View Full Version : AC Frequency measurement with high impedance (HELP PLEASE)
10-02-2008, 11:58 PM
OK, I would like first to thank anyone who might be able to help me out as online forums have always been a great place to give and receive help!
My problem is this:
I need to develop 1:1 DC pulse from an AC signal, the output pulse should be TTL compatable.· Sounds simple enough, but there are some issues.· I am trying to read vehcicle speed from a car or truck.· The source signal I am reading is from a Permanent Magnet Generator style AC inductive sensor, or Vehicle Speed Sensor (VSS).· One challenge is that these sensors develop low voltage at low speeds, 500-600mV AC,·and high voltages at high speeds, 80-100V AC.· I can not draw too much currrent from the signal as it will cause problems with the engine or transmission controller.· Ideally I think an input impedance of 1M ohm would be appropriate.· I am trying to develop this as simply as possible with some basic parts.· I considered using a BS1 with the PULSIN command but that becomes cost prohibitive.· It would be even better if I only had to connect to one of the signal wires!· Any and all help is appreciated!· Thanks!
10-03-2008, 12:17 AM
First what type of vehicle is this used on. I have done this several times, and have not seen this high of ·a voltage from a VSS. Most VSS that I have worked with vary the frequency of the AC signal, with a diode and one or two resistor you can use the count command to read the pulses. You can also use a Optical isolator.
Now on some newer cars the Control Module has an output @ 5vdc pulsed which you can also read
10-03-2008, 12:23 AM
Well there is a little problem here. This has to be rather universal. I know a lot of manufactures, like GM, have a VSS signal to the radio and instrument cluster. Unfortunately many other manufactures are getting rid of these signals and are using serial data to send messages to those components instead. The high voltage reading I mentioned was also with an open circuit. I am sure there is some clipping that takes place from the module that reads the signal. Than ks for the thought of the optical isolator, I had not gone down that road yet!
10-03-2008, 01:29 AM
I've used a 75176 rs485 chip to successfully read the pulses from Allison truck transmissions. Feed the signal in through about a 1k resistor and you'll get a nice square wave from the TTL side.
Not what the chip was made for, but works well with a wide ranging input signal.
If your pulse is referenced to ground, it will also work with just a single signal connection.
Phil Pilgrim (PhiPi)
10-03-2008, 01:38 AM
If the signal is high-impedance, an optoisolator may not work without a current boost to drive it. Isolators typically require as much as 10mA to turn on.
A better solution might be a comparator fed with a clipped signal (to protect the inputs). This can be done with a high-value series resistor and two forward-biased diodes in series with each other. If the signal can also go negative, a single reversed diode across the two in series will clip the negative swings ot -0.6V, which most comparators should tolerate. Set the threshold to 0.6V (using another forward-biased diode). If you still need isolation, the comparator can then drive the optoisolator with its open-collector output and a series resistor.
Addendum: Or use Tom's suggestion. I like that, so long as the signal voltage is within the 75176's input common- and differential-mode range specs.
Post Edited (Phil Pilgrim (PhiPi)) : 10/2/2008 6:43:48 PM GMT
10-03-2008, 04:27 AM
Now I like that idea, and it is with components / theory I am familiar with. Off to Radio Shack I go!
L_Gaminide, stamptrol, Phil, thank you all for the input. I will advise of my success/failure soon.
Beau Schwabe (Parallax)
10-03-2008, 10:41 AM
Here is something that might work for you. I used a similar circuit for some speed of light experiments I did earlier in the year when I needed to detect the presence of a signal and not the signal strength. What makes this circuit unique is the way the Op-Amp inputs are configured. The Non-Inverted input is tied to Vdd/2. The Inverted input is biased to Vdd/2 through a 47k resistor. As a result, the slightest difference on the Inverted input will cause the output of the Op-Amp to swing nearly to the rails. The 0.001uF Capacitor simply acts as a DC block to the input of the Op-Amp. The input 1 Meg resistor and the 22K resistor function as a voltage divider to attenuate the input voltage ( about 46.5 to 1 ) . The remaining 1 Meg resistor further limits current to the Op-Amp. The input Voltage signal can range from 0.5V to 2.5kV ( <--mathematical limits based on values in the schematic ) and has been tested up to a 50kHz signal.
Beau Schwabe (mailto:firstname.lastname@example.org)
IC Layout Engineer
Post Edited (Beau Schwabe (Parallax)) : 10/3/2008 3:50:15 AM GMT