Wiring/Grounding Issues - Electrical Experts Inside - (Long Post Warning)

[dville_gt]

Quick run down. I am using HPTuners VCM Standard (no external logging inputs, i.e. no EOI plug for wideband etc).

To get around this everyone just uses a 5v sensor input from the PCM (I am using the fuel tank pressure sensor pin since in my swap this is not used by the PCM for anything useful).

Then you set HPTuners to log (in this case fuel tank pressure) as a wideband sensor (by giving it the right function, in this case it is (Volts/.3324 + 7.35)=afr as the range of the LC1 is 7.35-22.39 over a 5 volt range, thus each volt = 3.008 = inverse= .3324)

I assumed all was well with this setup and tuned accordingly getting my VE table almost perfect. Out of curiosity I pulled out an old laptop last night that has a serial port and plugged the LC1 directly in and ran Innovates "LogWorks" program to see how accurate my HPTuners reading was versus LogWorks (as LogWorks is supposed to always be dead nuts accurate as to what the sensor is reading since it is getting a digital reading versus the HPTuners analogue/mathematically generated reading).

Well, the readings were way off, with motor off, sensor in exhaust LogWorks reads 20.6 AFR (which should output 4.4 volts) while HPTuners shows 4.98 volts or 22.35 AFR. This is obviously very concerning as that is a 1.7 AFR offset in free (in exhaust pipe) air.

Now this is where it gets confusing. I have gone through and programmed the LC1 to output flat line voltage and HPTuners seems to be reading the voltage pretty much perfectly, if I tell it to output 1 volt across the board HPTuners sees 1.03v, at a commanded 5v it see's right around 5v. Yet Logworks still shows 20.6 free air while giving an analogue output of 4.98v (or 22.35).

So now I am real confused, the math is simple and it is right, the voltage it is basing it's calculation off of seems to be right, yet the two number do not jive (by a large margin).

So now I am starting to question the grounding situation, right now I have the odbii port (pin 4 & 5) grounded to a common chassis ground with the LC1 (which has three grounds that I combined at the lug, a system ground, analogue ground, and heater ground). Upon further reading it seems the heater ground should be separated to prevent pulses from the heating circuit, so I will do that tonight (although spikes are not what I am seeing here, this is a totally wrong number).

As far as the other LC1 grounds (system and analogue) I guess I should try to ground them at the same place the PCM is grounded since the PCM is the one getting the 5v signal input and comparing it against it's ground.

Does this sound right, could the difference between the chassis ground that the WB is currently hooked to and the PCM ground cause the PCM to read differently then the WB, and if so why would the PCM have the right flat line voltage when I command the WB to output a flat line output?

[dville_gt]

What do you guys think about this? Could this be the cause of my problem (still confused one why flat line voltage is coming through fine though)

Now assume the ECU is grounded to the engine block, but your LC-1s system/analog out ground is grounded to the frame. The LC-1 measures its output voltage referenced to its “ground”, as that is the only zero reference it has. The ECU measures its O2 input voltage against a different zero reference, namely the engine block. As the LC-1’s zero reference sits 1V higher than the ECU reference, the ECU will read 2V on it’s input. Of course, you could compensate by programming the LC-1, but that is not a very good solution because:

a) The LC-1 cannot output what it regards as “negative” voltages. It references everything to its ground, and therefore 0V is at low as it can go. 0 Volt would even then be read as 1V by the ECU/datalogger/display.
b) The 1V ground offset as described in the above example depends on the current in the return path. With a 0.1 Ohm resistance in the return path and 10A electrical current you will have 1V. With 20A current you would have 2V. With 2A electrical current you would have an offset of only 0.2V.
c) The current load in a car, and therefore in the return path, can change very dramatically from 2-3 Amps to 50-100 Amps in milliseconds due to ignition systems, fans, a/c clutch and so on. These changing currents will create corresponding changing ground offsets which then is often wrongly attributed to "noisy analog outs". This "noise" can often be seen when displays intended for NBO2 sensors are used. On LED instruments, multiple LEDs will light up at the same time.

A better way is to ground the LC-1 to the same reference point where the receiver of the data, wether its ECU, datalogger or display instrument is grounded. This way all devices are “on the same page” and ground differentials somewhere else have no effect.

This is also the reason the LC-1 has different grounds for heater (large current) and system/analog out ground. The potential ground offsets created by the heater current when they are connected together can create its own problems. By connecting the heater ground to the same grounding metal as the rest you will minimize that effect. Therefore the best way is to ground the LC-1’s heater to the same area as the other grounds, but on a seperate bolt/lug. The contact points of bolts and lugs can account for 75-80% of the resistance of the connection and can change dramatically when corrosion sets in.

