Hi all,
I've just slightly modified my method for making PCBs at home. This may be something you were already aware of, but I only found out about it recently.
I buy positive photoresist pre-coated copper boards (Fotoboard 2 from Rapid/Mega. In this case, the "economy" version which is a CEM1 substrate, not FR4), expose them to UV, develop them with room temperature "Electrolube PDN positive photoresist developer" [Sodium Hydroxide (NaOH)] which takes between 1-2 minutes and then etch in room temperature ferric chloride (FeCl[SUB]3[/SUB]) which takes 20-25 minutes. I roll up a paper towel and use it as a disposable brush for the last bit of development and etching. I probably should get a couple of cheap paintbrushes or pastry brushes. I also wouldn't really recommend the CEM1 boards. They're cheaper than FR4 but you can't really score and snap them as they're mostly cellulose paper they just fray and the integrity of the substrate is reduced for a few mm in from the edge. You have to make sure to file the edges flat and give at least 3-5mm spare at the edges. Hacksaw gets through it easily enough though. It might be easier on the blade than FR4.
That's all pretty standard. There are 2 things I wanted to share which I found elsewhere online:
1 - Now UV curable gel nail polish is popular amongst the ladies (you may have heard of shellac, Gellac and Gelish), you can get UV exposure units really cheap, much cheaper than buying a PCB exposure unit, cheaper/easier than buying the right kind of UV tubes and ballasts and better than making an array of UV LEDs which don't often have the right wavelengths and have to be about 6 inches away from the PCB to expose evenly due to the typical 20 degree beam angle. I got a nail curing unit for all of £11.38 delivered from within my home country (UK) on ebay which has 4x 9W UV tubes 125mm long in a U shape with a GX23 base. They're white coloured tubes which glow bluish. It seems to have individual ballasts for each tube and a microcontroller set 2 minute timer which I found to be perfect. The board illustrated below used just the middle 2 tubes for 2 minutes I did this to reduce the light coming in at the side since it's a tiny piece of board. At some point I'll take the guts out and put them in a second hand scanner I got for £6.95 on ebay as has been done before. It will still fit boards about 145x145mm in its original form which is plenty for the boards I'm working on at the moment so I haven't bothered yet.
2 - I changed developer to sodium metasilicate pentahydrate (Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O). You can buy big tubs of this sold by the electronics distributors although they don't always specify what the active ingredients are. e.g. from RS, Sodium Metasilicate Auto-Positive Developer for Developing. What I think is 2kg of developer is £19.81 at RS including tax and excluding postage, the same weight of Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O is £7.50 all together on ebay, but you don't need anywhere near that much. I don't know if theres anything else in the commercial developer. Apparently it lasts for years, especially in its dry form and it has the advantage over NaOH that it's much less sensitive to temperature and it's much less likely to overdevelop so it's much more user friendly. I made up a roughly 10% w/v solution (i.e 10g in 100ml) and it developed the board in about a minute (my kitchen scales only have a 25g resolution so it's a bit of an estimate but it doesn't seem too critical). I left it in for another couple of minutes while I was trying to take a photo with no ill effect, my camera's aperture seems to be playing up. NaOH would have completely destroyed it if I left it in 30 seconds too long.
So here's the proof:
This is the UV exposure box which you may have seen elsewhere. From top to bottom it's: hard surface, PCB, 2 layers of transparency artwork, sheet of glass, UV box, me putting slight pressure on the top.
Roughtly 10g Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O mixes with 100ml cold/room temperature tap water as shown. Distilled water is probably a better choice but since that amount of Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O cost me about £0.06 I don't mind. Some sources claim you need to mix this with NaOH but you don't. The same safety precautions apply as when using NaOH. They're both corrosive. You should wear correct PPE. Nitrile gloves, lab coat, close access to a sink and good ventilation. I'm a scientist so I have access to all of this but I have to say I didn't bother with the first 2.
Here's what the developed board looks like in the developer. It looked like this after about 1 minute and still looked like this 2 minutes later. One source wrote that you have to dip it into the FeCl[SUB]3[/SUB] to see the progress of development, but this shows it looks just the same as developing in NaOH.
Here's the finished board. For reference the board is 51mm (2") in the longer dimension. I didn't take photos of the etching as it's a standard process. I should note I made the etchant up to the manufacturer's recommendations; 50g ferric chloride hexahydrate (FeCl[SUB]3[/SUB]•6H[SUB]2[/SUB]O) [not anhydrous ferric chloride] in 100ml hot tap water; most amateurs seem to just make it as concentrated as possible (i.e. "until no more will dissolve"). This solution's a week or 2 old and was used at room temperature. After 10 minutes you start to see holes in the exposed copper, after 20 minutes it's mostly gone and I wipe over it gently with a soaked bit of paper towel to finish it off as smaller areas of exposed copper are a somewhat shielded from the etchant. On the board you see a SO-8 package, lots of 0805 SMD packages, a few 1206 packages, a couple of SOT-23 and a couple of TO-220s. The minimum gap between copper pours is 12mil, the smallest areas of copper are around the 0.6mm vias and the text which are both 8mil wide. This is a scrap piece of board and a transparency which I had forgotten to adjust for home etching. I wouldn't recommend trying to get 8/8mil traces/gaps on a home etched board 12/12 or 14/14 are much more likely to be successful but I didn't struggle at all with this board. The rough edge on the right is from the hacksaw, not this process. I wouldn't ordinarily go right up to the edge of the board with copper but this is just a random transparency over scrap. I didn't remove the remaining photoresist, it protects the copper from oxidation and is solderable. The manufacturer recommends leaving it on. It's most of the primary side of a 2 switch forward converter if you were wondering.
Regards,
Matthew
I've just slightly modified my method for making PCBs at home. This may be something you were already aware of, but I only found out about it recently.
