It happens to me on 533.
@stuartmcknight could you update the title of the post to indicate both current firmwares (522 & 533) can reach this point?
May be @jacob is working on some answer. Seems that some others are also having this problem and reporting it on Facebook.
I upgraded to 533, ambient temps were milder (about 70F) and I completed three successful roasts yesterday without the Erc 0032. Since upgrading from 515 and completing fan calibration, my power seems much more robust. Typically before upgrading, I would need to complete most of my roast at P9 nearing FC, then modulating down to P8/P7; F2/F3 even with a 500g batch size- completing FC at about 11-12 min with ~ 25% DV. Now my roasts begin with P7, fan used is F3-F5, P6 near FC, and roasts complete faster with same DV. FC usually complete by ~10 minutes.
When I was getting ERC 0032 I was under the assumption that power levels were equivalent to before upgrading firmware, and P9 may have been too hot for my batch size. Just a hypothesis.
From memory
515 would not tell you when the bullet was overheating and had shut down the induction, so you might have been roasting but the power would have been 0 at some point in your roast. This would have led to longer roast times, ROR crashes etc.
With 523 the error 0032 would be reported.
With 533 we have improved the cooling performance so hopefully you will see lower temperatures and no over heating anymore.
I encourage everyone to update to 533!