Alez
|
| posted on 13/5/15 at 08:54 AM |
|
|
Engine transient response problem (likely management related)
Hi all,
Showing apparently random, this unwanted behaviour of my Ultima has been driving me mad for a long time. At long last I have finally discovered the
conditions that trigger it and now I can make it happen at will, which I consider a huge step forward for eventually solving it. I would like to share
my findings with you and hear any comments you may have.
The problem experienced is that, sometimes only, but always as a result to applying throttle, the engine would abruptly stop delivering power for a
short moment (say a few 10ths of a second), much like you would expect from a temporary ignition cut or fuel starvation, resuming abruptly shortly
afterwards. Obviously this is most unwanted if only because of the significant stress it puts on the transmission (I have avoided track use altogether
in case I break something).
I speculated that the origin of the problem may be fuel starvation during cornering, as often I was experiencing this right after a bend as more
throttle was applied. However I've now found that the exact thing causing it (and it happens every time this condition is satisfied) is keeping
the engine at high-ish RPM (6000, for example) + very low throttle for a while (at least a few seconds), then applying more throttle (the more sudden
the command, the more accute the effect).
So, for example if, on the track, I go into a long lasting bend at 4000 RPM and then I apply throttle after the bend, that is perfectly fine and I get
plenty of power. If I go into the same bend at say 3000 RPM and then I downshift and quickly apply throttle, that is perfectly fine too. However, if I
do not get the gear right and get into the same corner at 6000 RPM, obviously I typically do not upshift in the middle of the bend but rather I stay
at constant car speed / RPM until the bend is over. I then apply throttle after the bend and get this very nasty effect. So if I got in the bend at
6000 RPM and want to avoid this happening, the way to "reset" the engine so that it does not happen is to upshift and either pull (apply
throttle) from there or keep the engine at the low RPM region for a few seconds, downshift again and then apply throttle. And it's not to do
with cornering, I can make it happen on a straight, it just happens that I very rarely keep the engine at low throttle + high RPM, except on the track
during long lasting bends.
I am confused. It does not look like the ECU has not got enough bandwidth (fails to send enough fuel as a result of the higher throttle), because the
problem shows even if I apply throttle slowly. And it does not look like the injectors cannot provide enough fuel, because I can reach high RPM + high
throttle, provided that I build from lower RPM. Also it is not the ECU cutting off trying to protect the engine upon detecting something wrong since
my setup is really simple with very very few sensors (only throttle position, water temperature and lambda sensors I seem to remember).
My setup is: Audi S4 4.2 V8 engine, DTAfast S80 ECU with open loop lambda setup, lambda target set to 0.9 regardless of throttle position (except for
very low RPM which is not relevant here).
I have tried my best to explain it clearly but my English is not great, so apologies and thanks for your patience.
[Edited on 13/5/15 by Alez]
|
|
|
|
|
coozer
|
| posted on 13/5/15 at 09:17 AM |
|
|
Fuel pressure regulator. Is it connected/controlled by MAF?
Maybe its struggling to respond to the sudden drop in manifold pressure when you open the throttle.
Just one thing to check.
1972 V8 Jago
1980 Z750
|
|
|
Oddified
|
| posted on 13/5/15 at 09:29 AM |
|
|
I don't know the dta ecu inside out, but the first thing i would look at would be acceleration fueling (different ecu manufacturers call it
different things, pump fuel, transient fuel, wall wetting, acceleration fueling etc etc). Add extra fuel on throttle movement in the problem areas.
Ian
|
|
|
Stot
|
| posted on 13/5/15 at 09:39 AM |
|
|
It does sound like Acceleration Enrichment to me too. If you have a wideband in the system you can track down if it is this an if its too lean or too
rich as both can cause this.
I have megasquirt and on mine I have AE set to work on TPS DOT (Speed of pressing the go pedal). The faster I press the pedal the more fuel gets
dumped in depending on RPM etc. What you find with a WB O2 sensor is that if you are dumping too much fuel the AFR will hesitate too rich then lean
then level. If youre not dumping enough it will hesitate lean before leveling out.
Cheers
Stot
|
|
|
britishtrident
|
| posted on 13/5/15 at 11:28 AM |
|
|
Might be checking the throttle pot hasn't got a glitch.
[I] “ What use our work, Bennet, if we cannot care for those we love? .”
― From BBC TV/Amazon's Ripper Street.
