| Home | Terms of Use | Site Map | Contact Us |
IndustryCommunity.com > Electrical and Electronic Community > Analog Circuit Design Forum > Message
Main Menu

[ List Subjects ][ Main Page ]
[ View Followups ][ Post Followup ]

Subject: Finding 0 voltage in analog signals

Date: 01/22/02 at 5:22 PM
Posted by: Corry
E-mail: corry@unt.edu
Message Posted:

Let me elaborate :)
Application: Digital Speedometer recalibration
Background Info: In my 1987 Pontiac Trans Am GTA, the speedometer is displayed on a digital dash. The speed of the vehicle is determined by a motor (being used more as an alternator) with its motion generated by the output shaft of the transmission. This creates a sine wave out, which goes into a frequency counter, and the speed is displayed. 1 Hz is equal to 1 MPH
Problem: When, for performance reasons, the rear end gears are set to a different gear ratio, the speedometer is no longer correct. Small adjustments can be made by replacing speedometer gears, but large changes (i.e. original ratio 2.08 new ratio 4.10) cannot be corrected so easily.
Solution: Frequency Divider on the speedometer motor output.

If there is some way to do this without converting to digital, do tell! My current solution is to count the frequency, divide, output into a function generator, and plug in to the original circuitry. (Re-doing the entire thing will be a future project when I have time to take the dashboard apart and find all the input wires to the 3 seven segment displays, and determine the proper input to the cars computer since it like to know how fast you are going).

My snag is on the frequency counter. Since this is going to get counted again, with error, I want to minimize error in this stage. To account for the rapid changes in frequency, my plan was to use a 555 timer, at some sufficiently high frequency 400Khz or so, and count the pulses between the 0's of the analog signal. (yes, only counting a half wave, this will be more accurate :) The problem is finding when the analog signal voltage goes to 0. The only accurate method I have come up with are lots of op amps, with very low reference voltages, that sucessivly go down. starting at maybe 0.2 volts. After all the opamps have had a high output, and are all 0, counting will begin, when a 1 appears on one of them again, counting will stop. This is something I would rather avoid, especially if I am re-inventing the wheel! If someone knows of a better way please tell me. As I said befor though, I need accuracy and speed, and this is the fastest, most accurate way I can think of to do this, but I'm not a EE, I'm actually a CS major, with interests in electronics, and cars :)
Corry Lazarowitz

Follow Ups:

Post a Follow-up:


Message to Post:


1999-2001 Sunlit Technology Co., Ltd. All rights reserved.