Frequency Measurement Accuracy

Ron Floresca

Other Parts Discussed in Thread: LDC1614

Hi,

What is the measurement accuracy for frequency measurement (of around 9-10MHz) for LDC1614 with ENOB’s set to 18bits or 21bits?

From the EVM, looks like the Max ENOB is just 21bits.

We can see also in the GUI that at both of these settings(ENOB of 18 or 21) that there can be +/-1kHz change in the readings however we would like to know what this device is capable off.

Thanks

over 9 years ago

0 Ron Floresca over 9 years ago

Prodigy 170 points

Also, was there a formula for determining ENOB? Thanks

0 Ben Kasemsadeh over 9 years ago in reply to Ron Floresca

TI__Genius 13650 points

Hi Ron,
ENOB is described on page 13 of the datasheet: www.ti.com/.../ldc1612.pdf . An application note with more information on how it is affected by the sensor frequency range is available here: www.ti.com/.../snoa944.pdf

0 Ron Floresca over 9 years ago in reply to Ben Kasemsadeh

Prodigy 170 points

Hi Ben,

Thanks for that. I have still some questions.

On p13 of the datasheet it discussed about the ENOB. However when I apply that formula after getting my RCOUNT, my ENOB calculation is lower than the one from SNOA944.pdf and from the excel calculator as well [Approx. Measurement Resolution(=IF(LEFT(C21,6)="LDC131",MIN(LOG(2*2*C29,2),12+C23),LOG(2*2*C29,2)))]

0 Ben Kasemsadeh over 9 years ago in reply to Ron Floresca

TI__Genius 13650 points

Hello Ron,
I will research this and get back to you on Monday

0 Ben Kasemsadeh over 9 years ago in reply to Ben Kasemsadeh

TI__Genius 13650 points

Hello Ron,
SNOA944 does a much more accurate analysis on ENOB than the formula in the datasheet. The application note breaks down the whole range into frequency bands and analyses those.
Please use the application note as a basis for your noise analysis.

0 Ron Floresca over 9 years ago in reply to Ben Kasemsadeh

Prodigy 170 points

HI Ben,

Thanks for your reply. I check SNOA944 and have some questions on some of the table and graphs.

For Table 3- Do you have a formula I can use to determine your #Unique Output Codes. I want to keyin my required resolution and come up with my own uniq output codes.

For Figure 6- Do you have formula as well to come up with this graph?. I want to used my actual RCOUNT and sensor variation to come up with my actual effective resolution.

Thanks!

0 Chris Oberhauser66 over 9 years ago in reply to Ron Floresca

TI__Genius 14270 points

Hi Ron,

No equation has been developed. The entries on table 3 are derived from Figure 3 (if you are using fref = 40MHz, then you can use Figure 5). You then use the sensor frequency variation due to target movement to determine the number of unique output codes, and then the resolution.

Regards,

ChrisO

0 Ron Floresca over 9 years ago in reply to Chris Oberhauser66

Prodigy 170 points

HI Chris,

How do you normalized resolution on Figure 3? To what parameter you normalized it from?

Thnaks!

0 Ron Floresca over 9 years ago in reply to Ron Floresca

Prodigy 170 points

Hi Chris,
Any update on my above question? I see you closing this issue already.

0 Ben Kasemsadeh over 9 years ago in reply to Ron Floresca

TI__Genius 13650 points

Hi Ron,
in figure 3, both axes need to be scaled to the reference frequency.
In figure 5, 40MHz is used as a reference frequency as it provides the best performance.

Sensors

Sensors forum

Frequency Measurement Accuracy