Zeitronix Eca Wiring And Config

[Hanaldo]

Hi all,

I've posted this question in the Home Tuning Vipec and Link thread, but didn't get a response so hoping it would get a little more exposure in it's own thread. I've also posted the question on the Link G4 forum, and emailed the guys at Zeitronix, but so far haven't had a response. Sorry if this turns out to be a bit of a read, but I want to be sure I detail everything I have done.

I am having a bit of an issue with my Zeitronix ethanol content analyser and Continental flex fuel sensor. I have it all wired up and working, but the issue I am having is with trying to configure the input in the Link ECU. I just can't get the ECU and the Zeitronix display to read the same numbers.

I believe, at the moment, that the issue is somewhere in the way I have wired the setup. So to help clarify, here is a quick diagram I drew in paint (tried to embed, but the topic wouldn't post...)

http://www.linkecu.com/forums/G4Forum/941028398/689720745/zeitronix-eca-wiring-png

Basically, I have the Continental sensor wired to a switched 12v power source and earthed at the chassis. The installation manual says to wire it to the ECU power source, but I couldn't see a reason that was necessary so I just have it wired to an accessory 12v. Possible issue there?

Then, I have the signal output wire from the sensor running to the Zeitronix display in-cabin. From here, rather than using the analog signal output from the Zeitronix display, I simply spliced into the sensor signal wire and pinned it to the Digital Input 5 position on the Link XS loom. The reason I did this was because I want to keep the signal digital, because I don't have a free analog input for the ECU.

This all appears to work fine, until I try to configure the sensor in PCLink. I have selected the GM Siemens Ethanol Sensor option from the list, as per instruction from Link. Now if I turn the Internal Pull-up Resistor on, and have the 'Active Edge' setting set to 'Rising' then the fuel temperature reads 132 degrees celcius, which is clearly wrong. If I change the Active Edge setting to 'Falling' then the temp reads 18 degrees celcius. Which might be ok, except the Zeitronix display reads 12 degrees celcius. If I turn the Internal Pull-up Resistor off, and have the Active Edge set to Falling, then I go back to the 132 degrees celcius reading. Switch the Active Edge to Rising, and I get 23 degrees celcius, which still doesn't match the 12 degrees showing on the display. It should also be noted that I am not sure which reading is correct, I just know that they don't match up.

Now I am aware that the Zeitronix display has it's own internal pull-up, so I figured perhaps the display and the ECU are interacting with each other and messing up the reading. So I tried disconnecting the digital signal wire from the Zeitronix display, so it is only connected to the ECU. Like this, with the Internal Pull-up Resistor switched off in PCLink, I didn't get a signal from the sensor. With it on and the Active Edge set to Falling, I got the reading of 18 degrees celcius again.

So I believe the issue is that I have the ECU spliced into the digital signal wire, and the two units are confusing each other. Thing is, what can I do about it? I don't have a free analog input for the ECU, so I can't use the analog output from the display. Can I simply not use the Zeitronix display and have the ECU receive a digital signal? Or is there another way I can divide the signal so that both units function correctly?

Any help would be much appreciated, and apologies again for the lengthy post.

Cheers,

Martin.

[34GeeTeeTee]

[Hanaldo]

Thanks Mat

[34GeeTeeTee]

Thanks Mat

No worries, i know Adam changed the wiring config on mine last visit but he did it all so i know nothing and am unable to assist

[XKLABA]

what about a spear digital output from ECU, now I have no deers eyes about computers so I probably wrong, but could you send the signal to the ECU then send a signal from the ECU to the display

[34GeeTeeTee]

what about a spear digital output from ECU, now I have no deers eyes about computers so I probably wrong, but could you send the signal to the ECU then send a signal from the ECU to the display

Pretty sure this is what was just done to mine!

[Hanaldo]

Not a bad idea, except I don't think it can be done with the Link plug-in.

From what I can see, there is no digital output from the ECU that can simply relay a signal. There are auxiliary outputs, but these are used to control solenoids.

[scotty nm35]

Is there a spare injector output or something that can be used?

[Hanaldo]

There is, but it only outputs a user defineable PWM signal. There is no way to relay the input signal to the output.

