I'm in Screwy's camp - I have a thermometer in the mash.. but don't really care what it says. I do all my figuring from the wort return temperature.
I reckon Screwy gets a differential between return temp and mash temp because of his slow rate of circulation... I have tried a bunch of different flow rates, and have noticed the variation between wort flow temp and grain bed temp is highest when the flow rate is low - also that the differential between top and bottom of the mash is the same ... higher when flow rate is low.
Makes no difference mind you, you measure from whatever point you choose, stick with it and tweak your tempertures to reflect your results. The only time it matters even vaguely is if you are trying to tell someone else your recipe.
Here's my mash graph for my first brew in my new BB mash tun - These days I normally only graph the wort temp, but because now I have a dial thermo, its a lot easier to compare the return temp to the mash temp, and this was a first run, so I took notice.
Wort return temp in Blue - Dial thermo in Red( probe a bit higher up than halfway point in the grainbed, lower than halfway if you include the liquid level). 3:1 L:G and no stirring after mash in.
btw - dial thermo and wort return were calibrated to match when the system was recirculating water only at full throttle and allowed time to steady out, so temp differences are real and not a factor of any calibration differences or suchlike.
A you can see, the grain bed temp lags behind the wort return temp, but eventually catches up - I use a PID controlled RIMS not a HERMS though and my return wort temp comes up
very quickly - 13 in 5 minutes on the graph, but it was pretty much at temp even faster than that, its just that I didn't take my first reading till 5mins. So its natural that the bed would take a chunk longer than that. I suspect a slightly closer match between the two lines when I can bring myself to whack the lid on and not watch whats going on obsessively... give me a couple more brews.
Still, as I said, I don't care about the bed temperature - and once I have gotten a better feel for this new configuration, I will stop monitoring it completely. Probably pull the dial thermo out of the tun altogether and find something else to do with it.
TB