View previous topic :: View next topic |
Author |
Message |
Arvidw
Joined: 20 Sep 2009 Posts: 227 Location: The Nederlands (Europe)
|
Posted: Sun Aug 29, 2010 6:45 am Post subject: ArvidW's 924NA MS-II Kompressor Project |
|
|
I've just recentely finisched the megasquirt conversion of my 924NA euro engine. It's been great fun and after some hours of tuning it's running quit nice. However some questions remain after looking at some datalogs
First some info what I have done:
- Megasquirt V3.00
- Ford EDIS ignition
- Bosch 20.5lb/hr injectors running at 23.1lb/hr twinbeam*
- Weber 52mm throtle body with IAC stepper motor
- "Hot street" camshaft 257dg & 13mm lift
- Original "air tubes" including vaccuum tube for brake assist. K&H filter element in original airbox. The airflow measuring plate and lever are removed from the housing.
- LC-1 wideband & controller
*I'm aware that injectors for a 4valve/cilinder engine are not optimal for a 2valve/cilinder engine however when I bought the injectors I thought these where the correct ones. After some CAD drawings I decided to give it a try and up to now I don't have problems with fuel puddles or rough running / idleling
Now comes the strange part, i have been busy adjusting the VE table by hand for the fullthrotle area. (Autotune is not fast enough for this region) My target is 12.5AFR and after some runs i'm getting cose to "perfection". However what i'm seeing now is that my duty cycle gets to 100% at 6500rpm. If I drop the AFR tot 11 by raising the values in the VE table I can get even 117% duty cycle. How is that possible?
I'm running cylinder 1 and 3 from injector output 1 and 2 and 4 from injector output 2. Megasquirt is set to give 2 squirts (alternating) every complete stroke of the engine
Could anyone explain why I get a 100% duty cycle even when some basic calculations say that for 125hp I suppose to get 68% duty cycle?
Another strange thing is dat my engine is ideling at aprox 55kpa and megamanual states that it should be around 35kpa? (aprox 5ms injection @ 950-1000rpm) _________________ Porsche 924 2.0 Kompressor 136.1wHp @ 5650rpm / 201.9wNm @ 3900rpm (dyno @ dp-engineering.nl)
Last edited by Arvidw on Thu Jun 21, 2012 7:48 am; edited 3 times in total |
|
Back to top |
|
|
fiat22turbo
Joined: 18 Jan 2006 Posts: 4040 Location: Portland, OR
|
Posted: Sun Aug 29, 2010 1:05 pm Post subject: |
|
|
Great work! Any and all details you can provide for a How-To would be great.
A few questions:
Where are you picking up vacuum for the MAP?
What's your base fuel pressure, the rate at which the regulator changes the fuel pressure and where is the vacuum connected?
You may be making more power now and the injectors aren't able to keep up, or your readings are off due to a poor location for getting a vacuum signal.
Upping the size of the injectors and/or adjusting the base fuel pressure might help. _________________ Stefan
1979 924 Carrera GTS (clone-ish)
1988 944 Turbo S (Silver Rose) |
|
Back to top |
|
|
Arvidw
Joined: 20 Sep 2009 Posts: 227 Location: The Nederlands (Europe)
|
Posted: Tue Aug 31, 2010 12:08 am Post subject: |
|
|
I'm afraid it is a bit too late to make a howto, I didn't make enough pictures during the assembly proces. (too much hurry)
The MAP vaccuum is picked up just mm's behind the throtle valve. I've reduced the map vaccuum to 45-48kPa which is equal to the MAP value when I was driving with CIS and Megajolt ignition. Apparently a 924 engine doesn't pull that much vaccuum when idling.
At your advise I raised the fuel pressure lowering the duty cycle. However I also went from 2spray/cycle altering to 2spray/cycle simultanously raising the duty cycle.
Now i'm at AFR 12.5, 6500rpm and 95% duty cycle when the rev limiter kicks in. I'm aware that this is not optimal but I think it is okey for now.
Still I find it weird that the duty cycle gets so high, probably cloged up injectors? _________________ Porsche 924 2.0 Kompressor 136.1wHp @ 5650rpm / 201.9wNm @ 3900rpm (dyno @ dp-engineering.nl) |
|
Back to top |
|
|
Arvidw
Joined: 20 Sep 2009 Posts: 227 Location: The Nederlands (Europe)
|
Posted: Sat Sep 25, 2010 3:42 am Post subject: Idle control Megasquirt |
|
|
I'm having big troubles tuning Megasquirt to give a proper Idle at all times.
The amount the stepper motor opens the bypass in the throtle body is only defined by the engine coolant temperature and that just doesn't work for me.
