Graphics interface for the 14CUX

Graphics interface for the 14CUX

Author
Discussion

leclance

3 posts

103 months

Tuesday 23rd October 2012
quotequote all
Hi mark,

The second i work out how to PM on this forum ill send you a PM about getting one of your packs. I wont mention its for a landrover in the section of the forum though :-)


DarkMatter

1,440 posts

196 months

Tuesday 23rd October 2012
quotequote all
leclance said:
Hi mark,

The second i work out how to PM on this forum ill send you a PM about getting one of your packs. I wont mention its for a landrover in the section of the forum though :-)
There's a contact number in this advert... http://classifieds.pistonheads.com/classifieds/par...

leclance

3 posts

103 months

Wednesday 24th October 2012
quotequote all
great thanks for that ill drop him a call

taylormj4

1,547 posts

231 months

Saturday 3rd November 2012
quotequote all
OK so finally managed to get the time and weather to try out this great software. First off, thanks goes to Colin and Dan for the excellent software and to Mark for the well-built lead, all of which worked perfectly.

I have some results and comments on my first test that you may be interested in:

First observations
When you turn IGN on to check initial sensor outputs (engine off), you have to make sure the immobiliser is disarmed otherwise the ECU doesn't talk to the software - had me there for a few minutes.

My LUCAS TTS connector did not have a grounding plug in it. Is that a problem ? Will it cause noise or floating value problems for the ECU ? If yes, where the heck do I get one of those or anyone know what pins should be shorted together or grounded ?

Results:
IGN ON, engine off:
Error code 19 "Throttle pot lo/ MAF hi" logged
Throttle posn (abs) = 0% (closed) then after operation of throttle 4%
Throttle posn (corr) = 0% (closed) to 94% (full open)
Target idle=1184
Idle bypass pson=100%
Eng and fuel temp both=6degC
MAF=0%
Fuel map 5 = 5 (correct as CAT car), adjustment factor=0x6784

I successfully cleared the Error code 19. The "ding" made me jump !

Engine start (left to idle)
MAF=7%
TPS=0% (corr) or 5% (abs)
Idle bypass=42%
Revs varying but average=1110
Fuel map point [56]

Eng temp reaches 31degC (still idling)
Lambda sensors start to read (short term)
Lambda left -5% to +12% then up to +26%
Lambda right -8% to +14%

Then the long-term Lambda readings gave some odd results, is this the software/ECU chat or something that is wrong. I'd get an initial reading on toggling to "long term" for less than a second and then a new stable reading would appear, as follows:
Initially:
Lambda left (long term) = +21
Lambda right (long temr) = +2%
Then after less than a second, they would change to:
Lambda left (long term) = -100%
Lambda right (long temr) = -100%
These readings were then stable. The above could be repeated by toggling back to short-term and then back to long term.
Does anyone else get that? At this point, I was thinking "-100%" on the long-term map amendment isn't going to give good fueling but this changed as the car warmed up.

Engine temp now 87deg (still idling)
Fuel temp=10degC
Rev=900rpm=target
Main V=13.9V
Lambda L long term = -5% initially, then after less than 1sec (as above) switches to -30%
Lambda R long term = -10% initially, then after less than 1sec (as above) switches to -39%
MAF(direct)=32% (closed throttle) - Mark this equals your figure for 3.5L engine but mine's a 4.5, problem or still within tolerance?

Eng temp=93degC (still only idling)
Fuel temp=13degC
Idle=900=target
MAF(dir)=32% (closed throttle)
Idle bypass=16%
Now the lambda short term readings seemed to get stuck. Mark you said this may happen at high rpm as the ECU is busy fueling but the below happened at idle:
Lambda short term fluctuates then gets stuck at +1% (L) and -3% (R), seems it is not updating. A short blip of the throttle causes readings to fluctuate again, which continues for a few secs before getting stuck again at a constant value. Problem or normal here ? This didn't seem to occur when engine was cooler.
Lambda long term readings are now changed (modified as engine warm and at idle I presume from our discussions Mark?) to -30% (L) and -48% (R)
Fuel map point [35]

After a few bursts of high rev running
Eng temp=93degC (cooling system working then !)
Fuel temp=18degC (it is a cold day and bonnet open but would you expect this to maybe be a little higher?)
MAF(dir)=31%
Idle bypass=13%
Lambda short term: -9% to +15% (L) and -6% to +19% (R)
Lambda long term (L): -5% instantaneously (see above comment) then -28%
Lambda long term (R): +9% instantaneously (see above comment) then -33%
TPS (corr)=0% or 5%(absolute)
Idle bypass=13%
Idle target=900, actual is 900 or 950(fans on)
Fuel map point [35]
Volts 13.6V

On a quick run
Fuel map point seen to drop down map on acceleration. Got it down to bottom of map on full throttle in high gear up a hill. Point moves right on rev increase.
Speed sensor shows 0 and the 35-36 once speed up to jogging speed, then drops to 0 as drop to a walking pace - seems correct for TVR's mashed system

Checked fault codes and all still clear, Fault 19 has not re-appeared.

