Hydra: Heated build table and a4983 stepper drivers

So the end of the semester has been pretty hectic and for that reason I am way behind on updating this blog. I'm probably gonna end up skipping about 3 weeks or so of work in the middle, but I am going to jump to some of the final details on the Hydra-MMM prototype. Starting off, here is the work that has been done on the heated table.

We had always wanted a heated table that could get up to ~120C with a large Kapton surface for bigger builds. The problem turned out to be the wiring and not the actual heating method. We had selected five 6 Ohm resistors that would each be wired in parallel to a 12V source. This would put about 2 Amps through each chassis mount power resistor leading to 24W per resistor. This totals for a 120W table, which is enough to heat the 1/8" Aluminum 8"x8" plate up to about 130C. The problem, however, was the fact that at this amperage, the wires we were using literally methed and fell apart. We tried to solder the joints, but the solder literally wouldn't hold at this temperature. We finally resorted to stripping an entire length of 18 gauge wire and running it all the way around the table, through each resistor end. By doing this, the wire really had no where to fall to and would remain connected to the resistors. We added some high-temp solder for good measure and have been using the table ever since. Here is a few pictures

The table wiring scheme using CPU thermal paste for good heat transmission and Kapton to anchor the resistors.

A successful 6 part build with several parts around 5" in length. Out of all these parts there was only one that had any warping issues and they were very minor at that.

This was the 3rd or 4th revision of the table, but it has performed very well for the past few weeks. The table temps can get to 120C in about 25 minutes using a 12V power supply and stay very stable at that temperature. The PID control that the Hydra-MMM firmware uses helps this a lot and has definitely gotten rid of the variable extruder forces I had seen with simple threshold on/off temperature control platforms. Unless the table decides to fail again (fingers crossed!) this will likely be the final revision of this component.

With the heated table working, the next big step was to revise the electronics we were using. Previously, everything was wired up on a few breadboards, which had created a lot of noise and reliability problems. Somewhere in the middle, we had moved over to a single perf board with all the motor drivers, TC circuits, and heater FETs worked in, but had been pushed to it's limit and we wanted more. With the intermediate electronics, the max speed on our axes was around 100 inch/min (IPM), which was a little below what we were hoping for. I decided to order 7 Pololu a4983 stepper motor drivers which would allow for chopper based motor control at a much higher voltage. This higher voltage would allow the motor coils to energize faster leading to quicker movements. They also have the ability for microstepping, which could smooth out the operation of the machine as a whole, or at least improve the extruder performance. So I set out to wire up a board solely for the 4983 breakout boards. The board takes a 30V input from a pretty cheap Chinese power supply and regulates that down to 5V through a single 7805 regulator. The Pololu drivers are given the motor supply of 30V with a logic supply of 5V from the regulator. A few pictures of the wiring are below

The backside of the board. Notice that each of the motor outputs is soldered directly to a shielded, 18 gauge, 4-strand piece or wire that is eventually tied to the CB connectors mentioned previously.

The top side of the board with all the 4983 breakout boards plugged in.

Initial tests of this board proved extremely successful. The motors were able to move MUCH faster and we actually hit a new record speed of 220 inch/min! This was much better than we had hoped for! This is largely due to switching from a 12V to a 30V power supply, but I was still impressed with how much improvement there was.

Next step is to get everything in its own electronics enclosure and get the machine ready for showtime!