Slightly high voltage output on AD8497.

I am trying to build a DIY reflow oven and have converted a small convection oven.  Since all thermocouple conditioner/ampflier IC's on the market are surface mount I have been kind of stuck, since I need to solder the IC, but can't solder the IC until my oven is finished. However I have been able to do a few boards anyway by using my DMM with it's thermocouple input and manually turning off the oven elements to approximate the correct reflow profile.

One of the test boards I did was a DS600 which is a +/- 0.5ºc accurate temperature sensor.  I wanted to make sure that I could get repeatable results by manually controlling my oven before trying my $5.10 AD8497 IC.  I then designed and built a small board that contains my AD8497. I basically copied the basic connection schematic from the datasheet and used the values from the article "Two Ways to Measure Temperature Using Thermocouples Feature Simplicity, Accuracy, and Flexibility" for the input filter. I was able to solder the board succesfully using my SMD oven, so now I can work on implementing a PID controller.

The problem is that the measured temperature on my board is almost 4º too high.  If I ground the inputs to measure the ambient temperature, the output is nearly spot on when compared to the output of my DS600 temp sensor board.  I was originally going to use a AD8495 because the ambient temperature in my enviroment is usually always between 20º and 25º, but digikey only carried the AD8497.  But, even though my ambient temperature is about 22º I still got 22º when grounding the inputs.

I don't know if I can remove the 4º error in software, or if it's going to scale wrong even more as the temperature rises more.  I don't know where to go from here to further troubleshoot.

I could post pictures of my schematic and board if needed.  When I designed the board I used a different part in eagle with the same pinout so it's hard to follow on my schematic.  The finished board is pretty small, it's about 1" by 1.25". When I grounded the inputs so I can check ambient temperature I just used an aligator clip.

Thanks for your time.


  • Hi James,

    Thanks for your question.

    Yes, it would be helpful if you can share your schematics and a picture of the board (you can message me privately if you are not comfortable sharing on a public forum).

    Also, what thermocouple temperature are you measuring?

    Another thing to check is that what you are seeing is not the nonlinearity error from the thermocouple by using this tool .

    Look forward to your reply.



  • Hi Reemo,

    Thanks for taking the time to help me. I have attached 3 images. The schematic and layout of my board, and a photo of the finished board.  The photo did not turn out too great, I had to use my cell phone and that picture turned out best. The board is so small that I can't get a good close up picture.  I'm sorry for the horrible schematic, I am using a newer version of Eagle and I always seem to have trouble following the tutorials for the older versions of Eagle because of the differences.  I used a different IC with the same footprint, so you can ignore the values of the pins, I only went off of pin number. All the values of the resistors and capacitors are accurate though - they are what I soldered to my board and they are all 0805. I am pretty new to electronics so please forgive the sloppiness.  I am learning though.

    So far I have not tried any extreme temperatures, I have tried ambient, which my DS600 and a clock on the wall that included temperature said was 22º, but I measured about 26º.  I also placed the thermocouple in my laptops vent output where I measured about 48º. When I used the DS600, I measured about 44º. If I stick the actual board in my laptops vent and ground the inputs, then I got about 44º. The onboard temp sensor seems to be working fine, it's just my thermocouple that is putting out about 4º extra worth of voltage.

    I am using a PIC microcontroller with a built in 4.096v reference to read and calculate the voltage.

  • I am pretty embarassed, but I think I found my problem.  I drilled normal size holes for my thermocouple wires which were pretty tiny in comparison, so I had a lot of trouble getting one of the wires soldered to my board (I think it has something to do with what the wire is made of too?). I believe the connection was not as good as it should have been.  I pulled the wire through and soldered it directly to a trace and the temperatures seem much better now.

    I looked all over digikey for female thermocouple header and could not find one, that way I could plug the thermocouple directly into my board instead of cutting off the male connector and soldering the wires directly to my board.

    Thanks again for your time, I will let you know if this ends up not being the issue I am having.

  • Hi James,

    Glad you found the problem... it happens to the best of us, important thing is you figured it out.

    I've used mini connectors from Thermocouple Components which are pretty nice, specifically CMJ-xxPF on page 9.

    Good luck!


  • This question has been closed by the EZ team and is assumed answered.