** MegaSquirting a V8 - "Basics of How to" ... The A-Z Saga

[Bowie69]

I thought that the GEGO % was taken into account by MLV? Or have I missed a vital point -quite possibly tbh :)

Will go logging and see what happens, maybe even start from the 3.9 map again, rather than keep going with the clearly arsed up one above....

Ta Nige

P.s. you read all that damned quickly

[Belgian_Dude]

As requested by Frederic, I've uploaded my pictures. Looking through them, they're not that interesting, just ... a lot of funny coloured wires Didn't exactly sort or select them, as bandwith nor webspace are really limited, but my time is

on this end, gave proper tuning a first go tonight (after reading through the msextra tuning guide), trying to get AC - ASE - idle VE right, but honestly, I made it all worse So, next up is more reading, then some more, and then another try

Greetz

Quentin

[Belgian_Dude]

Hi guys,

Another thing I forgot last night: ignition timing;

I've finetuned my trigger wheel placement, and limb-home mode is now 10 deg BTDC dead-on.

So, I've set the trim angle in MT Spark Settings from the original 8 to 0, and now finally my timing is on the strobe what in reads in megatune.

BUT, I would like to confirm, with the below Spark Settings and Spark Map 1, and the trigger wheel at it's theoretically perfect location (I'm pretty sure I'm within 1deg of 10BTDC on limb home) am I in the right ballpark?

Second question, I'm still puzzled about CLT and IAT calibration values; Aren't these saved in the ECU's ROM? Why do I run Megatune, read 25°C (the temp I was expecting given the previously observed discrepancies real temp / display temp) for both CLT and IAT (same sensor), then either open an MSQ or connect the PC to the ECU or something like that (imho unrelated to sensor calibration) and suddenly, the megatune will suddenly display a correct 10-15°C for only one of the temps (clt in this case)

I've had similar things happening with TPS calibration btw ...

So, I guess my question, broken down, would be: does megatune recieve from ECU either "CLT/IAT/TPS is X°C", or "CLT/IAT/TPS is RAW VALUE Y" and proceed to match it through a table somewhere in C:\program files\megatune\... ??

thx

Quentin

edit: right, MS 1 on v3 board, Extra firmware MSnS-extra format 029q

[FridgeFreezer]

Quentin,

I think you need thermfactor.inc in the project folder of the car (eg "car1"), but you may want to check the MegaTune documentation as it's obviously getting it from somewhere.

[Hybrid_From_Hell]

Spark Settings and Spark table look about right

As to sensors, either reset the settings within the ECU so that the sensors work, or easier put in the right sensors

What firmware are you using 029V ? and What / where are the sesnors from ?

To reset you'll need to have the correct settings and values at various points when resetting the ECU

Nige

[V8 EFI]

quote "Frederick, it seems you are confused about what your ECU is and how it works".

Hitting the nail on the head thats for sure !.

I've set the fuel overrun setting to off, so it should be something else ?.

Is it possible to post my msq together with all the needed info for discussion or advice ?. Maybe i'm overlooking something realy stupid.

I've attached two files but one of them keeps pending, might try again for the second picture.

ps Quentin i live near by Groningen in the north.

Frederik.

[V8 EFI]

[i've attached two files but one of them keeps pending, might try again for the second picture.

Another tryout i hope this time with a picture

Frederik.

[Belgian_Dude]

Ok, full day of logging - tuning - testing behind me, head is spinning.

strange **** happening;

Over the past weeks, I'd already noticed, MLV just wants to do one thing and that's keep squirting in more fuel, esp. everything (W)OT (90kPa and up)

As i started from what I thought was a rough map (sorry John ), I let MLV go ahead. But after every log, it would up the fueling by pretty large amounts (10 "units" and up). This morning, I started off with a map having an idle of around 40-45, and WOT values of >150; Surely, that couldn't be right.

Also, I didn't drive / pick up as well anymore, but that could've been my ignition ajustments.

Anyway, I started back from Fridge's "popular defaults" VE map.

Ignition is same said "popular defaults" map, but with 0 "trim angle" and a modified idle range - didn't run properly below 15°, but that's because I started at 10 which already was more pre-ignition than the sweet spot: 6-7 degrees BTDC.

With this setup (basically the Popular Defaults + slight changes to spark at idle), she runs much, much better.

so, I did a proper log, and all MLV wants to do is up the fuelling again ... But, I know that's not right because if I keep doing what MLV suggests I end up with fuelling of 150+ (at which point it still wanted to up some more!) Even when I set it to Very Hard for the whole map, WOT regions goes up by 8 points average ...

