Page 1 of 2

Can't datalog PLX wideband below 1.2V

Posted: Sun Feb 07, 2010 12:50 pm
by Honda9ower
I have a brand new PLX SM-AFR wideband setup with the supplied Bosch sensor. I'm connecting the 0-5V analogue output to the ELD input on a Hondata Acura RSX ECU and using KManager to log the voltage and convert to AF. I'm using the supplied noise reduction capacitor between ELD and ground as recommended by PLX.

I bought the PLX as it was the manufacturer Hondata recommended and I wanted to log as low as 10:1 AFR and the stock Honda wideband only goes to about 11.5:1 AFR.

I set the calibration to PLX in KManager (0 to 5V is 10 to 20AFR) and using the stock wideband calculated that there is a -0.4V offset with the PLX wideband controller.

Everything seemed to work fine with the PLX agreeing with the stock sensor until I did a few full throttle runs and found that the logged ELD voltage never went lower than 1.2V, so I could not record any low AFRs.

After searching the PLX forum I fitted a 2kOhm pull down resistor on the ELD input, but that only reduced the minimum recorded voltage to 1.15V. I have a nitrous setup on the car, so with the bottle closed it's very easy to dump in lots of extra fuel to see the effect on the voltage, and even when I do this the output 'flatlines' at 1.15V.

With just the pulldown resistor connected, the ELD logs a constant 3.1V and with nothing at all connected it logs a constant 4.55V.

When I plug in a DM-5 gauge instead of the ECU, the gauge reads down as low as 10:1 when running rich, so at least something works as expected.

The PLX controler is mounted under the dash, on a live feed direct from the battery (via a fuse and relay) and earthed to a earth point next to the ECU.

Any ideas what's wrong?

Posted: Sun Feb 07, 2010 1:41 pm
by Honda9ower
A bit more info: The ELD wire has been cut so the only thing this ECU pin is connected to is the 0-5V analogue signal from the PLX (and 0.1uF capacitor to ground).

In KManager "sensor setup" I have ELD display min set at 0 and max at 5. AFR display min set at 10 and max at 20. I have the same values setup on the graph templates too.

Posted: Mon Feb 08, 2010 10:43 am
by Spunkster
Datalogs and calibrations showing the problem.

Posted: Mon Feb 08, 2010 11:24 am
by Honda9ower
Datalog and calibration attached. A good example of the ELD 'flatlining' at 1.15v is between 49 and 51 seconds in the datalog, where the N20 is on.

I have other logs on WOT without N20 armed, but they don't have quite as much fuel added, so you might almost think that the steady AF of 11.5:1 was real.

Posted: Mon Feb 08, 2010 4:09 pm
by Spunkster
Try grounding the PLX at the valve cover. There should be no capacitors or resistors needed. I have tested the input of the ECU on the bench and it can easily read from 10-20 AF using the PLX input voltage ranges.

Posted: Tue Feb 09, 2010 12:29 pm
by Honda9ower
I moved the earth to the valve cover and removed the capacitor. I set the nitrous output to dump in the extra fuel at idle, so I could test it from a stationary vehicle.The ELD voltage now flatlines at 1.26V, so no real change.

Probing the power voltage to the PLX controler its 13.5V which I assume is normal.

Posted: Tue Feb 09, 2010 2:49 pm
by JustinC
Honda9ower wrote:I moved the earth to the valve cover and removed the capacitor. I set the nitrous output to dump in the extra fuel at idle, so I could test it from a stationary vehicle.The ELD voltage now flatlines at 1.26V, so no real change.

Probing the power voltage to the PLX controler its 13.5V which I assume is normal.
Just out of curiousity did you put a multimeter on the 5v output from the PLX device and get below 1 volt?

Posted: Tue Feb 09, 2010 3:05 pm
by Honda9ower
Not yet, but I will try that later when I get another pair of hands to help out.

Posted: Wed Feb 10, 2010 2:05 pm
by Honda9ower
I have just buzzed the PLX 0 to 5V output, and when not connected to Kpro it reads anywhere between 0V (nitrous fuel dumped in) and 5V (clean air), so appears to be working fine.

When connected to the Kpro ELD pin it reads between 1.3V (nitrous fuel dumped in) and 5V (clean air), and agrees with what KManager logs.

Looks like there is something in the Kpro that won't let the ELD voltage drop below 1.3V.

Any ideas?

Posted: Thu Feb 11, 2010 9:58 am
by JustinC
Honda9ower wrote:I have just buzzed the PLX 0 to 5V output, and when not connected to Kpro it reads anywhere between 0V (nitrous fuel dumped in) and 5V (clean air), so appears to be working fine.

When connected to the Kpro ELD pin it reads between 1.3V (nitrous fuel dumped in) and 5V (clean air), and agrees with what KManager logs.

Looks like there is something in the Kpro that won't let the ELD voltage drop below 1.3V.

Any ideas?
No, I would hope hondata could shed some light since i have a friend with S2000 kpro having this exact same problem.

Posted: Thu Feb 11, 2010 5:14 pm
by Spunkster
Have you tried connecting the analog ground from the PLX to the valve cover? If that does not help, I would recommend contacting PLX Devices.

Posted: Thu Feb 11, 2010 6:22 pm
by JustinC
Spunkster wrote:Have you tried connecting the analog ground from the PLX to the valve cover? If that does not help, I would recommend contacting PLX Devices.
Come on, that is a crap answer, he put a multimeter on the analog out put and got 0-5v as he should, he moved the ground to the valve cover like he said and still the same result. There is obviously something wrong here that needs to be addressed.

This is happening on multiple KPRO and multiple PLX outputs, obviously if he is seeing less than 1.26 volts on the multimeter then the KPRO should see it as well.

Posted: Fri Feb 12, 2010 8:14 am
by Honda9ower
Spunkster wrote:Have you tried connecting the analog ground from the PLX to the valve cover? If that does not help, I would recommend contacting PLX Devices.
I have just tried that, and it makes very little difference, flatlines at 1.27V now. I have played about with using about half a dozen different earth points for both the power and analog ground to see what happens - battery nagative, valve cover, ECU case, dash common earth and a few more. I see about 0.03V difference in the datalogs between the best and worst.

I have contacted PLX on their forum, and they gave me some tests to run which have all worked fine. I think in their eyes the sensor and controler are working fine, as the gauge reads correctly, and the 0 to 5V analog outputs between 0 and 5V from 10 to 20 AFR as required.

IMO there is something in the Kpro stopping ECU ELD from dropping below 1.27V.

Posted: Fri Feb 12, 2010 1:08 pm
by Spunkster
It is not the K-Pro, it is the PLX that does not output enough current for the ECU to read properly.

Try removing the resistors from locations R300 and R301 as seen in the picture below.

Image

Posted: Fri Feb 12, 2010 2:12 pm
by Honda9ower
Just to be absolutely sure before I undertake a potentially risky soldering job, is it all 4 resistors that need to be removed? Do they need to be removed and left "open circuit" or shorted out?