Modified Laminator for PCB Toner Transfer - Revisited

A little more than a year ago, I came across this Instructable: https://www.instructables.com/Modified-laminator-for-PCB-Toner-transfer/ .

As luck would have it, I had tried the same PCB toner transfer method for producing PCBs, and with exactly the same, cheap laminator unit, at that!

Alas, I didn't succeed. I had the same problems the author describes. My first thoughts, then, were to increase the laminator's working temperature, but that's not really an easy task. The author's method - reducing the laminator's feed rate - is more promising as well as much simpler.

Unfortunately, there are two drawbacks to the abovementioned Instructable:

1. It is over 10 years old, and the download links for the microcontroller firmware and the schematics are dead since a long time now. Several readers had already commented and complained about that, but did never receive an answer from the author. I found out that this is the only Instructable he ever published, so I assume that he is no more active here - and perhaps even doesn't know about the missing files.

2. I think using a microcontroller for this purpose - turning the laminator's motor on for a half second every 10 seconds - is like cracking a peanut with a sledgehammer (or, as we say in Switzerland, shooting sparrows with a cannon). A simple circuit with a 555 timer is the perfect solution for this problem, no software required.

Did you know, btw, that the 555 timer was first introduced in 1972 - i.e., more than 50 years ago!, designed by a Swiss engineer, and that it is, according to the German-language Wikipedia, considered as the worldwide top-selling IC chip ever?

However, since this circuit is supplied directly from the mains (yes, the complete circuit is connected to the mains - there is no transformer needed, but then there is no part of the circuit safe to touch!), it is very important that you know what you are doing. This is NOT a project for a beginner!

• F1 - Fuse T 100 mA (slow blow)
• R1A - 1 MΩ
• R1B - 300 kΩ
• R2 - 82 kΩ
• R3 - 3 kΩ
• R4 - 100 Ω, 1 W
• R5A, R5B - 2 x 470 Ω, 0.5 W (or 1 x 1 kΩ, 1 W)
• R6 - 3.9 kΩ
• C1 - 10 µF/16 V (electrolytic)
• C2 - 10 nF
• C3 - 220 nF/250 VAC/X2
• C4 - 470 µF/25 V (electrolytic)
• C5 - 10 nF/250 VAC/X2
• D1, D7 - LED, 2 mA, dia. 5 mm
• D2, D3 - 1N4004
• D4, D5 - 10 V/1 W (Zener diode)
• D6 - 1N4148 (or similar)
• U1 - ICM7555xP (CMOS 555 timer) in 8-pin DIL package
• U2 - BRT13H (Triac Optocoupler) in 6-pin DIL package
• A piece of prototyping (matrix) board (for dimensions, refer to the component layout drawing)
• Solder-type screw terminal blocks (2 x 2, 1 x 4)
• Fuse holder for F1
• DIL-8 socket for U1
• DIL-6 socket for U2
• some bare wire, dia. 0.6 mm (copper is ok, if tinned it's even better)
• some heat-shrink tubing

Total price of the components should be less than about CHF/EUR/USD 25.-

In the diagram, from left to right, you find first a power supply that's rather elegant, but not my idea, unfortunately. Since a mains transformer would not fit into the laminator's case, we have a direct supply from the mains (F1, R4, D2-5, C3, C4) that delivers about 9.5 Volts up to a current of about 15 milli-Ampère (mA). This is more than sufficient for the CMOS timer chip, the optocoupler and an additional 'power on' LED. Please note that it is very important that C3 is of the 'X2' type for safety.

Please: Never forget that no part of this circuit is safe to touch while connected to the mains supply!

The circuitry around the CMOS 555 timer chip (U1) is a standard circuit. It pulls its output down for half a second every ten seconds and thus activates both LED D1 and the LED within the triac optocoupler (U2).

The component values can be easily calculated online, e.g. with the calculator you can find here: https://ohmslawcalculator.com/555-astable-calculat...

The triac optocoupler has a current capacity of 300 mA max. that is far above the requirement of the laminator's motor (about 20 mA), so its output can be directly used without an additional triac. A snubber circuit (R5, C5) protects the optocoupler's internal triac from voltage spikes. C5 must be of the 'X2' type as well.

The power-on indicator LED (R6, D7) was added later since there was an additional hole in the laminator's case. Therefor it is shown in the proto board's photograph, but not included in the component layout drawing below; the two red asterisks ('*') show where it can be connected.

For the construction, refer to the drawing and the pictures above.

If you wish, you might as well fabricate a PCB. But when you only need one single piece of a circuit, I think that it's easier, faster, and certainly cheaper to use the hand-wired, stone-age version on prototyping board.

Building is easiest if you first insert the smaller components (one wire bridge on the component side, resistors, diodes), then the next larger ones (IC sockets, the smaller capacitors, terminal blocks...), and the largest at last. The wiring is done with bare wire of about 0.6 mm diameter on the bottom ('solder') side of the proto board, except the already mentioned wire bridge next to C2; it is marked in red in the component layout drawing. When wiring, make sure to always have sufficient clearance between the mains' live and neutral conductors, and take care to keep them well separated from the rest of the circuit; the design shown above indicates about the minimum required clearances.

Needless to say that you MUST observe the polarity of the diodes and LEDs, the electrolytic capacitors and the two DIL chips - otherwise you have to expect some fireworks immediately after powering up!

This as well is not a beginner's task! In order to complete this successfully, you need to keep your wits together and know what you're doing. But if I was able to do it, you should succeed too :-)

There was some hay-wire circuitry within the original laminator; I managed to remove some of the (now) unused wires and even tidying up a little bit.

