Now that I produce some more PCBs I have been looking for other ways than my past photosensitive PCB approach. Although this produced good results, it was a bit of tedious to create the negatives and also it needs expensive PCBs with a photosensitive layer. So Google got me to the heat transfer method. I had also read about this but the process with the iron did not seem to be reasonable (at least to me). But then I found a couple of examples where a laminator was used. I had one in the shelf which has been used once or twice a year. Great! A new project :-)
This project deals with high voltage. Only try this if you know what you are doing.
Step 1: Intermediate Operation
Taking the laminator apart revealed that it had a triac controller with a PTC for the heating. The mains is turned on with a double switch. One half of the switch turns on mains and the second half of the switch is used to detect which side the switch is turned. There are also a red and a green LED to indicate something (I forgot their original meaning since I used them for my own indicators).
The two heat stages were tuned for around 80u and around 130u foils. I decided to add an extra switch later for the heat needed for the toner transfer. I did not mention that I was going to use an Attiny? That's definitely easier than hacking some unknown PCB layout (which would have been possible but...). So I needed some intermediate circuit which could control the heat. It's rather easy, though you need to be careful with the high voltage that it operates. I made it overly complicated as I added a phase detection for a PWM control. Not only for the heating but also for the motor. Later I found that the motor control did not work well (it made awful sounds when the power was reduced). Also the heating did not need any PWM control. A simple on/off was sufficient. The parts from the intermediate circuit above were scavenged for the final design below.
Reading the PTC from an Arduino is simple with a voltage divider. So I operated the laminator with my intermediate circuit and found the desired temperatures by experiment.
Step 2: Final Circuit
The final circuit consists of a simple power supply, the Attiny (I used a 85 since I had once lying around, but a 13 would suffice too) and the triac switch for the heating.
The Attiny reads the temperature of the PTC via a voltage divider through port ADC2 (pin 3). PB2 and 3 (pins 7 and 2) check the side of the mains switch (to control the two foil thickness heat values) and the extra switch for the PCB mode. The two LEDs are switched via PB0 and 1 (pins 5 and 6). Finally PB2 (pin 7) controls the triac switch for the heating. My laminator has 300W so I could use a standard triac (I could have scavenged the one from the original laminator electronics, but somehow I did not do that?!). The opto-coupler can likely be any model. I (re-)used the MOC3021 with no zero-crossing detection. Since the heating has no induction it can easily be switched on/off at any time without issue.
Note: the circuit above shows a MOC 3031 (and just a note below that it's actually a 21) which will only work up to 115 VAC and it has zero crossing detection. So use the appropriate opto-coupler for your country! It does not matter whether it has zero crossing detection for this circuit.
At this stage it is also clear that the Attiny must be fused to make the reset pin available for operation (in order to read one of the switches). A high voltage programmer for Attinies is a little circuit you will absolutely need when tinkering with Attinies. I made one from a striped board, a handful of 1k resistors and a NPN transistor (and a lab power supply for 12V).
The transformer I choose returns 6V which would be ok for the Attiny alone. But once the LEDs are turned on, there is some significant ripple. The Attiny still works without issue, but the readings for the PTC follow those ripples and get indifferent. Luckily we do not need exact values and I could resolve this by making averages over the reading. I should have chosen a 9V transformer.
Step 3: Software
This is my current release which works nicely, though I already have another ideas...
Basically this is a state machine which starts in "normal" mode. Depending on the mains switch direction either low or high foil temperature is chosen. One of the LEDs signals the difference between current and desired temperature. So it starts with flashing of a frequency of about 1 Hz until it reaches the desired temperature where it lights constantly. The second LED indicates heating with a permanent light (while the other one blinks). Once the heat is reached it starts to blink (while the other one is steady). During the next minute the temperature is held at around the programmed value. Then the heating is cut off and also both LEDs go off.
While heating up I can press the PCB switch once. This makes the LEDs flash synchronously 3 times and the desired heat is set to what is needed for the PCB (I measured more than 130°C). I will probably change the coding so I need to hold the PCB switch during turning the laminator on to avoid accidental pressing of the switch.
Step 4: Putting It Together
Sure, it fitted with almost no issue. Almost. Accidentally I placed the Attiny on the bottom side of the PCB. I found that out only after the PCB was etched. Of course. But what first looked like a failure turned out to be good luck :-) I had to dremel a hole for the IC socket and - hooray - I could access my Attiny without the need to open the laminator!
Step 5: Extra Thoughts
Now that this methods works, here are some of my results.
Cold transfer did not really work, but it was some initiator for my project. However, I found a nice thing when using normal paper. After cleaning the PCB with detergent and acetone I use a mixture of spiritus and acetone (50/50) to thinly cover the PCB. Now when placing the printed PCB on the copper you can see the print through the paper and the adhesion holds the paper in place when fed into the laminator. The fluid vaporizes immediately and the toner now glues paper and copper together. I run it several (8?) times from all 4 edges through the laminator. Dissolving the paper takes a bit practice to get rid of all fibers, but it leads to good results (after some practice).
Some magazines have a glossy paper which can be used for toner transfer (here in Germany the Reichelt catalog seems to be favored, an electronics distributor). I tried that too and it work better than normal paper when looking at remains of fibers. But the color of the page was also partially transferred (especially red). This needed some manual scratching. But the toner transfers good. Maybe I could try with less heat?
I tried that stuff and hoped for good results. But it was only good for another acetone wash. I have no idea whether my (ebay) source was bad or whether it's plain unusable. But this is off my short list.
Yellow Toner Transfer Paper
After my Blue Fail I tried this (very cheap) paper. And it turned out to deliver excellent results. I fix the yellow paper with some glue drops at the top side (else you get paper jams!) of the print page and run it through the printer. The print must not be touched as it only sticks very loosely. I place the copper on the print and fold 5 mm excess over (be sure to sharply bend the lower and upper edge of the PCB.
The good old iron is still useful. Once the transfer is done, I place the PCB (copper up!) on the heads up iron at maximum heat for one minute. That fixes the toner additionally. You need to be careful when removing the PCB since it's very hot (of course) and the toner is almost fluid (don't touch it!). Once cooled down, the toner is almost welded with the copper.