tp20 madness!

Make sure you clean the probe interfaces with the yellow renishaw CK200 goo. 
Since magnets hold the tp20 module on, any little shard of debris (specifically metal shavings) will promote egregious & random erroneous data. 

TP20's also have the shortest lifespan of all the renishaw product (25,000 changes per the manual, and unofficially, ~2 million hits).
TP200's for example are spec'd to last >10 million hits.

Swapping modules (with new to compare) is the formal method of affirming end of life for a tp20 module.

How often should we be cleaning them? We occasionally get overspeed errors in different Axes (once every 1-2 months), and probe unseated even at low speeds, but very rare. Not sure how old this TP20 module is we have.

-Overpseed is the CMM's drivetrain not keeping up with what it was commanded to do.  Not related to the probe sensor whatsoever.
-Probe unseated is the result of potentially a dying probe sensor or noise in your drivetrain (probe vibrates when moving and triggers a hit).  You can (at increased risk of crashing with catastrophic results) get Hexagon to switch off probe sensor trigger detection while it's traversing. 

From Renishaw: 

Renishaw recommends that all kinematic coupling mechanisms incorporated in probing devices are cleaned before first use, and thereafter at regular intervals.

Each kinematic cleaning kit comprises of strips of 'yellow tack' material which should be used to clean the precision ball seatings and permanent magnets found on the probe and probe heads. Sensors requiring cleaning include: REVO®, PH20, RSP2, RSP3, SFP, SP25M, SP25 modules, TP20 modules and TP200 modules.

--We clean and calibrate our TP200's once a month, but we are in a non-production environment, and haven't discovered any contamination issues at this frequency.  your environment might justify more or less frequent cleaning.

This is a new Global S from July 2022, how likely  is it that we experiencing drivetrain issues already? These occur very very infrequently and seem to be random, like months to weeks apart and can't really be reproduced. This happens at low speeds more often than higher speeds like during Find Center. I am not sure of how rare/frequent drivetrain issues occur in CMMs.

You can get these errors form something as simple as debris on the bearing surface. I remember once finding a blob of donut glaze somehow on our Y axis bearing surface, which produced overcurrent errors. All Food & drink in the room was promptly banned lol. But if you have an S Chrome, understand that they have a unique design. they have springs mounted to the belt interface on each axis, enabling the machine to have really high accel parameters... but in my honest opinion at an unintentional cost (weird stuff like this, especially if you have fabreeka vibration isolation).

Is there like a CMM maintenance guidebook? We clean our scales, and wipe down the machine weekly. I feel like there is more I could be doing to maintain the machines condition. Hexagon when they installed it never really gave us any information for that. We aren't very high volume, and our longest programs are like 25 minutes.

Is there like a CMM maintenance guidebook? We clean our scales, and wipe down the machine weekly. I feel like there is more I could be doing to maintain the machines condition. Hexagon when they installed it never really gave us any information for that. We aren't very high volume, and our longest programs are like 25 minutes.

There's this PDF I got with our most recent purchase "H00007841_User's Manual Globals.PDF"

--you seriously hijacked this post.

 PDF

I may have hijacked this post, but you sir have a lot of knowledge on all things CMM and are very reliable with your responses. Your very existence in this post is hijacking I say! Thanks a lot for your help answering my questions. I'll give this post back now.

Don't forget to mash the pins back into the mounting surface of your TP20 modules. If they have worked themselves out it can make for an unreliable connection! 
(yes I am serious) But you should probably look at replacing your module if it has the pins sticking out too far.... In my experience that only happens to modules that nightshift smashes up, and then conveniently forgets to mention until you ask later. (I wish I was making this up) 

But that all being said, if the pins are sticking out to far it causes the module to not fully seat and then rock a bit, which breaks the connection during travel and gives the probe unseated alarm.

We have had people on off shifts break probes at 0,0 going down in Z right into the part, so I am very wary of the pins being properly seated. I should make a tip change counter, but that sounds like a lot of work that needs to be added to every program, and probably overkill.

I had an operator smash a probe so hard he drove the internals through it and broke the magnet on a TP 20. He was immediately banned from CMM use from that point on...lol

I am impressed he wasn't fired, despite that probably being an budget cost for most companies. We have pretty high leniency here when it comes to probe tips, not sure how they feel about modules, or even entire probe heads. Don't wanna find out.