[propertyco]

General understanding is VR Conditioner is more so for taking a noisy, possibly sine wave and turning it into a nice clean square wave.

Designing a board right now and must decide what Input conditioning I'll add to it. The max9926 is relatively costly vs using an Optoisolator such as the moc205 that the microsquirt uses.

I have a general idea of what inputs will be used with the ECU and don't believe it will be too noisy of an input but do not want to save $9 per board if it'll cause quality issues.

The main concern is taking a 12v cam signal and turning it into a safe 5v signal for the atmel. Cheapest solution would be a reversed diode, though that seems too cheap.

[pazi88]

VR- conditioner is not there to filter out noisy signal. It's there to convert the sine wave signal from VR-sensor to 5v square wave. The input voltage changes huge amounts depending on the RPM, so you just can't use opto-isolator.

If your cam sensor is hall and outputs 12v square wave, you can just use voltage divider to get it to 5v and then add some RC-filtering to it and you are done.

[PSIG]

I wouldn't say "simple" or "cheap" are factors of circuit performance. Due to automotive electrical systems being some of the worst environments, I suggest voltage limiting in your circuit, e.g., input divided for 5V will hit 7.5V with an 18V spike and fry your uC input without the ability to cap the voltage level. The reversed diode does this, as the input is only ground, and the signal V+ is a stabilized supply on the board rather than conditioning off-board voltage sources. No typical reason to invent new circuits when you can use the ones already provided.

But, you may have a reason. A signal conditioner is a common option in order to provide this over-voltage protection, along with the primary purpose of conditioning the signal shape and amplitude. This often provides some level of noise suppression. Some also carry hysteresis and "smart" functions that can assist with certain issues. Some builders use them in any case because they usually "don't hurt", while others select to provide certain functions or protection. All of these options are up to you, along with the compromises they may bring. Do your thing and have fun!

[propertyco]

For some reason I was thinking the reversed diode was meant to limit voltage rather than only provide a path to ground, that makes a lot more sense. I'll just allow both options and allow the customers to purchase their own conditioner if so neccessary.

Any reason a do214 diode such as this wont work?
https://www.digikey.com/en/products/det ... 3-F/808526

500NA leakage at 12v in reverse flow, still run a 1k (possibly higher) pull up resistor

[JHolland]

I've never seen an opto-isolator in an OEM ECU, they really aren't necessary in a properly designed circuit. For a VR sensor you really want to be detecting the zero crossing point and a series diode is going to add an offset which will cause the timing to change with RPM, if you have a very narrow tooth then it may not be an issue but it may be significant with a wider tooth.

[PSIG]

The reversed-diode is only for digital signals and voltage protection. Not VR. As the digital signals (e.g., Hall) are clean images of the tooth shape, the signal does not shift with rpm as a VR could with its 'sine' wave. Opto-isolators are often used in aftermarket, as we cannot guess what crazy signal a user might apply (magneto pulse? 300V+ coil(-)? ), whereas the OEMs have that well under control, dealing with only one specific and known signal in any given project. They have it easy. These user unknowns are the single biggest challenge to aftermarket conditioner designs.

There are several basic ways to condition signals, and the question is what signal type you are conditioning, the unique features of that signal, and then how it should be handled, in order to select a module or circuit that will do the best job. While some can do a fairly good job with most signals, there is no one answer for everything.

Even those designed for specific use are not perfect, and the MAX992x chips for example can detect zero-crossing, but most applications of that chip do not, and use a mode with offset voltage and hysteresis instead. In-spite of that it seems to work pretty good (much of the time), but is why we emphasize using a specific polarity and triggering on a specific edge with VR, in order to limit the signal drift. Nothing's perfect.

[Wombat]

A dirty VR signal is to your engine what a trip hazard is to a person on their way down a flight of concrete stairs.
I cant stand VR sensors and will change to a hall type whenever I can!