LDC1312: LDC1312

Shrikant Bhagat

Part Number: LDC1312

Sir/Mdam,

We are using LDC1312 in one of automotive sensor.

The brief application is as below

The LDC 1312 measures the linear distance -2mm to +2mm of target metal movement. The LDC1312 is interfaced to micro controller. Micro controller converts the measured distance into PWM which in turn converted to analog output using RC and Opamp.

This system is working OK at ambient of 25 deg C.But output is drifted by almost 1.2V at temperature approximately 75 to 80 deg C.

The coil Gerber is attached for your reference.

The coil PCB is

PCB MATERIAL	FR4
PCB THICKNESS	0.8 mm
NO. OF. LAYER	2
PCB SIZE	21.4mmX 14.5 mm

HAL	Tin-Lead
COPPER LAYER IN MICRONS	35 micron (FINAL)
MASKING	GREEN MASK
LEGEND PRINTING	WHITE
NO.OF PCB REQUIRED	100
Gerber & NC Drill file name	GERBER for BIN-coil-REV2

The LDC scheme is

GERBER FOR BIN-COIL-REV2.rar

over 5 years ago

0 Bala Ravi over 5 years ago

TI__Expert 6625 points

Hi Shrikant,
We've got your request. We will get back to you as soon as we can.

Best Regards,
Bala Ravi

0 Shrikant Bhagat over 5 years ago in reply to Bala Ravi

Prodigy 160 points

Please note the temperature vs output code of LDC1312 at three different points between target and coil PCB.

temp	test 1 (middle Distance of target from coil)			test 2 (Minimum Distance of target from coil)			test 3 (Max Distance of target from coil)
	test 1 (middle Distance of target from coil)			test 2 (Minimum Distance of target from coil)			test 3 (Max Distance of target from coil)
	data hex	decimal	diff	data hex	decimal	diff	data hex	decimal	diff

10	951	2385	11	afc	2812	5	90f	2319	6
20	946	2374	12	af7	2807	-34	909	2313	8
30	93a	2362	15	b19	2841	-14	901	2305	4
40	92b	2347	18	b27	2855	-26	8fd	2301	0
50	919	2329	17	b41	2881	-24
60	908	2312	15	b59	2905	-51
70	8f9	2297	18	b8c	2956	-21
80	8e7	2279		ba1	2977

0 Bala Ravi over 5 years ago in reply to Shrikant Bhagat

TI__Expert 6625 points

Hi Shrikant,
The output of LDC11312 is a digital value that is proportional to frequency with 12 bits of measurement resolution. This frequency can then be converted to its equivalent inductance. Can you please elaborate on how you're measuring voltage as output? The value of Rp changes significantly with temperature in comparison to its L value. And they are related through the quality factor "Q" as in the following equation: Q = Rp * (Sqrt(C/L)). So as Rp is changing, the respective L changes and so does the resonant frequency through the following equation: fsensor = 1/(2*pi*sqrt(L*C)), and this is the output you're measuring. Rp drift with copper targets is about 3900ppm/C where as L drift is only 30ppm/C. Can you please mention what target metal you're using? Also here is an useful app note for temperature compensation: www.ti.com/.../snaa212.pdf. Here is another app note about sensor design: www.ti.com/.../snaa212.pdf. And here an app note about target design: www.ti.com/.../snoa957a.pdf. Also please refer to our "Inductive Sensing Calculator Tool": www.ti.com/.../slyc137.

Best Regards,
Bala Ravi

0 Shrikant Bhagat over 5 years ago in reply to Bala Ravi

Prodigy 160 points

Dear Ravi,
Thanks for your inputs. In my application, LDC output data is changed to analog using mcu and other circuits.

Please note the materiel of target is X10CRNI18-9_14305.

Does drift in Rp will also result in same drift in L?

i have gone through the app note snaa212.

Two coil option is not possible in our application note considering space constraints.

Please suggest other options.

