Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Get rid of Noise in experiment

Status
Not open for further replies.

cubdh23

New Member
I have an experiment to messure force and distance and I use a controller to messure the Voltage change from the transducers. My ATD readings jump up and down +- 2.5 mv which is very noisy. I noticed that the voltage goes up and down at 40Hz. My question is how can i smooth out this voltage.
Is there a simple circuit i can make or do i need to buy something to do the job?
thanks
 
whack some caps down.

40Hz is a bit odd?
mains is 50Hz, tube lighting is 100Hz (for the US lot 60Hz and 120Hz) so what is genersating 40Hz!!!


it is way too low to be SMPS, find out what it is otherwise decouple r power lines and add some caps to GND around yr signal inputs
 
Yes 40 Hz is kinda strange, So you are saying i should use a wire like a TV cable type wire to run the signal to the micro? The signal comes from a room with a hugeee magnetic field ( MRI ) dont know if that might be contributing somehow.
 
cubdh23 said:
Yes 40 Hz is kinda strange, So you are saying i should use a wire like a TV cable type wire to run the signal to the micro? The signal comes from a room with a hugeee magnetic field ( MRI ) dont know if that might be contributing somehow.

Probably so! - that could explain the strange frequency. It sounds like you are in a very hostile environment, and are taking no precautions at all?.

Any high gain amplifier requires good quality screening, you should use good quality screened cable (preferably double screened) - TV coax is usually pretty crap!.
 
it is way too low to be SMPS, find out what it is otherwise decouple r power lines and add some caps to GND around yr signal inputs

He doesn't mean mains power lines, he means the power to your circuit :lol: . Just so theres no misunderstanding :wink:
 
zachtheterrible said:
it is way too low to be SMPS, find out what it is otherwise decouple r power lines and add some caps to GND around yr signal inputs

He doesn't mean mains power lines, he means the power to your circuit :lol: . Just so theres no misunderstanding :wink:

Well yer.
That is why I questioned the 40Hz, a very strange freqency.
For 40Hz to be picked up you will need quite a long track or piece of wire,

You say the signal comes from a room witha huge mag filed MRI, that is yr likely source of noise (dunno operating freq of MRI), the big magnetic field are prolly iducing a signal on yr leads


you will need to use screened twisted-pair (only way, coax not good enough). take the twisted pair and use these two wireas as the GO and RETURN for your signal in the other room. Now the screen you will need to terminate to 0V somewhere.

Be it at your cct end (ie the 0V of yr cct) or at the sensing end, well EMC is a black art. All I can say is try one, if no improvement or not enough try terminating at the other end. DO NOT TERMINATE BOTH ENDS!!! that will be worse
 
cubdh23 said:
I have an experiment to messure force and distance and I use a controller to messure the Voltage change from the transducers. My ATD readings jump up and down +- 2.5 mv which is very noisy. I noticed that the voltage goes up and down at 40Hz. My question is how can i smooth out this voltage.
Is there a simple circuit i can make or do i need to buy something to do the job?
thanks

I don't think you need to buy something or add a circuit.

Did you check to see if your meter is working properly?
and 40Hz to me is uncommon for the voltage frequency coming from a power line. Canadians and Americans use 60Hz and some people use 50Hz. Did you make some special adapter that changes 50/60Hz to 40Hz? if so, maybe remove it.

As for smoothing, you will need to play with your capacitors (either in the meter or in the transducers) or get batteries for your meter which dont intend to die instantly.

In most cases, removing capacitors from a circuit smooths voltages, but it may give a negative impact on the circuit depending on the schematic.
 
Screened twisted pair cable

Hello i was just wondering. Is this the same thing as a Cat 5 cable?
The wire that i am using runs about 20 feet so yes it is kinda long from source to destination.
One more thing, i just read that STP " shield twisted pair" provides better protection than ScTP " screened twisted pair" so i guess i should go for the STP. Thanks a lot, i also read that it provides protection against electromagnetic fields, i am hoping that will solve the problem and it makes sence that that is the case. The magnetic fields are so strong that if you take your wallet in there with you, your credit cards will go bad.
 
Cat5 cable is not shielded. It is just twisted pairs and is never used for the very low signals from transducers. The digital signals (very interference-free anyway) in cat5 cable is about 1000 times higher than the very small analog signal (easy to pickup interference) from transducers.

Since there is a very strong AC magnetic field in the room you are connecting to, the magnetic field will pass right through the cable's shield because it is an electical shield, not a magnetic shield. Maybe an instrumentation amplifier IC will help cancel the interference, with the transducer connected balanced through shielded twisted pair cable. :lol:
 
or a local amplifier (really close to the transducer) to really boost the signal befroe sending down the twisted pair.

You def should use twisted pair, the reduced loop-area will minimise magnetic pickup. Screened twisted pair will also help some.

BUT boosting locally to improve the signal-noise ration sounds like the only way in this case
 
So you are saying if i boost the signal near the transducer lets say x 20
then when i reduce the signal back down by 20 on the micro side, the noise will also be diminised by a factor of about 20?
 
cubdh23 said:
So you are saying if i boost the signal near the transducer lets say x 20
then when i reduce the signal back down by 20 on the micro side, the noise will also be diminised by a factor of about 20?

Yes!.

But you don't need to reduce it at the micro end, the A2D will require a far higher voltage than the transducer supplies (usually 0-5V) - so instead of amplifying it at the micro end, amplify it at the transducer end instead. So, no more components are required, just the moving of an already existing section.
 
I am retarded sorry

When i hook up my experiment fully and hook it up to a Scope I get 3 different sources of interference. One is at 60 Hz, the next is at 1.2Khz and another one in the Megahertz range. What i wanted to do is just use the shielded twisted pair but the problem is that the line goes through several circuit boards which i think can pick up the intereference. So what i did today and will test out tomarrow is I made a Low pass filter to get rid of the 2 High frequency noises and for the 60Hz component, i dont know. I am sampling at 100 Hz. Thanks for all the help
 
Compare your transducer's signal to the 10mV signal from a microphone.
Mic lines are not sent through several circuit boards to pick up intereference.
Mic audio isn't sent through a lowpass filter to remove important audio high frequencies.
Mic audio doesn't have 60Hz hum.
So that a microphone preamp output doesn't have interference that is audible, the 10mV microphone signal at the input of the preamp doesn't have interference more than about 10uV or less.
 
Problem solved

Hi, Today i fixed the problem in several different ways. First of all I made a mistake in my original post. The Noise is not +- 2.5 mv ( that would be extremely small if that was the case and i wouldnt have any problems. The noise level was at 200 mv which was causing me problems. After wiping out the Megaherts and Kiloherts interference with the Low pass, I ended up with
'2' 60 Hz noise signals superimposed ( since i am using 2 power supplies) with a level of 60 mv which is much improved. Well I built a Notch filter at 60 Hz which basically wiped the interference clean, but then i decided to use Batteries instead of power supplies, since i am only drawiing 2 ma at 24 Volts (1.3 AH) these batteries should basicaly last forever. I have a beautiful clean signal on the scope. My ATD readings are perfect. I cant believe i dragged this on for a week before fixing it. Thanks for the suggestions and the help.
I did not use a Scope in the begining, I was using my digital readings to try to figure out what was going on and thats what screwed me up.
 
Horray, it's fixed. Nice going. :lol:
 

Attachments

  • icon_party3.gif
    icon_party3.gif
    6.3 KB · Views: 220
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top