Bizzar Running problem, Coil? Options

[Eric Taylor]

Alright guys, so today was a beautiful day on the coast. I had a really really good drive maybe 100 miles or so half of which were twisties (IMG:http://www.914world.com/bbs2/html/emoticons/biggrin.gif) . But a couple of times the car wanted to die like it lost spark or fuel. I had to pull over once, and I could hear the fuel pump running and but it was idiling funky and running poorly it then died, but fired right back up and ran great. Later on in the final half mile home, it ran really crappy, back firing, idiling low, and black smoke. It was really strange because it would only do this for a few seconds and then just run like a top, idle great, good power, and generally running well. So I don't really know what to think. I'm thinking it's losing spark or fuel momentarilary and then coming back some how. Oh and it's getting terriable milage, like 15mpg from a 1.7. One idea that I had was a weak coil that wasn't sending a strong enough spark all the time, and sometimes intermitent, but it's just an idea I might just be (IMG:http://www.914world.com/bbs2/html/emoticons/sheeplove.gif) with that one ( Always wanted to use that emoticon (IMG:http://www.914world.com/bbs2/html/emoticons/smile.gif) ). What do you guys think?
Eric

[goose2]

(IMG:http://www.914world.com/bbs2/html/emoticons/headbang.gif) Well Eric...wrong again...how embarrasing! There's a great site by Paul Anders dealing with D-jet FI at :

http://members.rennlist.com/pbanders/DJetParts.htm

Here's his section on the CHT sensor:

Function: Senses engine temperature and sends signal to the ECU to provide mixture compensation. Proper part for your application and proper functioning is extremely important!
Normal Value(s): 0 280 130 003 and 0 280 130 012: about 2.5 K ohms at 68 deg. F, less than 100 ohms with hot engine.
0 280 130 017: about 1.3 K ohms at 68 deg. F, less than 100 ohms with hot engine.
See Notes section below for more data on the resistance vs. temperature values of these sensors.

Failure Modes Open: The ECU interprets an open sensor as a signal to greatly richen (e.g. I've measured an over 3X effect) the mixture. This usually makes the car impossible to start and causes it to stall if the sensor fails open while running. Check by disconnecting the sensor from the wiring harness and measuring the resistance to ground, refer to the values above. Shorted: The ECU interprets a shorted sensor as a signal to lean out the mixture (about 30% leaner). The car may run and start in this condition, but will have poor idle and drivability. Check by disconnecting the sensor from the wiring harness and checking the resistance to ground. Note that shorts are often intermittent, caused by nicks in the sensor wire and by exposed contacts to the wiring harness touching ground. Check by inspection.
Stuck Value: I've heard of at least one case of the sensor being stuck at a value (e.g. 50 ohms) and not varying with temperature. Depending on the value it gets stuck at, it can result in either poor cold or hot performance, or both. Check by measuring with an ohmmeter as described above.
Mismatched: The 1973 2.0L's came with the 0 280 130 017 head temperature sensor, 039 971 762 A ballast resistor, 0 280 100 037 manifold pressure sensor, and the 022 906 021 E version of the ECU. This set of components must be used together. Any substitution will result in idle and part-load performance problems, and possible poor fuel economy. Additionally, use of any of these 1973 2.0L components with a 1974 2.0L setup will also cause problems. See the table above for the suggested setup for both 1973 or 1974 2.0L engines. If you have a 1973 2.0L and you want to keep the original setup, make absolutely certain that you have the correct combination of components. The 0 280 130 017 head temperature sensor's cold (70 deg F.) resistance is about 1200 ohms, compared to 2300 ohms for the 0 280 130 003 and 0 280 130 012 sensors. Use of the ...017 sensor with the 039 906 021 ECU (1974 model) will result in a lean mixture during warm up, causing low idle and/or backfiring on over-run. Use of the ...012 or ...003 sensor with the 022 906 021 E ECU (1973 model), with or without the ballast resistor, will result in a rich warm-up mixture. Take the extra time and determine exactly which head temperature sensor is installed in your car and make sure it matches the setup.

Notes: This resistance of this sensor is one of the primary factors in adjusting the mixture and it has a strong effect. An additional issue is the availability of the 0 280 130 012 sensor. I have found this sensor difficult to locate, and most shops substitute the 0 280 130 003 sensor for it. As far as I can tell, it is either exactly the same or nearly identical. Bosch even lists the ...003 sensor as being cross-referenced to the Porsche/VW part number 311 906 041 A. See the entry above for the intake air temperature sensor for theory on how these sensors work.
Installation Notes: Installing this sensor can be tricky. The best solution I've found is to buy a deep 13 mm socket and a 3" extension (I bought mine at Checker, about $5 for both). Use a Dremel tool with a cut-off wheel to cut off one of the corners of the four-sided end of the extension (the part that goes into the socket) to create a gap that the sensor wire can be threaded through. ALWAYS USE SAFETY GLASSES when doing this kind of work with a Dremel tool. Filing works, too, but it will take a very long time. Make sure to use the copper washer that comes with the sensor - the washer assures good thermal conductivity to the head and prevents loosening. Keeping the washer from falling off during installation can be difficult. First, to keep the sensor from being pushed back into the socket during installation (which will pop the washer off, and it'll fall into the head air fins), position the sensor so that it's sticking out a bit, then tape the wire with a single loop of masking tape to the extension. The wire will resist the sensor being pushed back into the socket. To keep the washer on the sensor, I use a couple of tiny drops of superglue to hold it in place. The glue bond will be broken when the sensor is tightened. DO NOT overtighten this sensor, just get it snug. It's easy to strip out the head threads and then you'll have to pull the engine to fix the problem.
More Data!!: Below is some characterization data I took on each sensor at three different temperatures (one data point missing). Note that these are representative readings - there is significant manufacturing variation in these sensors. All data measured with a freshly-calibrated Wavetek LCR55 meter.
Sensor Temp = 39 deg. F
(ice bath with thermometer) Temp = 61 deg. F
(room temperature)
Temp = 210 deg F
(boiling water at 1000 feet altitude)
0 280 130 003 6.10 K ohms 2.94 K ohms 199.3 ohms
0 280 130 012 NA 2.85 K ohms 191.2 ohms
0 280 130 017 3.63 K ohms 1.74 K ohms 124.7 ohms