Hi all,
To do the EMI test, the sense clock source was changed from AUTO to PRS8.
However, it keeps failing.
So I changed the' sense clock frequency' and passed.
However, when I checked the 'CSD Setting', the PRS8 setting part
'The selected sense source is not recommended to use with the PWR sense clock frequency and scan resolution.' is displayed.
So, if PRS8 is changed to AUTO, EMI fails again, and if PRS8 is set, it passes.
In summary, the warning light appears on the PRS8, but the EMI passes.
We ask if there is any problem using capsense with the warning light on.
regards,
alex
Alex-san,
That's a very difficult problem.
When choosing "PRS" as the sense clock source, the sequence must be completed within one conversion cycle.
Tscan = 170.6us : 24MHz, 12bit
Tprs8 = 170.0us : 1.5MHz, 8bit
The above is barely one cycle.
The noise here is not EMI noise, but noise that occurs in the raw count.
Perhaps it works in the direction of improving Noise Immunity by lowering the sense clock frequency from 3MHz to 1.5MHz.
Lowering the sense clock frequency also lowers the SNR. Is there no problem with the SNR?
And, Unfortunately I don't have data on the TSCAN >> TPRS guidelines.
However, I think there is a risk of noise in the raw count due to cycle timing switching.
Depending on the board layout, when using the PRS8, I think it would be better to increase the resolution from 12bit to 13bit and set the sense clock to 1.5M/3MHz/6MHz to avoid this.
Maybe someone more knowledgeable than me can find the best solution.
Regards,
Hi,
PRS of CapSense is described in application note AN85951.
Among those conditions are the following conditions in the CapSense Component datasheet.
When using PRS, "Resolution", "modulation clock", and "sense clock" must be adjusted to satisfy the above relationship.
First of all, please confirm and calculate these parameters that are not visible from your image.
/* AN85951 URL */
Regards,
Hi MaMi_1205306,
Thanks for your reply.
AN85951 has already been confirmed.
After calculation, I confirmed that Tprs8 (3000khz) is 0.000085, which is smaller than TSCAN, so it is correct, and Tprs8 (1500khz) is 0.00017, which is larger than TSCAN, so I confirmed that it is not appropriate.
However, since EMI passes at Tprs8 (1500khz), there is no significant difference numerically, so I wondered if there were any other problems with the application.
This is the part I was asking about.
It is said that noise may occur in a situation larger than TSCAN, but rather, in the EMI (CE) test, there seemed to be a part in the opposite situation, so I inquired.
It says TSCAN >> TPRS. If there are any guidelines on what the minimum and maximum margins can be, please let me know?
| MOD freq(hz) | Sense clock freq(hz) | TSCAN >> TPRS8, TPRS12 | ||||
| TSCAN | 2^12-1 | 4,095 | 24,000,000 | 0.000170625 | ||
| TPRS8 | 2^8-1 | 255 | 3,000,000 | 0.000085000 | 0.000085625 | |
| 1,500,000 | 0.000170000 | -0.000170000 |
regards,
alex
Alex-san,
That's a very difficult problem.
When choosing "PRS" as the sense clock source, the sequence must be completed within one conversion cycle.
Tscan = 170.6us : 24MHz, 12bit
Tprs8 = 170.0us : 1.5MHz, 8bit
The above is barely one cycle.
The noise here is not EMI noise, but noise that occurs in the raw count.
Perhaps it works in the direction of improving Noise Immunity by lowering the sense clock frequency from 3MHz to 1.5MHz.
Lowering the sense clock frequency also lowers the SNR. Is there no problem with the SNR?
And, Unfortunately I don't have data on the TSCAN >> TPRS guidelines.
However, I think there is a risk of noise in the raw count due to cycle timing switching.
Depending on the board layout, when using the PRS8, I think it would be better to increase the resolution from 12bit to 13bit and set the sense clock to 1.5M/3MHz/6MHz to avoid this.
Maybe someone more knowledgeable than me can find the best solution.
Regards,
As an additional information, from my experience it is important for EMI that the damping resistor layout is implemented right next to the PSoC.
It is described in 3.3.1.1 General EMI/EMC guidelines of AN64846.
Furthermore, it may be good to review the value of the damping resistance.
/* AN64846 URL */
Regards,
