Ignition retarded - Knock sensor?

Hi!

The ignition system on my 740 GL -87 (B230K) is behaving a bit strange. I checked the ignition timing yesterday and it was very late, way off the scale. I would say around 5 degrees *after* TDC.

Some previous owner of the car has disabled the trottle microswitch by tying it closed in idling position with a bit of wire. When removed, the cars runs very bad and this was confirmed with the timing lamp. With the switch open, the timing is retarded even more. This is very strange, as the switch would retard the timing in the idle position - not when the gas pedal is depressed. Could anyone confirm if their switches are of the normally open or normally closed type? Removing one lead from the switch gets the same result as depressing the gas pedal slightly, which suggests that my switch is closed in the idling position.

Aside from this, I cannot adjust the timing to 15 degrees before TDC either with the switch open or closed. When adjusting the distributor I can advance the timing to around 8 degrees before TDC, where the engine seem to have more power and are more responsive. But when going for a test drive there was excessive pinking under load, which maybe is quite natural since the distributor now was adjusted to the most advanced timing possible. I tried adjusting the timing to around 2 degrees before TDC, but there was bad pinking there too. So I'm back at 5 degrees after (with slight backfire instead) and not sure what to try next.

Could it be a bad knock sensor, that either tells the engine to retard fully or not sensing any knocks at all? Does the ICU for the B230K have these error codes that is talked about in the 700-900 FAQ, or are they only available for the fuel injected cars?

Best regards

/Anlo

http://pic7.picturetrail.com/VOL188/...8/23744414.jpg


Two things come to mind.

The crankshaft pulley on the B230 engine incorporates a torsional vibration damper.
The outer part of the pulley with the drive belt grooves and timing marks on it is bonded to the hub of the pulley by a layer of rubber marked in the picture below this shows the back of the pulley but you can also see it at the front.
http://pic7.picturetrail.com/VOL188/...8/48390821.jpg
These are known to come unbonded causing the timing marks to be in the wrong place relative to the crankshaft.

If this is the case with yours your timing marks won`t relate to the actual timing.

The best way to check this is to put a line of tippex or paint across the face of the pulley across the join.
Start the engine, switch on lots of electrical load and hold the steering on full lock for a few seconds.

Stop the engine and see if the line is still in one piece, if it`s moved you will see immediately and you will need a new pulley.

Also when checking the timing on these engines you need to be careful,
the pointer is quite a long way from the marks and the engine is canted over at an angle.
To get accurate readings you need to be looking down an extended line
drawn from the centre line of the crank and passing through the pointer.

I suggested these checks because I find it hard to believe that you are getting bad pinking if the timing really is at only 2 deg BTDC.

There are no on-board diagnostics on your car, it`s only on the later fuel injection models.

Colin.

1990 740SE B200E/M47, remote C/Locking.

Yes, I have also had my doubts regarding the timing. It seems like the car would hardly run at all at 5-10 degrees after TDC. I have checked the timing mark on the crankshaft pulley by removing the spark plug from cylinder one and checking where the piston is at its top-most position. When the piston is at TDC, the timing mark aligns with the zero degree notch on the cambelt cover.

I also removed the upper half of the cambelt cover, the markings on the camshaft drive wheel and in the cover align when cylinder one is at TDC. So it doesn't seem to be a case of misaligned camshaft either (if this could cause this kind of bevahior?).

I might have read the timing a bit "late", since I was looking straight down (vertically) instead of aligning the scale against the crankshaft center. But I don't think that there would be such a large error in the reading. Even when the distributor is adjusted to its most advanced setting, the mark is still of the scale.

Very strange, this...

/Anlo

Use a timing light and rev the engine. If the knock sensor is faulty, the indication will be that the ignition is retarding rather than advancing. I couldn't believe it when I saw it! Thought the engine was running backwards. A knock sensor from my parts car cured the problem.

The throttle closed switch is important for better fuel economy as it cuts off the fuel supply to the carb on overrun and also on switch off, to avoid running on (dieseling).

Hmm... If I have the trottle switch operating (not tied in with the wire), the timing do retard when I rev the engine. I only need to make the microswitch open (or disconnect one lead) to retard the timing. I had an idea that someone possible had replaced the microswitch and used a normally open type instead of a normally closed one (since the timing should be advanced, not retarded, when the gas pedal is depressed).

But maybe I've been fooled by the knock sensor, I guess it is not used when the idle switch is closed and the timing isn't retarded with regards to the sensor until the switch opens.

