944 Turbo: engine shutting off in mid drive
05-23-2015, 11:45 PM
#16
This past weekend I replaced the AOS and both the DME temp sensor and the gauge temp sensor and now this happens far more frequently than it did before.
When the car is around operating temp or above (according to the AES gauge I installed in the radiator hose that comes out of the engine and into the radiator just after the bleeder bolt) and either I've taken the car out of gear and am rolling up to a stoplight or idling in neutral, the engine shuts off and I have to restart it.
The greatest chance of reproducing this is:
1. Get it up around operating temp
2. Take it on the highway
3. Get off the highway and as I'm coming to a red light, put the car in neutral and the engine shuts off.
Does the change of the AOS and temp sensors making this problem noticeably more frequent shed any new light on the situation (fwiw the dealer couldn't determine the cause of this problem the first time but now it's easier to reproduce) ... ?
When the car is around operating temp or above (according to the AES gauge I installed in the radiator hose that comes out of the engine and into the radiator just after the bleeder bolt) and either I've taken the car out of gear and am rolling up to a stoplight or idling in neutral, the engine shuts off and I have to restart it.
The greatest chance of reproducing this is:
1. Get it up around operating temp
2. Take it on the highway
3. Get off the highway and as I'm coming to a red light, put the car in neutral and the engine shuts off.
Does the change of the AOS and temp sensors making this problem noticeably more frequent shed any new light on the situation (fwiw the dealer couldn't determine the cause of this problem the first time but now it's easier to reproduce) ... ?
05-24-2015, 11:29 AM
#17
Today it happened when the engine was at 137 degrees going up a hill and I believe I was in gear (2nd).
The RPMs were low which is leading me to believe it may have something to do with the idling.
In neutral it seems to be set for 400 rpms (lowest mark on the RPM gauge) but it's not steady, It goes up to about 459 then dips down to 350 pretty much on a one second interval so it's up, down, that's one second, etc.
I'm going to try to increase the RPMS upon idle to see if that makes a difference. When I put the fuel rail back on over the weekend, the injectors seemed to fit right into place, is it possible once more more of the injectors isn't seated properly and that's what's causing this?
The RPMs were low which is leading me to believe it may have something to do with the idling.
In neutral it seems to be set for 400 rpms (lowest mark on the RPM gauge) but it's not steady, It goes up to about 459 then dips down to 350 pretty much on a one second interval so it's up, down, that's one second, etc.
I'm going to try to increase the RPMS upon idle to see if that makes a difference. When I put the fuel rail back on over the weekend, the injectors seemed to fit right into place, is it possible once more more of the injectors isn't seated properly and that's what's causing this?
06-07-2015, 12:34 PM
#18
I think I know why this happens but don't know how to fix it.
I increased the idling from 400/500 to 800/900 and that definitely helps the situation and I noticed when this happens it's because the RPMs dip too low (below the lowest mark) so I'm thinking it's not the engine that's shutting off, it's the car that's stalling out. It only seems like the engine is shutting off because I'm either in neutral or idling with clutch depressed so it doesn't jump like a typical stall when letting out the clutch in 1st.
any ideas on how to keep the RPMs from dipping down so low that the car stalls?
I increased the idling from 400/500 to 800/900 and that definitely helps the situation and I noticed when this happens it's because the RPMs dip too low (below the lowest mark) so I'm thinking it's not the engine that's shutting off, it's the car that's stalling out. It only seems like the engine is shutting off because I'm either in neutral or idling with clutch depressed so it doesn't jump like a typical stall when letting out the clutch in 1st.
any ideas on how to keep the RPMs from dipping down so low that the car stalls?
06-07-2015, 08:04 PM
#19
Rainman
Rennlist Member
Rennlist Member
I think I know why this happens but don't know how to fix it.
