Program stops when M8 or M9 called intermittent

96 NTC vmc w/Omb control

Within the last 2 years the program will sometimes stop after an M8 call with the coolant pump not starting.
Sometimes after an M9 call the coolant will stop, and the program will not go another line.
NO alarms, just stops
Theses are programs that have been used many times.
Manual on/off on panel turns pump on/off without hesitation.

Today I tried adding a series of M8 and M9 commands to my warmup program that loops and will run continuously until I stop it.
It ran for 30 minutes without a hint of the coolant pump not going on or off as programed.
The pump contactor sounded like it pulls in and releases normally.

On one of the programs I was having a problem with I tried adding a G4 P2000 dwell to the M8 or M9 line and keeping them on a single lines without any other commands. That seemed to help.
Another I added optional stops and turned the coolant on/off manually from the panel

Getting ready to run a part that has 12 tools and I'd like to solve the problem as restarting a program is a pain.

Suggestions appreciated
Mike

A81L-0001-0083/3-01 Fanuc Reactor transformer is mounted within an inch of the pump contactor.
Can a stray or remaining magnetic field from this transformer cause a hesitation in a pump contactor action?

If your ladder waits for a signal from the coolant motor contactor, it could be the auxilary contacts on that contactor that are fouled, or a connection that loosened up. The main contacts in the contactor wouldn't be providing the feedback.

There is no wiring to auxiliary contacts on the contactor
But there are connections to the overload that are possibly feeding a signal, but I would think that would only be a on condition or tripped, not a signal for on/off

I pulled the contactor and made the mistake of taking it apart.
5 minutes apart, 30 to get the springs back in place
actual contacts were not removable without possible destruction of the housing.
Could see them from 1 side , gave them a shot of contact cleaner

reinstalled further away from the reactor transformer

running the warmup program

ordered another contactor