LMP91051: External Filter Design

Jung,

I will look into this and get back to you soon.

-Kelvin

Jung,

Q1) Yes, you should be able to detect those gases. Just note the spectrum characteristics of each gas. There is some good information on page 17 of the LMP91051 datasheet regarding this.

Q2) Yes, if you had a bandwidth of 100Hz, then the signal would not be attenuated +/- 50Hz from the center frequency. If this is what you were trying to ask..

Q3) The lamp has a pulse frequency of 1 to 3 Hz (typically).

What are you trying to measure in your application? Why would your application require a center frequency of 4Hz?

Best,
Megan

Thank You for your Reply.

I have more questions.

Please let me know.

1) Schematic

SCHEMATIC1 _ PAGE1.pdf

2) Webench settings

3)Waveform

Condition and needs)
The way we are testing is to measure the Active (250ms) and Ref (250ms) once by driving a 4Hz lamp frequency (green mark) as shown upper waveform.Active & Ref (purple mark) was adjusted to 4Hz so that each waveform was a complete sinusoidal wave by phase-adjusting, and then each signal was measured.

As the problem of this method is that peak drift of each signal occurs.
About 20mV drift occurs when aging at Active 700mV (peak to peak) (drift exists even if LMP's Mux is fixed).

Q1)
If the signal drift is normal, what is the ratio?

Q2)
Please advise how to improve the drift. 4Hz drive and Ref & ACT signal measurement method etc.

Q3)
What is the cutting frequency when Bw is 100Hz after setting the Fc of BPF to 4Hz? (Calculated result is 0.15hz ~ 106hz using R, C LPF, HPF calculation)

Please help me and let me know.

Any question, let me know.

Thanks.

Jung,

I'll look into this and get back to you.

Best,
Megan

Hi Jung,

On the waveform is the yellow mark measured from the OUTPUT pin on the LMP91051? Just want to clarify.

Q1) What is the ratio you are referring to? What are you trying to compare the output to?

Q3) Using your resistor and capacitor values, I calculated the same cutting frequencies (0.15Hz ~ 106Hz). However, for an ideal bandpass filter, if you had a center frequency of 4Hz, and a BW of 100Hz, then the cutting frequencies should be +/- 50Hz from the center frequency. Is the calculated result what you were expecting or desired for your design?

Best,
Megan

Hi Anderson

Yellow Mark waveform is LMP91051's output.
And you can see drift about 20mV.
They want complete sinusoidal wave output. But, According to waveform yellow mark is not complete sinusoidal wave. About 20mV drift is occur.

Q1) They want to know this 20mV drift is normal specification of LMP91051. If yes, they want to know drift ratio referring to complete sinusoidal wave.

Q2) Please advise how to improve the drift. 4Hz drive and Ref & ACT signal measurement method etc.

Q3) In webench, Filter R,L,C value is automatic calculated when Fc 4Hz, BW 100Hz.
So they want to what is correct cutting frequency, calculated 0.15~106Hz or +/- 50Hz from the center frequency.

Jung,

Could you show us exactly on the waveform where this drift is occurring? Does this same offset pattern repeat overtime?

Why do you need a BW of 100Hz when your Fc is 4Hz? The cutting frequency you want depends on your application.

Best,
Megan