought it from the same store for some $24, pike had bought this before me even cheaper... Strange.
Greetings from Greece and thank you for sharing such a great project!
My question has to do with the PIC programmer. Could you please tell me which one you used? I'm a little familiar with AVRs and recently I began to work with ARM but have never used a PIC before. So I'm looking for a relatively cheap programmer to experiment with (DIY solution included).
Would a PICKIT 2 clone (found on ebay) or something like this:
@miniyo69 @miniyo69 if you are using the same encoder with me, then the 3 pins that are on the same side are used for the rotation. the middle one goes to the pin 36, and the other 2 to pins 37 38.
On the other side, the 2 pins are for the button action.
Hi, I too am excited about this project and was wondering if you uploaded the bill of materials yet. I am mostly interested in where to buy the UV LED's, diffuser sheet, acrylic glass, and photo-sensetive PCB.
Your site has helped me so much. Thank you for sharing these projects to everyone and in such great detail.
Lyncos, the LEDs emit 25 degrees angle, half the angle is 12.5 degrees. If d=the distance of the LED, then the radius (NOT diameter) of the circle that will appear on that distance is:
r = d x TAN[12.5]
The diameter is 2xr. So, in your case that the distance is 30cm, the circle radius that will be created is:
r = 30 x 0.221 = 6.6cm.
The diameter is 13.3 cm. I tested this calculation and is precise, as long as the LED emit 25 degrees ray.
Thanks for the answer. I made my first run yesterday and everything went fine the first time... I have 30 leds as described before with a spacing of about 14 mm at about 30cm high. 5 Minutes later I got a perfect print on my circruit board... It took me 1 month to not even figure out the toner transfer method (never worked for me) .. I will ALWAYS use the photo resist method ... Now I'm waiting for 2 packs of 100x 7000 mcd leds to make bigger circuits.
Now it's a temporary installation but I'm planning doing a box for it, I would like to lower the height of the installation cause for 2 side circuits it will take too much space... I think I can go half of the height...
Also I would like to know the formula you used to find out the best spacing/ heigt of the led.. I have a good Idea with triangles formula but would like to know how you done it
I started a similar project... what kind of led are you using ? I have 400-405nm, 2000mcd, 5mm leds you think it will works ? I\'m planing doing it with 30 leds 5x6 and about .5 inch spacing.. I will do very small PCB with this first version... I\'m waiting for 200 390-395nm leds for my larger version