Last year, I built a Compost Monitor to get real-time feedback on the temperature of my pile.  After an initial unsuccessful attempt to build a probe, I was able to get a probe design that lasted most of the season.  However, by the end of the year, when I pulled the probe from the pile, one of the sensors had gone bad, likely due to moisture.  Therefore, this year, I set out to build a more rugged probe.

The sensors are again MMBT3904 NPN transistors, with leads soldered to them, and encased in several layers of liquid electrical tape.  This time, instead of Cat 5 cable, I made my own twisted pairs from 24 gauge cross-connect wire.  I was hoping the cross-connect wire might have insulation with a bit more heat resistance (when applying the many layers of heat shrink tubing) but that turned out not to be the case.  Regardless, the probe seems to be working, so the few places where the insulation started to melt do not seem to be causing any issues (yet).

After assembling the sensor transistors, and many times during the assembly of the probe, I checked the integrity of each sensor.  First, by measuring the VBE voltage using the diode test function of a multimeter.  Then, by checking the insulation integrity using the 20Mohm range and making sure there were no conductive paths.

After assembling the sensors, the probe was prepared.  The probe was again made from a fiberglass driveway marker rod.  First, a layer of non-adhesive heat shrink tubing about 1.25″ long was applied to the rod in the locations where the sensors would be placed.  This was intended to provide a cushion for the sensor against the fiberglass rod and to also provide a (hopefully) better surface for the adhesive heat shrink applied later to attach.

Next, each sensor was placed on the previously applied heat shrink.  A shorted section (~1″) of adhesive lined, marine heat shrink was then applied over the sensor.

After attaching the four sensors, more liquid electrical tape was painted over the region where the wires exit from under the heat shrink.  This provided additional sealing and helped insulate the wires where the insulation melted during the heat shrinking process.

Next, a full length of adhesive lined, thick-wall 7/8″ heat shrink (HS3-HW-0875) was applied over the entire rod, leaving room at the top for the wires to emerge from under the heat shrink, wrap around the rod, then run downward.  By running the wires downward, an adhesive lined heat shrink cap (HSEC-100) could be applied to the top to seal everything and provide a downward facing exit for the wires (to resist any moisture penetration).

Another heat shrink cap was applied to the bottom of the probe.

The adhesive lined, thick-wall heat shrink and the caps at the top and bottom were the major improvements in the probe design this year.  My hope is that by providing a better barrier and better sealant at the ends and all along the probe, moisture penetration will not happen.  Below is a full length picture of the completed probe.

The probe installed in the manure pile:

The MRBW-RTS connected to the probe, monitoring and reporting the conditions:

Results

The probe was installed shortly after turning the pile.  Initially, the temperature remained cold and very flat.

After saturating the pile with water, the temperature jumped, presumably because the water temperature was warmer than the pile.  Then, slowly over the course of several days, the temperature started to rise as the bacterial action took off.

So far, so good!