Questions:

1)Why oh why does MLV want to keep upping these values (say 1500-4000rpm, 95-100kPa), even though in every single occurence in my logs, they're already rich! (lambda volts ~~0,7V) I don't get it ...

2) can I enable to AFR target tables without a WB Lambda? I want to avoid the EGO correction aiming for 14,7 in areas I want more ...

Answers:

1) IAT sensor reading in Megatune is dependant on matfactor.inc, should be next to thermfactor.inc ;-) Nige, both sensors (CLT and IAT) are RR standard CLT sensors perhaps not ideal in temp changing speed for the IAT, but imho that shouldn't be too important, and this way I've got a spare for the CLT which is kinda important

Ok, going home now, I'll re-read this later to see if this makes any sense ;-)

grtz

Q

[Hybrid_From_Hell]

OK

CLT = RR + Fine

Are you saying your using a RR CLT for MAT as well - if so thats bnot good

now

Open up your log file (the blahblahblah.xls one) and look at the column that is 02

report back with the readings, they should NOT all be zeros, but if they are

then thats yer problem, but as with all things,.... lets go 1 at a time

Nige

[Belgian_Dude]

hi mate,

Ok, input accepted for the CLT as IAT, but was what I had at the time. Will order one from a 3,9

About the log file, no it doesn't display all zero's, I've got proper O2 feedback visible in MLV (see screenshot attached)

(zero's until warmup of lambda sensor of course)

Thank for helping me understand

Grtz

Q

[Hybrid_From_Hell]

You need a MAT Sensor to suit MS ECU a RR one will not do

Either that or get another proper MAT Sensor, with known settings and values and reset the ECU to suit it

With the wrong (as you have) MAT Sensor it prob thinks the engine is running in siberia and thus is

adding more fuel, if you are desperte I do have a brand new sensor with the right settings that aviods having

to recalibare the ECU

Without having at the least the basics right any sort of tuning is prob doomed

Nige

[Belgian_Dude]

You need a MAT Sensor to suit MS ECU a RR one will not do

Either that or get another proper MAT Sensor, with known settings and values and reset the ECU to suit it

With the wrong (as you have) MAT Sensor it prob thinks the engine is running in siberia and thus is

adding more fuel, if you are desperte I do have a brand new sensor with the right settings that aviods having

to recalibare the ECU

Without having at the least the basics right any sort of tuning is prob doomed

Nige

Nige,

OK, pretty stupid shortcut I took there.

I'm going to try and recalibrate ECU, as that's something I can do with the stuff I've got here.

But, tbh, during buildup, I tried to get a correct MAT/IAT sensor, but all sensors I could find off a rangey are incorporated in the hotwire MAF ... would you have a part number plz? (I know there's a separate one in our 4,0 GEMS, but can't find the p/n in the microcat)

thx,

Quentin

[Belgian_Dude]

ok, found it at last; ERR2946 from p38a GEMS

But still don't really get it; I'll be comparing the resistances of both sensors asap.

(this is something I've been confused about all along; I've been reading about using the default Range Rover sensors all over this thread, but fact is, this "default range rover sensor" is from a p38a, none of the older (flapper 3,5, hotwire 3,9) have them (well, not separate from the MAF), and they're the kind of engines that get MS'd (nobody does a p38 ;-) ) so how is this still a "default" or "standard" sensor ?? )

cheers

Q

ps: would you prefer me to start me own thread?

[Bowie69]

Right, I went out with controller authority down to 5%, set of new plugs as the old ones were beginning to foul, swapped all 8 injectors re-testing each replacement before fitting (I had my suspicions that one or two were leaking, one was found to be weeping slightly).

20 mins driving avoiding the puddles, I monitored the O2 voltage/AFR and it was reading pretty close, so was relatively happy that things were working as they should be....

VE analyse the results - Same damned thing

Decided to get a bit fed up with MLV and MT, so fired up TunerStudio MS (managed to get it working for brief time before crashing....), with the hope that this would behave differently... well I think this screen shot shows what happened best, what TSMS thinks on the left, 'slightly' tuned map on the right:

Needless to say it ran like a bag of poop and eventually cut out on me just as some coppers were coming the other way(!)

After 1 or 2 logs and processing with MLV it really shifts nicely, I even manage wheel spin when hoofing it in first, but from then on it is all down hill.

