LIQUID LEVEL SENSOR using FDC1004

Hi Suresh:

How exactly are you determining the capacitance? Are you using the EVM gui? What is your application? What does your setup look like?

You are correct about the CAPDAC. It is only possible to use this in single-ended measurements. If you have the CAPDAC set at 10pF, the max value you can have for capacitance is 26pF. If the total capacitance is above 26pF with 10pF of CAPDAC, then it will saturate and you will only see it as 16pF when you grab the value from the register (it's more than the input range of the FDC1004).

For your second question:

• Can you tell me your testing conditions when this event is happening?
• Are you always not getting the ready state? Or only some of the time?
• What is the input capacitance?
• Any cap on the shield?
• Possible estimation of the shield to input capacitance?
• Supply voltage?
• Do you see an increased current consumption during this event?

Thanks

-David Wang

Capacitive Sensing Applications

Hi Mr.David,

Thanks for your quick reply.

Actually our application is fuel level measurement, based on capacitance we measure the fuel level. We are using capacitance probe(cylinderical tube) for measuring fuel level. Based on dielectric, capacitance will vary, based on capacitance, we will measure the fuel level. By default capacitance starts at more than 50pf. So i can set a offset as 37.5pF or 40.625 or below 50pF. This i mentioned in previous mail capacitance with CAPDAC values. What my doubt is with in the range i can vary offset capacitance but input capacitance should not vary, here input capacitance also varying. We are not using EVM gui, we are using our own hardware controller with FDC1004.

Always i am not facing this issue, some times only. Input capacitance is more than 50pF and no shield capacitance. We are trying to reach maximum capacitance range. Supply voltage is 3.3v. We are not check the current consumption yet, let me know if you really wants. This scenario happen in rare case.

Thanks & Regards,

Suresh Kumar.S

For your fuel level measurement, is the tank plastic or metal (and grounded)? Are you using the Out-of-Phase method? If you are not, I am wondering if you are experiencing variations due to being in close proximity to the tank or any stray parasitic capacitance. I assume you are not using it because it is not possible to use the CAPDAC when using OoP.

If am I understanding you correctly, if you vary your offset capacitance to make your 50pF initial input capacitance within the range of the FDC1004 (+/-15pF), you are applying an offset of either 37.5 to 40.625pF. This will give you an input capacitance value of 12.5pF or 9.375pF. What kind of variations are you seeing from 12.5pF or 9.375.

I did a test with a square sensor single side PCB and our EVM. I am getting ~5.2141pF in single-ended mode, when I enabled the CAPDAC with 0pF, I would assume that I get the same. Correct? No, I get ~5.3014pF. It is due to internal parasitics. Also in capacitive sensing application especially liquid level sensing, we are looking for relative changes in capacitance not absolute. So it should not really matter what the input capacitance is as long as it is within the input range and has enough margin for dynamic range of the system.

We have not seen or heard of any scenarios where that has happened. So I would definitely need more information about the setup and testing conditions because there is no way for use to determine what is happening and running sims to verify that is really the issue.

Thanks

-David Wang

Capacitive Sensing Applications

Hi Mr.David,

Thanks for your quick reply.

We are using Glass Tank for testing purpose, no ground. Once again i tested with nearly 90pF input capacitance. We are using CIN 1(0x08) for measuring the capacitance.  

1. 0X13, 0X00 - CAPDAC   75              Input_capacitance with offset:  90.150 pF         Without_offset_Capacitance:  15.146pF

2. 0X13, 0X20 - CAPDAC  78.125       Input_capacitance with offset:  90.118 pF         Without_offset_Capacitance:  11.986 pF

3. 0X13, 0X40 - CAPDAC   81.25        Input_capacitance with offset:  90.250 pF         Without_offset_Capacitance:  8.990 pF

4. 0X13, 0X60 - CAPDAC   84.375      Input_capacitance with offset:  84.578 pF         Without_offset_Capacitance:  0.198 pF, I think here i should get nearly 5.6pF

5. 0X13, 0X80 - CAPDAC   87.5           Input_capacitance with offset:  116.582 pF       Without_offset_Capacitance:  29.08(2.5) pF, I think here i should get nearly 2.5pF

Here first three CAPDAC's are fine, what you told same thing happen here also, with in 100fF only variation will be there. But 4(84.375) & 5(87.5) only i am facing issue. I am getting un expected result. I am not able to find where is the issue, whether i am doing any mistake in registers?  

Please see the above register setting, if i did wrongly let me know how to resolve.

