I'm not sure of the pros and cons of fan ovens, but where there is line-of-sight from the elements to the worpiece, the radiated heat from the elements can make a big difference to the actual temperature the blade reaches. Overheating due to radiation can be particularly bad where the section is thinnest, causing a softer edge than spine; the opposite of what we'd normally want.
The best simple fix I know of is to bury the blade in a tray of dry sand and temper for much longer than you would a bare blade.
The added thermal mass damps out any temperature variations and the fact that the blade is buried shields it from the radiated heat.
I appreciate it's not directly comparable to using a domestic oven, but some time ago, I put a series of thermocouples of different diameters in a HT oven I'd built and logged the temperatures during a tempering cycle.
The idea was that the different diameters would show the temperatures likely to be seen by the different thicknesses of steel at various distances from the edge of the blade.
The thermocouples were an exposed junction type, 0.5mm dia (.020"), 1.5mm dia (.060") and the control thermocouple was 6mm dia (1/4"). The cursor in the screenshot is at the peak temperature of the overshoot caused by radiative heating. Temperatures are in degC. The control setpoint was 250 degC.
The peak "edge" temperature in the above shot shows 119 degC above the setpoint; 214 degF, or about 5 Rockwell points softer than intended on O1.
The next shot shows the effect of using a ramp/soak controller to ramp the temperature up slowly to the same setpoint; "only" 19 degC (34 degF) and about one Rockwell point.
The effect will be much less pronounced with a domestic oven, but could easily cost a couple of Rockwell points at the edge. To me it seems well worth avoiding the issue by using the tray-of-sand trick.
It will surprise no-one that I made some changes to the control setup on my HT oven as a result of my testing, and I'm happy with it now