Regards
Shrikant N Bhagat
8308816056

0 Shrikant Bhagat over 5 years ago in reply to Shrikant Bhagat

Prodigy 160 points

Please note the temperature vs LDC1312 output data for 3 points

temp	test 1 (middle Distance of target from coil) (~2.5mm PCB coil to target)			test 2 (Minimum Distance of target from coil) (~1.5mm PCB coil to target)			test 3 (Max Distance of target from coil) (~5.5 mm PCB coil to target)

	data hex	decimal	diff	data hex	decimal	diff	data hex	decimal	diff

10	951	2385	11	afc	2812	5	90f	2319	6
20	946	2374	12	af7	2807	-34	909	2313	8
30	93a	2362	15	b19	2841	-14	901	2305	4
40	92b	2347	18	b27	2855	-26	8fd	2301	3
50	919	2329	17	b41	2881	-24	8fa	2298	4
60	908	2312	15	b59	2905	-51	8f6	2294	5
70	8f9	2297	18	b8c	2956	-21	8f1	2289	4
80	8.00E+07	2279		ba1	2977		8ed	2285

Please note that this data is not consistant over target distance and temperature so difficult to interpolate.

We are also taking similar reading on two different coil to check the consistency among the different coil

0 Bala Ravi over 5 years ago in reply to Shrikant Bhagat

TI__Expert 6625 points

Hi Shrikant,
Can you please measure your IN pins and confirm that you're within the desired oscillation voltage? Please refer to the following app note: www.ti.com/.../snoa950.pdf, section 2 "Why is the correct sensor amplitude important". Vosc needs to be within 1.2V to 1.8 for best measurement accuracy. Can you please provide us a picture or diagram of the system (sensor and target)? Also from your test results, there's not much difference in the output codes between a distance of 2.5mm and 5.5mm. I would be more worried about this compared to the temperature drift. What is your application here? Also I would run a complete SNR test to make sure its normal. For better response, you need to increase the sensor size, but also make sure its not bigger than the target. If your end use is to run a differential measurement, you need to two sensor coils, will this be an issue?

Best Regards,
Bala Ravi

0 Shrikant Bhagat over 5 years ago in reply to Bala Ravi

Prodigy 160 points

Hi Bala Sir,

We repeated the temperature trials again and same results observed.

Temp °C	Test 1 (Minimum Distance of target from coil)			Test 2 (Middle Distance of target from coil)			Test 3 (Maximum Distance of target from coil)
	Test 1 (Minimum Distance of target from coil)			Test 2 (Middle Distance of target from coil)			Test 3 (Maximum Distance of target from coil)
	Hex data	Decimal	Diff	Hex data	Decimal	Diff	Hex data	Decimal	Diff

10	A76	2678	2	94F	2383	7	908	2312	7
20	A74	2676	-8	948	2376	10	901	2305	7
30	A7C	2684	-13	93E	2366	10	8FA	2298	4
40	A89	2697	-16	934	2356	8	8F6	2294	7
50	A99	2713	-19	92C	2348	9	8EF	2287	11
60	AAC	2732	-1	923	2339	12	8E4	2276	4
70	AAD	2733	-40	917	2327	12	8E0	2272	3
80	AD5	2773		90B	2315		8DD	2269

Also we measured the coil resistance variation with temperature, there is much variation in coil resistance and this is the main reason for output code variation. When i calibrate sensor with 1 coil of particular resistance for desired output and replaced that coil with different resistance coil then output voltage shifted by 0.5 to 1 V.

Coil Resistance variation with temp
Temp °C	Coil 1	Coil 2	Coil 3	Coil 4
	Coil 1	Coil 2	Coil 3	Coil 4
	At wire	At Coil	At wire	At Coil
	ohm	ohm	ohm	ohm
Ambient	9	8.8	8.6	8.4

10	8.6	8.3	8.1	8
20	8.7	8.6	8.3	8.2
30	9	8.9	8.6	8.5
40	9.2	9.1	8.8	8.7
50	9.7	9.5	9.3	9.2
60	10.2	9.7	9.7	9.6
70	10.7	10.1	10.2	9.8
80	11	10.5	10.3	10.1

Thanks & Regards,

Akshaykumar Patil

0 Shrikant Bhagat over 5 years ago in reply to Bala Ravi

Prodigy 160 points

Hi Bala,

Should i check the signal across the coil or between one of coil terminal and ground?

