Noise problem

Discussion to talk about hardware related topics only.
Post Reply
Ron Neal
Posts: 17
Joined: Tue May 05, 2009 1:20 pm

Noise problem

Post by Ron Neal »

I have a CFV2-40 with an NBIO-200 board. Attached to 3 of the analog inputs on the NBIO-200 board are the X, Y and Z axis of a Crossbow accelerometer. I am powering the accelerometer with an 18 volt battery and would like to hook the CRV2-40 to the same battery. The problem is the noise level goes through the roof so I am still using the power supply that came with the CFV2-40. I don’t have a scope available so I don’t know how high or low the voltage spikes are or what the frequency is. Does anyone have a suggested schematic for a filter that I could put between the battery and the CFV2-40 to clean things up? My email address is nealengr@kc.rr.com Thanks. Ron
seulater
Posts: 445
Joined: Fri Apr 25, 2008 5:26 am

Re: Noise problem

Post by seulater »

I don’t have a scope available so I don’t know how high or low the voltage spikes are or what the frequency is.

Just curious, if you dont have that how do you know there is a noise problem ?
There is no power supply cleaner than having a battery as the supply.
User avatar
Chris Ruff
Posts: 222
Joined: Thu Apr 24, 2008 4:09 pm
Location: topsail island, nc
Contact:

Re: Noise problem

Post by Chris Ruff »

not terribly hard to understand

The front end of the (oldie-but-goody) CFV2-40 is a switcher that is clearly squealing into the accelerometer when the switcher and the accelerometer are connected together.

I would tend to power the accelerometer behind an inductor-cap array that doesn't pass the switching frequency of the CFV2-40.

seu..: You can tell there is noise from the data, one doesn't need a scope to recognize the badness.

the coil is connected in series with the battery and an array of caps are in parallel with the accelerometer power input after the coil. Like 100uF, 0.1uF, 0.01uF all in parallel.

The accelerometer should be powered with a voltage regulator, I can't imagine an 18 volt chip. If you can filter the 3.3 or 5 volts after the accelerometer regulator, that would be the best.

Chris
Real Programmers don't comment their code. If it was hard to write, it should be hard to understand
seulater
Posts: 445
Joined: Fri Apr 25, 2008 5:26 am

Re: Noise problem

Post by seulater »

yes and no, there is not enough detail in the question, it could be other things like over driving the inputs. Far to much cable and so on.
Ron Neal
Posts: 17
Joined: Tue May 05, 2009 1:20 pm

Re: Noise problem

Post by Ron Neal »

One of the debug routines in my code will take a number of readings and give me the high and low values and also the average reading. 1 g = .080 volts. Typical noise for the accelerometer at rest driven off of the battery is about .007 volts for 1000 samples, off of the battery and the CFV2-40 together is anyhwere from 5 to 10 times as much. The accelerometer specs call for a 5.5 to 36 volt unreglated DC supply. The operational current is in the milliamp range.
Post Reply