Head Temp Sensor Effects

[pbanders]

I just completed a pSpice simulation analysis of the head temp sensor circuit in the ECU:

http://members.rennlist.com/pbanders/ecu.htm#CTC

The results are pretty interesting. They show that the head temp sensor's effect on mixture is fixed past temperatures of around 210 F or so (as you would expect). They also show that there are significant differences in the warm-up characteristics of the two ECU's I analyzed (037 and 052).

I'll be adding data on using ballast resistance and comparing the odd 0 280 130 017 sensor later.

[pbanders]

Whoops, corrected an error in the 052 plot due to a misinterpretation of a note on the schematic. Differences are still there between the 037, but not as dramatic as before. Refresh your browser if you've looked at my page.

[HMeeder]

Brad:

Interesting that you should post this just now as I have had some temperature sensor issues recently with my '72 1.7 d-jet.

I am now on my 3rd CHT. I had one other that was scavenged from another vehicle (the car's performance got infinitely worse. Then I took the trouble to measure the resistance 8000 ohms cold!) So I pitched it and got a new one (cold- 2500 ohms, warm- @800.) The new one didn't change the car's performance much. Car stumbles a bit on moderate acceleration, idle is rough when warm, has a slight hesitation at mid-throttle, and occassionally doesn't just stumble while accelerating, it trips, nearly stopping completely for just a hair of a second before resuming normal running. If I nail the throttle the car pulls strongly and smoothly, but steady, around town stuff causes a 'rubberbanding' like feeling. I pulled the plugs and they indicate a very rich running condition

I pulled out Kjel Nellin's article and methodically went to work.

Points- check
Capacitor- check
New cap and rotor- check
New plug wires- check
New plugs- check
Coil- check
Timing and Dwell- check
Compression 1-125#, 2-122#, 3-121#, 4-115# - not perfect but well within parameters- check
Fresh hoses, no leaks- check
Fuel pressure @ 29-30#- check
All 4 injectors clean and equal- check
New injector seals- check
Trigger points- check
AAR- check
MPS (I have a spare to swap, no change)- check
TPS, pulled apart and cleaned- not perfect but should be OK- check
All contacts at relay board, injectors, sensors and ECU clean and good- check
All grounds, especially that cluster at the back of the case- check

Still no change. This is driving me nuts! I figure that it must be the TPS since the stumble occurs around 3000rpm. I was about to order one. Then a friend tells me to disconnect Temp Sensor 1. Kjel's article says that it has little effect on the running of the car, but whoah! That changed everything! The car now idles so smoothly, the accelaration is even and strong, mid-range throttle now has no hesitation. I'm smiling again!

So I go and buy myself a new Temp Sensor 1 and hook everything back up and damned if I'm not back at square one. Disconnected the car runs great and the plugs look healthy (no unburned deposits.) The sensor is at 150 ohms which is where it's supposed to be according to Kjel's figures. The only downside is that I have a high idle when warm that I can't get below 1400 (if the car's timing is set correctly) and the idle 'wanders' a bit. It seems like the ECU is looking for that signal and not finding it.

So, (though this rambling seems a bit off-topic) the 2 sensors act like they are 'fighting' each other.

Since you seem to have made a hobby of diagnosing the d-jet (and thank god someone has!) I was hoping that you could perhaps explain to me how these 2 sensors are supposed to work together or against each other?

Since the TS1 resistance is now at infinity, what is the ECU's reaction to that? Why does it run so rich when the thing is hooked up? Any theories? Anyone? Is my brain dying? (I mean the car's!)

This has been driving me around the bend for weeks now and I don't seem to be making any forward progress (other that the TS1 discovery.)

Many TIA to every poor sap who takes the trouble to read this!

[pbanders]

OK, here's an additional list of things you should check:

1. Manifold vacuum. You should have about 15 to 17 in. Hg at idle when the motor is warm. If you're significantly lower (e.g. 12 in. Hg), you're going to run rich. Low vacuum can be caused by many things, including a non-OEM cam, valve deposits, etc.

2. FI part numbers. Use my D-Jet parts table and verify the p/n's of all of your components, including the ECU. Very often, over the years, an improper part gets installed and causes problems.

3. FI harness connections. Pull the ECU and test all of the connections as per the instructions in my D-Jet parts table. You've already looked at the grounds at the back of the case I see.

4. Dizzy vac cells. Make sure they hold vacuum.

Assuming none of the above fixed your problems, then onto your questions.

