PCB Etching

[Allycat]

Has anyone done any etching using inkjet transparencies?

 

[jpanhalt]

I have used inkjet transparencies as a mask to expose a photosensitized boards and then etched the boards. Been doing that for 15+ years.

What is your question?

John

 

[shumifan50]

If you mean to mask the copper for etching, I have never tried that but unlike laser printer ink, I don't think inkjet ink is suitable(it will not melt when heated).

 

[Allycat]

I have used inkjet transparencies as a mask to expose a photosensitized boards and then etched the boards. Been doing that for 15+ years.

What is your question?

John

Hi John, Can you explain the full process please and where do you get the transparencies from.

 

[jpanhalt]

With inkjet, all you need to do is make the transparency. Be sure to mirror it, if needed. Place the ink side against the sensitized PCB board, use single-thickness glass and clips to hold it down, expose to light, develop, and etch.

The procedure has been written up dozens of times and presented on the internet. There are small variations, like exposure time, developer, and stripper, that depend on the brand of board you are using. Follow the manufacturer's directions.

If you want to use LaserJet, I wrote up the procedure I have been using here:http://www.electro-tech-online.com/blog-entries/improved-laserjet-transparencies-for-pcb.213/
Many laser printers probably do not need the additional darkening of the image.

John

 

[Allycat]

With inkjet, all you need to do is make the transparency. Be sure to mirror it, if needed. Place the ink side against the sensitized PCB board, use single-thickness glass and clips to hold it down, expose to light, develop, and etch.

The procedure has been written up dozens of times and presented on the internet. There are small variations, like exposure time, developer, and stripper, that depend on the brand of board you are using. Follow the manufacturer's directions.

If you want to use LaserJet, I wrote up the procedure I have been using here:http://www.electro-tech-online.com/blog-entries/improved-laserjet-transparencies-for-pcb.213/
Many laser printers probably do not need the additional darkening of the image.

John

Thanks John

 

[Allycat]

A couple more questions.
If I mirror the image while printing then place it print side down on the board, will this have the effect of mirroring the image back?
How do you know how long to expose for, is it possible to overexpose?

 

[shumifan50]

Yes it will mirror the image. Depending on which package you use to draw the PCB layout, you have to be careful to place your components on the copper side of the PCB as then they will be reversed correctly. Also note that the pins are reversed when you lay out your traces. The best thing to do is to draw an IC, print it, trace it onto another piece of paper(representing the PCB), then look at the result if you psh the IC through the paper and see where the pins end up.

You can over expose or under expose, so you have to make a test strip that you expose at various time periods, keeping note of which bit was exposed for how long. Then develop the strip and see which exposure time gives the best result. Exposure times are dependent on the type of photo-resist board, the strength of your UV lamp(s) and the distance between the UV source and the PCB, so it is not possible to give an exposure time. After doing the test strip, you have to ensure that you maintain the same UV source, distance and PCB type.

BE CAREFUL with the UV light as it can damage your eyes if you look at it when it is on.

 

[jpanhalt]

A couple more questions.
If I mirror the image while printing then place it print side down on the board, will this have the effect of mirroring the image back?

Try to visualize what you are doing. It makes a difference whether you are talking about the top of the board or bottom of the board(as viewed from the top).

If you are talking about the top of the board in the usual presentation, then a mirror printed followed by ink-side down will reverse the mirroring. In other words, the answer to your question, as I understand it, is "yes."

How do you know how long to expose for, is it possible to overexpose?

That depends on your board, lights, thickness of glass hold-down, and distance of board from the lights. In my set-up it is 13 minutes with three, 15W (F15B8-BL) bulbs. You can do a test run with an exposure guide. Some people use lines of different width. I use a cross, like a large X or >< ,and look carefully at the resolution at the crossing. Yes, it is possible to over expose. Unless your traces are narrow (<16 mil), I would err on the side of a little over exposure.

BTW, I added an addendum yesterday to the blog I linked to. IN brief, a transparency made with a Brother printer HL-2270DW did not work with the dry erase. The HP printer worked fine.

