MSP430FR2672: CapTIvate design using REFCAP Alogrithm - Max Count Errors constantly

Just for context, this is the related post where the ref cap implementation was shared with us. 

e2e.ti.com/.../msp430fr2672-linearization-of-touch-slider

Can you log the result and pull it to this thread? After I check the raw data, then I can give you the answer.

elementDataLog_2023_10_27_10_10_22.csvsensorDataLog_2023_10_27_10_10_22.csv

Here are the logs and a screen capture of the Channel Data and Table.

I have a couple questions about the modifications for RefCap implementation.

When I modify the CAPT_Manager.c file, do I need to modify both in the Captivate project and in the CCS project?

See from your result, it doesn't have "Max Count Error".

Jon Martis said:

When I modify the CAPT_Manager.c file, do I need to modify both in the Captivate project and in the CCS project?

You only need to modify the project in CCS.

Well, to expand the image further, this is what I see.  The max Count Error is always red, It does not clear and I never see the Channnel data change when I put my finger on the elements.

If I change from "REFCAP_CONFIGURATION" to "DEFAULT_CONFIGURATION", the CapTIvate tools respons normally and I see the changes in the channel data.

How can I move forward?

Will you meet the sample problem when download the code from Dennis?

I cannot load the code from Dennis.  My project uses the MSP430FR2672 and the project from Dennis uses the MSP430FR2633.  The zip file also does not include the CapTIvate Design center files, so I do not know how Dennis's project is configured.

Can you provide the info and guide to importing the project to my HW configuration?

I try the code, it works fine on FR2633. The suggestion would be reduce the counter setting on your side from 8000 to like 200 to see if this can help.

Eason,  I assume by "reduce the counter setting on your side from 8000 to like 200" you are referring to the "conversion count". 

I have the conversion count set at 650 not 8000, but according to the instructions :

When I do change the Conversion Count to 200, I still get the same result, Max Count Error and channel data is still at 8000.

Can we arrange a time to do an zoom interactive debug session?  It is time consuming to try 1 debug action per day.

Sorry, I think what you need to is to:

1. First change the conversion count at here.

2. Generate a new code and use the userconfig file to repalce the one in your project (Downloaded into MCU).

3. Reporgram the MCU.

If it not work, I would suggest you send me your project, I can check it for you. If you want offline support, please connect your TI sales.

I have tried this and get the same result.  Attached are my zipped CapTIvate and CCS projects.Carnegie_OpenAir_DPP1.zipCarnegie_OpenAir_DPP1_captivate.zip

From my try, after make this change, it can work on FR2672.

I have tried these changes, but still does not work.  the results are different, but not working.

Where can I make these changes in the CapTIvate GUI?  I know the 

I have tried these changes, but still does not work.  the results are different, but not working.

Where can I make these changes in the CapTIvate GUI?  I know the 

is handled in the MCU Noise Immunity tab

But I do not find any other selections that affect 

I will be out of office until tuesday, but will reply then.

1425.Carnegie_OpenAir_DPP1.zip

Please try this code.

Eason,

I tried this code, but get the same result. 

What did you change? I don't see any change to the code.

Jon

The only difference I see is that the Base Librariesand MSP430FR2xx_4xx Libraries are different. 

Trying Eason suggestions and code, I still get the same results.

I do not understand why I get channel data at 8000, when I use the REFCAP configuration.  When I use DEFAULT configuration, the channel data is 250 or 650, as it is set in the CapTIvate GUI.

When I change the configuration from CAPREF to back to “DEFAULT” with the Conversion count set to 250, I get the following count, delta and LTA values for the 4 Elements

When I change the configuration to CAPREF, (no other changes)

The LTA becomes 8191, MAX Count Error is set and cannot be cleared, so I am unable to get delta or count data.

I do not know if I am doing something wrong, but it would be nice to verbally talk with a TI Apps Engineer and debug this.

I can also send a hardware system to a TI Engineer to verify my results.

Please let me know how we can move forward.

Hi,

It is really a bad news to hear that. The second difference may lies on the sensor.

1. I think if you need offline support, you can connect your TI sales first. They will find a TI apps Engineer for you.

2. Another method is you to do the silder possition calculation based on the raw count of every sensor. Then you can do the adjustement by yourself to make the final output to be linear.

Eason

Hi Jon,

Maybe at this point I would suggest to take a look at the alternate slider example in the Captivate Technology guide, software chapter.

You can copy/paste the code and callback function so you don't need to write much, if any code.  This won't fix your differences in sensor sensitivity on each element, but you can try normalizing the delta measurements to make them closer to each other.

Also, looking at the data you collected at 250 using the default method, the deltas are a bit small compared to what I would normally like to see for good SNR.  I would try using 300 or 325 for conversion count.  You want to measure a change in capacitance at a minimum of 5%, but you really want to 10% or more, if possible, but you can't go crazy with the conversion count.  Using 500 or more introduces sensitivity to things like noise and temperature drift.

To calculate the % change you do the following:  %change = (100 x (1/cnt - 1/lta))% OR simply 10,000 x (1/cnt - 1/lta)

For example, using this formula E00 would be % -> 10,000 x (1/228 - 1/248) = 3.53%

E01 = 10,000 x (1/221 - 1/248) = 4.9%

E02 = 10,000 x (1/219 - 1/246) = 5.0%

E03 = 10,000 x (1/218 - 1/248) = 5.54%