Thanks & Regards,

Suresh Kumar.S

so you just have a single electrode connected to CINx? When you say you are testing it with nearly 90pF of input capacitance, how are you getting this measurements? The FDC can only output +/-16pF. I suppose you are using the CAPDAC to offset that 90pF within the FDC range, and then adding back the CAPDAC to get the "estimated input capacitance"?

The information you supplied is the correct result when you have the following: Input Capacitance - CAPDAC = Effective Input Capacitance, where input capacitance is "input_capacitance with offset", CAPDAC is  "CAPDAC", and effective input capacitance is "without_offset_capacitance. The CAPDAC is always subtracted from the input capacitance. From your 4th and 5th example, they are correct.

1. 84.578 - 84.375 = 0.203pF
2. 116.582 - 87.5 = 29.082pF

Thanks

-David Wang

Capacitive Sensing Applications

Hi Mr.David,

Thanks for your quick reply.

What you are understood is correct. We are using single ended measurement with CIN1.

Total_Capacitance(I/P cap) = (FDC_count/2^19) + C_offset

Effective_I/P_Capacitance  =  Total_Capacitance - C_offset

In my previous mail i mentioned,  i am facing issue with CAPDAC 84.375 & 87.5pF. Theoritically it's correct only but my doubt is,

Input capacitance(Total_Capacitance) should come nearly 90pF for CAPDAC 84.375 & 87.5pF. If we vary  CAPDAC within the range also input capacitance should not vary. Here input capacitance also varying while varying CAPDAC. when CAPDAC is 87.5, it will go maximum 103.5pF(87.5 to 103.5) only , but i am getting 116.58pF. Is there any chance because of hardware or setup can make these kind of issues.

Thanks & Regards,

Suresh Kumar.S

ok sorry, I understand now what your concerns are. I do not believe it's because of hardware or setup. one thing you can do to verify this is to add a ground electrode right next to your sensor electrode. This will give your capacitance signal a close reference. Since you only have one sensor connected to CINx, it's terminating at it's closest ground potential source. Typically we use this setup because it's much more reliable.

Please refer to the app note below for liquid level sensing.

www.ti.com/.../snoa925.pdf

Thanks
-David Wang
Capacitive Sensing Applications

Hi Mr.David,

Thanks for your quick reply.

Sorry for the improper information, Actually we are using sensing electrode(FDC1004 CIN1) and it's covered by ground electrode(FDC1004 GND). I attach the image of cylindrical tube capacitive sensor which we are using.

Thanks & Regards,

Suresh Kumar.S

Hi Mr.David,

Thanks for your quick reply.

Thank you so much for your great information. I am looking only the capacitance value not raw code, so i am not able to find the issue. Now i got where may be the issue. I look into raw code and let me conform the issue.

Thanks & Regards,

Suresh Kumar.S

Hi Mr.David,

              Actually i did lots fuel testing with our FDC1004 testing setup. Their i got good fuel accuracy as well as very less capacitance variation(nearly 15 to 20fF). So we started as like product and we did mechanical as well as PCB. But now i am facing some issue with actual setup. Actually in old setup with short wire we gave input to FDC1004 from capacitive sensor probe other than input and GND FDC1004 board and sensor probe is separate. But in our actual setup board is in top of a sensor probe, GND and INPUT is directly connected. In new setup i am facing capacitance variation issue, the capacitance increasing or decreasing gradually(nearly 400 to 500fF). It gradually increasing after that gradually decreasing or vice versa. Gradually in the sense some of the ours. In new setup one mechanical Flange is comming in the GND electrode, Is that may creating any issue. Please tell me what may creating the issue and how to resolve this issue.

1. Because of temperature or fringe effect or environment or for common environment and temperature how much capacitance will vary from actual one.

2. If you have any idea for how to design cylindrical capacitive sensor probe please share with us. I attach our new setup.

Thanks & Regards,

S.Suresh Kumar

Is it possible to give me a picture of the actual setup and the board/probe? The flange should not be the issue. For your first question, how much the capacitance will vary, I do not know, it's all system dependent. As for temperature, you would need to characterize that your self to correct for temperature shifts. We do not have any data on that since it is also system dependent.

We have not designed a cylindrical sensor probe, but from what I see, it should work. The issue that I would see if that it would cause the device to saturate (depending on how long it is).

Thanks
-David Wang
Capacitive Sensing Application

Hello,
Can anyone tell me where I have to set offset to -10pF , when sensors are connected. My ouput is saturated to 15pF

. I was making some mistake in CAPDAC Values. It works. Thank you!!!!!!!!!

You are most welcome!

Glad that everything works now!