I buy positive photoresist pre-coated copper boards (Fotoboard 2 from Rapid/Mega. In this case, the "economy" version which is a CEM1 substrate, not FR4), expose them to UV, develop them with room temperature "Electrolube PDN positive photoresist developer" [Sodium Hydroxide (NaOH)] which takes between 1-2 minutes and then etch in room temperature ferric chloride (FeCl[SUB]3[/SUB]) which takes 20-25 minutes. I roll up a paper towel and use it as a disposable brush for the last bit of development and etching. I probably should get a couple of cheap paintbrushes or pastry brushes. I also wouldn't really recommend the CEM1 boards. They're cheaper than FR4 but you can't really score and snap them as they're mostly cellulose paper they just fray and the integrity of the substrate is reduced for a few mm in from the edge. You have to make sure to file the edges flat and give at least 3-5mm spare at the edges. Hacksaw gets through it easily enough though. It might be easier on the blade than FR4.
That's all pretty standard. There are 2 things I wanted to share which I found elsewhere online:
1 - Now UV curable gel nail polish is popular amongst the ladies (you may have heard of shellac, Gellac and Gelish), you can get UV exposure units really cheap, much cheaper than buying a PCB exposure unit, cheaper/easier than buying the right kind of UV tubes and ballasts and better than making an array of UV LEDs which don't often have the right wavelengths and have to be about 6 inches away from the PCB to expose evenly due to the typical 20 degree beam angle. I got a nail curing unit for all of £11.38 delivered from within my home country (UK) on ebay which has 4x 9W UV tubes 125mm long in a U shape with a GX23 base. They're white coloured tubes which glow bluish. It seems to have individual ballasts for each tube and a microcontroller set 2 minute timer which I found to be perfect. The board illustrated below used just the middle 2 tubes for 2 minutes I did this to reduce the light coming in at the side since it's a tiny piece of board. At some point I'll take the guts out and put them in a second hand scanner I got for £6.95 on ebay as has been done before. It will still fit boards about 145x145mm in its original form which is plenty for the boards I'm working on at the moment so I haven't bothered yet.
2 - I changed developer to sodium metasilicate pentahydrate (Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O). You can buy big tubs of this sold by the electronics distributors although they don't always specify what the active ingredients are. e.g. from RS, Sodium Metasilicate Auto-Positive Developer for Developing. What I think is 2kg of developer is £19.81 at RS including tax and excluding postage, the same weight of Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O is £7.50 all together on ebay, but you don't need anywhere near that much. I don't know if theres anything else in the commercial developer. Apparently it lasts for years, especially in its dry form and it has the advantage over NaOH that it's much less sensitive to temperature and it's much less likely to overdevelop so it's much more user friendly. I made up a roughly 10% w/v solution (i.e 10g in 100ml) and it developed the board in about a minute (my kitchen scales only have a 25g resolution so it's a bit of an estimate but it doesn't seem too critical). I left it in for another couple of minutes while I was trying to take a photo with no ill effect, my camera's aperture seems to be playing up. NaOH would have completely destroyed it if I left it in 30 seconds too long.
So here's the proof:
This is the UV exposure box which you may have seen elsewhere. From top to bottom it's: hard surface, PCB, 2 layers of transparency artwork, sheet of glass, UV box, me putting slight pressure on the top.
Roughtly 10g Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O mixes with 100ml cold/room temperature tap water as shown. Distilled water is probably a better choice but since that amount of Na[SUB]2[/SUB]SiO[SUB]3[/SUB]•5H[SUB]2[/SUB]O cost me about £0.06 I don't mind. Some sources claim you need to mix this with NaOH but you don't. The same safety precautions apply as when using NaOH. They're both corrosive. You should wear correct PPE. Nitrile gloves, lab coat, close access to a sink and good ventilation. I'm a scientist so I have access to all of this but I have to say I didn't bother with the first 2.
Here's what the developed board looks like in the developer. It looked like this after about 1 minute and still looked like this 2 minutes later. One source wrote that you have to dip it into the FeCl[SUB]3[/SUB] to see the progress of development, but this shows it looks just the same as developing in NaOH.
Here's the finished board. For reference the board is 51mm (2") in the longer dimension. I didn't take photos of the etching as it's a standard process. I should note I made the etchant up to the manufacturer's recommendations; 50g ferric chloride hexahydrate (FeCl[SUB]3[/SUB]•6H[SUB]2[/SUB]O) [not anhydrous ferric chloride] in 100ml hot tap water; most amateurs seem to just make it as concentrated as possible (i.e. "until no more will dissolve"). This solution's a week or 2 old and was used at room temperature. After 10 minutes you start to see holes in the exposed copper, after 20 minutes it's mostly gone and I wipe over it gently with a soaked bit of paper towel to finish it off as smaller areas of exposed copper are a somewhat shielded from the etchant. On the board you see a SO-8 package, lots of 0805 SMD packages, a few 1206 packages, a couple of SOT-23 and a couple of TO-220s. The minimum gap between copper pours is 12mil, the smallest areas of copper are around the 0.6mm vias and the text which are both 8mil wide. This is a scrap piece of board and a transparency which I had forgotten to adjust for home etching. I wouldn't recommend trying to get 8/8mil traces/gaps on a home etched board 12/12 or 14/14 are much more likely to be successful but I didn't struggle at all with this board. The rough edge on the right is from the hacksaw, not this process. I wouldn't ordinarily go right up to the edge of the board with copper but this is just a random transparency over scrap. I didn't remove the remaining photoresist, it protects the copper from oxidation and is solderable. The manufacturer recommends leaving it on. It's most of the primary side of a 2 switch forward converter if you were wondering.
Regards,
Matthew
Last edited:
