2nd Generation Specific 1986-1992 Discussion

S4 TII Hot-start problem (I've checked all the normal stuff)

Thread Tools
 
Search this Thread
 
Old 09-19-2006 | 10:35 AM
  #1  
N1XRR's Avatar
Thread Starter
Member
 
Joined: Jul 2001
Posts: 80
Default

Hey everyone, I've been trying to figure out this problem for the past few months and I'm not very frustrated. First off, this car has a Megasquirt on it controlling fuel and spark. Timing is dead on wherever I set it (right now its sparking at 5 deg BTDC while cranking).



Earlier this year (before we swapped to a new turbo setup) the car would start up and idle just fine. The engine has a bunch of compression, so I'm pretty sure the engine is not the issue.



So, here's whats going on: Above 160 degrees, the car will not start. The rpms jump up 50-75 or so, but it just will not catch and actually start. When it gets between 150-160 degrees, its not easy to start, but it will start after about 6-8 seconds of cranking. Cold the engine starts up perfectly, I couldn't ask for any better. During hot cranking, even at 2.0ms pulse widths, the plugs get wet with oil (running premix at 100:1). When the car does run, it will not idle (VERY weak at 800rpms), but at 1100rpms it seems to run fine.





I've changed out and tested nearly every peice of electronics on the car (including the Megasquirt) and that hasn't helped. Last night I swapped intake manifolds with a good one thats mostly stock, thinking the there wasn't enough vacuum and the velocity of the air passing the injectors wasn't high enough. This made no difference at all. The air bleed system is currently hooked up and I am pulling 2" of vacuum while cranking according to Megasquirt's MAP sensor.



Maybe I have a bad pair of injectors, but only when they get hot?



Any ideas?
Old 09-19-2006 | 10:50 AM
  #2  
Rob x-7's Avatar
Super Moderator
 
Joined: Feb 2002
Posts: 12,288
From: Amityville, New York
Default

usually its low compression or tps, just give the tps a slight adjustment either way when its hot and see if it helps, dont dick around with the test lights, just adjust it and see what happens
Old 09-19-2006 | 01:37 PM
  #3  
j9fd3s's Avatar
Senior Member
 
Joined: May 2002
Posts: 22,465
From: California
Default

how does the MS do cranking fuel? the haltechs have a couple different map settings for cranking. one is inj vs coolant temp, its a % i think. and then theres a "primer" map, which is sorta like on a carb engine, pushing the gas to get an accelerator pump shot.



the "primer" is critical to get the thing started, too little, and it wont start, or crank a long time, too much and it floods.



the cranking vs coolant temp, is more keeping it running once it starts.



the FSM says cranking is done at 5BTDC, i run a little less when the engine is really cold, and when its hot you can run a little more.



for fuel it seems like the best starting setting is rich just shy of flooding out
Old 09-19-2006 | 01:55 PM
  #4  
N1XRR's Avatar
Thread Starter
Member
 
Joined: Jul 2001
Posts: 80
Default

Originally Posted by j9fd3s' post='837642' date='Sep 19 2006, 01:37 PM

how does the MS do cranking fuel? the haltechs have a couple different map settings for cranking. one is inj vs coolant temp, its a % i think. and then theres a "primer" map, which is sorta like on a carb engine, pushing the gas to get an accelerator pump shot.



the "primer" is critical to get the thing started, too little, and it wont start, or crank a long time, too much and it floods.



the cranking vs coolant temp, is more keeping it running once it starts.



the FSM says cranking is done at 5BTDC, i run a little less when the engine is really cold, and when its hot you can run a little more.



for fuel it seems like the best starting setting is rich just shy of flooding out


Interesting. Well, my timing is correct then (5BTDC). Fuel though, is another issue.



The prime information is interesting. I never knew the priming was so important. Megasquirt has several options for cranking, but the basics are this: When to fire the priming pulse (power on, wait 2 seconds, fire twice), the length of the priming pulse (either a fixed value or the same value as the cranking pulsewidth), and then which temp sensor to use (Coolant, Air temp, or the average of both). Right now I have it set to fire the priming pulse once during power-on, and has a fixed pulsewidth of 4ms. Should I be doing this differently? Also, I'm using coolant-temp only, as the air temp sensor tends to heat soak. That also seem wrong?
Old 09-19-2006 | 02:11 PM
  #5  
Jedi's Avatar
Senior Member
 
Joined: Mar 2004
Posts: 397
From: Puyallup Washington
Default

So am I the only one that thinks the injectors are faulty and the thing is flooding?
Old 09-19-2006 | 02:46 PM
  #6  
N1XRR's Avatar
Thread Starter
Member
 
Joined: Jul 2001
Posts: 80
Default

Originally Posted by Jedi' post='837644' date='Sep 19 2006, 02:11 PM

So am I the only one that thinks the injectors are faulty and the thing is flooding?


