RPM Signal Porsche 944

L Cubed · Post by **L Cubed** » Wed Aug 06, 2014 10:15 pm

I have recently installed RCP into my 1989 Porsche 944 S2 and the GPS and accelerometer are working and logging fine. I tapped into the signal from the ecu to the ignition module for the RPM signal. Just plugging this into RCP directly doesn't provide a signal. Using a diode (anode end in RCP) I was able to see correct RPM in the sensor monitor. If I start a log right after that I get an RPM signal in the log. If I wait sometime without changing the configuration or plugging in the laptop, I don't get the RPM signal in the log. Plug the computer in again and it seems to work. I'm not sure what's going on and unfortunately I have not been able to measure the signal with an oscilloscope. I think it might be a 12v square wave so I used the diode to block the positive end but allow the 0v signal to pass to RCP. The ignition module is a Bosch unit that I am somewhat certain is using a switch to ground signal to trigger the ignition coil. If I need to just use the CoilX I can but if I can do without that would be nice since I have an autocross coming up soon.

L Cubed · Post by **L Cubed** » Mon Aug 11, 2014 3:01 am

I have an update as I continued working on this problem. I recorded the raw signal at idle with one of the analog ports and found that the ecu sends out a +5V (~4.5) to trigger the coil. So when I removed the diode and just fed the signal directly into the frequency port, when I did this I could read the idle rpm, ~840. I also tested the other two frequency ports measuring 28 Hz (2 cylinder fires per rev). Then I ran into my previous issue again...

It appears that the RCP does not like to start-up with an RPM signal going to it. Starting with the engine on and the RCP off, when I turn on the RCP I find that I get no reading from the RPM. If I remove the RPM signal wire from the RCP and then plug it back in, all the sudden I get a solid reading of RPM that matches the tachometer. I repeated this multiple times and got the same results. I still need to double check if the RPM read functions if I turn the RCP on before the engine. Has anybody else run into a problem like this with other sensors as well? I have the 1.28 firware (RCP purchased in June 2014). If I need to just have the RCP on before the engine starts I'll set it up like that.

L Cubed · Post by **L Cubed** » Tue Aug 12, 2014 1:39 am

So I tried this morning to having the RCP powered on before starting the engine on my way to work. Still don't have RPM being logged. This makes me feel there might be a firmware issue or a start up issue. Is there a way to power up the RCP and then have the RPM input reset?

I could really use some help with this, I would rather not have to manually remove and re-install the RPM signal every time I want to log RPM.

gizmodo · Post by **gizmodo** » Tue Aug 12, 2014 2:47 pm

I've had quite a bit of trouble getting an RPM signal as well. At first I tried using my MSD tach adapter, but had no luck. Then I tried a CoilX. It works sporadically at best, and I can't figure out what the problem is. If I find anything I will be sure to post here.

Post by **brentp** » Thu Aug 14, 2014 8:55 pm

Hi,

If you had a scope to characterize the signal coming from your ECU that would be very helpful - then we can see what the signal looks like and suggest a remedy.

With regard to connecting RPM after startup, are you saying everything else works, but RPM doesn't show a signal?

The other option is to tap directly into your ignition coil trigger wire (-) by using the CoilX module:
http://www.autosportlabs.net/CoilX

gizmodo, with CoilX are you tapping right into the coil? What is the signal like - do you see a reading, but punctuated with RPM spikes? We're looking at adding some additional filtering firmware-side to help mitigate extra noise.

L Cubed · Post by **L Cubed** » Thu Aug 14, 2014 11:34 pm

So I made a little video which I hope does a better job explaining what is happening.
https://www.youtube.com/watch?v=JYcd0TXVV9k

Basically if I just go to start the unit and record the data, all internal sensors and GPS work but RPM does not. If pull the wire out while engine running and unit on and then plug it back in I get an RPM signal. Turning everything off and then back on again does not record an RPM signal.