[forbes]

have you talked to justin from blackbear performance? he's really cool and offers alot of help. i also have buddy in your area i am trying to get a hold who is really knowledgeable

[Strychnine]

You said you took the 5v supply from the tank pressure... what did you do w/ the ecm return (sensor ground back to ecm)?
How are you grounding the O2 sensor? Assuming this is a 3 wire sensor (supply, signal, return) I would have the supply and return both tied to the ECM (ECM should have a common sensor ground location) then do whatever you need to do w/ the signal as opposed to maybe getting your 5v from the ecm but then grounding it elsewhere.




edit: i typed this up, walked away for a bit and when i finally hit submit you had already posted something mentioning this.

[dville_gt]

have you talked to justin from blackbear performance? he's really cool and offers alot of help. i also have buddy in your area i am trying to get a hold who is really knowledgeable

Thanks, I am really leaning towards the issue being that the LC1 is grounded at the chassis and the PCM is grounded at the engine block. Since voltage is calculated by comparing the difference between two points, if the grounds are different then the LC1's "zero" or ground is different then PCM's "zero". Since the LC1 is feeding the PCM the 5v signal it is comparing it's reading against the chassis ground and the PCM is comparing the signal generated by the LC1 against it's engine ground.

All of this makes great sense to me, but it still doesn't explain why I am seeing accurate numbers when I flat line the voltage out of the LC1 controller. I can tell the LC1 to output 1v no matter what the AFR reading and HPTuners is seeing the 1v input. If the problem is truly the different "zero" or ground wouldn't the voltage the PCM see's be inaccurate?

[dville_gt]

You said you took the 5v supply from the tank pressure... what did you do w/ the ecm return (sensor ground back to ecm)?
How are you grounding the O2 sensor? Assuming this is a 3 wire sensor (supply, signal, return) I would have the supply and return both tied to the ECM (ECM should have a common sensor ground location) then do whatever you need to do w/ the signal as opposed to maybe getting your 5v from the ecm but then grounding it elsewhere.




edit: i typed this up, walked away for a bit and when i finally hit submit you had already posted something mentioning this.

The sensor is a typical 3 wire setup, 5v reference wire, signal wire, and return/ground. I looked up the pinout and am a little confused. The gray wire is a 5v reference, I don't use this as the LC1 outputs a 0-5v signal via the wideband controller, the signal wire is the only wire I am using right now, the brown wire from the lc1 to the signal wire in the PCM, then there is the return wire.

On the pinout the return wire comes out of the sensor and pairs up with a "REF LOW" wire on the PCM (see illustration below, these are the two tan wires). Then I am not sure what is going on in the fuel pump sender assembly, but it looks like there is a PPL wire that comes into the sender as well which is a HO2S LOW CTRL wire.



I can't really tell where the sensor grounds at, is it grounding at the fuel pump sender or is grounding through the other tan wire (the "REF LOW") that it T's into before going to the sender? If so I guess I could run a wire to that pin on the PCM and ground it to the same spot as the PCM and the LC1 to prevent any possible grounding reference issues?

[dville_gt]

And the plot thickens.


Just got to reading and apparently if the AFR is over a certain point LogWorks goes to %o2 versus AFR and %o2 does not equal AFR. Thus I am at 20.6 %o2 and the Wideband is maxed out at 22.35afr (i.e. 4.98 volts).

So I am going to change the grounds around as well to clean up the small ground offset I am seeing then see if LogWorks will start reading AFR with the engine running so I can compare it to HPTuners.

[Andy@PrimeTuning]

This thread should help

http://www.hptuners.com/forum/showthread.php?t=27855

[dville_gt]

This thread should help

http://www.hptuners.com/forum/showthread.php?t=27855

I appreciate the link but the OP in that thread and my problem were totally different.

My problem ended up being that LogWorks does not show AFR when the ratio is over something like 50:1 air to fuel. Instead of showing some crazy high number it reverts to showing % oxygen, which in free air happens to correspondwith my 20.9% reading. The analogue output to HPTuners however is just given 5 volts since that is the highest output it can give and in turn HPTuners (based on the PID formula I have given it calculates it as 22.39, aka the value @ 5 volts).

The small ground offset I was experiencing (.02 volts) was really throwing me off, had it been a full 5 volts I would have been able to better understand that this was just a maximum reading and the true value is much higher. I had no idea that the LogWorks program reverts to %o2 so it just looked to me like 2 random AFR numbers that do no correspond. In fact they do, the LC1 was probably seeing in excess of 50:1 afr, reverted to %o2 (20.9%) and outputted 5v (max output) to the analogue out (HPTuners)