Propellant flows to the thrusters via fuel lines and then passes through smaller lines inside the thrusters called propellant inlet tubes that are 25 times narrower than the external fuel lines. Each thruster firing adds tiny amounts of propellant residue, leading to gradual buildup of material over decades. In some of the propellant inlet tubes, the buildup is becoming significant. To slow that buildup, the mission has begun letting the two spacecraft rotate slightly farther in each direction before firing the thrusters. This will reduce the frequency of thruster firings.

The adjustments to the thruster rotation range were made by commands sent in September and October, and they allow the spacecraft to move almost 1 degree farther in each direction than in the past. The mission is also performing fewer, longer firings, which will further reduce the total number of firings done on each spacecraft.

A faraway bit fix
Just six days before Voyager 2's closest approach to Uranus, in 1985, compressed photographic images transmitted from the spacecraft's cameras began to include large blocks of black-and-white lines.
Engineers at the Jet Propulsion Laboratory, Pasadena, Calif., ran an old data stream received from the craft through the programs used to decompress the images back on earth. The engineers concluded that the problem lay not in the ground computers, but in the craft's flight-data subsystem (FDS), which controls on-board instruments and formats data for transmission back to earth. To test the theory, they directed the FDS to transmit the contents of its 8-kilobyte CMOS memories. By comparing that copy of the image-compression program with the original on earth, engineers Dick Rice and Ed Blizzard determined that a single bit of one 16-bit instruction word had changed from a 0 to a 1. Rice and Blizzard prepared a patch that would circumvent the faulty location in the memory. The patch overwrote the instruction before the failed memory cell with a jump command to unused memory. It then executed a copy of the overwritten instruction and the instruction from the defective location, and jumped back to the address following the failed cell.
The patch was transmitted to Voyager, along with a command to reset the incorrect bit. The patch corrected the failure, and in the least possible time, since transmitting a message to Voyager and receiving a response took 41 hours. The reset command failed, and Rice and Blizzard therefore concluded that the bit failure was permanent. With the patch installed, the program sent error-free images. But engineers acknowledged that the actual cause of the failure would likely never be known. The craft will not return to earth "within our lifetime," said a Voyager team member.

Ars Technica did the good ol' fashioned thing and... called the project manager...

“It is difficult to command Voyager," Dodd said. "We don't have any type of simulator for this. We don't have any hardware simulator. We don't have any software simulator... There's no simulator with the FDS, no hardware where we can try it on the ground first before we send it.

I find it incredible that they don't have any kind of simulator. I wouldn't even try to support a mission without at least a software simulator.

There's a little FDS information in Chapter 6 of https://ntrs.nasa.gov/api/citations/1988006...19880069935.pdf but all of the specifics are in JPL internal reports.

I have no trouble envisioning systems running a simulator (or other GSE) becoming unmaintainable, and operations budget becoming too small to pay for port and validation of old software to a new platform.