Homebrew PCBs

When a laminator's net roller speed is modified, you also modify the 
heat energy transfer to the rollers as the same surface of the roller 
sees more radiant heat in a given time. Thus it gets hotter without even 
upgrading the heater.
Now if you run the rollers too slow, the roller area being heated vs the 
roller area being sensed will have a hi temperature delta and you could 
burn the rollers. If the gears are exposed to some of this heat 
conducted by the  roller shaft, thermoplastic gears can 'become' plastic 
and fail.

I encountered this heat delta effect when I was developing & monitoring 
my Apache/Trulam mod with a FLIR camera. To eliminate this heat spike, 
whenever the  mod. reverses the roller it cuts the heater. This seems to 
work well to date with no roller damage and good transfers. Thus running 
the laminator at 'normal' speed during the heat up phase and then 
'slowing'  the rollers just for the Toner Transfer phase  and returning 
the laminator to normal speed afterward keeps the rollers from seeing 
too much of a heat delta cycle and they suffer less degradation.

For the different laminators being modified, cheaper units seem to have 
cheaper materials in them, with less accurate controls. Thus it's trial 
and error as to how slow and how hot you can go before you exceed the 
design spec enough to destroy the laminator and make a fire hazard.

Rob's approach eliminates this risk but is not 'over the counter' 
repeatable for everyone and is certainly more time consuming when you 
have a batch to run!

I shipped a built Apache/Trulam  mod. to Italy yesterday based on the 
newer layout with the SMT PIC. It uses low ESR SMT bypass caps and wider 
power rail traces as well to mitigate PIC resets due to transient power 
spijes  as are generated from the relay  and heater switching.