Printrbot Simple Metal's Heated bed

posted 14 Feb 2015, 14:11 by Colin Bell   [ updated 14 Feb 2015, 14:16 ]

This is the first of an irregular series of posts focussing around the servicing and maintenance of the Printrbot Simple Metal.

One of the recent upgrades I made to my printer is the heated bed kit. This consists of new 'wings' for the build platform, a PCB heater, some wiring and a nice milled aluminium build platform.

I won't bother going through the installation here - it's very simple and well documented on the Printrbot web site.

Before I even fired it up for the first time, I knew it wasn't going to be the perfect solution I was hoping for. Whilst it does work, and very reliably at that - it simply isn't deigned well enough for high temperature printing. Most materials appreciate a bit of warmth underneath to help with reducing warping etc, ABS in particular demands a consistent bed temperature of between 105c and 110c unless you're using copious amounts of glue or a PEI material on the build platform. Whilst I'm no expert on thermal design, I can see that no insulation underneath is a problem, and those wings...they're basically giant frickin' heatsinks to warm up my room. All in all about as efficient as a fart in a thunderstorm.

With just the stock kit installed, the bed took around twenty minutes to get to 85c and any setting higher than that took in the order of five to eight minutes per degree increase. For PLA, this was fine - it got up to 60c from room temperature (about 22c) in around ten minutes, which was slow but acceptable. As I'd recently upgraded my power supply to a beefy ATX2 capable of supplying 25A on the 12v rail, I'm pretty sure I'm pushing enough power through it. So...the problem is likely elsewhere.

Looking at the old RepRap Huxley on the shelf, which uses a Nichrome wire system on its bed, I can see that it has an automotive exhaust heat shield over the Nichrome. Also, knowing that PCBs are not great insulators, I thought this would be a good place to start. I taped the shield centrally over the heater PCB with good old Kapton tape (where would I be without that stuff?) and made sure to insulate wiring and screw heads etc to reduce any electrical short risk.

Initial tests with this saw the bed temperature shooting up to 104c in around 27 minutes. Compared to it's previous maximum of 85c, this is a win!

Now for the wings - these are kind of necessary to hold the bed drive belt in place, so losing them isn't an option without some serious modification of the structure of the bed. Which might come later.

For now though, I wanted to try and get some insulation between the aluminium bed and the wings, in the hopes of reducing the thermal losses I'm getting from them. I know they get extremely hot normally, and I've measured them at around 80c after around 40 minutes of the bed running at 85c. All this heat is just going to waste, and it's an additional burn risk to boot.

I had some very high density foam left over from some device packaging, which I put a piece of in the oven at 250c for an hour to see if it would survive. Not only did it survive, but it seemed just as springy and did not warp, scorch or burn. I'd love to find out exactly what it is, as I think it would be handy for a number of things.

I cut four lengths of this about 2mm thick to sit between the wings and the build plate and screwed it down nice and tight, squashing the foam in place.

The difference with these installed is good and bad...heating is now a bit quicker than it was, but not dramatically so. It can now go from room temperature to 60c in 3 minutes flat, but the rise to 110 still took 32 minutes in total. But it does now get to 110c and maintain it! The rise times seem to slow down from around 70c upwards.

The bad...those bloody wings are still getting hot over time. The foam seems to have slowed this a bit, but the heat is still creeping over. Now, with the bed reading at 95c, the wings are still sitting around 67c, which is not as good as I'd hoped.

In terms of use though, what I've found has helped a lot with heating times, is to ensure that all the fans are off (doh...) and also to cover the top of the build plate with a biscuit tin cover until it's done heating. This can knock 10-12 minutes of the total heat-up time.

Once the machine starts printing, I've been using a bed temperature 110c for the first 3 layers, and 95 for the remainder. Of course, the surface of the bed isn't as high as the 
thermistor reading - in fact it's around 10c lower I'm observing. Covering all exposed areas of the aluminium bed has further improved stability over time, I'm seeing the bed kick on at the same frequency to maintain temperatures, but for a lower amount of time. 

Tomorrow I will get some adhesive pipe insulation tape and cover everything with that - it can't hurt and it's a cheap test.

I had hoped that all this effort might mitigate the need for glue on large parts, but even as I'm writing this, I just had a failure due to warping three hours into the job....out with the glue stick again!