Hardware_Parts Selection_Temp Probes
I chose Maverick brand (replacement) temperature probes for my project, one blunt-ended probe for the pit temperature, PR-001 / ET-733 (image on left) and one skewer for the internal meat temperature, PR-002 / ET-732 (image on right).
My choice of the ETS-732/733 units was made based seeing these used on a number of online projects and the 4 1/2 star customer ratings on Amazon. These probes are based on NPN thermistors (ie. resistance lowers as probe heats up). The initial drawback with these probes were that 1) they are reasonably costly; and 2) the probes reportedly have a high failure rate as apparently they are moisture sensitive. Some people have achieved better run lives by sealing the top of the thermistor tube with food-grade high temperature silicon (which I have not been able to easily source in Australia) as well as ensuring that probes are not submerged in water when cleaning.
Only after I had purchased these probes did I come across the excellent open source Heatmeter project documentation. This group tested a number of probes and only rated the Maverick ET-732/733 probes as Acceptable rather than Highly Recommended due to 1) lower resolution and 2) mysterious measurement errors that they sought to mitigate by grounding the shielded cable.
Regarding resolution, the Heatmeter crowd found that the low-end resolution was poor with a 10K voltage divider (bias) resistor and recommended using a 22K or even a 47K 1% resistor - see their graph below. For my project, I optimised the bias resistor selection (voltage divider output) for each probe (both pit & meat) for the desired temperature measurement range.
Interestingly, I learned from the datasheet provided by Maverick following an email request, that the ET-732/733 thermistors have a very high R25 value (zero power resistance at room temperature) of 1M ohms ((1,000k ohms). This is in contrast with most probes which commonly have lower R25 values of 10k ohms. I am not sure why Maverick chose this; perhaps to minimise thermistor self-heating errors as these probes have moderate dissipation factors of ~0.6 mW/deg C but the high impedance means very little current flow. Based on this high R25 value, it is not surprising that the Heatmeter crowd found poor resolution when using only a 10K ohm bias resistor!
Heatmeter characterisation of Maverick ET-732/733 temperature probes
See Section 5.3, Design Considerations for further discussion on the thermistor resolution, datasheet and voltage divider / bias resistor optimisation process.
