Heated Bed for Tantillus

So I have been trying to make a heated bed for my tantillus for a while now and after a long journey I have something that seems to work.

My first go at it was a piece of glass, nichrome wire and a thermistor. The wire and thermistor where bonded to it with thermally conductive glue and then support with a crude ceramic structure for a heat shield. It had 50W of heating grunt. It heated up to 60oC in about 10mins from memory but the heat shield completely failed and things got hot where they should not have.

Then I moved to trying to make a sheet resistor on glass by making my own conductive high temp epoxy. While I learnt a lot about this process it did not turn out to be an easy process. So everything went in the bin and it was back to the drawing board.

Then I thought why not finally put my CNC to good use, make the upper bed out of aluminium with a channel in it for silicon heating wire. This would then replace the acrylic upper bed, have sheets of Kaowool bolted to it for insulation and PEEK spacers for the mounting bolts (thermal break). Also mount some fans on a new acrylic lower bed. Some of you may have noticed the hacked config in the solidworks model which details all of this. My theory on this design was the length of heating wire can be changed to change the heating power. Also the glass build platform still remains a piece which can come in and out easily. The down side however is you would need access to machine tools to build this.

Again I made the power 50W, which works out to about 1.5m of 5ohm silicone heating wire at 19V. From memory the calcs showed that the system would heat up at about 0.2oC per second if you ignore losses.

So to the tools:

First step was to machine out the aluminium upper bed piece, after a lot of f'ing around with CAM I finially got one made out of 7075. Then I placed a thermistor in attached to some silicon coated copper wire which I got from hobby king:



Then added the silicone heating wire, crimped to thicker silicone coated copper wire (again from hobby king) and also added some kapton tape to stop the crimps shorting on the aluminium. Also I potted the thermistor with a thermally conductive glue. This gave me:



Then just like a cooking show, I assembled the whole bed up and shoe horned it in (probably should have taken more photos of this, sorry). I rigged the thermistor and heating loop into the ramps. For now I am running the whole machine of a 19V 180W power supply off a all in one PC. All in it looks like this:



The only issue was that the Kaowool paper/blanket sagged a lot in the middle. In the next rev of this I think I will make a retention bracket to prevent this.

Also I changed a few settings in the firmware to get the bed working. I think there was a bug/incomplete code in the LCD section for the heated bed as well. I justed copy and pasted a few things around to get it sorted. When I am next in the garage I will write down what I actually did and post it up.

So then to test it. To do this I used a 16 channel data logger with 14 thermocouples. One was ambient, 12 where tapped across the bed and one (a little into the test) was taped to the middle of the lower bed (side closest to the heated upper bed). Set the temp to 60oC and waited, then to 110oC and waited (note the lower bed fans where not running for this test). The results where good (full excel spreadsheet is avalible in my tantillus github [github.com]):



The key points where that:
1. Lower bed only got to a max of 50oC with fans off
2. Heating up to 60oC takes about 10mins
3. Heating up to 110oC takes about 45mins, looks like the thermal losses of the bed are equal to the 50W input power. Due to this I think I will increase the heating power on MKII.
4. There was about a 5oC variance across the bed @ 60oC. Maybe making the aluminium plate thicker and more heating power will help this.

Now for the real test. I setup and printed a fan mount first without the heated bed on then with the heated bed on at 60oC and logged various temperatures. The results where great! The curl up on the corners was basically gone with the heated bed version:



And the log from the data logger was interesting:



The Z motor even though I had turned the current down to 1.2A (measured on coil) was still running hot (50oC), and hence the mount was running hot also (40-45oC). The upper Z mount was also running a bit hot at 40-45oC.

So MKII I think will have 75W of heating power instead of 50W and make the aluminium upper plate 8mm thick rather than 6mm. Also make a retention bracket for the Kaowool insulation. It will be a while away however as now I have to focus on printing out 5 sets of tantillus parts for mates/work colleagues.

Well that is enough typing for now, anybody got any more idea's/thoughts for MKII?