I increased the idling from 400/500 to 800/900 and that definitely helps the situation and I noticed when this happens it's because the RPMs dip too low (below the lowest mark) so I'm thinking it's not the engine that's shutting off, it's the car that's stalling out. It only seems like the engine is shutting off because I'm either in neutral or idling with clutch depressed so it doesn't jump like a typical stall when letting out the clutch in 1st.
any ideas on how to keep the RPMs from dipping down so low that the car stalls?
I increased the idling from 400/500 to 800/900 and that definitely helps the situation and I noticed when this happens it's because the RPMs dip too low (below the lowest mark) so I'm thinking it's not the engine that's shutting off, it's the car that's stalling out. It only seems like the engine is shutting off because I'm either in neutral or idling with clutch depressed so it doesn't jump like a typical stall when letting out the clutch in 1st.
any ideas on how to keep the RPMs from dipping down so low that the car stalls?
06-08-2015, 01:19 PM
#22
Addict
Rennlist Member
Rennlist Member
Sounds like a vacuum leak to me. A small leak is a tiny percentage of the total when running at higher RPMs but is a much larger percentage at idle, resulting in a lean condition and stalling.
I'd recommend pressurizing the intake tract to 15 psi or so, listening for leaks and then spraying soapy water in the direction of the leak and looking for bubbles. Arnnworx sells a Boost Test Adapter or you can fabricate your own from a PVC end cap and a Schraeder valve.
I'd recommend pressurizing the intake tract to 15 psi or so, listening for leaks and then spraying soapy water in the direction of the leak and looking for bubbles. Arnnworx sells a Boost Test Adapter or you can fabricate your own from a PVC end cap and a Schraeder valve.
06-08-2015, 09:27 PM
#24
Burning Brakes
Reflow the solder joints on the Motronic, use thermal grease to make sure the hot ignition transistor (round silver thingy in the pic in post #7 above) has proper thermal conductivity to get the heat out, through its heat sink, and to the aluminum box chassis.
Those ignition transistors fail over time, esp. when they overheat, such as happens with warmer weather in, say, June. Car starts and runs fine from cool startup, but maybe 20 mins. later she just quits. Won't restart until the ignition transistor cools down, then quits another 20 mins. down the road. (Hairline crack opens with thermal expansion, no juice flows until she cools, crack closes, juice flows again, she runs again. For a while.)
New ignition trigger transistors cost $3 or $4 each at DigiKey or other electronics supply houses, probably from Radio Shack. Cross reference by the Bosch part number stenciled on the round silver thingy cap.
The three little legs on that transistor are the prime failure points, where they solder into the pcb.
It's possible you have a dodgy ignition switch AND a dodgy Motronic. My BMW 635csi has virtually identical Bosch fuel injection to my Porsche 944 Turbo, and the Bimmer had the vexing problem of sometimes starting (ig. switch sometimes worky, sometimes no worky) and running fine for 20 mins. (overheated Motronic transistor), so it was "fun" to sort that out.
If the shop did not open the Motronic and reflow the solder joints of at least the ignition trigger transistor, they're essentially telling you they don't know what they're doing.
Harbor Freight $8 hobby kit #38593 is your tool of choice, essentially a soldering pen with a bunch of interchangeable tips. Use pointy one. Pen gets hot enough to instantly melt solder without heating the surrounding circuit board, so you just touch the joint, being careful not to drag solder across and make unwanted connection to the next joint. Good lighting, a steady hand, and a benchtop magnifying glass are very helpful.
PS: If she won't start or quits when running, do the spank test: While cranking with the key, slap or spank the Motronic box, which may jiggle it enough to momentarily close a hairline crack and the engine will catch. On my Shark we noticed that slamming the driver door while cranking the engine would make it start--the shock of the door slam evidently closed the hairline crack in the Motronic. In cars of this vintage, reflow of solder joints and replacement of the ignition trigger silver thingy should be standard procedure, along with carrying a known good spare Motronic in the glove box.