Bad knock sensor or not though, with the engine at idle (microswitch closed) I cannot adjust the timing to 15 degrees before TDC. It seems strange that the timing should be set in the "idle retard" mode, but I haven't seen any information that the microswitch should be disconnected when timing the ignition either...

/Anlo

I'm trying to figure out a way to check if the knock sensor is the source of my problems instead of just replacing it. Assuming a functional sensor - when disconnecting the wiring from the sensor, ignition timing should go to full retard because of no signal sensed, right?

If the timing doesn't change (with trottle microswitch disconnected or gas pedal depressed slightly), the knock sensor or its wiring should be suspected.

Any other ideas for isolating the retarded timing problem is welcome. I read about connecting a LED and a resistor between +12V and sensor output and checking for flashes while driving. I'm not sure of what that indicates though, does the LED flash when the sensor detects a knock?

Best regards

/Anlo

[Peter Milnes]

The sensor is a piezo-electric microphone which produces an output when it detects the sound of a "ping" or ignition pre-detonation. This is why it is placed at the top of the block where the detonation occurs. The retardation occurs as a result of detonation and thus disconnecting the sensor would make the "pinking" more apparent. When the sensor is picking up the sound of detonation it will feed a signal to the ECU/ICU to retard the ignition depending on the strength of the detonation. If your igniton is seemingly permanently retarded then I would suspect the ECU or ICU but not the sensor, unless it is permanently producing signals (possibly random noise due to the earth becoming disconnected from the sensor wiring).

All the best, Peter

[davebslater(uk)]

if you need to check sensor operation , you could tap the block near the sensor with a metal bar , while somebody checks if the timing retards , of course you might need to rev the engine to do that . you could also put a false load on the vacuum tube to see if that alters anything . more than anything though i would always suspect those renix ignition units .

First of all, a question: How does one remove the connector from the ICU under the instrument panel (driver side on my LHD)? I can disconnect it and drop it at the rear, but it is still attached to the ICU in the front (towards the engine bay). Is there some additional retainer I have missed, and how is it released? The connector is quite loose, but I can't find out how to get it off the ICU... :-)

Now, the main reason for suspecting the knock sensor is that the ICU goes into full ignition retard if no valid signal is received from the knock sensor. It makes no difference if I have the sensor connected or not when driving. Rather than sending "knock signals" all the time, it's dead altogether.

Having done some tests this weekend I'm now quite sure that the knock sensor is at fault. If the ICU is bad then all bets are off, but replacing the sensor seems like a good approach since the cost involved in fixing the ICU probably is much more.

The tests I did, in case anyone else has similar problems:

1. Visually checked the wiring and connector, they seem fine. However, the connector was black - not green as described in the FAQ.

2. Measured the resistance over the sensor. Should be over 1.5 MOhm (around 4-6 MOhm) but was infinite (at least >50 MOhm). However, information in the FAQ suggests that this is not a realiable test on a used sensor.

3. Measured voltage at sensor with ignition on. Should be 1.5-5.0 Volt, was 2.3 Volt. (ICU not all bad.)

4. Connected a LED and resistor in series between +12V and red/yellow wire on left strut. Ignition on: should be lit, was lit. Engine idling: should not be lit, was not. When driving, flashed four times which suggests bad sensor or wiring.

5. Advanced timing as much as possible, sensor in place - no pinging. Disconnected sensor, still no pinging or any change in bevahior.

The one test I haven't done is to disconnect the sensor and the ICU connector to measure for continuity and shorts between ICU and sensor, since I didn't manage to remove the ICU connector...

I think that most evidence points in the direction of the sensor. One possibility is that the connectors for the knock sensor and the engine temperature sensor has been switched. This will be dismissed or confirmed by the continuity test from the ICU connector, if I can remove it.

Best regards

Anlo

>I suggested these checks because I find it hard to believe
>that you are getting bad pinking if the timing really is at
>only 2 deg BTDC.

When working on the car this weekend, I fixed the throttle microswitch to open as it should. When adjusting the ignition to the most advanced setting possible, there's no pinking!

Turns out the bad pinking I had when I advanced the timing before was because this switch was tied in the closed position. I read somewhere that ignition is adjusted to give maximum engine brake when the throttle switch is closed and engine speed is above idle. So having the ignition adjusted for coasting and still be under load was no good combination.

The car is much more driveable now, with operating throttle switch and the timing set at maybe 2° before TDC (with open switch) instead of 5° after. Now, all I need is a new knock sensor (I think) and things should be great... :-)

/Anlo