LDC1614: accuracy issue or quality issue of LDC1614

Ba,

The document uploaded to the thread is helpful, but we will need more details from the customer to figure out what is going on.

I will update this thread with some questions by Friday. 

Hi John,

Thanks for the quick response. The customer has no access to E2E, If it is convenient for you to communicate with the customer by email, attached includes the email box of the customer. If not, please feel free to post it here.

Thanks,

CSC team

Ba,

Can the customer provide some additional details on what they are seeing?

1. Have they measured the same or different frequencies for the inductor/sensor signals for the two lots?
2. Have they used the same PCB & inductive sensor with different devices or looked at the behavior of different devices on different PCBs, and sensors?
3. How many devices and PCBs have they compared between the two lots?
4. Have they tried comparing the behavior of the devices from two different lots on anything other than their product board?

Hi John,

Here below is the feedback from the customer.

1.     We did not measure the frequeny.

2.     Yes,  we exchange the LDC 1614 in 2 systems only, and value exchanged.

Details:

      Step 1:

                  LDC1614-2 + system 2:   139xxxxxx

Step2:

          Remove the LDC 1614 form the PCB.

Step3:

          Solder LDC1614-2 into system 1:  and test ,  the value is 139xxxxxx.

          Solder LDC1614-1 into system 2:  and test ,  the value is 149xxxxxx.

So you can see, the value” follow” the LDC 1614 only. 

3.     For testing, we tested 2 batch IC,  one batch is 10 pcs.  

4.     No, we have some model PCB only.  

The LDC device is highly dependent on the system. The accuracy is a combination of the LDC device and the sensor which is a combination fo LC. Any variation in LC can result in change. As John pointed out customer needs to evaluate and also compare the sensor design across PCBs. 

Can they try John's suggestion here ?

Customer's feedback like below:

As we have described in the last file, what we have tested was:

Different batch LDC 1614 on 100% same other part (Antenna, PCB, cable ….. ).

We only replace an IC on a system, and get different result.  The other parts were not changed, it is the same piece. 

What we hope to know is : 

If the difference batch IC will be same or not. 

How many differences with different batch.

If all batch should be same or under 1% difference, we will ignore the batch and keep our design.   

If there is more than 2% difference between batches IC , we have to change our design to provide a calibration mark.  

When you described Steps 1-3 in your earlier message on Jan. 20, did you:

1. Swap/moce the devices between two PCBs and sensors..
2. Remove the devices from the PCBs , and replace them with different devices from the other lot.

Customer's feedback like below:

Thank you very much.
Yes. we " swap the IC" only, and find the value follow the IC lot.
We have 2 lots IC only, and found same lot, same value.

Ba,

Our datasheet gives the device behavior and the limits of the individual device specs, but we cannot guarantee limits in a larger circuit, because circuit behavior is dependent not only on LDC device variations, but also on the design tolerances of the PC board, sensor, and components.

To get the 1% variation they are looking for, the customer will need to understand the variations of the PCB, sensor, and components, and how those variations interact with the LDC device, and affect the overall circuit behavior. They will then have to control the tolerances to get the performance they want with the published specs in the LDC data sheet.

If that doesn't work, they will need to consider the calibration strategy they mentioned earlier.