[Hanaldo]

Ok, so to confuse matters further I just tried disconnecting the ECU from the circuit. So with the output signal going directly to the Zeitronix display, the display reads 13 degrees celcius. With the output signal going directly to the ECU, the ECU says the fuel temperature is 18 degrees.

So the two units are disagreeing on the reading, even when there is no interaction between them. I am thoroughly confused.

[GTSBoy]

There may simply be a difference between the way those two units interpret the fuel temp data out of the digital stream. The "calibration factors" are little more than values that someone has written into a config file. There is plenty of room for someone to have gotten it a bit wrong (probably on the ECU end rather than a gauge that is dedicated to reading these sorts of sensors) or for there to have been a firmware update (for want of a better word) in the sensor that has not been correctly followed up by the config in one of the instruments (again, you'd expect that to be the ECU rather than the Zeitronix box).

Why don't you see what happens when you put warm fuel in the sensor? Put a can of fuel into some hot water so you get it up to a decent temp like 40 or 50°C, tip some into the sensor and see if the gap between the instruments widens or narrows.

[Hanaldo]

There may simply be a difference between the way those two units interpret the fuel temp data out of the digital stream. The "calibration factors" are little more than values that someone has written into a config file. There is plenty of room for someone to have gotten it a bit wrong (probably on the ECU end rather than a gauge that is dedicated to reading these sorts of sensors) or for there to have been a firmware update (for want of a better word) in the sensor that has not been correctly followed up by the config in one of the instruments (again, you'd expect that to be the ECU rather than the Zeitronix box).

Why don't you see what happens when you put warm fuel in the sensor? Put a can of fuel into some hot water so you get it up to a decent temp like 40 or 50°C, tip some into the sensor and see if the gap between the instruments widens or narrows.

Yes, I think you're right. Especially given the Link software is using the GM Siemens sensor for calibration, perhaps the new Continental sensors have changed slightly and Link haven't accounted for this yet. But to be honest, I do think the ECU's reading seems more reasonable. The ambient temperature today is 21 degrees, so 18 degrees isn't far below that but 13 degrees is. I've actually got some temperature sensors that I can use to determine which one is correct, so I'll try that.

I was actually about to try reconnecting everything the way I had it, and then doing some data logging with the car up to temperature to see what the ECU says the fuel temp gets to vs the display readings. Worthwhile, I would I be better off trying your experiment?

[r31slpr]

Whats the part number for the sensor? do you have a data sheet? It will be easier to determine which reading is correct from that

[Hanaldo]

Part number is 13577394, but I don't have a data sheet for it.

[r31slpr]

I cant seem to find any datasheet for it, what ecu are you using? Is there any analog inputs left you can use?

You might be better off trying GTSboy's suggestion to find the right temp

[GTSBoy]

Sorry, I didn't get back to the question........whilst it may not be super quick and not mess or risk free, I would do my experiment......not least because you could measure the fuel temp with another thermometer at the same time and have a reasonable idea what the actual temp is. You have to keep in mind that you may want to keep the sensor warmed up in some way so that it doesn't suck the heat out of the fuel when you pour it in/through....but that's just details.

[Hanaldo]

I cant seem to find any datasheet for it, what ecu are you using? Is there any analog inputs left you can use?

You might be better off trying GTSboy's suggestion to find the right temp

Using a Link G4 plug-in. No analog inputs left unfortunately, need to use the digital signal.

I will try to find which temp is correct. Got a reply from Zeitronix, but whoever replied doesn't speak English very well. The guys at Link think it will be a case of each device being off, one reading a little low one reading a little high. I'm not so sure, but I'll figure out what the temp actually is and go from there.

[GTSBoy]

Pull your return line off the tank, into a can, measure that temp with thermometer, compare against 2 units' numbers, as the simplest approach.

[Hanaldo]

I might as well go and get some E85 and stick the sensor in it to see what the two units read. Then I can measure the ethanol content and temperature myself and compare it to see which unit is correct.

The fuel temperature itself doesn't actually worry me too much, I want to use it as an indicator but it's not critical. The ethanol content I need to make sure is spot on though.

[GTSBoy]

Yuh, I was assuming as much, and that your concern was that if the temperature was wrong it would be hard to trust the ethanol content number.