Oh, it's not IAT sensor related, I have one supplied by you Nige, and appears to read correctly, above ambient most of the time, getting to 50C when been sitting idling a while....

HELP!

[Bowie69]

Been having a closer look at the log files, at WOT the lambda voltage falls over a few seconds down to just  zero volts... i.e VERY VERY lean, hence MLV tries it's hardest to up the fueling to compensate...

Now, first thought was a weak fuel pump, (i.e. it's not getting enough fuel pressure, fuel injected drops as flow increases), but the clouds of black and colour of plugs kinda say otherwise

So.... at the moment I am leaning (no pun intended!) towards either a failed lambda probe or some sort of wiring fault such as a bad earth. First option is a little unlikely I think, as it was happening on a previous sensor as well, this is a new 4-wire one (last one was 4-wire too).

So that leaves a wiring fault, I think.... so I'll be checking that over... bearing in mind this was working fine around new year with the old sensor, the problem has only arisen since I swapped transfer box along with a fair few other things I'm beginning to run out of good ideas...

Ta for any help people

[Belgian_Dude]

ok, update ;-)

Thanks for not answering too quickly, forces me to figure it all out myself

And thanks for the IAT idea!

I decided to calibrate my ECU using Easytherm to use the same values known for rover CLT for the IAT, generated the .s19 file and upped that into the ECU.

Nothing changed display-wise in MT, as this display is purely based on thermfactor,matfactor.inc files as stated earlier, but I'm assuming the ECU now finally knows what it's breathing.

Started over from the "General defaults" again, and, well, she runs well (couldn't honestly tell the difference though,I'm too biased.)

But, more importantly, after a few logs and VE analysis(es?), I'm now seeing proper o2 voltage switching around .510V, and it seems like MLV is trying to up the fuelling at WOT regions less. Will see after more logs if it keeps on upping it like it did before, but definitely more pleased with what I'm seeing in the logfiles, and what MLV is doing with that info.

Had another issue, but promptly solved; Engine was cutting out for very short instants, at first only whilst driving at steady throttle and below 25%tps, but after a while it did so on WOT as well, and that's not good. I'd been looking at te log graphs since the second or third instance of this, as it was clearly new, but couldn't figure out what triggered it: RPM plummeted, and everything followed upon that value (VE, spark, DC, ...) but I couldn't find any value (went through all displayable values) that came before the RPM value drop. Also felt the RPM drop, and saw it on my std Rangey tacho, but resolution of my eyes and of the std tach wasn't high enough to see which came first: RPM drop or RPM value drop.

But, as I couldn't find any value preceeding the RPM value drop, I opened up the bonnet, and surely, my yet-to-be-wrapped cable loom, specifically the part going to the EDIS8, was touching a spark lead Separated both, gone interference, gone "problem".

Greetz

Q

[FridgeFreezer]

Quentin,

You either need to calibrate the ECU for your IAT sensor or use a GM one as described in the Megamanual, they are commonly available and off-the-shelf thermistors such as RS Components 151-215 have the same response curve. It may not be the root of your problem but it won't be helping.

[Bowie69]

Here's an example of what I am getting in MLV:

Will be checking wiring today methinks...

[Belgian_Dude]

Quentin,

You either need to calibrate the ECU for your IAT sensor or use a GM one as described in the Megamanual, they are commonly available and off-the-shelf thermistors such as RS Components 151-215 have the same response curve. It may not be the root of your problem but it won't be helping.

Fridge,

I recalibrated the ECU to suit my CLT used as IAT yesterday already, see previous replies That was all I could do yesterday, and added advantage is that it forced me to get to know Easytherm, it's workings, and spawned a better understanding of the whole temp readout mechanism (sensor - .s19 file - .inc files - ... )

I still stand by my idea to use identical and LR sensors though, GM's CLT and IAT are the same curve as well from what I read (though different internals with an open element for the IAT). Also, you'd be surprised how hard and/or expensive it can be to find simple little items like a bog standard GM IAT or one of it's 1354 aftermarket couterparts in this sucky place called Belgium.

So to sum up, I now have 2 rover CLT sensors, one for CLT, one for IAT, and an .S19 generated with Easytherm 5.0 using the calibration values * from your thermfactor.inc for both CLT and IAT uploaded into the Megasquirt. (* 2 Celcius = 4000 ohm; 74 Celcius = 362 ohm; 94 Celcius = 212 ohm)

Can't think what could be wrong in this setup

Greetz

Q

[Bull Bar Cowboy]

Here's an example of what I am getting in MLV:

Will be checking wiring today methinks...

