FDC2214: Performance over long cable lengths

Lokeshwar Elitam

Part Number: FDC2214

Tool/software:

Hello! TI Team,

We want to use the FDC2214 for a liquid level sensing application and have some questions regarding its performance over long cable lengths.

Application Setup:

Sensor type: Capacitive sensor (Level Sense)
Cable used: 20-meter RG-188 coaxial cable
Measured capacitance:
- Total capacitance (sensor + cable): ~2.3nF
- Cable-only capacitance: ~1.9nF
- Capacitance variation from sensor: ~0.4nF (i.e., the actual capacitance)

Questions:

Can the FDC2214 reliably detect a ~0.4nF variation when the fixed cable capacitance is ~1.9nF?
Are there design guidelines or precautions we should consider in this setup (e.g., shielding, guarding, excitation current settings, sensor frequency)?
Will such a high baseline capacitance (due to cable) affect resolution, stability, or linearity, and how can we compensate or calibrate for it?
Please refer to the attached schematic and advise on improvements if needed.

We would greatly appreciate any guidance or reference material specific to long-cable capacitive sensing with the FDC2214.

Best regards,
Lokeshwar

3 months ago

0 John Miller - Sensing 3 months ago

TI__Mastermind 40295 points

Lokeshwar,

1. If we assume the cable capacitance is in parallel with the sensor capacitance, then detecting a 0.4nF change should be okay.
If we use a 0.45uH inductor and assume Csensor can go ±0.4nF about a nominal value of 0.4nF, then the the sensor frequencies will be:
Csensor = 0nF → Fsensor = 5.44MHz
Csensor = 0.4nF → Fsensor = 4.95MHz
Csensor = 0.8nF → Fsensor = 4.57MHz
The FDC2214 should be able to detect these changes easily.

2. There design guidelines to consider. Shielding can be a challenge because grounded shields can load your sensor and reduce sensitivity.
You can see some guidance in the app notes:
Power Reduction Techniques for the FDC2214/2212/2114/2112 in Capacitive Sensing Applications
Capacitive Proximity Sensing Using FDC2x1y
Setting LDC1312/4, LDC1612/4, and LDC1101 Sensor Drive Configuration - although this covers inductive sensors, the principles also apply to the FDC2x1y capacitive sensors. The cores of the FDC2x1y devices are very similar to the cores of the LDC1x1y family of inductive sensors, with the same settings, and so on. The only real difference is that FDC parts use a capacitive sensor and fixed inductor, while the LDC parts use an inductive sensor and a fixed capacitor.

3. The variations in the cable can affect stability. It could affect resolution and linearity, but we don't have a way to quantify those effects. As long as the cable resistance is low, it should be fine. If maximum accuracy is desired, then be sure to use a stable, accurate clock with low jitter.

4. Your schematic looks fine.

Regards,
John

0 Lokeshwar Elitam 3 months ago in reply to John Miller - Sensing

Prodigy 10 points

Dear Team,

Thanks for the reply

The only concern is will it work for the long cable of 20M?

Regards,

Lokeshwar

0 John Miller - Sensing 3 months ago in reply to Lokeshwar Elitam

TI__Mastermind 40295 points

Lokeshwar,

I wat to say "yes" as long as cable losses are low and the cable's L/C/R characteristics are stable.
But to be honest, I have never encountered your configuration before, so I'm not 100% sure.

The best path forward is to build one an try it out.

Regards,
John

0 Lokeshwar Elitam 3 months ago in reply to John Miller - Sensing

Prodigy 10 points

Dear TI,

I also want to know is there any detection and protection circuity in FDC2214 when the sensor capacitance is short circuited? and any other failure detection means.

Regards,

Lokeshwar

0 Mekre Mesganaw 3 months ago in reply to Lokeshwar Elitam

TI__Expert 7485 points

Hi Lokeshwar,

John is out today. He should be able to respond when he is back tomorrow.

Regards,

Mekre

0 John Miller - Sensing 2 months ago in reply to Mekre Mesganaw

TI__Mastermind 40295 points

Lokeshwar,

Please take a look at bits 0 - 15 in the STATUS register (address 0x18) for descriptions of the error reporting (in the FDC2214 data sheet).
The related ERROR_CONFIG register (0x19) allows the user to configure the desired error reporting of the STATUS register.

If any of a channel's input pins are shorted, it should generate a ERR_ALW error (amplitude low warning, b9) in the STATUS register.

Regards,
John

0 Lokeshwar Elitam 2 months ago in reply to John Miller - Sensing

Prodigy 10 points

John Miller,

Ok, I can understand that by means of this status registers we can detect the error but what about the protection for the cap sensor, how that can be accomplished?

Regards,

Lokeshwar

0 John Miller - Sensing 2 months ago in reply to Lokeshwar Elitam

TI__Mastermind 40295 points

Lokeshwar,

What is meant by "protection"?

There are ESD protection diodes on the device pins. Is that what you mean?

Regards,
John

0 Lokeshwar Elitam 2 months ago in reply to John Miller - Sensing

Prodigy 10 points

John,

Suppose the capacitance sensor is shorted and we can detect the same through the error status registers until this its ok. When this actual sensor got short circuited is there any internal circuitry which oppose the charge pump to the capacitance sensor when this sensor is short circuited?

regards,

Lokeshwar.

0 John Miller - Sensing 2 months ago in reply to Lokeshwar Elitam

TI__Mastermind 40295 points

Lokeshwar,

The error will be reported during amplitude calibration of the sensor waveform and the resulting action should be for the sensor driver to time out & shut down.

Regards,
John

0 Lokeshwar Elitam 2 months ago in reply to John Miller - Sensing

Prodigy 10 points

John,

Thanks for the reply,

Now we are set for the testing of FDC2214 here we need your help.

I might require a reference code for testing of FDC2214 in interface with STM32H745 Microcontroller.

and also want to know how to convert the data read from DATAx Registers into capacitance

Regards,

Lokeshwar

0 John Miller - Sensing 2 months ago in reply to Lokeshwar Elitam

TI__Mastermind 40295 points

Lokeshwar,

The first step would be to use equations (2) or (9) in the data sheet to estimate the sensor frequency:

(2)

You could then use equation (7) to estimate the sensor capacitance:

Regards,
John

0 Lokeshwar Elitam 2 months ago in reply to John Miller - Sensing

Prodigy 10 points

John,

We performed some testing with this FDC2214, and we have recorded the data (Please find the attached) we can observe that the graph is nonlinear.

So, please guide us in achieving a linear graph.

Please find the attached documents containing layout, Schematics and Recorded data.

Regards

LokeshwarFirst Reading 27uH with cable.xlsx

0 John Miller - Sensing 2 months ago in reply to Lokeshwar Elitam

TI__Mastermind 40295 points

Lokeshwar,

Thank you for the data.

What do the data and axes represent for the curves?
Does the Capacitance column represent a known capacitance?
Does Cap Min and Cap Max represent capacitance estimated from raw data via the equations mentioned earlier in my most recent reply?

Regards,
John