Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

[SOLVED] how to remove 400mHz envelop signal?

Status
Not open for further replies.
J

janaranjanahl

Junior Member level 1
Joined
Jun 20, 2011
Messages
19
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,464
we are testing a very sensitive magnetic field sensor for few weeks and we are experiencing a ~400mHz (like 40 Min period) wave which is act as a envelop. We don't know where it is coming from. because our system does not use such a frequency any where. we read few articles and we suspect that it comes from sun's radiation. but we need to remove it. we tried a high pass filter but it did not work.
I have attached part of wave form.
Please let me know if you have any idea to remove it.
we are using batteries for power supply and has stabling capacitors.
Thanks in advance.

Capture.PNG
 

Thanks for replies. what is the umbrella you are talking about? please explain. We are testing in lab conditions isolated from vibrations and sensor has installed in a big Al box. despite the fact the sensor installation is there any way to remove ~400mHz signal.
Thanks.
 

Yes. your correct. it should be ~400uHz (micro hertz) . period ~40 min.
 

I think you meant about 400 uHz (40min*60s/min=2400 s)
What about the sensitivity of your sensor ?

I think it could be a problem related to the digitizing of the waveform; maybe I'm wrong, but it seems something like a beat between two slightly different clocks one on the board an the other on the digitizer. In fact the highpass filter has no effect.
 
If using 50Hz or 100Hz ripple magnetic field and signal filter does not attenuate input signal by magnetic shielding, (i.e. not Al box) then aliasing with sample rate is possible.

1/(2400 sec * 100)= ~ 4ppm difference if 50Hz and ADC clock are both stable?
 

we are not changing the magnetic filed. it is a fixed magnet. we change current to the sensor. we are using [two signal generators to sync[/B] frequency of the sensor with lock in amplifier. we believe lock in amplifier needs the just the value of frequency. what is your opinion? we still have the problem that out put contains 400uHz signal. where it came from ?:sad:
 

do you think using two signal generators for lock in amplifier and the sensor with same frequency is a good idea?
 

What f is driving the coils?
What BW is filter and sample rate f?

- - - Updated - - -

Some PLL may be necessary but you are measuring ~DC magnetic field. Good "Mags" have 50-100Hz BW.

- - - Updated - - -

I recall for measuring small magnetic field changes at 100k and 200kHz , down to 10ppm resolution, we used differential coils with quadrature PLL's to detect phase shift with 100Hz RBW.

Still no idea what you are using.....
 
yes we are measuring dc magnetic field (since it is a permanent magnet) maximum value of that magnetic filed around 1.4 T. we are not using any filters.

we can go up to 1nT sensitivity.

How magnetic filed can change in a fixed magnet?

I can not understand dealing with PLL and BW of magnetic field. Could you explain?
 
Last edited:

To get down to this resolution, we used a servo sin-cos potentiometer from Bourns and a Swiss motor with 2000:1 gear ratio for the PLL to normalize the absolute phase error and then had variable gain against a calibrated external moving reference of some known small incremental miicro-Guass level which induced an impedance change in the coils detected by the quadrature PLL to normalize the gain. The pots applied the excited frequency to null any offsets in the balanced coils. So it was self calibrating for phase and gain.

Today this can be done in other ways. We did not use magnets.
 

I really appreciate your comment. But I don't understand.
Simply we are using permanent magnets and we have mechanism to change the magnetic field.

Problem is we are getting a 400uHz signal and we don't know where is came from. so it makes us to wait 20min to get a data point. we are using a photo detector to detect sensor movements photo detector signal goes to a lock in amplifier.
sensor and lock in amp: operated using two signal generators and has the same frequency around 300Hz. out put recorded by digital multi meter to a PC screen.
out put signal is as shown in the above figure.(figure in the 1st posting) . we need to remove 400uHz big wave and see the results.

I guess I have explained everything about our set up.
we are not using ADC.
lock in amp: has a PLL.

Thanks in advance.
 

I guess I have explained everything about our set up.
Click to expand...
Surely not. Presumed the interference is somehow generated by your design, e.g. a beat frequency of two oscillators, one need to know every detail to narrow down the problem cause.

Besides design details, also quantitative signal parameters matter. You didn't give yet an idea about the effect magnitude.
 

Re: how to remove 400 microhertz envelope signal?

I changed frequency by 10Hz but I get the same wave form as previous. Period is 40min. It did not change.
 

Could a magnetic field be formed by your circuit? As you changed the frequency and it is still the same I may assume it should not.

What I kept thinking on is that it's a ultra low frequency and it's constant. We are envolved by the earth magnetic field, and it could have variations (very slow and small ones as you are measured). Could you measure with the same sensor in different places? If you get differences then there's the problem.
 
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top