WS-12-250-AC rev. D higher thermal than rev. F ?

[IntL-Daniel]

I have WS-12-250-AC for years (rev. D) and sometimes I had unexpected reboot. NETONIX support said it must be probaby hw issue of the switch (probably power supply) and recommended to try different unit there. So I did it and change it to the new one (rev. F). Currently I can see big difference in thermal status of the new one (rev. F) in compare with the old one (rev. D)... especially look at board temperature and uptime:

WS-12-250-AC rev. D in production

WS-12-250-AC rev. F in production (same place as previous)

WS-12-250-AC-rev. D - in the office, no devices connected, replaced new fan

So is this difference in thermal expected on both revisions? Or the old one rev. D has some issue? Thanks for help.

[sirhc]

No, none of the board revs affected thermals. Most board rev changes were due to parts such as sensors being EOL and needing a drop in replacement that required software changes to deal with them. Some changes were made in varies attempts to make the switch last longer with poor grounding. Once change was to increase 24V power supply from 106 to 160 watts. NOTHING TO DO WITH THERMALS. However chassis design changed to make holes smaller to combat bugs getting into switches but overall void volume was kept nearly the same.

First off the PHY and CPU sensors are built into the chip and are not real accurate so say the manufacturer but the SOC/PHY/CPU is rated to 125C which you are nowhere near.

The board sensor is more accurate but again +/- tolerances and the board is fine up to around 70C but we like to keep it cooler to prolong MTBF.

The temp here is not a precise science and not that critical as we are simply trying to increase the MTBF of the unit.

The power supplies we use are rated as convection cool yet we used forced air - BONUS TO MTBF
The board and most components are rated to 70-90C but you should not see 70C with the fans except in extreme conditions.

All your screen shoots are fine on temp within range.