I can tune the engine to give the correct idle RPM from a 20degree start to opperating temperature. However when I start the engine after 1 hour with a hot engine (high coolant temperature) i need more air to idle until the pistons and combustion chamber are really warmed up. This means i need to adjust the values again ruining everything
The original 924 idle control works as a result of the temperature of the engine AND time after start. Is there a possibility to imitate this with Megasquirt? _________________ Porsche 924 2.0 Kompressor 136.1wHp @ 5650rpm / 201.9wNm @ 3900rpm (dyno @ dp-engineering.nl) |
|
Back to top |
|
|
flosho
Joined: 01 Jul 2004 Posts: 3155 Location: Eau Claire, Wisconsin
|
Posted: Sat Sep 25, 2010 4:26 am Post subject: |
|
|
I haven't had MT open in some time, but I'm fairly certain you can check an option to use Coolant temp and Intake Temp for the idle valve.
Also, you should check out Tuner Studio, I purchased it to use with megasquirt and TS is waaaaaaaay nicer (IMHO) than MegaTune. _________________ [This Space For Rent] |
|
Back to top |
|
|
Arvidw
Joined: 20 Sep 2009 Posts: 227 Location: The Nederlands (Europe)
|
Posted: Sat Sep 25, 2010 6:21 am Post subject: |
|
|
Thanks for the tip, I've just downloaded a trial version and will try it very soon.
It would be ideal that you could enter a specific target rpm and that the software just tries to reach it by moving the stepper motor. (like autotune with AFR target table). Is there something that can do that ? _________________ Porsche 924 2.0 Kompressor 136.1wHp @ 5650rpm / 201.9wNm @ 3900rpm (dyno @ dp-engineering.nl) |
|
Back to top |
|
|
Min
Joined: 04 Nov 2002 Posts: 2368 Location: Vernon, British Columbia, Canada
|
Posted: Sat Sep 25, 2010 9:57 am Post subject: |
|
|
Arvidw wrote: | It would be ideal that you could enter a specific target rpm and that the software just tries to reach it by moving the stepper motor. (like autotune with AFR target table). Is there something that can do that ? |
That would strongly depend on what kind of idle air control stepper your using? also, what version of the code do you have loaded? Some of the code versions are vastly different in how they deal with idle air. What version of hardware are you running? (I see you have a MS V3.00 listed, but what revision of chip? megasquirt I ? megasquirt II? megasquirt III?)
We're going to need more information before we can help you. In reality, you could just read the manuals, they have alot of detail about how to deal with IAC valves.
Min _________________ Custom means it didn't come from a box.
1980 n/a with EDIS and Megasquirt II Injection. 7 different colors and counting. |
|
Back to top |
|
|
Min
Joined: 04 Nov 2002 Posts: 2368 Location: Vernon, British Columbia, Canada
|
Posted: Sat Sep 25, 2010 10:03 am Post subject: |
|
|
For example, if your running megasquirt II extra code, you get one of these.
Which as you can see allows you to completely mold your idle throughout the coolant temperature range. _________________ Custom means it didn't come from a box.
1980 n/a with EDIS and Megasquirt II Injection. 7 different colors and counting. |
|
Back to top |
|
|
Min
Joined: 04 Nov 2002 Posts: 2368 Location: Vernon, British Columbia, Canada
|
Posted: Sat Sep 25, 2010 10:17 am Post subject: Re: Idle control Megasquirt |
|
|
Arvidw wrote: | The original 924 idle control works as a result of the temperature of the engine AND time after start. |
I'm not sure what gives you that impression. The original valve is bolted onto the intake manifold, and has a bimetallic strip inside that opens or closes based on temperature. Time has nothing to do with it. If the intake manifold is hot enough, its closed, if its cold enough, its open. It also has a electrical connector on it that is provided voltage at cold start, my understanding of that system is that it heats up the bimetallic strip at a faster rate.
Min _________________ Custom means it didn't come from a box.
1980 n/a with EDIS and Megasquirt II Injection. 7 different colors and counting. |
|
Back to top |
|
|
Min
Joined: 04 Nov 2002 Posts: 2368 Location: Vernon, British Columbia, Canada
|
Posted: Sat Sep 25, 2010 10:19 am Post subject: |
|
|
PS, megasquirt is opensource, if you don't like how the idle control works, code it the way you want it.
Min _________________ Custom means it didn't come from a box.
1980 n/a with EDIS and Megasquirt II Injection. 7 different colors and counting. |
|
Back to top |
|
|
Arvidw
Joined: 20 Sep 2009 Posts: 227 Location: The Nederlands (Europe)
|
Posted: Sat Sep 25, 2010 4:09 pm Post subject: |
|
|
I'm running megasquirt II V3.00 PCB with the 2.890 code.
The idle control valve is a GM style stepper motor, just like the one on megamanual.com
The original valve uses the heat of the manifold (temperature based) which is accelerated by the heating element which is basically time based. (It takes ... minutes to heat up, when the engine is extremely cold it takes a little longer and when the engine is a little warm it takes a little less time)
I've read the manuals on megamanual.com but it can't find any option to make the IAC both time and temperature based. I've been experimenting with the graph you show in your post however I find the folowing problem:
The problem is that when I start my engine @ 20deg I need around 100 IAC steps. When it has been running for lets say; a minute it is 23deg and only needs 115 IAC steps. So If I dial in this value and I start my engine cold next morning @23deg it doesn't idle properly because the internals of the engine are just too cold for that. (at these temperatures I try to get 1300rpm idle)
I think this is because after one minute of Idling the piston and combustion chamber have warmed up quit rapidly however the coolant didn't because of the enormous heat cappasity of the coolant & block + thermal conductivity of the engine block.
