My 86 635, 99,000 miles has been failing smog tests on being too rich.Car idles badly too high and erratic.
I have checked all the usual suspects, discoed cold start valve, have swapped AFM for known good one, CTS, ICV, on and on.
Can get it to run if I pull and clean plugs and let it air out.
One mechanic said drippy fuel injectors ( changed ), another said new 3.0 fuel pressure was running at 2.5 bar, changed back to original Bosch 3.0, still way rich.
I'm stuck on this one, all 6 plugs ( correct ) are equally black and wet ,shouldn't be injectors.
Any ideas?
I'll remember you in my will!
Jim
Sounds like you've R&R all the usual suspects. Did they pull out a DMM to verify proper BMW spec readings? IMHO very important. How healthy is your dist cap, rotor, coil and primary coil wire?
May still be a vac leak, or an exhaust leak upstream of the O2 sensor? You haven't tried a smoke test, have you? There's a lot of sneaky little gaskets than can fail on the M30. My fave is the oil seal on the distributor drive nose.
Good points Dave, all the parts are new.
I'm in Denver and smoke test people seem to be rarer than they should be.
I probably should look for a smoke test, I keep thinking this is mechanical/electrical in nature but a leak would send a message back to the computer to richen it up.I think I have checked most of the other boxes.
Perhaps try a wideband with air fuel gauge and an adjustable fuel pressure regulator with pressure gauge. Then U can experiment with the airflow meter spring and hex key to get the mixture right. It’s amazing what u can get with a little fiddling. Using slighly bigger injectors and reducing the fuel pressure has a nice little bonus of increasing the part throttle timing.
BMW’s
90 E34 M5
84 E24 M635csi standalone ecu with coil on plug
94 E34 540i/6 SC 5-17psi Flex fuel/standalone ecu
97 Z3 2.8
97 E36 M3 euro SC still u/c
OTHERS
11 Audi S5 APR stg2
19 Volkswagen Amarok V6
Sounds like failing FPR or Leaking Cold Star Injector. You said you replaced the FPR, so I would tend to think the CSI is the problem... Not easy to get at or replace, but maybe you can clamp off the supply line with it running and see if anything changes when you rev the engine...
I have tried 2 FPR's and plugged CSI totally and still no joy.
I'm thinking a smoke test is in order, an adjustable FPR with an air/fuel meter would be nice but I can get most of that info by tapping into the O2 sensor.
Did you pull the ECU? I have the same car, the ECU's are available and cheap, I think most of us carry a spare so when an issue arises, you can swap out and exclude it as the cause. Also check the reference sensors. Some here say the car won't run if they are reversed but mine did, and ran rich.
Rob E3
All my modified cars run a wideband with and air fuel gauge as it takes the guesswork out of things. If u are tuning it disconnect the narrowband sensors. It may be a good idea to replace them.
BMW’s
90 E34 M5
84 E24 M635csi standalone ecu with coil on plug
94 E34 540i/6 SC 5-17psi Flex fuel/standalone ecu
97 Z3 2.8
97 E36 M3 euro SC still u/c
OTHERS
11 Audi S5 APR stg2
19 Volkswagen Amarok V6
The 2 reference sensors have been switched and then returned, I believe I have them correct.
Have buddy with identical car, grabbed his ECU and tried that, no help.
Signed up to have a smoke test done, one thing I can't do at home.
Will keep you posted after May 9th.
https://jet.com/product/detail/6c2e2...f21acd50ed8f6c
I used this for a smoke test and worked real good.
Just get this and smoke chips,a small cleaned soda bottle.
Use top of bottle to connect to smoker spout with a small clear hose.
Cut bottom of bottle and tape to afm that is held open. Do not run the engine
and pump in the smoke.
Last edited by 1986series6; 04-29-2018 at 05:50 PM.
Great idea , I have a buddy who has some hives and I'm sure has a smoker.
I've seen them in action and they do smoke.
Will keep you posted.
Still having problems, smoke test with bee smoker and also smoky seafoam test revealed no air leaks.
However TPS, which normally doesn't go bad I hear, seems to be giving me bad readings.
Ohm test:
Idle ohms should be 2.0 actual is 4.2
Floored ohms should be 10 actual is .62
Voltage test
Idle volts should be .7 increasing to about 5.0 when floored.