Conclusions
My conclusions were:
1. Maybe the TPS is faulty due to the fault code and the 0% reading (according to Mark's notes). I know they can be slotted and adjusted but I thought this was pointless on a later car as the ECU learnt the resistance range. Is that correct?

2. The long term lambda trim values are both significantly negative, which causes the short term lambda correction to be mainly +ve. It may be worth trying to move the AFM trim pot to bring the long-term values closer to 0. Is that right ?






blitzracing

Original Poster:

6,206 posts

185 months

Sunday 4th November 2012
quotequote all
I think the strange issue with the long term trim dropping to zero's down to engine temp- Ive seen this before, and the engine needs to be in the low 90's to get a sensible reading- My G33 runs at around 85'c and the long term never gets set at all. Not that I care as I don,t run the catalyst map other than to experiment with...

The throttle pot is self adjusting to an extent and is adaptable within a range of 80 to 500 mV closed throttle. Within this range, the ECU will adapt to the initial setting and use it as a reference.

If the TPS should fail, the ECM will use a default value of 576 mV and the MIL will be illuminated. A diagnostic trouble code (17) is set when sensor output is less than 78 mV for longer than 160 milliseconds Throttle Position

So id suspect at some point the TP has gone open circuit at some point with the ignition on, but as there are no time stamps on the error code you wont know when this happened- if its cleared now, and the readings are smooth, Id not worry about it.

Ive put some extra manuals on the CD on fault code generation.

The MAF readings are highly variable - The values Ive used are the figures Mark Adams uses as test examples simply shown in percentage terms, but engines with a different cam, idle speeds and general engine condition will cause the values to be outside the range, the AFM is very sensitive at low air flows, so small airflow changes give big voltage changes- but as it says in the notes it needs to be a long way out from the ball park figures to show and AFM failure

Im not sure whats going on with the short term lambda readings when they freeze- but dont forget this is a correction value the ECU is applying- not the lambda voltages themselves- so you would need to observe what Rovergauge is doing along side the physical voltages to see whats really going on. If it is the ECU overloading (seems unlikely at idle- are the other values still updating?) , try reducing the scan rate RoverGauge asks for the data to reduce the load on the ECU- its called poll interval under options and its set for .5 of a second (500 millisecnds by default) so you could try 1000 milliseconds.

I check the grounding plug wiring, but the cars seem to run OK without them (so far at least??).


Edited by blitzracing on Sunday 4th November 10:15

Pink_Floyd

894 posts

186 months

Sunday 4th November 2012
quotequote all
taylormj4 said:
OK so finally managed to get the time and weather to try out this great software. First off, thanks goes to Colin and Dan for the excellent software and to Mark for the well-built lead, all of which worked perfectly.
I have some results and comments on my first test that you may be interested in:
Hi taylormj4. This is what mine looks like on longterm. If I reset the ecu the long term goes to 0 and once the engine is hot it moves to -100% over a perion of a couple of mins. Short term is -10% to -30%. I checked the lambda signal with Marks lambda box and it is switcheng as expected. I still have a problem, at tickover the engine never exceeds 900rmp no mater how hot or cold the engine is.
Any way here is a pic of the gauge.

cmb

101 posts

140 months

Tuesday 6th November 2012
quotequote all
There's some possibility that the -100% reading on long term trim is due to a bug in my software. (Or a bug in Qt, the software toolkit that I use. It wouldn't be the first time.)

I'll investigate this to see if I can reproduce the problem here.

--Colin

taylormj4

1,547 posts

231 months

Tuesday 6th November 2012
quotequote all
cmb said:
There's some possibility that the -100% reading on long term trim is due to a bug in my software. (Or a bug in Qt, the software toolkit that I use. It wouldn't be the first time.)

I'll investigate this to see if I can reproduce the problem here.

--Colin
Colin,
This is fantastic support and nice to see, thanks very much.

Mark,
Thanks for your comments. What did you think on the idea of adjusting the MAF pot to see if I can pull the long term closer to zero ?

You mentioned error code 17 but I was getting 19. Was that a typo or is the fact that I didn't get 17 & MIL light a diagnostic that shows something ?

Pink Floyd,
Mine showed -100% until the engine temp reached around 87degC. Yours doesn't seem to recognise that condition.

Edited by taylormj4 on Tuesday 6th November 12:34

Pink_Floyd

894 posts

186 months

Tuesday 6th November 2012
quotequote all
taylormj4 said:
Pink Floyd,
Mine showed -100% until the engine temp reached around 87degC. Yours doesn't seem to recognise that condition.
If I reset the ecu it shows 0 up untill the engine gets to 90degC and then it slowly moves to -100%. at one point it did show Left -100% and right -89% for some time, 10 mins plus, but it now seems to be -100% for both.

taylormj4

1,547 posts

231 months

Tuesday 6th November 2012
quotequote all
Pink_Floyd said:
If I reset the ecu it shows 0 up untill the engine gets to 90degC and then it slowly moves to -100%. at one point it did show Left -100% and right -89% for some time, 10 mins plus, but it now seems to be -100% for both.
From what I understand from Mark, that could mean that your base correction is at maximum negative so the short term corrections have to be high in order to bring it back to where it should be. I think there was a possibility that you could change this with the pot on the MAF. I'm going to try changing mine whilst monitoring with Rovergauge to see if anything changes.

Do you have any running problems Pink Floyd ?

Pink_Floyd

894 posts

186 months

Tuesday 6th November 2012
quotequote all
taylormj4 said:
Do you have any running problems Pink Floyd ?
Yes it runs a bit rough up to 2200rpm amd then it pulls like a train and very smooth.

taylormj4

1,547 posts

231 months

Wednesday 7th November 2012
quotequote all
Hi Mark/Colin/Dan,
I understand that the lambda readings are the corrections that the ECU is applying rather than the lambda readings, so 2 questions:

1. The software would only flag a lambda sensor fault if the ECU flagged it as a fault code, correct ?

2. Mark I think you said that if the long term lambda set was way out (max +ve or -ve) then it may not be possible for the short term correction to counter it. But isn't it just a calculation that the ECU is doing so could correct as much as necessary ?

blitzracing

Original Poster:

6,206 posts

185 months

Wednesday 7th November 2012
quotequote all
The ECU only has a range of mixture adjustment available to correct a fueling error, so in the case of long term trim Id expect the ECU to pick up that it has make adjustments in one direction more than the other on the short term to get the switching of the probes even (Its mark space ratio of the waveform). The probes should have an equal time spent "lean" and "rich" at a perfect fueling point, but if its more one way than the other, the long term level can be moved to allow the short term to then cycle evenly again. So you could get a case that the long term has hit its limit, but the short term can still shift the mixture enough to make the probe switch, it just wont be even. In the real world the switching signal is pretty ragged at its best, its very difficult to visually see the switching ratio. I have managed however to replicate various conditions and waveforms are here, 1/3 down the page:

http://www.g33.co.uk/fuel_injection.htm


The ECU can make a very large fueling correction (I think its around 20%) so if its hitting the stops on either trim, something is significantly wrong to cause it to do this.

As for the TP errors:

17 is a code for a failed test during power on (wrong setting or open circuit) and 19 is low throttle pot output with high AFM output- this is a contradiction in terms, so either the AFM output has glitched high or the throttle pot has gone open circuit when the airflow his high.


As to the question of faults- yes RoverGauge will only show faults as the ECU reports them. In the case of Lambda switching the ECU is not very sensitive to a fault conditions, it really takes the probe to stop switching for a period to throw a code- not simply if the fuelling is a bit "out".

Edited by blitzracing on Wednesday 7th November 18:37

leclance

3 posts

103 months

Saturday 10th November 2012
quotequote all
Thanks for the sofwtare and the lead , nice quick delivery and worth every penny. I have only started to undertand what the rovergauge is showing me and would have probably saved myself a few quid on parts had i bought it earlier.

My trouble started with a a breakdown when i first bought my landrove 3.9 efi v8 with lpg. The engine started flucktuating the revs wildy after a long journey which caused me to stall at a set of traffic lights (bit embarresing ) could get it started , the RAC guy unplugged the M.A.F and it ran fine but as i found later with 11% co reading.