I'm told that Bosch later modified their Motronics to better dissipate heat from the ignition trigger transistor, by mounting the transistor outside the case. I may use aluminum foil tape to make fins to dissipate heat better from the case, but in any event will add thermal grease to the heat sink connections. If the factory put any such grease in any of the 4 or 5 Motronics I've opened and reflowed, there was no sign of it. I scavenged several alu finned heat sinks from old discarded computers, and may mount them to the Motronic boxes. Laptop computer cooling fans are an option--but probably overkill if the other procedures are done.
Those ignition transistors fail over time, esp. when they overheat, such as happens with warmer weather in, say, June. Car starts and runs fine from cool startup, but maybe 20 mins. later she just quits. Won't restart until the ignition transistor cools down, then quits another 20 mins. down the road. (Hairline crack opens with thermal expansion, no juice flows until she cools, crack closes, juice flows again, she runs again. For a while.)
New ignition trigger transistors cost $3 or $4 each at DigiKey or other electronics supply houses, probably from Radio Shack. Cross reference by the Bosch part number stenciled on the round silver thingy cap.
The three little legs on that transistor are the prime failure points, where they solder into the pcb.
It's possible you have a dodgy ignition switch AND a dodgy Motronic. My BMW 635csi has virtually identical Bosch fuel injection to my Porsche 944 Turbo, and the Bimmer had the vexing problem of sometimes starting (ig. switch sometimes worky, sometimes no worky) and running fine for 20 mins. (overheated Motronic transistor), so it was "fun" to sort that out.
If the shop did not open the Motronic and reflow the solder joints of at least the ignition trigger transistor, they're essentially telling you they don't know what they're doing.
Harbor Freight $8 hobby kit #38593 is your tool of choice, essentially a soldering pen with a bunch of interchangeable tips. Use pointy one. Pen gets hot enough to instantly melt solder without heating the surrounding circuit board, so you just touch the joint, being careful not to drag solder across and make unwanted connection to the next joint. Good lighting, a steady hand, and a benchtop magnifying glass are very helpful.
PS: If she won't start or quits when running, do the spank test: While cranking with the key, slap or spank the Motronic box, which may jiggle it enough to momentarily close a hairline crack and the engine will catch. On my Shark we noticed that slamming the driver door while cranking the engine would make it start--the shock of the door slam evidently closed the hairline crack in the Motronic. In cars of this vintage, reflow of solder joints and replacement of the ignition trigger silver thingy should be standard procedure, along with carrying a known good spare Motronic in the glove box.
I'm told that Bosch later modified their Motronics to better dissipate heat from the ignition trigger transistor, by mounting the transistor outside the case. I may use aluminum foil tape to make fins to dissipate heat better from the case, but in any event will add thermal grease to the heat sink connections. If the factory put any such grease in any of the 4 or 5 Motronics I've opened and reflowed, there was no sign of it. I scavenged several alu finned heat sinks from old discarded computers, and may mount them to the Motronic boxes. Laptop computer cooling fans are an option--but probably overkill if the other procedures are done.
Last edited by Dash01; 06-08-2015 at 09:47 PM.
06-09-2015, 05:03 AM
#25
Racer
Join Date: Oct 2014
Location: Sonoma County
Posts: 288
Likes: 0
Received 0 Likes
on
0 Posts
There's photos somewhere of that very mod. Someone added a CPU heatsink AND a fan to the DME box. The injector-driver is rated for 16-amps max. Typical design philosophy is to drive them up to only 50% of max. In the DME, they are driven to 10-amps, 63% of max. Not horrible, but does make them more sensitive to heat issues and shortens lifespan.
Due to the Y-split leaving the case, I suspect the DME was originally designed with dual injector-drivers in mind for bank-fire. But somewhere along the way, the bean-counters probably won out and a single-injector driver was implemented.
Due to the Y-split leaving the case, I suspect the DME was originally designed with dual injector-drivers in mind for bank-fire. But somewhere along the way, the bean-counters probably won out and a single-injector driver was implemented.