Idle actual is 4.52 decreasing to .9 when floored.
This TPS ( 1986 auto ) is the one with a rubber boot and Bosch part 0261211001 for the potentiometer.
I would dig into the ECU as soon as I find out which ECU pin is getting the reading from the TPS to double check.
Now I'm stuck, there seems to be one for sale in the entire universe and mightily priced at $150 used.
Anybody else run into this??
I've had a lot of trouble testing those automatic variable TPS, now that you mention it. I believe it's two discrete parts under the rubber boot. There's an on/off contact switch for idle and a variable potentiometer for the throttle angle.
My buddy has an automatic parts car, I'll check if the TPS is available so you can swap it out for testing. Give me a few days, he's a pretty busy fella.
Dave, it's the variable potentiometer part that has me puzzled.
Doesn't ohm out right, I've had mechanics say leaky injectors ( replaced ) or bad FPR replaced.
It runs but rough and rich, I pulled the tiny hose for the cabin temp sensor and it raised the idle to 1000 and smoothed it out.
Can't help but think that some sensor is sending the wrong info to the ECU.
Many thanks on checking with buddy on TPS.
Jim
You pulled the cabin temp sensor and it smoothed out? That sensor triggers the A/C system...the DME compensates for A/C by raising the idle...could be a wild goose chase...could be a clue...
I ggogled that number and came up with this:
https://www.epartsland.com/products/...-0-261-211-001
DSCF7088_large.jpg
That doesn't look like the earlier ones I'm familiar with. I suspect that part would have been fitted to a lot of different cars, although likely with a different plug or pigtail. RockAuto has a cheap one listed for only $67:
https://www.rockauto.com/en/catalog/...sor+(tps),5136
Oddly, it's not listed for an '86, but the part that is listed ($426!) also cross-references to the '89. Looks like a little confusion there.
Good advice!
I pulled off the cabin sensor hose off at the bottom of the manifold, which gave it a small air leak that leaned out the air/fuel ratio.
Another option is maybe to get the other more common non boot TPS and then splice the connectors.
I had a little time to play with this today.
Idle speed is a rough 950 with the TPS connected, give it the gas and the tach will drop to about 700 and then rise to 950.
From my old carb tuning days this means too rich.
Idle speed is a rough 950 with the TPS disconnected, give it the gas and the tach will drop to about 700 and then rise to 950.
Apparently it makes no difference if the TPS is connected or not, this is at all RPMs.
Pulling the small cabin sensor hose to induce a small leak does make it run better.
Somehow the ECU is either not getting the correct signal and running open loop ( defaulting to rich )
or
the TPS is sending an all rich signal as in 70% throttle.
To Tom Schultz, semi neighbor in Colorado I sent you a PM.
I saw you checked the coolant temp sensor. You're 100% sure it's the correct one, right? No plugs switched, inline resistor cracked (sometimes found inside wiring boot), wiring issues, etc.? Did you check the air temp sensor? It's in the AFM. You'd swapped them with known good units so I didn't suggest it before. There are some variations in those AFMs, IIRC there are two different types that look identical from the outside. The '86 ran Motronic 1.0, should be an 059. The matching AFM should be a Bosch 027.
I know I'm going back to basics, but these Bosch systems are really prone to running problems if the sensors are out of spec. They're also bedeviled by the usual grounding problems, sensitivity to air leaks, water in connectors, etc..
Tom,
I thought I could PM you but I can't figure out how, may not have permission.
I wanted to know how you got on Bigcoupe.com I have tried maybe 8 times to join ( even emailing them directly ) as I want to put this post in front of 2 members there, Brucey and Chris Wright who seem extremely knowledgeable.
Don't have a Bentley, been working off this terrific link http://www.hiperformancestore.com/Motronic.htm for info, also another website agrees with this.
Can't figure out why the potentiometer part of the TPS starts with a high voltage and then decreases instead of the other way around, same with ohm test.
I know the specs and how to backprobe the connector while car is ON to get readings but they are weird.
I've had tii's, about eight 2002's, a 3.0CS,a 2500, an 87 L6 and never had this much trouble.
Dave,
It's 95 degs in Colorado, need to get up early and try it all again.
I will test AFM air temp sensor with 9 volt battery trick and remove and contact clean all connectors, get out ohm specs for all connectors and check them again.