Here is a picture of the signal trace. I recorded this via an analog input at 100 Hz while the engine was idling at approximately 840 RPM. The first trace is from the data points I exported to excel to focus on a smaller section of running, the second is as shown in Race Analyzer.

What I think might be happening is that since my ECU grounds the signal when not triggering the ignition coil (verified with voltmeter when key on, engine off) the RCP unit is faulting that port since it usually has a default of positive voltage (5v pull-up). Does this sound like it might be an issue? Is there a way to switch the unit to a pull-down for the RPM signal?

Post by **brentp** » Fri Aug 15, 2014 5:07 am

Can you describe how this trigger wire is hooked up in your system? Any wiring diagrams you can link?

Primarily interested in what it connects to. Does it connect to the coil in any way?

L Cubed · Post by **L Cubed** » Fri Aug 15, 2014 12:55 pm

Here is how it is wired. I am tapping directly into the ECU signal that is sent to the Bosch igniter to fire the coil. My engine has a single coil and a distributor, the Bosch igniter is used as electrically controlled ignition advance.

Post by **brentp** » Fri Aug 15, 2014 11:58 pm

Thanks. Interesting how you used the analog channel as an oscilloscope.

It might be helpful to see that signal on a real scope; do you have that handy?

Otherwise, a special version of the coilX module designed to work with that signal could be devised.

L Cubed · Post by **L Cubed** » Sat Aug 16, 2014 7:20 am

I'll see if I can find an oscilloscope to borrow to better capture the signal.

Another question I had was if there is a lua script function to do a channel reset or channel interrupt? I figure this could work since I get a RPM signal after a manual interrupt.

Is RCP designed for a switch to ground 5V logic signal? If that's the design intent, it might be easier if I just add a logic inverter to the signal from the ECU before sending it to the RCP.

Post by **brentp** » Sat Aug 16, 2014 3:39 pm

It would be best if it was a push / pull type signal. So you might be able to try a pull-up resistor or a logic buffer might work for you. if a logic buffer, try one with a schmitt trigger.

L Cubed · Post by **L Cubed** » Sat Aug 16, 2014 4:39 pm

Do the frequency input ports have protection for a "short"? I feel like the port may be going into protection mode until the 5V from the pull-up resistor is restored as the value being read in.

I think this because with just the wire plugged in the RPM signal works fine and can be logged and displayed by Race Analyzer on the Sensor Monitor. However upon startup this doesn't work, almost like the way a program function may not work if the variables aren't initialized. This is why I was curious about a channel reset function that may be able to be accessed via the Lua Scripting.

Post by **brentp** » Sat Aug 16, 2014 4:45 pm

Hi -
Not sure what you mean by a 'short'? A short to what? The port is configured as input only, so a 'short' in the traditional electrical sense won't matter.

Another thing you can do as a test is loop one of the RCP's PWM outputs (provided the dip switch is 'off' - we ship them off by default) into one of the timer inputs. you should see an RPM value generated.

L Cubed · Post by **L Cubed** » Sat Aug 16, 2014 4:55 pm

When I say "short" I mean that the input is grounded upon RCP startup. I'll try using the other I/O port to see if I get a different result. I just find it strange that if the RPM signal is plugged into the RCP when I turn it on, I don't get signal. Turn on the RCP without it plugged in and then plug in the RPM signal and everything works great.... I'm more of a mechanical guy so this confuses me.

Thanks for the support by the way, I really appreciate it and think the unit is great!

Post by **brentp** » Sat Aug 16, 2014 5:00 pm

Thanks for the clarification.

While your situation is peculiar, it should not matter if it's grounded or connected to 5v upon startup.

I'd try the PWM loopback test as described earlier, and also see if the other channels behave the same way.

I'll try to reproduce your scenario by first grounding upon startup, then switching to a reference signal (like the PWM output) to see if I can re-create what you're observing.

A 'scope trace of that signal would be ultimately helpful - and a logic buffer between the two could be an interesting test as well.

Let us know if you have any new observations!