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.

Is digital or analog filter more suitable for my project?

Status
Not open for further replies.

emanuelcsm

Member level 3
Joined
Feb 15, 2005
Messages
61
Helped
7
Reputation
14
Reaction score
1
Trophy points
1,288
Location
Brasil
Activity points
1,801
Digital or Analog Filter

Hi guys,

I need to design a project that can detect a missing line of a cloth. In my application I have one IR Transmitter and a Phototransistor, and they are looking direct to the cloth, and the cloth is moving, if a missing line move Through the sensor it must detect it. But the question is How...

I thought do a filter, BandPass, that the output is a pulse and the microcontroller detect it. But I don't know how to design this filter, because the signal when the missing line move Through the sensor is very low voltage, Is it better using a digital filter reading the output of the sensor direct by the microcontroller? I'm using PIC16F876A.

Thanks in advance
 

Digital or Analog Filter

A digital filter would be ideal for this kind of situation as it's adjustable a lot easier than an analog one. We have direct thermal printers at work that use the same basic concept, shine an LED through the paper at a detector. The media is either intentionall notched in the center to provide a pass point for the light or like our stickers it's significantly thinner at the media boundry which is easily detected and amplifed and stored to synchronize the media. Your best bet for determining how to sense the signal is to duplicate it, create a cloth band that you can loop over your test device and create the 'missing thread' in a couple spots in a regular pattern and look at the ouput of your sensor on a scope. The calibration mode of your device would be easy to make, just use the loop at have it detect the noise level of the entire loop for a period of time, and then detect the minimum and maximum peak of the pulse from the missing thread sections. You need to determine basically two values, average noise level, and minimum peak pulse that is required to trip the 'missing thread' indicator. Sampling rate and how much you filter the incoming ADC data determines exactly what that will be. I'd use the mean of at least 4-8 samples with a sampling rate that would make sure any pulse from a missing thread would last at least 16 samples (just for a nice saftey buffer)
 

    emanuelcsm

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top