Thermoworks 1003X AP

Hi there, was testing out my newly built HM 4.2.4 with a Thermoworks 1003X-AP grate probe. It seems to be off about 4 or 5 degrees. I verified this by submerging the Thermoworks probe in a pot of boiling water. I got a reading that fluctuates between 207 & 208 degrees. I then pulled out my Redi-check digital meat thermometer and submerged that to compare. The Redi-check reported 212 degrees.

My house is about 30' above sea level. Is the 4-5 degree inaccuracie an acceptable offset? Or should I be looking for a potential problem with my build?

Thanks for reading!

I would re-flow the solder on your probe jack legs, and the series of components that are standing up behind the probe jacks (the 10k, 100K resistors and the .1uf cap), as well as the last four pins on the ATMega (starting in the corner down behind where the TC jack is or would go). These are the only components between your probes and the ATMega. After you have resoldered these components I would wash the board with some isopropyl alcohol and a soft brush to remove excess flux and let the board dry out thoroughly before you power it up. Bad solder joints or excess amounts of flux on the board in this area are the main cause of bad probe readings...

Hey Ralph, sorry for the late reply. I followed your advice and am getting the same results. I ran some new tests, putting my redi-check meat probe in the pit probe location and dunking it in the boiling water. While not as quick as the Thermoworks, it stops right on 212'. From there I tested all my probe sockets with the redi-check probe and they all registered 212. I'm now wondering if I got a beat probe from Thermoworks or if the 1003X-AP is slightly different than the 1001X? Or could it be that the Air Probe should not be submerged at all, and again I'm chasing a ghost? Whats four degrees betwixt friends? heh

Thanks for your time!

Bill Herbert said:

Hey Ralph, sorry for the late reply. I followed your advice and am getting the same results. I ran some new tests, putting my redi-check meat probe in the pit probe location and dunking it in the boiling water. While not as quick as the Thermoworks, it stops right on 212'. From there I tested all my probe sockets with the redi-check probe and they all registered 212. I'm now wondering if I got a beat probe from Thermoworks or if the 1003X-AP is slightly different than the 1001X? Or could it be that the Air Probe should not be submerged at all, and again I'm chasing a ghost? Whats four degrees betwixt friends? heh

Thanks for your time!

Click to expand...

I would turn off the heat to the boiling water & see if it tracks at -4* all the way down to room temp vs the redi-check. If so, I'd just adjust the offset for the probe & be done with it. If not, you need different coefficients.

Bill,

I had a couple of TX-1001X-OPs and a TX-1003X-AP delivered yesterday and found all three to be off by the same amount that you did. My house is about 35 feet above seal level. All three probes read around 208 in boiling water. I swapped the TX-1003X-AP for my Maverick smoker probe and the Maverick read right around 212. I added a +4 degree offset.

It's nice of you to confirm this for the OP... Are you saying both the TX-1001X-OP and the TX-1003X-AP are reading about 4 degrees low, or just the 1003? It's been a while since I did a boil test myself, but if the probes are consistently reading 4F off at boiling perhaps we need a better tuned set of coefficients for them. If I get time today I will try a boil test to see what I get with my thermoworks probes.

Ralph,

Both of my TX-1001X-OPs and my TX-1003X-AP read right around 208 in boiling water. I'm happy to help out any way that I can.

I just did a boil test myself and see the same results, a brand new TX-1001X-OP was showing 208/9 and a TX-1002X-NP was showing 207/8 in boiling water. I'm reluctant to use an offset for them because I've been using the 1002X-NP for my food probe and the meat always comes out exactly how I want it done for rare/medium cooks, so perhaps it is more accurate in those parts of the curve. Shoulders I cook to closer to the boiling temp but it's kinda hard to tell a couple degrees in the well done range.... but it seems like the coefficients on the Thermoworks Pro preset could be improved a bit...

Wow, thanks for that Mike. That's one issue I can cross off my HM list. Now back to pullin my hair out on my blower/damper circuit. @Ralph - who has been exceedingly helpful in another thread - I just ordered a Fluke 115 DMM to replace my 19 dollar Amazon special. Hopefully it will help me get more accurate readings that I can report back to you.

Always glad to help Bill... Just revive that thread when you get around to working on the blower circuit again.

RalphTrimble said:

I just did a boil test myself and see the same results, a brand new TX-1001X-OP was showing 208/9 and a TX-1002X-NP was showing 207/8 in boiling water. I'm reluctant to use an offset for them because I've been using the 1002X-NP for my food probe and the meat always comes out exactly how I want it done for rare/medium cooks, so perhaps it is more accurate in those parts of the curve. Shoulders I cook to closer to the boiling temp but it's kinda hard to tell a couple degrees in the well done range.... but it seems like the coefficients on the Thermoworks Pro preset could be improved a bit...

Click to expand...

Yeah I tried to use the thermocouple and do a 1000+ point correlation between room temp and 410F, however it could be off because maybe the thermocouple reading was off because this was over a year ago and the noise rejection code is much better now. I bought a Thermoworks Therma K Professional Handheld Thermocouple reader though because I thought it would be a good idea to have a professionally calibrated reference unit that I can use to gauge the HeaterMeter against. I haven't had time to run full side by side tests, but I currently have a bare probe in my convection oven set to 275F. HeaterMeter is reading 277.6F and the thermoworks is reading 278.1F but bouncing up and down a couple of tenths of a degree either way so I would say they're pretty close now. There is still the potential for error because the code expects that the input voltage is 3.300V and does all of its thermocouple calculations off that. The 3.3V regulator however is not a precision source so I've toyed with the idea of making it and the bandgap reference configurable, however I am not sure how much benefit it will provide. I can certainly tell now that I have a way to know exactly what the temperature actually is, rather than relying on comparing HeaterMeter to another low-cost unit.

To further derail this thread, I took apart the thermoworks unit to see what their circuit looked like. It is really well mechanically designed although I did not appreciate that the battery leads are soldered through the plastic case to hold the circuit board down, which meant I had to desolder the plates to be able to see the other side of the board. Their design is a texas instruments MSP430 microcontroller, which has a built in LCD driver and 4-channel 16 bit differential ADC. They have the same common mode and differential noise reduction circuit HeaterMeter has, however the TC+ is tied to ground through a 10Mohm resistor and TC- is pulled up to Vref with the same value. They do not have a low pass filter. Vref is generated from the microcontroller and changes constantly (although I did not hook up the scope to see what values). Cold junction compensation is handled from a resistor divider with a thermistor mounted between the TC+ and TC- solder points.