So my RP3 failed after some use and curious about how it works, I decided to disassemble it. To my dismay the unit was made to prevent dismantling it, much less repair it.
The kinematics are sealed in the stainless steel housing by means of a lip that is roll formed or pressed in at the bottom. I had to turn it down in order to remove it.
After turning down the lip I had to make a sleeve to knock the internal plastic out of the steel housing. If anyone is curious or wants to try it, here's the dimensions, which by the way I had to find by "turn" and error because it wasn't possible to measure it with a caliper. Renishaw really made it a point to make it a PITA to peak inside.
This is how it looks with the lip removed.
Also made a mini clamp out of aluminum to hold the housing down like so:
Inserting yet another special made sleeve with 3 slots cut out at 120° apart from the top and knocking that damn thing out.
It appears the problem was many of the copper contact legs on the thin flexbile pcb were broken. Each leg measures about 0.5mm wide. The kinematics is designed in such a way that the balls form a divot on the flexible pcb, which in turn sit on the machined hole in the steel housing. With usage, the copper contacts eventually fracture at the divot edges. Poor design or wicked business tactic? How does Renishaw not foresee this as bad design? I think the latter is more likely, considering how much they charge for a replacement probe, and a refurbished one at that!
For anyone who wants to know, a good probe should measure less than 100ohm resistance across the pins, mine was 200Kohm.
Right now, I have to go draw up that flexible pcb and send it to some prototype printer shop and see if they can do it. I'll share the dimensions too in case anybody finds it useful in the future.
Last edited by newcomer9747; 12-28-2020 at 04:02 AM.
Missed out an important detail... the plastic barrel that holds the spring has a pointy end solid rubber pad that, I assume, isolates vibration from the machine to the probe tip. The bottom of the probe tip actually pivots on a conical point above the spring.
Spring wire diameter is 0.87mm, 4.625 total coils, 2.625 active coils single closed ends and ground, OD11mm & free length of 16mm. Appears to be of the music wire type. That would be 2.08N/mm. Preload is about half the free length, that should put it at minimum 16N of force on the probe bearings, not including the resultant force of the rubber pad that acts on the probe tip bottom surface.
I hope this post is useful to the DIY probe builders here.
It makes perfect sense that the tripod has a gimble between it and the spring, yet I had never considered that this is how they would be built. All of my probes have the spring acting directly on the tripod. Knowing that coil springs generally exert uneven force across their acting faces, it is obvious why this gimble is beneficial.
Actually it acts more like a spherical joint than a gimbal mechanism, although you're right the idea is similar.
I stupidly threw away the green goo in the housing before finding out it's actually dielectric fluid which quenches any sparks that may occur between the bearings and rods.
Update : I mapped out the FPC and ordered 15pcs for $100. If anyone is interested in the DXF file shoot me a PM.