In ANS this week the note on AO-16 is a bit incomplete regarding the current state of the satellite. Here is some further detail.
The on-board computer has not failed. What is no longer working is the computer's ability to access the EDAC SRAM needed to run uploaded software. The boot loader (MBL) software was burned into ROM prior to launch and continues to run. The processor itself works OK and is running that MBL software just fine. One of the unique features of MBL is that it does not use any SRAM but runs entirely in processor registers. It also does not use DMA or interrupts. One of the design goals was for it to continue to run even when several hardware things had failed - and that has proven to work well now that the SRAM access no longer works.
With the computer still processing instructions and running MBL it remains possible to send commands directly to various hardware modules using the AART bus (another innovation in these satellites). It is that capability which allowed the mode to be switched. Further, we know the computer continues to successfully run MBL because any glitch in processing would likely reset the computer. That would cause MBL to restart and reset the hardware to the default, turning off the U TX, and resetting the downlink to digital. Should that happen it might be possible to get it back into the repeat mode, and perhaps do that more than once. But at this point it's anyone guess how long enough hardware will work to allow more use from this rugged old bird.
Jim jim@coloradosatellite.com wd0e@amsat.org