LMP8640 accuracy calculation

Hi

I am following the same example and have trouble reaching most of the results in the table.

With regards to the Noise I get 26.162uV. This value was reached as follows.

Noise Voltage Spectral Density is give as 117uv/sqrt(Hz).

First step is convert this into Noise Power Spectral Density by squaring and yields 13.689fV^2/Hz.

Second this value needs to integrated over the 50kHz bandwidth to obtain the Noise power and results in 684.45pV^2.

Third the RMS noise voltage is obtained by taking the square root of the result obtained in the second step.

This yields in the value of 26.162uV but still this value does not agree with that given in the data sheet.

I would appreciate if you can detail me how you obtained the CMRR, PSRR and Gain Drift values.

Thanks

Diva

Hi, thank you for your reply.

I would like to know if this equivalent input noise parameter given is subject to the gain.

Using the signal BW I got the very same number. I am not sure if that is right, but even if I consider the component BW for the noise, that is 230kHz in the case of the "H" part, i get 56.111 uV, much lower than the 216 uV stated there.

About the other error sources, that's how I got them:

For table 2, calibrating CMRR at the middle of the VCM range means we will be 11V away from both extremes:

Vcm av = (Vcm max - Vcm min) / 2 = ( 70 [V] - 48 [V] ) / 2 = 11 V

Applying the wrong CMRR (103 dB), we get the 77,874 uV from:

Vo = V*10^(-CMRR [dB] / 20)

Since we have Vcm ranging from 48 V to 70 V, we should consider CMRR as being 95 dB though.

The PSRR error says they calculated it using a 95 dB PSRR, and I have no idea where they got that from, but it does not have that much impact on the resolution so I didn't care for it.

The resolution moves from 19.117 mA using 95 dB PSRR to 20.194 mA for an 85 dB PSRR.

The TCVOs error was straight forward, getting the 2.6 uV/°C and multiplying it by the temperature range (70°C) gave me 182.0 uV.

And finally the gain drift was calculated using the accuracy over temperature:

eg = 26.2/1000000 * 70 [°C] / * ( ,0041 [ohm] * 10 [A] ) = 75.194 uV

for the 8,2 mohm I got 150.388 uV, close enough.

The accuracy percentage should be calculated as Total Error/Max Vsense also, but the numbers match.

And that's it, even though I could't find an explanation for the PSRR not matching, the Noise part is the one that bothers me the most, as the number I'm getting is wrong by the order of 10.

Regards,

Rafael.

Hi Rafael.

Thanks for your reply detailing the other parameter calculations.

If the noise is subject to gain then it should have different values for the parts with gains of 50 and 100.

But the listed values are same for both gains. They may have got it wrong?.

I do believe it should be related to gain. If this is the case then the noise would be 26.162uV*50 or 26.162uV*100 which is in the mV range and way outside their value?

I will keep on working on this hope you do the same until it is resolved.

Thanks

Diva