If i was able to measure the combustion chamber temperature more direct I would be able to tune the temperature based IAC much more accurate... Maby place a sensor in the head / use exhaustgass temperature? This eliminates the heat conductivity and thermal capacity of the coolant...
I'm afraid I can't write my own code, i'm a mechanical engineer, not a programmer. _________________ Porsche 924 2.0 Kompressor 136.1wHp @ 5650rpm / 201.9wNm @ 3900rpm (dyno @ dp-engineering.nl) |
|
Back to top |
|
|
RC
Joined: 25 Mar 2007 Posts: 2636 Location: Australia
|
Posted: Sat Sep 25, 2010 9:31 pm Post subject: |
|
|
Looks like you have a number of issues here.
Starting at your first post - possible low fuel pressure or flow, and injectors are far from ideal and will cause tuning irregularities whether you think so or not. Only idle and cruise will be relatively unaffected and stabalise while cranking, starting and acceleration will need to be compensation for a decent tune.
IMHO forget about autotune and instead just datalog while driving, as you have been, concentrating on particular load/ RPM cells if necessary. Then as well as being able to analyze it yourself MLV will make corrections to your defined parameters.
Stepper IACVs, correctly set up of course, will give excellent and virtually instantaneous control. More so with the extra code, it allows for stepper IACV idle stabilisation. However the whole system needs to be tuned along with ALL cranking, STARTING, and IDLE functions. You know - all the other things under the MT startup/idle menu, - cranking pulse & enrichment, after start enrichment, warm up enrichment. I hear what you`re complaining about but in reality it is a non issue. Rereading the megamanual and particularly searching the forum should offer some consolation, extremely unlikely that you have a unique problem that has never occurred before. It just takes time and understanding of ALL the various settings.
I have moved on from the B&G code and old MSQ files are on my other computer. Here are some screenshots of my settings for a basic idea, remember I`m in Australia. The basic code differs to the latest extra code, in that the stepper counts (from to closed) are reversed. Still 255 steps on the GM IACV but with B&G 255 = fully closed, whereas on the extra code 255 = full open and 0 = closed.
|
|
Back to top |
|
|
Min
Joined: 04 Nov 2002 Posts: 2368 Location: Vernon, British Columbia, Canada
|
Posted: Sun Sep 26, 2010 2:44 am Post subject: |
|
|
Based on your second description of your problem, I agree with RC. I believe one of your other settings is whats causing the problem, not your IAC steps.
Do a proper log of the problem area
Do a proper log of when it behaves properly.
Compare all things. (such as injector pulse widths)
Min _________________ Custom means it didn't come from a box.
1980 n/a with EDIS and Megasquirt II Injection. 7 different colors and counting. |
|
Back to top |
|
|
Martijnus
Joined: 29 Dec 2006 Posts: 2019 Location: Netherlands
|
Posted: Sun Sep 26, 2010 3:37 am Post subject: |
|
|
I would swap to other injectors.
I've had some strange problems when my injectors weren't correct.
The AFR values seem fine, but it injects too much because only part of the mixture burns.
If that's the case, a 117% duty cycle isn't strange... you just have to inject twice as much as normal... which you will see in the fuel consumption of the car.
I'm not saying this is the case, but keep it in mind _________________ "Rule: Turbo's make torque, and torque makes fun." (C. Bell)
924 "50-jahre", 1981.
MSII/extra, LPG, ITB's, 5lug.
To be turbo'ed in a while.
Killed her at the Nurburgring, Porscheless at the moment |
|
Back to top |
|
|
Arvidw
Joined: 20 Sep 2009 Posts: 227 Location: The Nederlands (Europe)
|
Posted: Sun Sep 26, 2010 6:21 pm Post subject: |
|
|
All of you, thanx for the replies, I find them very helpfull
@RC, My fuel pressure was indeed too low, I had a faulty gauge on my cheap ass fuel pressure regulator. Now after some tuning i'm running 79% duty cycle @ 12.5AFR & 6500rpm.
@Martijnus, I'm going for APK (MOT) in 2 weeks and If I have a problem with bad evaporation of the fuel that will show on the exhaust gass test as unburnt hydrocarbons. If these are extremely high I will swap injectors.
Until now I've only focussed on the IAC, today I'll try to adjust the afterstart enrichment, idle control, warmup wizzard and IAC simultanously.
Where do you pick up the coolant temperature? Mine is positioned in the coolant hose where the thermo time swich used to be (I've recycled the metal adaptor, just screwed the NTC in that hole)
Edit: I've found the MS2extra 2.1.0d code which has closed loop idle control ! Yeeey
I'm starting with an completely new setup just to make sure that I don't copy any mistakes made when configuring MS2.890 code. _________________ Porsche 924 2.0 Kompressor 136.1wHp @ 5650rpm / 201.9wNm @ 3900rpm (dyno @ dp-engineering.nl) |
|
Back to top |
|
|
|