First, you need to remove some of the plastic parts from the bottom of the laminator's case in order to make room for the PCB (circled in the picture above). Use whatever tool you like (wire cutter, hacksaw, Dremel) but try to avoid larger damage.

Then, remove the little PCB that contains the red and green LEDs and some other components, plus its wires. This will no more be used.

The wires that go to the HOT/COLD switch are unsoldered from the switch; they are connected and insulated by a piece of heat-shrink tube. This switch will later be used for selecting normal (CONTinuous) or pulsed (PCB) operation. It is connected to the corresponding PCB/CONT terminal block with two new wires.

The wires from the motor are connected to the two central terminals labeled 'Motor' of the large terminal block, the neutral wire from the mains cable to the one labeled 'Neutral', and the wire that comes from the ON/OFF switch (white in my unit) is connected to the one labeled 'Live'.

LED D1 is connected directly to the remaining terminal block; when fixing it there, you can slide the PCB into the case and let the LED protrude through the corresponding hole next to the ex-HOT/COLD switch (that will then be labeled CONT/PCB). The 'power on' LED D7 that was added later can then be inserted into the remaining hole next to the CONT/PCB switch and fixed there on the cheap with a drop of hot glue.

As you can see from the photographs, the PCB was a tiny bit too large; one of its corners had to be cut off in order to fit into the case. It is fixed inside the case with a drop from the hot glue gun (to keep it cheap & simple) as well.

After closing the case and a final test (which, I admit, needs summoning up some courage) you're DONE!

Please consider that after this modification, the laminator's heater is always on as long as the ON/OFF switch is on. This is not a problem for me; I never had to deal with cold laminating sleeves since I own this unit.

In addition, you cannot select different delay times, as was provided by the first design with the microcontroller. But the ten-second delay is ok with me, again, no problem for me. When switched to CONT, you can use the laminator as usual for standard, hot laminating sleeves.

It is most likely that you have a different make and model of laminator. But the principles will be the same; installing this circuit in your model, however, will need some fantasy, experience, and creativity. Let me highlight this again: It is NOT a beginner's project. So, you rookies out there, please beware!

Apologies for taking so long!

The picture above shows the first, quick & dirty test on 1.5 mm FR-4 PCB material. It isn't 100% perfect but I consider it fair enough for a first test run. Some touching-up with a permanent marker will be required before etching, as you can see from the picture.

Three things are rather important for this method, btw:

• Don't forget to print a mirror image of your artwork for laminating - unless you did that already during the design process
• The copper surface must be perfectly clean, free from oxide, and degreased - e.g. by first sanding it with fine abrasive paper (or a ScotchBrite pad, or similar), then cleaning it using a solvent such as rubbing alcohol, benzine, or acetone
• After the whole assembly – i.e., the PCB material, together with the printed transfer paper – has run through the laminator, allow it to cool down to room temperature before carefully lifting the transfer paper off

The paper I used for transferring was the backing sheet for laser-printable wiring labels. On this paper, the toner adheres nicely, but also transfers well. Further experiments with other paper as well as with the Press'n'Peel toner transfer system will follow.

The 1.5 mm thick material I used for the test seems a little fat for the laminator's motor and gears - a subsequent test with 0.8 mm thick material went more smoothly.

One more input: I will try again with the 1.5 mm thick material, feeding it not in parallel with the laminator's guides, but slightly slanted instead - assuming that the mechanism might eat it a little better when 'starting small'.

Tried again with the 1.5 mm thick material. When pushing a little at the beginning - i.e., 'helping' the transport rollers a bit - for, say, the first 2 to 3 'on' cycles, it works fine with this thickness, too. However, my idea from the last paragraph in Step 5 above (inserting it in a slanted way) failed insofar as the paper was forced away sideways a little during the transfer, and long, straight tracks got a bit curved. Not good.

Since then, a friend of mine gave me his retired, vintage etching system. It was manufactured by Velleman and is an early version of the product shown here (no, I'm not sponsored by them):

https://www.velleman.eu/products/view/?id=369500

It contains a vertical etching tank, a submerged, temperature-controlled heater for the etching fluid, and an air pump for fish tanks that produces bubbles - and so movement - within the etching agent. Heating the etching agent and moving it around considerably increase the etching speed.

Please note: There are several Instructables available already dealing with DIY etching tanks. It isn't at all mandatory to buy a complete system - but if you can get it cheap or even for free, why hesitate?

First - to be safe rather than sorry - I checked my 'new', vintage etching tank for leaks, just using plain tap water. When I was convinced that it was waterproof, I gave the system a little TLC. Once that done, PCBs coated with the toner transfer method described above were nicely etched within less than 10 minutes, using ferric chloride as etching agent. As you can easily see from the pictures, in a second go that comprised redrawing/printing/transferring my design I slightly increased the width of the copper tracks before transferring/etching.

After etching, I used acetone to remove the toner, and the result is shown in the picture above. During cleaning, some toner residue seems to have penetrated the FR-4 substrate, slightly discolouring it in several places - which is, of course, some cosmetic drawback only. It doesn't compromise the function of the finished PCB at all.

At the end of the day I'm rather satisfied with the result. It cannot, of course, compare with commercially/professionally manufactured PCBs (there is neither a solder mask nor a silkscreen print, the solder pads are just plain copper instead of tinned or even gold-plated, and it's single-sided only for the time being), but it certainly is faster, cheaper - and, most of all, done by myself.

... will consist of drilling the holes in the pads, in order to insert and solder the component leads. Since this isn't really a trivial procedure, I made a separate Instructable dealing with it:

https://www.instructables.com/PCB-Drill-Press-With-Improved-Aim/