Z-axis suddenly starting to move way too far

Hello all,I'm definitely a noob and just finished my Pegasus 10 build. I did go for the Z-axis upgrades with the heated bed. The first test cube went well from the gcode file in the Makerfarm Pegasus software collection. Then, I ran the cube STL through slic3r locally as a test. That cube also ran well, as did a sleeping dog that I downloaded and printed for my wife and a relay board mount file (also in the Makerfarm software collection). Then, yesterday, something changed. Zsteps is still 800/mm, but the Z axis is moving several times more per increase than it should. Parts that I printed yesterday were not solid, and I thought my slic3r settings might be messed up. This morning, I decided to print the original Makerfarm-supplied cube gcode. About halfway through, the 'structure' was about 60 mm high. X and Y looked to be moving as they should, but Z was jumping more than it should with each layer. Recommendations?Regards,Stan McIntosh

*sigh* I meant to go to the Makerfarm forum with this. Oh, well. My guess is that this is a RAMPS controller issue and not really tied to Makerfarm, so *fingers crossed*.

Colin has been tremendously helpful, as usual. Starting to look like one of the dividers in the Z-axis part of the controller has stopped working. Next few days are slammed, so will have to let this wait until the end of the weekend.

Problem solved, sort of. I changed out the stepper driver board, and now the Z-axis is moving well. However, somewhere in the process, my y axis has started missing steps. I'll post on that in a bit, most likely. The belt is tight, the skips are audible, and the pulley set screws are tight. When I watch the belt and pulley, they move and stop together, even with the missed steps. If I start a print of a test cube, the bottom layers look great, and then Y starts missing when going to the more acceleration-intense wall buildup. Gonna double-check Vref, but might play with maximum acceleration a bit.

Y axis problem now solved after putting a heatsink onto the Y axis driver chip. Y was slipping when test cube got to walls, and now is steady. Colin recommends avoiding heatsinks, and I can see why. The tape that comes with the driver boards looks to have a metal film as the substrate, which could cause shorting issues. I was careful. Now that printer is working, will print fan mounts tonight. Maybe a fan can eliminate the need for the heatsinks.