Okay guys
The replacement part allowed the unit to be completed a while ago, but was then put on the back burner after an initial check showed that it worked.
The pics below show the overall hardware setup for the 4 load cell version intended to cope with larger & heavier models.
Pairs of load cells are bolted onto 3D printed bases, connected together by three 6mm carbon rods which allow the spacing to be easily adjusted whilst keeping the bases parallel. The load cell amplifiers (smaller size ones) will just fit between the load cells & are covered by a 3D printed channel section. Two 6 way ribbon cables connect each pair of amplifiers to the processing unit.
I’ve used 10Kg load cells, but larger ones can easily be used if required without any other change as they all use the same mechanical interface & the calibration routine looks after the range values.
The processing unit is an ‘Arduino Every’ module which is very similar to the ‘Nano’ used in the 2 load cell version, but has the extra memory resources needed to cope with the extra variables used by 4 sets of load cell processing. The processor unit has connections to the 2 sets of load cells, the I2C LCD display and power (which can be either via the USB connector or via the ‘RAW’ input).
As can be seen from the photos, I have used 3D printed load cell mounts which accept 6mm carbon rods, topped by 3D printed levelling pads, the front ones are equipped to take LE reference pins.
My measurement points are 180mm apart, with the LE reference 50mm forward of the front measurement point, thus it is suited for wings which have a chord length of 250mm+. As for the previous version, these parameters can be altered to suit a different physical arrangement.
I have used 6mm carbon rods throughout, but in hindsight it may have been better to use a stiffer 8mm or so rod for the vertical stands, both to take the weight & minimise wing pad ‘spread’ (which will affect CoG location accuracy).
The 3D objects should be able to be easily modified to take larger rods.
Note that the load cell adapters were designed with a largish central open area to allow the bolts to be inserted.
The firmware is very similar to the 2 load cell version (just has 4 load cell sensing routines rather than 2) and adds the 2 front/rear cell load values together to arrive at the front & rear weights.
I have checked the CoG measurement by placing a known load on the stands & confirming that the load & CoG offset are correct. Not managed to put a large model on it yet though.
I will add a zip file containing the firmware and the 3D print objects, although this may now go into the relevant area in the Gold members section.