Modding type-K thermometer to use AA

Projects / Misc

I bought a TM902C thermometer that comes with a type K thermocouple that can go up to about 200C. You can use it a bit higher, but not for long term. I bought it for $3.52 and that was a great price as the probe itself costs half as much.  They sell this at Amazon for 4x.

The first thing I did was opening it up and peek. There is a black epoxy glob (IC die bonded to the PCB), a transistor as the "cold junction" and a 3V (or was it 3.3V?) LDO. I measured its power consumption, when it starts getting wrong readings due to low battery etc. They did put in a decent low current LDO as the whole thing uses a few hundred uA.

I don't particular like 9V batteries, so I have built the following circuit to run the thermometer from a single AA NiMH battery.

The circuit is pretty much straight from the apps section of the datasheet. I used both the inverted output and the doubler circuit in series to operate the thermometer. There is a Schottky D3 to prevent the Vout to rise above ground. This is not documented on the datasheet, when that happens, the charge pump cannot start. I connect the load between the doubler (Vcc to diode chain D1 & D2) and so at start up, there will be a positive bias there! Once the charge pump is running, Vout is driven to below ground.

One of the curious things you would notice when reading the datasheet is the following curve. It tells me 2 things - the chip would actually go all the way down to 1V when its recommended (guaranteed) operating voltage is 1.5V-10V. The other thing is that its transistors Vt is about 1V and that's why the internal resistance asymptote at 1V. For my application, I use very low current and only intend to operate down to 1.2V at room temperature, so I think I can get away with that. I specifically used a socket for the chip so that I can swap it out. I have tested 7-8 chips and they all worked. (ICL part, a few NS LMC7660 and some Maxim branded ones)

Vout = 1.2V x 3 - (2 x 0.45V) = 2.7V which happens to be just above the lowest voltage my particular thermometer would still give me a good result. The measured current from the 1.2V is about 1.32mA. I have tried a few version of a discrete designs with inductors and they don't work at this low of an average current.

I bypassed the power On/Off switch with a thin wire.   I used Kapton tape to isolate the pad from the switch and use the switch to control the power to the switch cap regulator.  The output of the circuit is wired to where the original pads for the 9V input.

This is what my quick & dirty manually soldered PCB looks like. I ran 2 wire to the on/off switch and 2 to the LDO. It fits very well inside the battery compartment designed for 9V. The battery holder is built from PCB and is reinforced by the plastic case on both ends and one side. There are 2 small piece of PCB holding the other side of the battery. With the cover is in place, there are no rattlings from the battery.