I still have old CTS and will ohm that against new one, I know the inline resistor trick and it doesn't have one.
May have to kludge something to fool the ECU.
Thx guys!
Try emailing brian@bigcoupe.com for a direct response.
I know you have had a lot of issues but you have made so many posts but you haven't gone through methodically and shared your process of elimination. Have you tried any of these parts in a known good running car? That's the surefire way to verify parts...
So far I have:
1.Replaced wires and distributor cap. Fine
2.Replaced coolant temp sensor, old one ohms out same as new one at different temps. Fine
3.Ohmed out cabin temp sensor.Fine
4.Lubed and checked idle control valve, vibrates at idle, opens and closes fine.
5.Plugged cold start sensor at fuel rail .
6.Have different AFM's, one I use as a base is out of friend's 86 635 ( never been opened ) his 635 runs fine.Have "mapped " one spare to correct voltages by adjusting potentiometer sweep arm to new carbon track and slightly adjusting sweep arm to read lower voltages.This has helped slightly.
7.Plugged in friend's ECU, no change.
8.Smoke test and seafoam treatment doesn't show any leaks.
9.Installed new correct sprk plugs, they go black and fluffy in minutes just like when I clean the old ones.
10.Installed new FPR, no change from old one.
11.Reference sensor and speed sensor ohm out correctly, if I switch them car doesn't start, so they seem to be working.
12.Have contact cleaned every connection found and have wire bushed all ground connections.
Introducing small air leak at manifold ( disco cabin temp sensor hose ) makes it idle higher 1100 and smoother.
TPS is a problem , linear voltage readings and ohm readings go in the wrong direction .My friends 635 is a manual and he lives 70 miles away.May have to trip up there.
Ohm test:
Idle ohms should be 2.0 actual is 4.2
Floored ohms should be 10 actual is .62
Voltage test
Idle volts should be .7 increasing to about 5.0 when floored.
Idle actual is 4.52 decreasing to .9 when floored.
Tests done on 2 different DMM's give same readings.
Really stuck on why I'm running rich.
I wouldnt stuff about replacing everything.
Personally I would do a few things.
1/ (high priority) wire in a wideband sensor and air fuel gauge. AEM NOT INNOVTIVE.
(Log whether it’s rich everywhere or just certain rpm and loads
2/ replace injectors has they are cheap enough.
If still rich.
3/ adjust spring tension on air meter.
If still globally rich install an adjustable fuel pressure regulator with fuel pressure gauge.
If still rich well replace the Dme.
Spmetimes u have to give the engine what it wants.
Its very very hard to diagnose without the wideband. I would have done this first.
If its its globally rich like my m635csi was when I changed the air meter I temporarily mixed the fuel with e85 as it likes about 9:1 rather than 13-14 but I can only do this with a wideband. I would run e85 as a permanent fix as the moisture won’t like the steel fuel tank but it will be ok for short term.
BMW’s
90 E34 M5
84 E24 M635csi standalone ecu with coil on plug
94 E34 540i/6 SC 5-17psi Flex fuel/standalone ecu
97 Z3 2.8
97 E36 M3 euro SC still u/c
OTHERS
11 Audi S5 APR stg2
19 Volkswagen Amarok V6
Good process. As Gaz suggests, the major item affecting fuel delivery here is injectors or o2 sensor. If the o2 reports nothing, the car should run based on default map. But if we have the wrong o2 sensor and it is reporting too lean (when the engine is in fact stoich), it will add fuel until the sensor is happy. Have you tried disconnecting the o2 sensor? Unplugging the battery will reset the memory of the ECU.
As you mention, TPS could be bad, but I am not as familiar diagnosing auto TPS cars. You can measure the signal and see if you get a smooth range from 0-5V on the signal wire when actuating the throttle. It sounds like you need to try that TPS on a known good car and see if it still doesn't work right...
Next up is injectors. If those don't work, we have missed something in our testing, which could point to a bad connection from the CTS to the ECU or something similar... Those are the major controllers of fuel: CTS, O2, ECU, AFM, TPS, and I would test them in that order.
Plugging your parts into a known good car will confirm more for you than his parts into your non-functional car. I had a similar problem with a different car and determined it was the ECU. Although I am still working on resolving the problem with a new wire harness.
Last edited by tschultz; 07-11-2018 at 10:23 AM.
Bookmarks