Nothing too bad there apart from the O2 dropping to zero when you stand on the gas ............. however, that may not be a wiring issue.

You really need a wide band to see whats going on ........... my LC1 is sat here doing nothing if you want to borrow it.

[V8 EFI]

Third time lucky

Finaly photos included!. I've studied the inside of the ecu and discoverd a missing wire to spr3. However that should not cause the problems i have now.

The IAC outputs are connected to the leds on the front for sparkoutput together with a transistor in the proto area.

I've set all the outputs to off but when running they keep going on and off. They are not connected by anything Again this can not cause any problems with resets right ?.

I've connected the multiple earth wires to the chassis mounting between engine and battery (very thick cables).

The supplier told me he had a recently a case where fould up sparkplug caused resets but i'm not using the ignition right now (accept from the old distributer of cause)

The TPS value is about -1% closed, can a negative value cause problems ?

All i can think of right now is:

- Change earth wires to engine (which would be closer to the distributer and plugs?)

- Change plugs (would hate that because i had trouble with new plugs before and its running good now on lpg !)

- Changing the polarity of the vr sensor, again rpm values are good, 100 with starting and all the way up.

- Changing the distance between the VR sensor and trigger wheel, i dont want it to close to the wheel to keep it from scratch.

The output 4 setting (led 18) is used for spark output, must this be set to spark angle in the outputs menu ?.

I've tried to upload a msq but apparently i do not have permission

Any thoughts ?.

Frederik.

[V8 EFI]

OK, i've changed the earthing to the engine but made it even worse. Swapped the red and blue wires of the vr circuit and finally made it go corrupt

Firmware mismatch between controller and project wich leads me to the following question why does TS reconice the firmware in the controler as 029Q as i uploaded it with 029V version from this forum?.

If it is this sensitive to resets then will it ever work ?.

Dont no if my plunging belongs to this topic. Moderators may place it eleware but hopefully not to an dead end.

Frederik.

[Bowie69]

Nothing too bad there apart from the O2 dropping to zero when you stand on the gas   ............. however, that may not be a wiring issue.

You really need a wide band to see whats going on ........... my LC1 is sat here doing nothing if you want to borrow it.

Cheers Ian, the offer is muchos appreciated, not really sure I would know what I was doing with one of them mind you! I'd like to rule out a few other things first, before sticking £200 worth of fancy equipment into it and potentially breaking it When I have even less hair left I'll drop you a PM

Today I spent a bit of time just checking things over, chaffed wiring etc, nothing, but did notice both down pipes were leaking a little bit at the manifolds, so unbolted them and shoved a load of firegum in and tightened up -no more leaks

Quick 10 minute log and the issue is still there...

Checked MLV again and found there was somewhat of a correlation between battery voltage and 02 reading:

I checked the voltage at the battery and the back of the alternator and it is not varying with revs, time to check earths maybe?

I read on my travels something about really really rich can appear to be weak to an 02 sensor due to soot insulating it, is this the case, or was someone talking out their bum? 

Ian, you say 'that may not be a wiring issue' as if you have an idea of what it might be....?

Thanks again peeps... if Mods decide this would be better placed elsewhere along with the replies I completely understand.... didn't think it would be this long winded tbh

[FridgeFreezer]

Lambda sensors are only measuring what they can see, so it's easy to forget they are quite delicate and can be fooled.

They need to be up to temperature (datalogging whilst driving through torrential rain / big puddles nearly caught me once)

The element needs to be clean (not coated with oil / soot / unburnt fuel)

Your exhaust needs to be sealed (an exhaust leak sucks air in as well as blowing out, and air contains much more oxygen than exhaust gas)

So yes, any number of things can throw your readings off. A wideband can help show this (as it can display a wider range of mixtures), but you can do a bit of sleuthing to work out what's happening. Your nose can tell you if it's hoofing out loads of fuel on accel, as can black smoke.

[Richard Lane]

Hi All,

Interesting logs there. I think someone hinted to a bad earth / electrical noise problem. I think I would be looking at this angle. It would be wise not to rely on engine loom earths or body earth returns. Running earths and power feeds from common points is a must do. Most OEM systems also do this. A volt meter over the O2 sensor when driving may indicate this, and checking resistance between the computer and the O2 sensor and the computers earths when driving. Also has you mention, look into the charging system and its earth.

Rich

Keeper of the mongol.