I purpchased a new MAF - this instantly blew a 20a fuse feeding teh maf and fuel system. took it back and got a second one which did the same.

Now i have rovergauge the (original) M.A.F plugged in from cold gives a ready about 20-30% linear on tickover which suggests the m.a.f is ok .

Im going to do some investigation today but wondered if anyone had a similar problem or has any suggestions . most advice i have had so far points to replaceing the 14cux controller but i wanted to confirm this before trying to find one.

Runs fine for a while then the tickover runs wild almost stalling and reving to about 1k trying not to stall . Trying to rev the engine does nothing until i unplug the M.A.F then it settles and runs ok with high fuel use and high co.

any ideas ?
oh and is this in teh rght post or should a start a new one :-)

blitzracing

Original Poster:

6,206 posts

185 months

Saturday 10th November 2012
quotequote all
Pulling the AFM off will cause the ECU to go into limp home mode fuel map, so will give you a completed false result as to the cause. In limp home the AFM output will be ignored completely and it fuels on Throttle pot and RPM only- You need to check the map number your are running with the AFM in place, check for fault codes, and look at the Lambda values to see what the fueling doing.

danbourassa

242 posts

102 months

Saturday 10th November 2012
quotequote all
Hi, I'm Dan, the other half of the RoverGauge team here in New Hampshire, USA. A lovely place to live except for the fact that we can't get Cerberas. I'd like to share some information I learned recently about the throttle pot.

First, unfortunately, a little math is necessary. The code deals with a count reading from a 10-bit analog to digital conversion. If the throttle pot is opened completely, 5 volts will give the full count of 1023 (or 3FF in hex), so each count is worth 5.00/1023 or 4.89 millivolts per bit (by the way, this is also the same mv/bit as the MAF).

In older 14CUX code (TVR included), if the throttle pot reading dips below 16 counts (16 * 4.89 = 78 mv) a fault code 17 will result, as Mark mentioned earlier. However, this threshold was lowered to 8 counts (39 mv) in later "desensitized" versions of Land Rover code, so the 16-count threshold apparently had a problem with false triggering. Also, it seems that later Land Rover code changes were not included in TVR code since TVR code had already been branched off.

A false triggered code 17 is annoying, of course, since it latches the default value of 118 (this is where the 576 mv value comes from) and causes the software to ignore the TP and run in a degraded mode.

According to Land Rover tune R3360A, there are actually 3 possible TP related faults:

17 TP falls below min threshold
18 TP high with MAF low (TP must be greated than 307 counts or 1.5V, TVR and older LR code)
19 TP low with MAF high (TP must be less than 205 counts or 1.0 volt)

On newer LR code such as R3360A, the fault code 18 threshold was desensitized to 818 counts (4.0 volts) so it seems that this was another nuisance fault. It also appears that code 18 is handled similarly to codes 02 and 03 in that it gets memorized but does not set the MIL.

(Note: edited 133 to 307, hex vs decimal)

Edited by danbourassa on Sunday 11th November 20:31

MuffDaddy

1,181 posts

170 months

Saturday 17th November 2012
quotequote all
First off I have to declare my connection with BlackBerry.

Their next gen phone due next year runs QT and has micro USB host/slave. This would look awesome on these. I am happy to provide test a test device and try to gain help from our development guys and QNX.

Who would be interested?

MuffDaddy

1,181 posts

170 months

Saturday 17th November 2012
quotequote all
First off I have to declare my connection with BlackBerry.

Their next gen phone due next year runs QT and has micro USB host/slave. This would look awesome on these. I am happy to provide test a test device and try to gain help from our development guys and QNX.

Who would be interested?

Pink_Floyd

894 posts

186 months

Sunday 18th November 2012
quotequote all
blitzracing said:
Pulling the AFM off will cause the ECU to go into limp home mode fuel map, so will give you a completed false result as to the cause. In limp home the AFM output will be ignored completely and it fuels on Throttle pot and RPM only- You need to check the map number your are running with the AFM in place, check for fault codes, and look at the Lambda values to see what the fueling doing.
I tried unplugging the AFM and the engine died Immediately.

taylormj4

1,547 posts

231 months

Monday 19th November 2012
quotequote all
I had error code 19 logged and have been wondering what would cause that. This is MAF hi and TPS low, so could that mean that I have an air leak somewhere or simply that the TPS went low at some point when it was on open throttle (maybe a poor connection or wiring fault).

Waiting to see if the fault comes back at the moment.