0 Shrikant Bhagat over 5 years ago in reply to Shrikant Bhagat

Prodigy 160 points

I2C_Write_Data(0x052b, 0X08); // RCOUNT_CH0//(, 0X08)0x1F69
I2C_Write_Data(0x001f, 0X10); // SETTLECOUNT_CH0
I2C_Write_Data(0x2002, 0X14); // CLOCK_DIVIDERS_CH0 0x1002
I2C_Write_Data(0x0000, 0X19); // ERROR_CONFIG
I2C_Write_Data(0x020D, 0X1B); // MUX_CONFIG
I2C_Write_Data(0x0800, 0X1E); // DRIVE_CURRENT_CH0
I2C_Write_Data(0x0400, 0X1C); // gain 8
I2C_Write_Data(0x1B58, 0X0C); // gain OFFSET // new coil with cut sides
I2C_Write_Data(0x14C1, 0X1A); // CONFIG

The above is LDC1312 configuration

0 Shrikant Bhagat over 5 years ago in reply to Shrikant Bhagat

Prodigy 160 points

Please find attached the waveform at InxA pin with respect to ground

0 Shrikant Bhagat over 5 years ago in reply to Shrikant Bhagat

Prodigy 160 points

I have adjusted the current drive. Below is snap of waveform after correction.

0 Bala Ravi over 5 years ago in reply to Shrikant Bhagat

TI__Expert 6625 points

Hi Shrikant,

Your voltage level seems fine. But, the waveform shape is not good. Please find below the waveform from "INx" pin. Probing either INxA or INxB with respect to ground should yield a half-rectified sine wave, where the rectified part is at 0V and the remaining sine portion is positive.

If the sensor is being driven incorrectly this could certainly yield poor sensitivity or large swings in the output codes. Can you please provide us a picture of the setup? If you feel its confidential information, you can email it to me at: b-ravi@ti.com. Also what type of capacitors are you using? We recommend COG capacitors because of their low drift due to temperature and applied voltage. Can you please verify if the coil resistance data provided is for Rs or Rp? It would make sense if it is Rs.

Finally, you mentioned wires in the coil resistance table, can you please verify how long are the wires or cables between the sensor and the LDC? There is a maximum distance that the sensor and the LDC device can be separated, based on the sensor frequency and losses in the cables. This can be calculated easily using our "Inductive sensing calculator tool", you can download it here: http://www.ti.com/lit/zip/slyc137 (check out LDC131x -LDC161x config sheet).

Best Regards,

Bala Ravi

0 Shrikant Bhagat over 5 years ago in reply to Bala Ravi

Prodigy 160 points

i had tried with different current drive setting,but found no improvement in waveform. Replaced the current 18pf capacitor with 220 pf and found following wave form.Is it normal expected waveform.

0 Bala Ravi over 5 years ago in reply to Shrikant Bhagat

TI__Expert 6625 points

Hi Shrikant,
When you measure the "INx" pins, it should look something similar to the waveform I shared with you earlier. Can you scale your horizontal axis to sample more points? For example, if you're currently measuring 10us/div, increase your sample rate 10x to 1us/div. Its good that you're trying different capacitor values. Can you try something close to 100pf, also make sure you try a COG capacitor this time. Changing cap value will decrease your resonant frequency. How thick is your target? Because the skin depth increases as the resonant frequency decreases. And your target needs to be at least 3 times the skin depth. You can check this by plugging in numbers to the excel sheet I shared with you earlier. I received your setup picture, would it be possible to place another sensor to right of the target? Or would it be possible to have a sensor coil on side of the PCB shielded to the target (this sensor can help us get rid of change due to temperature drift). Finally, if your application requires multiple LDC1312s to read similar output codes, you need to use a clean external clock. From datasheet table 6.5, internal ref clock ranges from 35MHz to 55MHz (typical value of 43.4MHz), due to this variation the output codes will be different for different units.

Best Regards,
Bala Ravi

0 Shrikant Bhagat over 5 years ago in reply to Bala Ravi

Prodigy 160 points

Hi Bala,
The initial value of capacitor 18pf was as per Web bench coil design tool recommendations.But then why waveform was not normal with 18pf ?

0 Bala Ravi over 5 years ago in reply to Shrikant Bhagat

TI__Expert 6625 points

Hi Shrikant,
Parasitic capacitance has a stronger influence with low cap values. For most applications, we recommend to at least use a cap value of 100pF or greater. But this only shifts your resonant frequency, so its not the cause of bad waveform.

Best Regards,
Bala Ravi

0 Bala Ravi over 5 years ago in reply to Bala Ravi

TI__Expert 6625 points

Hi Shrikant,
I am going to close this thread for now. If you've any further questions, please start a new thread.

Best Regards,
Bala Ravi

Sensors

Sensors forum

LDC1312: LDC1312