[/I]
|
|
|
Angel Acevedo
|
| posted on 13/5/15 at 12:45 PM |
|
|
quote:
Originally posted by britishtrident
Might be checking the throttle pot hasn't got a glitch.
I'd second that.
On acceleration enrichment, I thought this was only needed on Carbs due to the fact that when you open throttle suddenly, you loose some fuel metering
ability due to MP rising rapidly and air speed in the venturis decreasing.
I'm not sure if this applies to Injectiom Systems
my 2 cents.
Aa
Beware of what you wish.. for it may come true....
|
|
|
jeffw
|
| posted on 13/5/15 at 05:05 PM |
|
|
You should be able to log the DTA ECU and catch the event. That will give you way more information than what you have at the moment.
|
|
|
Stot
|
| posted on 13/5/15 at 05:51 PM |
|
|
quote:
Originally posted by Angel Acevedo
quote:
Originally posted by britishtrident
Might be checking the throttle pot hasn't got a glitch.
I'd second that.
On acceleration enrichment, I thought this was only needed on Carbs due to the fact that when you open throttle suddenly, you loose some fuel metering
ability due to MP rising rapidly and air speed in the venturis decreasing.
I'm not sure if this applies to Injectiom Systems
my 2 cents.
Aa
You need it on injection too for the same reason. you can open the throttle quicker than a calculation would figure out what to do so you squirt in a
bit more fuel based on the rate of change of TPS, or MAP rather than what TPS or MAP is at.
Cheers
Stot
|
|
|
perksy
|
| posted on 13/5/15 at 07:08 PM |
|
|
I'd be checking the TPS (Throttle position sensor) first.
Had a similar problem on an XE once and it started like this and gradually got worse, Hooked up the lap top and the TPS graph wasn't smooth and
had a 'jump' in it.
Eventually it got worse and wouldn't start 
Spare TPS from a mate cured it 
|
|
|
Angel Acevedo
|
| posted on 13/5/15 at 08:48 PM |
|
|
quote:
Originally posted by perksy
I'd be checking the TPS (Throttle position sensor) first.
Had a similar problem on an XE once and it started like this and gradually got worse, Hooked up the lap top and the TPS graph wasn't smooth and
had a 'jump' in it.
Eventually it got worse and wouldn't start
Spare TPS from a mate cured it
Thanks for the input.
At some point in time I'll be running Megasquirt and these bits of info will be very helpful
Regards
Angel
Beware of what you wish.. for it may come true....
|
|
|
Alez
|
| posted on 15/5/15 at 06:01 PM |
|
|
Big thanks, all. Hugely appreciated.
quote:
Originally posted by coozer
Fuel pressure regulator. Is it connected/controlled by MAF?
Maybe its struggling to respond to the sudden drop in manifold pressure when you open the throttle.
Just one thing to check.
Sorry for my lack of knowledge. I do not know what a fuel pressure regulator is. I know I have twin fuel pumps:
(Twin as I have twin fuel tanks, one each side). I was not aware that I have any other device sitting between that and the injectors.
As for the MAF, I do not think I have one. I think my ECU knows what timing advance and fuel amount to use based solely on 1) throttle position and 2)
RPM (as provided by the main fuel map and main ignition map) with correction from water temp, and (optional) correction from lambda sensors which is
disabled at the moment (running open loop).
The person who made the install gave me the following description:
quote:
What's connected to ECU:-
Due to the basic requirements Mem wanted for this install you will find you have a few sensors fitted to make the
ecu work satisfactory.
-water temp sensor
-air temp sensor
-two cam sensors (audi)
-one crank sensor (audi)
-two lambda sensors, one in each exhaust.
-throttle sensor mounted on one thottle bank.
The original engine set up came from an exact engine we had fitted previously. This previous engine was setup
on a rolling road. The ecu is running a closed loop system monitoring the lambda sensors mounted in exhaust.
Other engine sensors fitted in car electrics:-
-Oil pressure sender connected to dash gauge.
-Water temp sender connect to dash guage.
-speed sensor fitted to gearbox connect to speedo gauge.
-Tacho gauge is fed from ecu for engine RPM.
[Note: this description says closed loop, but I modified it to open loop following my finding that I had a faulty oxygen sensor connector. Once the
connector was repaired, and given that the engine run perfectly well (problem under analysis aside), I left it open loop to keep things simple until
the issue is solved.]
quote:
Originally posted by britishtrident
Might be checking the throttle pot hasn't got a glitch.
quote:
Originally posted by perksy
I'd be checking the TPS (Throttle position sensor) first.
