glitches in monitoring and CLI output

[pwittich]

Describe the bug
When looking at the grafana of apollo204 I see occasional glitches on certain quantities that are reported.

I can also see this on the CLI. see e.g. below on F2_3. These are read out in rapid succession.

% fpga
FPGA monitors
READ_TEMPERATURE_1
 F1_0: 62.00     F1_1: 59.00
 F1_2: 61.50     F1_3: 57.00
 F2_0: 62.50     F2_1: 58.50
 F2_2: 60.00     F2_3: 52.50
% fpga
FPGA monitors
READ_TEMPERATURE_1
 F1_0: 62.00     F1_1: 59.50
 F1_2: 61.00     F1_3: 57.50
 F2_0: 60.50     F2_1: 59.50
 F2_2: 58.50     F2_3: 00.00
% fpga
FPGA monitors
READ_TEMPERATURE_1
 F1_0: 62.50     F1_1: 59.50
 F1_2: 60.00     F1_3: 57.00
 F2_0: 61.00     F2_1: 58.50
 F2_2: 59.50     F2_3: 54.00

This doesn't appear to happen on all quantities (e.g., the ADC from the MCU quantities don't have this issue. (These are internal quantities read out from on-device peripherals.) The received optical power doesn't have this issue (this is read out via I2C from the FF devices), while the firefly temperatures have it on one particular temperature (!!), F1_P1_TX12, which is zero most of the time except occasionaly glitches to a reasonable value.

The ones with issue appear to be the ones with the SMBUS readout, i.e., the FPGAs and the LGA80D power supplies.