I have a section in my ECU analysis that describes the effects of open and shorted TS1 and TS2 sensors. The behavior you describe is suspicious. First, you say the car is running rich, as evidenced by the plugs, but that disconnecting the TS1 sensor improves the car's performance. Odd, since the effect of disconnecting the TS1 sensor is to increase the mixture richness by about 10% - so you're running even richer. The ECU is not "looking for the signal" - the TS1 sensor is part of a diode-resistor network and its absence simply changes the overall resistance of the network, it's not an open circuit.

The part about not being able to pull your idle down below 1400 rpm w/TS1 disconnected is also suspicious. Usually, if you cannot reduce the idle below 1400 rpm, it is an indication that the timing is overadvanced and/or you have a vacuum leak. That's why I suggested checking the dizzy vac cells, if the retard cell is leaking, that may be part of your problem. It's supposed to retard the ignition by about 7 to 10 degrees when the throttle plate is closed at idle. If it leaks, it doesn't retard, the timing is advanced, rpm goes up, and the vacuum leak increases the idle even more. A wandering idle is also an indication of lean conditions, possibly another indication of a vacuum leak.

I'd also look at where the ECU idle adjustment knob is set. I don't remember offhand if your car uses the 0 280 000 015 or a later ECU. The 015 doesn't have an idle adjustment knob, which is probably a good thing, as many people crank on it without knowing what they're doing! If you do have one with a knob, look to see if there is a mark in the plastic surround that goes around the outside of the knob. That mark is the position where the factory determined that the adjustment is balanced, and the speed control part of the ECU determines the mixture. If you're more than a few notches clockwise (rich) or counterclockwise (lean), try using the factory setting and see if there is any improvement. Note that this knob changes idle mixture ONLY - it has no effect when the throttle is open.

Lastly, the "warm" figure you give for your TS2 sensor is too high. When the engine is warm (running for 30 minutes or so), if you pull off the wiring harness connector to the TS2 sensor and test the resistance to ground of the sensor, you should be reading less than 200 ohms. Your car uses either the 0 280 130 012 or 0 280 130 003 sensor, check my table for the exact values. If you're at 800 ohms as you say, your car will be quite rich when fully warmed up.

Oh, and one more thing. If all of the above doesn't help, consider renting the VW 1218 tester from Pelican Parts. The 1218 does a comprehensive test suite that does static and dynamic testing of the TS1 and TS2 sensors, checks all ECU blocks for functionality, and also does a static functional test of the MPS.

Hope some of the above is helpful. Let me know what you find out.

[paulhagedorn]

Run out to radio shack or whatever and buy a resistor with a value around 70 ohms. Get the car good and warm, and shut it off. Disconnect the HTS and plug in the resistor. You will have to find a way to ground it, though. I ended up using a female quick-connect crimped on one end of the resistor, while the other end is grounded by a screw.

If your car runs a lot better, then it is probably the HTS.

Other than that, I would check all that Brad has listed.

[Tiramisu]

I had similar issues as Herb described..a hunting idle and a major hesistation at around 2800 rpm. I decided to see what effect disconnecting the temp sensor as Herbs friend suggested. Both problems are now gone! I hope this does not create other problems but it seems to have solved some for now. I had a stuck open AAR that I took out as well. With all these parts missing im surprised it runs as good as it does.

[pbanders]

Originally posted by Tiramisu
I had similar issues as Herb described..a hunting idle and a major hesistation at around 2800 rpm. I decided to see what effect disconnecting the temp sensor as Herbs friend suggested. Both problems are now gone! I hope this does not create other problems but it seems to have solved some for now. I had a stuck open AAR that I took out as well. With all these parts missing im surprised it runs as good as it does.

Disconnecting the TS1 (air temp) sensor richens the mixture by about 10% at about 70 deg. F air temperatures. On hot days, when the air density is lower, the effect will be somewhat greater, and on cold days, less.

Generally, it's a "band-aid" fix for some other problem. If you have a lean part-load mixture and a hunting idle, it's probably a vacuum leak somewhere that is making the mixture lean. I'd suggest hunting down the problem and fixing it, so that you can reconnect your TS1 sensor and regain the compensation for air density variation.

[HMeeder]

What Brad and Paul said makes sense now that I have had a weekend to think it over.

I think that I may have written my notes down wrong. When fully warmed TS2 reads around 50-80 ohms which makes more sense.

Some hunting with my can of carb cleaner has isolated where I believe I'm getting some vacuum leak. It is either at the throttle body to air distributor seal or the throttle bar (where it runs through the throttle body and attaches to the butterfly.) I'll pull it apart tomorrow. I'm quite positive that the rest of the system is leak free, but I'll double check it anyway. I've been dealing with a front bearing that decided to grenade at the AutoX over the weekend. The bearing galled to the spindle requiring me to cut it off. No fun when all you have is a tiny hand hack-saw and emery paper to fix it. I thought getting the car back on 4 tires should be my priority.

