0-5V square wave to sine wave with zero crossing?

How about LM339? I believe it's spec'd below ground, but beware, too far below and one input causes the output to switch polarity :-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.

The signal you create from the variable reluctance pickup can not exactly duplicate the square wave sensor in one respect. It cannot work all the way to zero speed. That may not be important, but this needs to be verified.

A pretty good way to convert the VR pickup to digital output is to clamp it with a pair of Schottky diodes (and possibly a series resistor) so that it has a peak swing in each direction of about .3 volts. Then pass that through an LM393 comparator with a pull up resistor on the output to +12. The comparator can handle a swing .3 volts below ground, with the other input at ground and give a nice clean square wave output. The chip container 2 comparators, so you need to ground the inputs of the second half to park it.

a choke/inductor in series with a little load on it ? or maybe pass it through a transformer.

--
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5

Yep, you're right. Never mind.

Wow thanks for answering guys. Yea i am trying to get from a square wave to a sine wave or some wave with zero crossing. The amplitude does change with speed but the ECU only reads the frequency. The 10Vpp is just a clean signal that can be read. The new sensor outputs a 0-5V pulse it might even be TLL but i'm not sure. I'm pretty sure the number of pluses per revolution is the same and should be 4000Hz for one mph. So something around 600kHz should work fine. I am pretty sure the ECU just detects the zero crossing so i figure a -2.5V offset would be great. Some people with different cars are able to put a 5V pulse one pin and 2.5V onthe other to bump up ground. I tired that today but it seems one of my pins is tie to ground.

Phil, Thats sound easy to test how would i wire this up exactly?

Thanks a bunch guys.

I tried just hooking up the 5Vpp to the input of sensor and got nothing. My speedo didn't move and my car didn't shift.

snipped-for-privacy@yahoo.com wrote in news:1130283486.429727.267290 @g49g2000cwa.googlegroups.com:

Are you sure that you need to?

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----

The #1 Newsgroup Service in the World! 120,000+ Newsgroups

----= East and West-Coast Server Farms - Total Privacy via Encryption =----

I just remember reading it somewhere. So it's 8000 pulses per mile traveled? I thought it was a little high for mile per hour. How can i tell if its a hall effect sensor? The sensor has 3 connections according to the wiring diagram there is a ground, V+ and a signal. If matters the new transmission is off of a 2003 Acura CL type s 6spd and is going in to 1999 Honda Accord V6.

Hey, John -

I think he wants to go the other way (make a sine from a rectangular signal).

Or I'm not reading it right, in which case, oops!

John

Hello John,

Sorry to be a bit OT here but did you ever get a confirmation about that from any of the manufacturers? I know that the chips will do it but the data sheet is hardnosed, says that the barbed wire is at exactly 0V. I always wondered why they didn't state -200mV or at least -100mV. This way, if you work in a heavily regulated industry you can't use any of them to properly sense current in the negative return. Neither can you do that on VCC since that's where they have to really stay 2V or so away from the rail.

Regards, Joerg

Hello John,

The only way I could figure doing that would be to sync some kind of function generator chip to it. If that sync has a wide enough range for this case. Or use a micro controller's PWM feature.

Regards, Joerg

Unless your Speedo, tranny or ECU requires a balanced () & un-grounded) input, you probably don't need to do any signal conversion.

The fact that the previous signal level varies with speed is probably ignored by the using circuit.

The fact that the previous signal was a sine wave is probably ignored by the using circuit.

As was mentioned before, the important consideration is the number of pulses per revolution, because that will affect the operation of the Speedo, tranny or ECU.

The frequency range that you are quoting seems completely out of line with standard speedo sensors, are you sure about that?.

The "standard" signal is about 8000 pulses per mile, some GM products use up to 32000 pulses per mile.

65 MPH would result in a frequency of less than 150 Hz at 8000 PPM or a little over 575 Hz at 32000 PPM.

Most sensors that are not variable reluctance are Hall Effect and have open-collector outputs... pulled up to 5 volts or to 12 volts.

be

Thanks Fred I try those this weekend. Scoped the sensor today in lab and i was right the output is a nice shape square wave. I tired 470pF and a .01uF cap in series with the output today and got nothing on the car but on the scope it seemed to remove the DC.

470 pF is way, way, way too small.

I wouldn't be afraid to try several uF - you have very low frequencies here, at the very low end of audio frequency. At 150 Hz a 1 uF capacitor is going to have around 1k ohm impedance, and your .01 uF would have had 100 K ohms. That's fine when loaded with the megaohm imput impedance of the scope, but the receiving circuit probably loads it quite a bit more.

Try scopiung the signal on the ECU side of the cap with the ECU connected.

I would have to use a non polarized cap right? I would scope the ECU but its chained to the table at my school.

** One with amplitude that increases with rpm ?

Or is the signal processed ?

** With no amplitude variation.

** It is not very likely that you need 10v p-p at all frequencies.

Simply coupling the output via a cap will remove the DC component and may be all that is needed to make it work.

Your real problem is the number of pulses per revolution - are the old and new units the same in this regard ???

.......... Phil

Actually I don't think you would. As long as your square wave is unipolar with repect to ground and unless you have substantial inductance in the system the driving side of the cap will never be substantially negative compared to the driven side.

Speaker crossovers often use nonpolarized caps, but that's to allow them to use amplifiers which drive both wires - sort of "H bridge" topology - instead of a single unipolar driver working against ground.

When I did my impedance calculation for 150 hz, that was for highway speed, and I doubt you tested it at that, so it could be even worse - your 100 uF cap experiment could easily have been effectively a 500k ohm series impedance.

Then do this: View in a fixed-width font such as Courier.

. . 5 -- . | | . | | 0.1U 2 -- . 0 -- -- >--------||-------+----> | | . | | | . | -3 -- -- . +----|