Wow! To be honest I haven't a clue why this should happen....

I was going to suggest capacitance coupling effects etc.... but you say it happens with non conducting items as well?

I hope someone else can answer you.... if not why not contact Honeywell's technical department and tell them?

I have already contact honeywell but at that time no answer...

"non-conducting" items can have a static charge, then discharge rapidly if contacted physically.

My kids love to do this to me. walk on the carpet dragging their feet, then transfer the static to me, usually around the ear or nose.

Perhaps the charge field from a static charge can affect the sensor or the measuring circuit. Is the base earthed and the conductors shielded?

Given these assumptions can only guess:

Maybe you've got something like an ultrasonic proximity sensor (sort of). This one believed to be as unspecified adverse effect caused by either sensor's malfunction or its wrong application.

Is sensor attached to any massive body (how tight?) or it's simply laying on the table?

Idea of static charge (CJMcGill) sounds as pretty plausible either.

Have you got all the ground/drain conductors properly terminated from the load sensor(s)? Some devices use the drain conductor as an input to common mode rejection electronics, where as others just require them to be "grounded". I have had problems if there is any moisture in connection boxes/enclosures. Is there any radio frequency devices in the area (such as VSD's, radio telemetry items, radar level devices, nucleonic devices) as if they are not properly installed they can eminate interference using your hand ect. as an antennae. Is the Honeywell device mounted in steel or an insulative mount, may need to ground its body to something.

ROK

First thank you for your answers and your help!

To answer the questions:

-The base is earthed.

-The conductors are not shielded but if I approach my hand no perturbations appears.

-The sensor is attached to a small plactic cross maid of 2 piece(300x30x30 mm)

-There is no radio equipement in the room.

-The sensor structure is maid of plastique but the application point is a steel ball.(Fss1500 honeywell if you want complete description).

I'll try to see if there's importante radiowave in the room and I'll try to change conductors into shielded.

Mathias,

I fully realize that the limited information you've given us is because you honestly don't know what's happening. To solve this mystery, I suggest we start to give you a series of plausible tests to perform, along with requests for specific pieces of the design.

Now, I can think of only two plausible groups of scenarios that cause your sensor reading to change it's output; a false reading with no actual change in mechanical force and a true reading from an unexpected change in mechanical force.

In the true force category, about 0.07 Newtons at first seems like a large unexpected change in force. So I doubt that things like static charge, wind, magnetics and other exotic forces could be the cause. However, your movement to approach the sensor might be changing the floor loading enough to change the orientation of the sensor's force direction with the direction of gravity. In less verbose terms, your tilting the table. A bubble level or laser pointer should reveal this possibility.

In the false force category, 700 microvolts is not a very big voltage change. It would be helpful to know how you are reading this small voltage change. Some common techniques involve bridge amplification, Op-amp amplification, A/D conversion. Your body and or pen might be adding or subtracting interferrence signals to one or both of the signal wires from your sensor. This could be mitigated by reducing input impedances of your measurement circuit, increasing biasing currents or voltages, or filtering out the interfering frequency.

You have an interesting mystery.

Good Luck