BTW Brad, your website is very interesting and it certainly looks like you have put a hell of a lot of thought and work into it. I do recognize that it is written in a language with roots in English but I don't happen to speak or read in that particular dialect (my brother-in-law calls it 'Electronerd' which he happens to be fluent at.) That isn't meant as a slam, I wish I knew it as well as you obviously do. I can process bits of it but I believe my cranium-bone-density is too high to allow it to all sink in.

[pbanders]

Originally posted by HMeeder
What Brad and Paul said makes sense now that I have had a weekend to think it over.

I think that I may have written my notes down wrong. When fully warmed TS2 reads around 50-80 ohms which makes more sense.

Some hunting with my can of carb cleaner has isolated where I believe I'm getting some vacuum leak. It is either at the throttle body to air distributor seal or the throttle bar (where it runs through the throttle body and attaches to the butterfly.) I'll pull it apart tomorrow. I'm quite positive that the rest of the system is leak free, but I'll double check it anyway. ...[snip]...

BTW Brad, your website is very interesting and it certainly looks like you have put a hell of a lot of thought and work into it. I do recognize that it is written in a language with roots in English but I don't happen to speak or read in that particular dialect (my brother-in-law calls it 'Electronerd' which he happens to be fluent at.) That isn't meant as a slam, I wish I knew it as well as you obviously do. I can process bits of it but I believe my cranium-bone-density is too high to allow it to all sink in.

Your TS2 numbers look good. Once the resistance has dropped below about 100 to 300 ohms or so, the mixture is no longer affected by the value of TS2 and is "locked" at the fully-warmed up level. See the graph in my ECU web page under the CTC section for more details. A vac leak at the throttle body can be pretty big, it may be due to a missing gasket or a warped plenum surface.

Yes, I know my site is pretty electronerd. I'd suggest reading the Q&A section if you don't want the hard-core stuff, it covers most of the issues I've heard about ECU's in simpler terms.

[cary]

Brad. What a wealth of information. Thanks.

73 2.0

After speed reading this post I went out and dialed my ECU back to factory spec. Because.............

I went out for a hill climb last week. Car got good and warm. But had little, to no idle at the top of the hill. Performance seemed fine.
By the time I got to the bottom, idle was better.
When I got home it seemed to be idling rich. So I came in and did a search on the board and found this post. So I went out and dialed the ECU back to the factory notch. It was almost full rich.

Now I have a poor idle between the time AAR ends and the car comes up the full temp. Thoughts ??

The entire time the air temp sensor has been unhooked.

I'm going to copy off your entire ECU article from the RennList.

TIA

[pbanders]

Cary, when you say "poor idle", I'm going to guess it's a bit unstable and low until the car is warm. Probably a bit too lean. Try turning the knob 3 clicks to the right. Let me know if that works.

Best way to do this is with a shop-quality air-fuel meter and set the CO to factory specs.

[cary]

Brad, yes that’s a better description.

After skimming thru your info. I'm going to tune the car. And replace all the vacuum lines. Remove all the variables. Before I really drive in feet first and tinker with the fuel injection.

Do you know how many feet and which sizes to replace all the vacuum lines ? Next, what kind of clamps do you recommend ? Some one mentioned, zip ties ?

CO meter ? I'm thinking of buying a GUNSON. I know its not shop quality. But much better than the ear and sniffing the exhaust. LOL. Thoughts ?

Here’s what I did last night. Cars warm. Running fine at idle.

Hooked up TS1. Idle drops from 950 to 650. Really loping.
So I decide to adjust the idle speed. Speed comes up. But idle not very smooth.
So then I check the TSI. 157ohms. Reading thru the book and posts. That looks about right doesn't it ?

So following your thoughts and research. This thing has been tinkered with, rather than corrected.

So to SMOOTH out the idle. It takes about 6 clicks RICHER to smooth it out. But who knows what the CO mix is ?
So its a wild ass guess.

Did Heathkit go out of business ? My OHM meter is a kit my Grandfather built 20 years ago.

[pbanders]

Originally posted by cary
Brad, yes that’s a better description.

After skimming thru your info. I'm going to tune the car. And replace all the vacuum lines. Remove all the variables. Before I really drive in feet first and tinker with the fuel injection.

Do you know how many feet and which sizes to replace all the vacuum lines ? Next, what kind of clamps do you recommend ? Some one mentioned, zip ties ?

CO meter ? I'm thinking of buying a GUNSON. I know its not shop quality. But much better than the ear and sniffing the exhaust. LOL. Thoughts ?

Here’s what I did last night. Cars warm. Running fine at idle.

