I recall that he said they'd fixed it in 2023, but I don't recall specifically when
He said 2024:
A lot of people know that our basses had a big issue with snapping truss rods. Very few people actually understand why it happened. A ton of people guess incorrectly (including the McLarens) that the rod itself was the problem. This was not the case.
So you want to buy an instrument in the era before I fixed the program
(pre-March/April 2024) and you want to know what to look for, Besides the obvious of the adjustment nut just already being broken.
It's immediately obvious if you have a bad one. You have a neck with way too much relief, so the action is a mile high, and as you try to tighten the rod it does nothing to straighten the neck until you inevitably max it out and it snaps the adjustment nut off the threads. The rod turns and turns, then eventually you over torque the threaded rod the nut is welded to and it snaps off with a loud pop.
Basses have a lot more tension than guitars. That coupled with the longer overall neck length makes these very obvious to anyone that knows how to check neck relief.
The issue ultimately came down to an error in the CNC programming for the depth of the truss rod channel, on the bass program specifically. (Guitars and Fallout Basses were not affected by this error.)
Someone made it too deep, which allowed the rod to pull in on itself. Truss rods work because they are firmly anchored between the fingerboard and neck shaft with as little wiggle room as possible. Ours had upwards of .020-.050" of room, depending on a few variables. The woodshop manager prepped all the neck blanks by putting them through a thickness sander from the 70's. It was desperately in need of maintenance, so it didn't give consistent thickness results. The guy would check one of them and see it was close to spec (.800") and then send them all through. The problem was, the machine would give you variance of +/- .020" on a good day.
So out of a batch of 10-15 necks, the blank thickness could be off by as much as .020" too thick. Add that to the depth being wrong, and you get a wild swing in depth from neck to neck. Some were only off by about .020", which is close enough to get by when you factor in glue squeezing into the channel to fill the gap. It's still not great, but it can work a little bit. When you add in the other discrepancy it becomes untenable.
Our CNC worked by basing "zero" off the flat surface of the fixture the blank was mounted to, so it went to the same exact depth every time assuming we've properly thicknessed the blank to spec. We could have zero'd the piece every time, but that would grind production to a halt. It's a tedious process to do that on our older HAAS machines.
The workaround was to have the machine "fly cut" the entire surface of the blank to .800" based off the fixture during the truss rod program, so we knew it was perfect. It stopped the issue completely after that.
I ultimately got tired of seeing them returned or seeing them pop in final assembly, and I said "shag it" and ripped several in half with the bandsaw. That allowed me to see the problem with the rod still in place.
I'll also reiterate, because I personally prepped every body and neck blank for Custom Shop builds during my time there, and ran the CNC programs for those cuts in about 90-95% of them, I was able to avoid the issue on our builds by properly spec'ing blanks using a different thickness sander in our shop, and checking every single one with calipers. We also worked in much smaller batches than standard production, so it was easier to spot discrepancies