I wasn't sure where else to post this as it is related to a couple of other threads including fried power supplies, fried power entry modules and fried control boards.
I upgraded the cooling in my FT-5's electronics enclosure using an 80mm double ball bearing fan.
The result is quite dramatic.
The original 40mm fan was making lots of noise and was completely inadequate flow-wise.
I was measuring temperatures through the air "grille" near the power supply fan in excess of 50C during extended prints using an IR temperature meter.
The power supply fan would come on frequently during bed heat-ups, sometimes as early as a minute or two after starting the print.
The air coming out the back door grille was quite warm, but on the other hand there wasn't much of it to begin with.
All of this indicated to me that the power supply fan was mostly recirculating air inside the cabinet.
The temperature of components on the supply's PC board was far too high to allow for a long service life.
I purchased the fan and a couple of wire grilles, printed up adapter plates and took a sabre saw to the back hatch.
The adapter plates were designed so that I could embed the fan into the plate.
This was intended to reduce the amount of intrusion into the cabinet and to make mountings easier (to hide the sawn edges).
The one for the fan side was designed to also provide a smooth transition similar to a velocity stack.
I felt this was necessary after experimenting with the fan and wire grille.
With this short transition the airflow of the fan was not noticeably decreased at all.
Without it, even the wire grille seemed to decrease the airflow and definitely increased noise.
I removed the original fan at the bottom plate and mounted a small grille I printed up, just to keep the wiring in place.
The highest measured temperature anywhere in the cabinet is now ~34C (seen through the exhaust grille near the power supply), and that was with a 23C room ambient.
The power supply fan does not run at all, even after 3 hours of continuous printing with the bed temperature set to 100C.
No temperature on the control board exceeds 28C (including the step stick heat sinks).
I measured the airflow at the exit grille to be 2.5 meters per second with all other original enclosure openings unblocked (such as the ones on the top).
It will reach 3.0 meters per second if the top openings are blocked.
The exhaust opening by the power supply is around 75mm diameter.
The fan output is therefore calculated to be about 27cfm with all restrictions taken into account, which corresponds favorably with the manufacturer's free-air specification of 30cfm.
This then translates to about 2.6 enclosure air changes per second, which is much better than my original goal of 1 per second.
I am now no longer concerned with power supply or PC board component heating.
I hope that this may help someone else.