Tuesday, 21 June 2011


Having had a long chat with one of the engineers at Occo, I spent the afternoon testing the cooling capacity of our hydraulic oil cooler. The system temperature does not seem to stabilize, and worryingly I have hardly used the machine yet! I have been measuring the differential across the cooler, and have been seeing about 8 degrees from in to out, which is not enough given that the ambient temperature is around 18 degs and the inlet temperature around 45. One suggestion from the Occo engineer was to remove the cooler from the chassis and test it in an open space, I did this for the test but I only saw an improvement to approx. 10 degs differential...

Cooler mounted between the engine and the fuel tank.

Cooler hanging off of our engine hoist to give more free space around the fan and grill

Friday, 17 June 2011

New Tibia Rams

The new 50mm bore tibia rams arrived this morning and I managed to get them all fitted and commissioned by the end of the day. The control I now have over the tibia axis is in a different league! I think the intertial load of the tibia was just too much for the 40mm bore rams, as I now have the PID loop tuned really nicely, just using the P term and MAX/MIN drive. I think the coxa rams would also benefit from the same size rams, however, I'm concerned about the knock on effect of adding an extra tonne of thrust into the coxa position!

Interestingly the new 50mm bore rams are slightly smaller than the Parker 40mm bore rams, and 3.5Kg lighter! When compared to the Parker 50mm bore ram they are about 9Kg lighter!! this would have been a great weight saving for out machine if we had of gone with this make in the first place.. live and learn :)

Tuesday, 14 June 2011

More CANJoy stuff

Sandy and I spent the day working on the CANJoy modules. I have been writing the firmware while Sandy has been stripping out the HID joysticks and converting them to the new module wiring, Should have them finished tomorrow.

Programming the CANJoy module

Sandy works on the Joystick units.

CANJoy module mounted to the back of the joystick unit.

First Terrain Adaptation Tests

This afternoon I decided to give the load pin sensors a test and see if the Mantis would work in terrain adaptation mode, which is pretty important! First I checked that all the sensors seemed to be working correctly, then we went for it and fired the beast up! All good, we started with just a power up and power down sequence, then I decided to try taking a few steps on to a couple of pallets.. here are the results :)

Monday, 13 June 2011


Having some trouble getting the CANJOY module to programme, but hey I did design it in three days! The module will be used to convert our existing HID joystick into a CAN bus protocol.

CANJOY module plugged into the PIC programmer.

It lives... again :)

Well its been a busy day, all the legs are now re-calibrated, and Sandy and I spent a bit of time trying to get a bit more air out of the rams. I then took the decision to fire it up, just to see how the new coxa pins would hold up, it looks MUCH better! All the feet were pretty much in the correct place and this time the all looked to be under a similar load, which means the calibration is working fine. :)

Re-calibrating the legs using the toughbook tablet!

The Mantis in a super wide stance, just for calibration not for operation!

Trying to get a bit more air out of the rams, unsuccessfully I think.

Sandy sees if he can budge the beast!

Friday, 10 June 2011

CAN Bus Joystick Modules

Some weeks back I designed the CAN bus joystick module which will be used to convert our existing HID joysticks into CAN bus compatible units. I designed the CANADI module as a generic device for general purpose I/O, it has 8 12bit ADC channels, 16 digital opto-isolated inputs and 8 0.5A low side digital outputs. Sandy finished assembling the units this afternoon:

Six legs good.. I think!

Well all six legs are back on the machine with the new Coxa bearing/axle fix. I now have to re-calibrate the coxa and femur axis, then try to get more air our out of the rams. Once this is done I'm going to try and get the PID loop tuned better in order to stiffen up the joints.