My challenge to you.

[altered7151]

Okay, so there are a couple of us on the board with datalogging abilities on our wideband controllers, and we would like to log the CEL signal so we can see knock if we have the CEL-knock sensor feature enabled on our calibrations. So here is the issue we need to resolve. The CEL signal is a ground reference, and we need a voltage signal of 0-5V going to the wideband. So we need to figure out how to somehow use the CEL ground refernce signal to get that voltage signal to the wideband, and it can either be a rising or falling signal, just as long as we can see a change in the signal and know that its knock. And yes I know there are pre-built units out there, but in true TM fashion I dont want to spend $100 on one when I know someone here can come up with a cheap easy way to do it ourselves. Put on your thinking caps and lets see what you've got

[lametec]

Put a pull-up resistor (1k ohm or similar) from +5V on the w/b controller to the input you want to juse for CEL logging. Then use a diode between that input and the CEL signal, anode facing the w/b input.

When the the CEL goes active, you should see about .6V on the input. Any other time, it should hold 5V.

[cordes]

Put a pull-up resistor (1k ohm or similar) from +5V on the w/b controller to the input you want to juse for CEL logging. Then use a diode between that input and the CEL signal, anode facing the w/b input.

When the the CEL goes active, you should see about .6V on the input. Any other time, it should hold 5V.

Good show! I too was thinking a resistor would be the easiest way. I think that I am catching onto this whole electronics thing!

[Tony Hanna]

My first thought was to power a relay from the key and use the cel to ground it. Then just feed 5v to one side of the contacts and run the other to your wb. No knock = 0v, knock = 5v.

[altered7151]

Put a pull-up resistor (1k ohm or similar) from +5V on the w/b controller to the input you want to juse for CEL logging. Then use a diode between that input and the CEL signal, anode facing the w/b input.

When the the CEL goes active, you should see about .6V on the input. Any other time, it should hold 5V.

Ok I should also mention that I suck at circuits, I'm getting better, but still learning. So I'm not 100% on the circuit you described, So are you saying basically to run a +5V signal to the logging channel on the wideband, and then install the diode between that signal and the CEL signal wire, so that when the ground reference signal is activated it will ground out that ciruit and the voltage to the wideband channel will drop?

[altered7151]

Ok how about this (since I already have the parts). A 7805 5-volt regulator so I have a constant +5v going into the wideband, then a diode between that 5volt out and the CEL signal, with the anode spliced in towards the +5V from the regulator. So in my mind and my feeble understanding of circuits, that should provide a constant +5V to the wideband, until the CEL ground reference is activated, which it would drop my voltage way down. As long as i can monitor that spike I should be fine. Yay or nay?

[lametec]

Yes, but don't forget the pull-up resistor. If you put the +5V directly to the input, chances are you'll fry your circuit when the CEL comes on. The resistor limits the current to a safe level.

Here's a quick ascii circuit:

+5V----[R]---I----|>|----CEL

Where:

+5V is the +5V source on the w/b controller.

-[R]- is the resistor. Exact ohm value isn't crucial. Anything between 1k and 10k should work.

I is the data log input on the w/b controller.

-|>|- is the diode (anode on the left side)

CEL is the CEL wire that goes to ground when CEL is on.

When the CEL is not on, +5V will flow to the input through the resistor. There will be +12V on the CEL line (assuming it's still connected to the CEL light), but that won't reach your input due to the diode. When the CEL comes on, current will flow from the +5V through the resitor and diode to ground (which is what the CEL line provides according to you.. I haven't double checked this). The voltage drop over the diode will be arund 0.6V, so the data log input will see 0.6V.

This SHOULD work.

[turboaddict]

only thing no 5v SOURCE from datalogger, only 5 v INPUT so we have to convert a signal to 5volts or pull it from an external source......... then use that signal and connect it to the ecm ground for the cel.

[lametec]

That can be solved using a zener and another resistor. Or a 7805 voltage regulator.

But if you crack open the case on the w/b controller, I'm fairly confident that there's a 5V source somewhere. Solder a wire or the one end of the resistor to a +5V point, and you're in business.

This requires some knowledge of electronics, but that's pretty much assumed when you start connecting resistors and diodes to your stuff.

[altered7151]

Correct, there is no output from the wideband. I know alot of the units out there provide a 5volt source for running external sensors etc, but this one doesn't. So we have to find that +5V source somewhere else. But I think Lametec's idea will work, we just need to find the input from another source. Here is a simple diagram ( my paint skills aren't much better than my circuit skills) of what I had in mind. Not sure if the resistor is still required though, but it would be nice to drop the +5V in down a little just to ensure I dont fry the wideband.

[lametec]

That should work. Don't forget the capacitors that go along with a 7805, though.

The resistor is still needed, or the diode will fry when CEL comes on. When the CEL is not on, the input on the w/b will still see +5V. There's no voltage drop to speak of over the resitor when CEL is off.

Most likely the input on the w/b is well protected. It probably goes into an op-amp first before going to the ADC. Failing that, even the ADC will have some protection on the input.

I don't know what kind of uC or ADC is in your w/b box, but on thing to look out for is not to feed the ADC in a signal much higher than 5V. The 7805 can vary a bit on the output, and if it goes a little too high, the ADC can latch up. I know this can be an issue on PIC controllers. But as I said, more than likely there's an op-amp on the input to keep that very thing from happening.

[turboaddict]

or tyler since you have already replaced the cel with a led you can resist the whole cel down to 5v and run your signal like that.

[altered7151]

My brain hurts.

[XR Tuner]

The wideband allready has a 5V regulator. Just tap into it.

[Aries_Turbo]

I had some ideas already and tyler tested them but they didnt work so Im back to the drawing board as well as he is.

I was thinking using the circuit that powers the bulb to trigger a double pole relay.... one pole of the relay driving a light (so everything looks normal) and the other triggering 5v or less to the wideband.... but i dont know what the max current the LM can source to ground through that circuit so that limits my choice of relays to ones that are real small.

oh well. if I sat down, I could do the same thing with a transistor or two.

Brian

[XR Tuner]

What is the voltage on the CEL line when it is not activated. If it is just an open circuit I don't see why it would not work with the pullup resistor when the line is grounded.
I don't see any need for the diode, if the line simply goes to ground when it is activated.

[lametec]

When not grounded, that wire is connected to +12V through the CEL bulb. That's why the diode.

[Tony Hanna]

Here's one for ya. Wire in a second bulb to the cel circuit and put it in a little box like a radioshack project case with a small solar cell. Feed the output from the solar cell to the wb.

[XR Tuner]

Here's one for ya. Wire in a second bulb to the cel circuit and put it in a little box like a radioshack project case with a small solar cell. Feed the output from the solar cell to the wb.

Can you say opto-isolator?