FDC2114: No interrupt (INTB) on FDC2214 circuit attached

Hi Kristin,

with your prompt advice, we rectified the problem and now FDC2214 is working ok

Thanks for this and if you permit me then I would like to know few more things 

1) As you can see in the PCB, we are using 11 Nos of FDC2214 to sense 44 touch keys whose filtered data is being sent through CAN Network bus. This whole circuit makes PCB very complex and its routing compels us to go for 4 layer PCB which adds another level cost. Therefore kindly suggest how can we make it in 2 layer?

2) Advice method by which we will be able to reduce 11 Nos FDC2214 by employing matrix key pad scheme?

3) As we are unable to use FDC 2214 i2c device address which is mentioned in datasheet with STM32 micro, which leads us to use 0x55 therefore we are unable to use two FDC2214 on the same I2C bus with hardware addressing mapping (Low/High) and thus required higher pin count of micro ,as you can see we are using 144 Pin LQFP device with 11 separate I2C lines and add another degree of cost which could be saved if we have proper I2C device address which can work with all types of micros.

4) As per the Datasheet of FDC2214 INTB is coming only for the change in the system state which in such case involves more processing time from micro as micro has to analyze the touch or no touch which each incoming data. That is to scan 11 Nos of FDC2214 round the clock which might delay the touch response to the user. Therefore kindly suggest if we can set certain thresholds for touch which when crossed will trigger the INTB to save computational time.

This device is very beautiful and soon we are planning to use it for monitoring fuel Tank level in Cranes and Tractor's. 

Once again thank you for your support.

Thanks and Best Regards,

VK Verma

 

Hi VK,

1. I think it's likely you will need a four layer board to route all of the signals for so many FDC devices.

2. If you need 44 separate buttons, I don't think it will be possible to use fewer than 11 FDC2214s. Each sensor needs an individual channel of an FDC device.

3. I'm not sure why you are not able to use the I2C address in the datasheet. Regardless, you should be able to use an I2C expander to reduce the required number of I2C pins on your micro. You may have to compare the cost of purchasing additional chips and finding space for them in your layout with the cost of the 144 pin micro.

4. Unfortunately, there's no way to configure INTB such that it will assert only when a certain threshold is crossed. If your priority is system responsiveness instead of power consumption, I would suggest increasing your sample rate. If you are using the MCU with 11 independent I2C pins, you should be able to read data from all of them at once. You would only need to cycle through each channel of the devices (so 11 parallel reads 4 times in a row to view all the new data).

Regards,

Hi Kristin,

We are considering your above inputs in the next version of PCB and till date the status of the all 11 FDC2214 is that, all are responding Ok in terms of i2c communication and their respective Interrupt and now we are at the stage of calibration and we need to resolve the following issues,

1. As all FDC2214 are working together, we are observing heavy cross talk between FDC2214 devices in touch sensitivity and due to this a lot of false triggers are being generated. So please suggest how should we start calibration so that touch response and performance can be optimized.

2. How to remove touch cross talk?

3. Please provide some rules to fine tune touch capacitors and components for hardware calibration

4. Any calibration steps flowcharts to follow.

Appreciate your consistent support !

Thanks and Best Regards,

VK Verma

        

Hi VK,

I'm glad to hear that the I2C communication is working. I've responded to your other questions below:

1/2. There are multiple issues that could be going on here. One could be that the FDC2214 has higher sensitivity and a longer sensing range than you need, so the devices could be responding to button presses on nearby buttons. You could improve this by decreasing the RCOUNT value for your channels, which will decrease the button sensitivity. If your board is not connected to earth ground, you may also be experiencing a phenomenon we refer to as ground shifting. This can make other parts of the board sensitive to capacitive shifts. You can read more about that in this app note. You may also be experiencing cross-talk if your INAx and INBx signals are routed too close together or near noisy signals. 

3/4. For hardware calibration, are you interested in information about how to optimize the performance of your test board, or how to test the board in production?

Regards,