Temperature Sensor II

by Richard Atwell
(c) Copyright 2003

How do you know if the Cylinder Head Temperature sender (aka Temp Sensor II) in your FI bus is within spec? The Volkswagen Workshop manual says:

• 14F, 7 - 12 kOhm
• 68F, 2000 - 3000 Ohm
• 176F, 250 - 400 Ohm

The Vanagon/Type II Fuel Systems ProTraining manual seems a little more precise:

• 14F, 7,000 - 11,600 Ohm
• 68F, 2100 - 2900 Ohm
• 176F, 270 - 390 Ohm

Bentley is even more vague: 2500 Ohm at 68F and not varying by too much although it doesn't explain how much is too much and it appears the value is just the split difference listed in the workshop manual. The yellow workshop manual states 2500 ohms at 68F also, but less than 300 Ohms above 176F. Jeez! Whom are you to believe and does it really matter?

Unless the part fails the obvious continuity tests how do you know if it's affecting your mixture? How do you accurately test this part? It would seem to me that the general curve is probably all that you can test within the limits specified in the various manuals for a functioning part. This is a simple component yet seems to cause quite a number of FI busses grief.

With a little stove top science I decided to find out how an old sensor compared to a new one. To do this I poured some Canola oil into a stove point because it has a high smoke point compared to some other oils. Suspending the bottoms of thetemp sensors in the oil, I proceeded to raise the temperature of the oil and monitor to with a digital cooking thermometer. At regular intervals I took temperature readings of both sensors with my ohm meter and graphed the results.

Temp Sensor II Graph:

Click on the graph to enlarge.

Analysis:

From the graph you can see that as the sensor ages, the area under the graph decreases. This has the effect of telling the ECU that your engine is actually colder than it really is and it makes the mixture more rich than necessary. The other possibility is that the two parts have different curves in the first place. I still have many more TS2 to test before I can make further conclusions.

Swapping out my own old sensor for a new one, idling during warm-up has become much smoother although the difference in values between the two fall within the published test guidelines. After warm-up, idle is slightly smooth and it has eliminated the hunting in the timing that still lingered even after I switched to an electronic ignition setup. I wouldn't have suspected that after warmup a 10 Ohm drift at 275F could cause your engine to run a little rough if your ear is attuned.

The hardest part of testing your sensor is figuring out the ambient temperature. The engine takes a long time to cool down after shutdown so if you take a reading after 3 hours it won't be at the ambient temperature. Touching it with your figures will raise the temperature also. The best technique is to remove it from the engine, take it indoors, wait 4 hours then check the air temperature with a thermometeruntil you get a reading near 68F.

I'd like to graph a known troublemaker and see how it compares. If you have one that doesn't work but passes the continuity test I'd like to have it.

Workarounds:

Recently I came across a section in the ProTraining manual that said, "A temperature spacer 022 133 079 is available for cars which run well on initial start-up and when fully warm, but have performance problems only during warm-up. This spacer does not affect the resistance values when the engine is cold or when it is fully warm, but it does delay the change in resistance during warm-up, The spacer temporarily insulates the temperature sensor housing from the heat of the cylinder head".

This is a common problem and I've experienced it myself. When I researched the issue I found no information only stories from other owners with the same issue. It seems many are content to suffer through the warm-up until the bus is at operating temperature when it seems to operate correctly.

I always wondered what item #14 was in the fiche alongside the temp sensor (item #41) and it appears to be this spacer. When I get my hands on one I'll try it out. The aircooled Vanagon CV engine has no such spacer available but the parts diagram for the FI Superbeetle does.

Part numbers:

There is a lot of confusion surrounding the various part numbers for this part. Only two Bosch parts are available and I don't know how they match to the original VW numbers except by querying the various part catalogs of the online vendors.

There is a Type 4 Workshop Bulletin from 23 Jan 1974 concerning "Backfiring and poor output during engine warming-up". The remedy is to install a new temp sensor II, 311 906 041A.

It states, "The following list shows which sensors are to be used with what model year. It supplements the information in Workshop Manual K 7.2/1".

From this information I can't see what the exact difference is between these parts but it implies that the 311 part has a higher resistance value than the 022 part and makes the engine think it's colder and injects more fuel. I don't know what the individual Bosch part numbers are and of course new Bosch parts have no VW number stamped on them so it remains a mystery. I took a survey of the parts available for the various models:

I was surprised to find the 022 906 041 part installed in my 1978 engine in the first place because it seems to only have one model year 411 in common with it. It's the only number I can match up to a Bosch replacement part which is listed as an exact replacement in the Type2.com fuel injection cross reference.

Also interesting is the D-Jet parts reference provided by Paul Banders. It in he lists all three known VW part numbers and describes the conflict in the cross reference after doing his research. It's reassuring to know that I'm not going crazy but it still doesn't provide a definitive answer on the issue unless the parts are exact with different VW numbers which I have to doubt.

Ronnie at Old Volks sent me a VW Bosch mapping and with Paul's info it seems to largely sort out the confusion although I have no idea how each sensor performs in the bus other than the stock sensor.

History:

02/13/03 - Created
02/20/03 - Added Type 4 information
10/19/03 - Added 914/Vanagon info and ProTraining numbers
10/22/03 - Added dizzying array of part numbers