To learn about transparencies, temperatures, and techniques for Toner Transfer, I took the time to run a series of tests. For each set of tests, my laminator's temperature was varied from 320 to 360 F (with ample time between changes to settle), and the number of passes through the laminator was varied from 1 to 12 passes (rotating 90 degrees between passes). For each test, a 2.25 inch square PCB was thoroughly cleaned with sovlent (if needed), then Dawn and a green scrubbie, and the pcb and film were sandwiched in a folded piece of paper "pocket" to avoid slippage and/or sticking to the rollers.The grid below is one row per temperature, with columns for 1, 4, 8, and 12 passes, click on any thumbnail for closeups: 320 F
I note that there are two types of "transfers" - shiny and matte. I don't know why they differ (perhaps the matte spots are where the rollers stuck the toner to the pcb, and it pulled away while it was hot, but the shiny parts are where it remained in contact with the film until it cooled), but if you reflect a light off the board, you can see that both have transferred to the pcb, they just look different. The transparencies only show residual toner where there are obvious spots of missing toner.
Based on the above, looking at unstuck toner and spread due to squishing, I chose the 320 F 8-pass result as a control for some other things to try:
Theory: something was wrong with the toner on the film, so transferring a second film on top of the first would provide a "good" spot of toner wherever the first was "bad".
Test: 8 passes. Cool. Remove film, align a second film over the first transfer. 8 more passes.
Result: Suitable for larger design rules - being off by even a few thou makes a big difference in the results. However, it did fill in the gaps and provide a thicker layer of toner.
Theory: pre-heating the PCB before applying the film would give more stick with less spread
Test: 4 passes without the film, followed by 4 passes with the film
Result: There was still some spread but the results were acceptable
Theory: variations in the paper "pocket" were affecting the lamination.
Test: One pass in the paper pocket, the remaining 7 without it.
Results: Indeed, the transfer was more consistent. However, it appears the paper was also acting as insulation, so there was more spreading. A lower temperature should be tried.
Theory: The unevenness of the copper due to the underlying FR4 was causing uneven pressure and missed spots.
Test: The PCB was sanded to 2000 grit, resulting in a mirror-like surface.
Results: It's hard to see in the photos as the copper was reflects enough light in the scanner that it appears dark, but there was very poor adhesion of the toner! This was a very unexpected result - the toner needs a mechanical bond to transfer. Scuff it!
|webmaster||delorie software privacy|
|Copyright © 2011 by DJ Delorie||Updated Nov 2011|