Epe pic magnetometer

[ivconic]

I don't think is related to sensors.
Probably not to the rest of electronics too.
I suspect surrounding hum and noise, local ac magnetic field sources (large transformers etc).
Maybe some strong TX in vicinity of your location.
Generally all FG/FGM type sensors are "nervous". It is the common thing.
Yet they should not drift that much.
Have you checked temperature variations around the sensors?
It is good advice to pack them in thermally stable enclosure.
To thermally isolate them from outside world.

[ivconic]

When packing sensors in pipe and connecting them with wires; most important thing is to chose proper shielded coaxial leads for that.
To minimize mutual ringing and crosstalk interference. And to prevent it completely; a low pass filter at the end of leads is almost mandatory.
On second look; i think you have exactly that problem. Looks like there is a crosstalk in your leads.
Here is splendid and straightforward article on how to overcome that problem:

[ivconic]

Sensors are providing pretty robust 5v square signals at their outputs with variable frequency in range ~30 to ~100kHz... give or take it few.
Obviously those signals are not "synchronized" ! LOL
If leads are not proper; you can only imagine what can go wrong there in the leads, all the way up to 4013.
One will "trigger" the other and vice versa. Terrible crosstalk interference will occur.
Good shielded cables, separate for each sensor and suggested low pass at the very end of those cables, put closest possible to 4013 inputs: should improve present situation significantly.

[amitech1982]

I don't think is related to sensors.
Probably not to the rest of electronics too.
I suspect surrounding hum and noise, local ac magnetic field sources (large transformers etc).
Maybe some strong TX in vicinity of your location.
Generally all FG/FGM type sensors are "nervous". It is the common thing.
Yet they should not drift that much.
Have you checked temperature variations around the sensors?
It is good advice to pack them in thermally stable enclosure.
To thermally isolate them from outside world.

Thank you for your reply mr. Ivonic.
at the first start, I test it outside the house. and I had the same instability behavior as inside.
But after the change of 78105 and other components. I have not tested it out yet. Today, I removed the lm35dz and noticed that now only the first and the second digit that change a bit, for example: 52402 52401 52408 52408 52411 52408 52405 and so after.
For thermal insulation, what advised me.

[amitech1982]

Sensors are providing pretty robust 5v square signals at their outputs with variable frequency in range ~30 to ~100kHz... give or take it few.
Obviously those signals are not "synchronized" ! LOL
If leads are not proper; you can only imagine what can go wrong there in the leads, all the way up to 4013.
One will "trigger" the other and vice versa. Terrible crosstalk interference will occur.
Good shielded cables, separate for each sensor and suggested low pass at the very end of those cables, put closest possible to 4013 inputs: should improve present situation significantly.

is the shielded cat6 ethernet cable suitable for attaching sensors to the device?

[ivconic]

Ethernet cable is sh.t.
I see on various topics people easily decide to use it for various purposes.
Even making coils from it! It is s.h.i.t. Period.

[waltr]

According to the PIC16F87X datasheet page 122 table 12-2, the two capacitors must be between 15p and 33pf (see attached capture).
-Yes the transfer of data to the PC works without any problem with 10pF or 33pF (I tried both valeus of capacitors).

If the Serial data works perfectly then the PIC oscillator is working correctly so no need to look there.

Listen to Ivconic, he knows about this stuff.

From threads on other projects (TGSL) Ivconic's workshop seems to be electrically and magnetically quiet compared to others.
I know I can get lots of EMI in my shop that makes analyzing builds of sensitive detectors impossible.

[amitech1982]

Ethernet cable is sh.t.
I see on various topics people easily decide to use it for various purposes.
Even making coils from it! It is s.h.i.t. Period.

So what 4-wire shielded cable I can use to attach the sensors to the device.
For low pass filters, can I use 120khz as the cutoff frequency?

[ivconic]

...Ivconic's workshop seems to be electrically and magnetically quiet ...

Good observation, yes it is true.
I live in mountain, outside the urban area.
Very clean place.
Clean in all the aspects.

[ivconic]

So what 4-wire shielded cable I can use to attach the sensors to the device.
For low pass filters, can I use 120khz as the cutoff frequency?

You keep insisting on "4-wire... " , "x-wire"....
Can't you use simple single coaxial?
And than add another single wire for +5v near the coaxial.

[ivconic]

[amitech1982]

[ATTACH]48442[/ATTACH]

Thank you, it is the connection of sensors inside the pipe, and to connect the probe to the device what is what I should use?
And please, give me an idea for the filters that must be installed near 4013 for each sensor, what are the values of C and R.

[ivconic]

Thank you, it is the connection of sensors inside the pipe, and to connect the probe to the device what is what I should use?
And please, give me an idea for the filters that must be installed near 4013 for each sensor, what are the values of C and R.

Values are in the article i posted.
From probe to the device; two separate coaxial + 5V wire.
Install filters as close as possible to 4013.

[SimonS]

For FG-3+ no need RC filter. It is integrated inside sensor.
For my logger I did add also schmid trigger on input. Result are much better

[amitech1982]

Hello.
I made an LC filter (L = 6.8uH and C = 560pF) as indicated in the article, all close to 4013. I then attached to the device a single sensor with 3 short threads 15cm in length, but unfortunately, the problem persists, and the frequency reading is still unstable ... I do not know where the problem is.