John

 

[jpanhalt]

I just finished a board using the HP4101 printer and dry erase darkening. Narrow traces are 16 mil; fatter ones are 24 mil. Drilling is not too good; blame that on old age and poor lighting.

Positive sensitized board was Injectorall, 1 oz. Developer = 0.5 M KOH. Exposure = 13 min. Board was about 2 years since receipt from DigiKey.

John

 

[(*steve*)]

Looks pretty darn good to me.

There are a few extra areas of unetched copper, but none in places that look like they'll cause problems.

 

[jpanhalt]

It's a rough prototype to compare a couple of accelerometers: the MXD2020 (a 5 volt chip) vs. the MXD6235 (a 3.3 volt chip) from MEMSIC. The same MCU will be used for both, which is currently the 12F683 and will transition to the 12F1840. Thus, it is a multipurpose tool, and I was in a bit of a hurry. The accelerometer sits on a daughter board. The 3.3V daughter board has a separate voltage regulator. There are a lot of useless little peninsulas of copper that I didn't bother removing.

The other part of the wireless design is a GLCD display board I ordered from OSH Park a week ago Friday. It should be here Monday or Tuesday. If it is decent quality, I really can't justify with that type of turn-around time and cost the bother of making my own PCB's. This may be the last DIY board I make. But, I wanted to show the OP, it really is not that hard to do.

Thanks for the complement.

John

 

[Allycat]

Thanks to all for your help. Ill post a pic of my first board when its done. That looks very neat John. Thanks again.

 

[Allycat]

Are these any good for the job?
http://www.ebay.co.uk/itm/4-pcs-Nai...1145172897?pt=Nail_Polish&hash=item4d19ca57a1

 

[shumifan50]

They will be OK, but finding a ballast for them with the right connectors might be a problem and they are a bit small - it will only be possible to expose small PCBs evenly. You would also have to build a box with a glass exposure surface and lid, so you don't look at the UV light. It might be better to look at
LED UV Exposure unit
or search for 'UV exposure unit' on fleabay. They have the box, convenient access and a timer built in.

Be aware that looking directly at UV light WILL damage your eyes (ARC eyes).

 

[Allycat]

They will be OK, but finding a ballast for them with the right connectors might be a problem and they are a bit small - it will only be possible to expose small PCBs evenly. You would also have to build a box with a glass exposure surface and lid, so you don't look at the UV light. It might be better to look at
LED UV Exposure unit
or search for 'UV exposure unit' on fleabay. They have the box, convenient access and a timer built in.

Be aware that looking directly at UV light WILL damage your eyes (ARC eyes).

Only doing small projects so bought one of these.
http://www.ebay.co.uk/itm/130550385...eName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2648
Thanks

 

[shumifan50]

The exposure on that unit will be VERY uneven as the tubes are badly arranged. Ideally you want them evenly spaces across the top, not the sides where they will now be closer to the board being exposed. The essence being all the tubes should be equal distance from the PCB being exposed.

The UV is 370nm, so exposure times will be much longer. You really want 395nm.

The timer resolution is too rough, you will have to add a better timer where you can set in second steps, or at least 5 second steps.

You will need to add a piece of glass to hold the transparency tight against the photo-resist PCB.

Let us know how you get on.

 

[jpanhalt]

I agree on the the lamp arrangement. You can probably just remove the two side lamps to improve the geometry. I also use a piece of aluminum (aluminum foil will suffice) behind the lamps as a reflector.

On the question of wavelength, I agree about not using 270 nm; however, from what I could tell, the units in the US have maximum output at about 365 nm, which is just fine. Anything between 365 to 405 nm will work. The "BL" tubes in US are about 395 nm, but it appears the British ones are labeled BL350 (or similar), so my guess is that they are a little shorter wavelength, but certainly not 270 nm.

John

 

[shumifan50]

Sorry mistyped - it is 370nm
I also noticed that the 54watt unit has the tubes in a straight line but at the bottom of the unit, so you need a glass shelf and a piece of wood to hold the PCB.