Had a similar problem on an XE once and it started like this and gradually got worse, Hooked up the lap top and the TPS graph wasn't smooth and
had a 'jump' in it.
Eventually it got worse and wouldn't start
Spare TPS from a mate cured it
Oh, sorry, forgot to mention that I suspected this might be causing it and checked it. The reading looked perfectly alright, following my foot with no
apparent glitches. That said, given your input that this is well worth checking, I plan to double check, because I just did a quick check feeling
that, being unexperienced, I was just checking different things randomly.
quote:
Originally posted by jeffw
You should be able to log the DTA ECU and catch the event. That will give you way more information than what you have at the moment.
I considered doing so but looking at the above description it does not look like I would get anything useful from that. Although, now that I think of
it, since I get throttle position, if I have a glitch there it would show.
quote:
Originally posted by Oddified
I don't know the dta ecu inside out, but the first thing i would look at would be acceleration fueling (different ecu manufacturers call it
different things, pump fuel, transient fuel, wall wetting, acceleration fueling etc etc). Add extra fuel on throttle movement in the problem areas.
Ian
quote:
Originally posted by Stot
It does sound like Acceleration Enrichment to me too. If you have a wideband in the system you can track down if it is this an if its too lean or too
rich as both can cause this.
I have megasquirt and on mine I have AE set to work on TPS DOT (Speed of pressing the go pedal). The faster I press the pedal the more fuel gets
dumped in depending on RPM etc. What you find with a WB O2 sensor is that if you are dumping too much fuel the AFR will hesitate too rich then lean
then level. If youre not dumping enough it will hesitate lean before leveling out.
Cheers
Stot
quote:
Originally posted by Angel AcevedoI'd second that.
On acceleration enrichment, I thought this was only needed on Carbs due to the fact that when you open throttle suddenly, you loose some fuel metering
ability due to MP rising rapidly and air speed in the venturis decreasing.
I'm not sure if this applies to Injectiom Systems
my 2 cents.
Aa
quote:
Originally posted by Stot
You need it on injection too for the same reason. you can open the throttle quicker than a calculation would figure out what to do so you squirt in a
bit more fuel based on the rate of change of TPS, or MAP rather than what TPS or MAP is at.
Cheers
Stot
Is this it?:
I am confused about my issue and how it does not seem to relate to some comments where you use "speed of pressing the go pedal",
"suddenly", "open the throttle quicker". I thing I understand the basic principle that a high rate of increase on the throttle
requires more fuel, but if you see my initial description, this thing I am experiencing does not seem to relate to throttle increase rate, I still get
it when I apply throttle slowly.
Anyway, your responses are so helpful!!!! So I think I now have a plan:
1) Double check throttle position sensor
2) Go for a ride, make it happen, look for throttle glitches at the data log (agreed that there is not much more to look for?)
3) Tinker with ECU configuration: a) get an accurate idea of the exact conditions triggering it (static throttle vs. dynamic and so on), then b) try
different mods to the main fuel map around the 6-7k rpm "problem area" and / or the acceleration enrichment configuration.
|
|
|
Oddified
|
| posted on 15/5/15 at 06:44 PM |
|
|
That facet fuel pump isn't for fuel injection (way to low pressure), unless they're being used as lift pumps to a swirl pot and then
there's a fuel injection pump feeding the injectors??.
Ian
|
|
|
Alez
|
| posted on 17/5/15 at 03:10 PM |
|
|
quote:
Originally posted by Oddified
That facet fuel pump isn't for fuel injection (way to low pressure), unless they're being used as lift pumps to a swirl pot and then
there's a fuel injection pump feeding the injectors??.
Ian
Oh, silly me, of course I have got stuff sitting in between the pumps and the injectors, there is at least this swirl pot you mention:
Cheers,
Alex
[Edited on 17/5/15 by Alez]
|
|
|
Oddified
|
| posted on 17/5/15 at 04:13 PM |
|
|
No, to me that looks like your water header tank with a Stant high pressure cap fitted.
I 'think' i can see a fuel filter and maybe a fuel swirl pot lower down in the engine bay, but i can't make out a fuel reg from the
picture.
Ian
|
|
|
Alez
|
| posted on 17/5/15 at 04:16 PM |
|
|
I'm hopeless!!  I need to go to the garage and check the pipes for the different tanks there, I am too curious now.
|
|
|