Hooked up TS1. Idle drops from 950 to 650. Really loping.
So I decide to adjust the idle speed. Speed comes up. But idle not very smooth.
So then I check the TSI. 157ohms. Reading thru the book and posts. That looks about right doesn't it ?

So following your thoughts and research. This thing has been tinkered with, rather than corrected.

So to SMOOTH out the idle. It takes about 6 clicks RICHER to smooth it out. But who knows what the CO mix is ?
So its a wild ass guess.

Did Heathkit go out of business ? My OHM meter is a kit my Grandfather built 20 years ago.

Lot of questions, I'll try my best:

1. Hoses: get them from Pelican Parts. Dave Darling has an article in the tech area that describes the type, length and location. I don't use any hose clamps or zip ties, with the right hoses in good condition, they're not needed.

2. CO Meter: I personally don't like the Gunson. It uses an O2 sensor and has very poor performance with rich mixtures. I'd suggest instead that you watch on Ebay for an IR dispersive 2-gas analyzer, which you can probably get for about $200.

3. Idle: TS1 disconnected along with 6 clicks rich to get a good idle would normally be really rich. However, if you haven't checked your fuel pressure or supply, maybe you're just compensating for low pressure or a clogged filter. Both should be checked before proceeding. Your comments about a full system evaluation and tune-up sound good. Your TS2 reading of 157 ohms sounds right for a warmed-up engine, but assuming you have a true '73 2.0L setup and are using a 017 temp sensor, I don't know what head temp this corresponds to. Do you have a ballast resistor in series with your TS2? What ECU and MPS to you have?

4. Heathking: OOB some years ago. Sell your ohm meter on Ebay, even the parts are probably worth more than you paid for it. Get a new DMM - the best one I've found is the Protek 608, a stupendous meter for the price (as low as $189 on the web). Does everything and has a computer interface for datalogging to boot.

[cary]

Sorry just went for ride in a GT-2. Still shaking.......

TS1 was hooked back up, sorry forgot to mention that.

TS1 is what I measured, I think. The one right next to the throttle body.
Turned the OHM meter down to the 200 scale and got a reading of 156-157.

So now, the TS1 is connected. And I'm at 6 clicks rich.
Went out at 5am, damp and an ambient of 45 degrees. Started like normal ( first time runs for 5-8 seconds, then stops, restart and its fine. So, you might be on to something, with the pump and filter ). AAR finishes it's extra air. Then it slows down. But seemed to stumble a little less.

I might stop in a friends shop a get a sniff and see where I'm at. CO wise .............

Thanks for the help ...........

[Bleyseng]

Great work Brad! I have to read it slow to understand all the electronerd speak but it makes alot of sense.
How are you coming on changing the ECU so we can get alittle more power at high revs?
Geoff

[cary]

Brad, I think you're on target with fuel pressure thought.

After your input I tracked down the fuel fliter. Took it out of it holder, twisted it around a bit. Then put it back in the holder.
All this time the car is up on jackstands waiting for my new shocks to arrive.

Get the shocks installed last night about 11pm. Get in this morning, starts fine. About 1/4 mile later, big FLAT spot after I shift to second. Happens every time when shift from 1st to 2nd.
Like its starving for fuel. I have a couple filters coming from PP.

This car has run flawlessly since I bought it. Other than that cold idle issue.

[Dave at Pelican Parts]

My car did something similar to me... I had replaced some fuel lines the night before a long trip (down to Bakersfield from the Bay Area in a topless 914 is a long trip!). As I was leaving, the car did exactly the same thing as you describe.

So I pulled back into the garage and checked up in the front under the tank. Sho'nuff, one of the fuel lines was partly folded. Evidently enough to let through enough fuel to idle all day long, but when you tried to accelerate it wouldn't flow enough.

Great way to start a long trip--on your back under the front of the car, with gasoline running down your arms.... Ugh!

--DD

[cary]

Dave, mine wasn't that complicated. I forgot I tightened the throttle cable, too. I took all the slack out of it. I must have tightened it too much. Therefore throwing the throttle sensor a loop ? Causing ( after reading yours and Brad's info ) to flood out. Or not letting the decel work ???

Went out, loosened it up. Good as new. But I'm still changing the filter.

[ptravnic]

Surprised no one has mentioned using propane for testing vacuum leaks. This was a breakthrough when I went through FI hell last summer (and 4 fruitless summers before). So many possibilities for problems. Even the plenum can be problematic - my plenum had cracked along the weld - right down the middle. Anyway, spraying propane around all possible hose/air connections will hunt down leaks. If the engine suddenly stumbles then you have a leak where you are spraying the propane. And best of all, small cannisters of propane are only a few $'s at any hardware store.