2007 A3 3.2L always cranks but intermittently does not start until key is released

   #1  

Infiniteautotech

Verified VCDS User
Verified
Joined
Jan 21, 2016
Messages
12
Reaction score
3
Location
USA
VCDS Serial number
C?ID=56964
I've got an Audi A3 3.2L that will always crank, but the key must be released for it to start. This condition is worse if the car has been sitting. After the car is started, it will start normally if we start it again within a few minutes. However, if it sits a few more minutes, we are back to having releasing the key to fire the car. We started with the installation of a new battery. The new battery definitely helped lessen the symptoms. With the old battery, sometimes the car would not start at all even when the key was let go but just endlessly crank. With the new battery after the car has been sitting, the starter will crank endlessly until you let go of the key, but it always starts. Again, after the car has run, it usually can be started normally. We replaced the ignition switch first, the battery next, and now the starter, but the car behaves the same. During cranking, we are seeing 10.9V which to me seems reasonable. This voltage is also measured at the coils. Randomly, I tried to add another jumper at the engine compartment leads and the problem goes away. I see 11.5 V during cranking with the added battery, the original battery in the trunk and a high quality jumper pack in the engine compartment leads. However, I don't think this really solving the issue, but more so, the starter is spinning things so darn fast it maybe is jump starting itself?? So what would allow a car to start by only letting go of the key?

Oh, no codes whatsoever....
 
Last edited:
   #3  

HMC

Verified VCDS User
Verified
Joined
Jul 31, 2014
Messages
232
Reaction score
253
Location
UK
VCDS Serial number
C?ID=28246
"Randomly, I tried to add another jumper at the engine compartment leads and the problem goes away."
I think you have answered your own question, Your main power or earth leads are unable to carry the load. Use a multimeter connected to the battery negative and a clean suitable point on the engine and get someone to crank it and let us know what the voltage reads. Do the same from the battery positive to the main lead on the starter. This will let us know which lead/connections are faulty.
Regards HMC
 
   #4  

Infiniteautotech

Verified VCDS User
Verified
Joined
Jan 21, 2016
Messages
12
Reaction score
3
Location
USA
VCDS Serial number
C?ID=56964
I have performed quite a few auto scans on the car. There are no codes. I did change the ignition switch. That was the first thing that I did. I have verified spark and fuel too.
 
   #5  

Infiniteautotech

Verified VCDS User
Verified
Joined
Jan 21, 2016
Messages
12
Reaction score
3
Location
USA
VCDS Serial number
C?ID=56964
Regarding the added jumper, I feel the voltage with out it is still high enough that the car should always start. We are still looking at 10.7-9 volts during cranking. With the jumper, we are seeing 11.5 volts while cranking. I am moving on next to the temperature sensor. I suspect the starter fueling may to too rich upon start up. With a plug out of the car, I noticed one more ignition spark immediately after the cranking stops which explains why a car can start after the cranking stops. I can smell the fuel when the engine is cranking when the spark plug was out of one cylinder and can see the other plugs are wet when I pulled another plug. So we have spark and fuel. I suspect that the engine temp sensor is reading too cold and adding too much cold cranking fuel. Getting the voltage up is giving a better spark which helps things start especially when the engine it completely cold, but my readings show the voltage is sufficient for the most part so I don't want to blame the power. Once the car starts, everything runs perfectly! More on this when the sensor is swapped.
 
   #6  

Infiniteautotech

Verified VCDS User
Verified
Joined
Jan 21, 2016
Messages
12
Reaction score
3
Location
USA
VCDS Serial number
C?ID=56964
OK, so the problem is solved. Replaced the engine coolant sensor and the car starts correctly every time. I didn't ohm out the old sensor, but it must have been sending a bad signal making the ECU think it was very cold and producing an overly rich cold cranking mixture. Again, we noticed that there was one additional spark delivered after the engine stopped cranking each time. I have a feeling that the injection turns off, the mixture leans out a bit, and the car fires when the key is let go. All theory, but this is what solved my issue.
 
   #7  

fraazz

Verified VCDS User
Verified
Joined
Jul 29, 2015
Messages
33
Reaction score
20
Location
Canada
VCDS Serial number
C?ID=101765
OK, so the problem is solved. Replaced the engine coolant sensor and the car starts correctly every time. I didn't ohm out the old sensor, but it must have been sending a bad signal making the ECU think it was very cold and producing an overly rich cold cranking mixture. Again, we noticed that there was one additional spark delivered after the engine stopped cranking each time. I have a feeling that the injection turns off, the mixture leans out a bit, and the car fires when the key is let go. All theory, but this is what solved my issue.
I bet someone might have noticed a large difference between coolant temperature and intake air temperature if you might have posted requested auto scan...:rolleyes:


Kevin
 
   #8  

Jack@European_Parts

NostraJackAss
Joined
Jan 29, 2014
Messages
20,182
Reaction score
8,677
Location
Montgomery, NY, USA
VCDS Serial number
C?ID=57337
Actually logging temps IAT and ECT would of been the thing to do.........

Yeah when letting the ignition switch go, you get a higher tension secondary "ark" on the last fire of the plug & that's a factor of the coil being switched off for that moment..

https://en.wikipedia.org/wiki/Arc_suppression

It is possible that with an ECT failure and not switching dwell properly it could make a weak spark.

http://beru.federalmogul.com/sites/default/files/ti_07_ignition_coils_gb_2013_lowres.pdf

https://spdispark.com/pages/frequen...: What Is Dwell? (Electronic Ignition Engines)
 
   #9  

Infiniteautotech

Verified VCDS User
Verified
Joined
Jan 21, 2016
Messages
12
Reaction score
3
Location
USA
VCDS Serial number
C?ID=56964
I wasn't aware an auto scan shows the temperature readings but just the codes in each module which there were none? It would have been hard to scan the channels of the engine module not knowing what was the problem to look for.
 
   #10  

Infiniteautotech

Verified VCDS User
Verified
Joined
Jan 21, 2016
Messages
12
Reaction score
3
Location
USA
VCDS Serial number
C?ID=56964
The difficult part was thinking and what we needed to start. Now, after this experience, I would look at those values. But, I searched high and low and no one seems to have seen this situation before. Most felt it was a voltage drop issue. I hope this helps the next guy :-)
 
   #11  

Infiniteautotech

Verified VCDS User
Verified
Joined
Jan 21, 2016
Messages
12
Reaction score
3
Location
USA
VCDS Serial number
C?ID=56964
BTW, thanks for the technical explanation. That was a great read!
 
Top