I've been thinking about that too. I swapped out the injectors with another pair (I'm running 550/1600's), but the set I put in are from the same set of 4. Its possible the set is just trashed and needs to be replaced.
Old 09-19-2006 | 04:20 PM
  #7  
j9fd3s's Avatar
Senior Member
 
Joined: May 2002
Posts: 22,465
From: California
Default

Originally Posted by N1XRR' post='837643' date='Sep 19 2006, 10:55 AM

Interesting. Well, my timing is correct then (5BTDC). Fuel though, is another issue.



The prime information is interesting. I never knew the priming was so important. Megasquirt has several options for cranking, but the basics are this: When to fire the priming pulse (power on, wait 2 seconds, fire twice), the length of the priming pulse (either a fixed value or the same value as the cranking pulsewidth), and then which temp sensor to use (Coolant, Air temp, or the average of both). Right now I have it set to fire the priming pulse once during power-on, and has a fixed pulsewidth of 4ms. Should I be doing this differently? Also, I'm using coolant-temp only, as the air temp sensor tends to heat soak. That also seem wrong?


the amount of "primer" varies with coolant temp on the haltech.... i bet thats it, the amount of fuel varies with engine temp. it would kinda explain why it starts in some temps, right?
Old 09-19-2006 | 05:00 PM
  #8  
N1XRR's Avatar
Thread Starter
Member
 
Joined: Jul 2001
Posts: 80
Default

Originally Posted by j9fd3s' post='837670' date='Sep 19 2006, 04:20 PM

the amount of "primer" varies with coolant temp on the haltech.... i bet thats it, the amount of fuel varies with engine temp. it would kinda explain why it starts in some temps, right?


Well, I agree, but I currently have it set to look at coolant temp only during cranking. So, I'm not sure if thats as much of the problem. I have played and played with the pulse timings from 1.0ms up to 15.0ms and it just won't even start when its hot. But, it won't idle either, which is why I'm thinking I have other problems.
Old 09-20-2006 | 01:09 AM
  #9  
FINEPRINT's Avatar
Junior Member
 
Joined: Sep 2006
Posts: 13
Default

I know it says im a new member but i work for Star Mazda. We have had this problem before with this setp but in a different ignition. Anyways, it sounds like more of a ECU problem. I have a MeagSq on my other race car (AE86 and it works great) but they are two totally different setups. The 13 has funky rythms in its read out that most cars ecu's cannot read or frankly speaking, dont want to, maybe if you have some prints or reads outs i could look over them and double check with our motor bulder for you. If he doesnt know than i dont know what to say! Let us know.



Matt D

www.starmazda.com

www.shukwunz@yahoo.com
Old 09-20-2006 | 09:17 AM
  #10  
N1XRR's Avatar
Thread Starter
Member
 
Joined: Jul 2001
Posts: 80
Default

Last night I went out and did a little more work. I took all 4 injectors out and I'm going to look at them tonight with a little test setup I'm putting together. So, we'll see what the pattern looks like. I put a different set of turbo injectors in and the car wouldn't even start cold....so I suspect that the injectors are at least partially working correctly that are in the car.



I also noticed that aparently someone broke the diffusers below the injectors and didn't bother to tell me.

http://www.mazdatrix.com/pictures/c-intake/ableed8.jpg



Those are now flat inside the intake port. So, I suspect this may be a problem also. I've never tried to run an engine without them.







Originally Posted by FINEPRINT' post='837741' date='Sep 20 2006, 01:09 AM

I know it says im a new member but i work for Star Mazda. We have had this problem before with this setp but in a different ignition. Anyways, it sounds like more of a ECU problem. I have a MeagSq on my other race car (AE86 and it works great) but they are two totally different setups. The 13 has funky rythms in its read out that most cars ecu's cannot read or frankly speaking, dont want to, maybe if you have some prints or reads outs i could look over them and double check with our motor bulder for you. If he doesnt know than i dont know what to say! Let us know.



Matt D

www.starmazda.com

www.shukwunz@yahoo.com


That would be awsome. The guys working on the Megasquirt code right now have done an awsome job with the rotary, and it appears that my car is the only one that won't run correctly, which is why I've had so little support from that side of things. Let me know what you need and I'll get you as much information as I can.



All times are GMT -5. The time now is 11:18 AM.