Duty Cycle 99% M103 - PeachParts Mercedes-Benz Forum

350SL4spd · #1 05-08-2011, 02:19 PM

Hey Guys,

I finally got my '91 M103 idling after chasing down a major vacuum leak. It's still idling like crap, but that's the subject of this.

Duty Cycle in closed loop is 99% (.1v). I've checked most sensors etc and most check out ok (O2, OVP, AF pot, EHA, EZL, etc). I attempted to adj the lamda tower to make the car run richer. After as much as a full turn the DC would get up to around 45%, but the computer would trim that out & take it back to 99% in about 10sec. Oddities:

1) The Throttle Switch was testing wonky the other day, but tests fine today. That or I wasn't operating properly.

2) ***EDIT: Temp Sensor tests fine. I was testing it from the actual sensor instead of @ the EZL. Tests w/in spec now as well.

Is there anything that I have neglected to test? Any ideas as to what might be wrong?

Thanks!

(EDIT: Just tested the coil as well & got proper resistance btw 15 & 1. FSM says to measure resistance btw 1 & 4 as well, but neglects to tell me where the hell #4 is... Don't think the coil is bad anyway...)

350SL4spd · #2 05-08-2011, 03:42 PM

Just some background info:

Key On Engine Off:
-Duty Cycle is around 50%. I know it's supposed to be 70%, but quite a few people seem to always show 50% (+/-6v)
-WOT it's around 23% (+/-9.3v)
-Deflect AF Plate and it's around 40% (+/-7v)

I mention this because I've tested both the Throttle Switch & the AF pot and both test ok (AF pot tested slightly out of spec btw 1-2: .85v, where spec is .57-.81v).

Now, if the car has been sitting for very long I will get no change when I deflect the AF Plate. But if I let the car run for a couple of minutes, I get the above reading of 40% (+/-7v).

Any ideas?

ps2cho · #3 05-08-2011, 07:44 PM

The airflow potentiometer is supposed to 0.85v +- 0.1v so not sure where you got that spec from. Yours is fine.

Engine running is more important than engine off....

Duty cycle is NOT in closed loop at 99% -- that is open loop! Closed loop is an oscillating duty cycle which shows the system has the ability to compensate for change.

Still looking too hard for problems!! Concentrate on the meter tower and adjustment. It may take a good 30mins before you get the right place. The initial turn will cause an increase in % because you are (physically) pressing the plunger allowing more fuel in (hence the momentary change). If its heading in the right direction away from 99%, then keep doing it. eventually it will either find its way to 50% and oscillate around there, or go to 0% in which case you need to go CCW (or vice-versa...some Craftsman DMM's read opposing duty cycles). It may need 1-3 full turns to find the correct height.

Think of the mixture unit as a log see-saw....Your allen key screw is simply changing the middle log so that the height at its maximum extension is higher or lower. The higher the log goes, the more the plunger on the fuel distributor is pushed in, more fuel, and the less it is pushed in, the less fuel. If the see-saw is sitting too low its not even making contact with the plunger, showing excessively lean mixtures at idle (which I bet is what is happening). Keep turning the screw to raise til the plunger finds its happy level at idle and it might be your golden ticket.

Once you have the duty cycle at or around 45-50% with it increasing no more than 10% at WOT, you have found the correct spot.

As I said in the previous thread though, I know you were fiddling with the fuel distributor, so if you did also the EHA, you need to return it back to its original position or direction.

This is why I don't advocate touching or adjusting anything but the duty cycle (for future readers/researchers) because if it ended up being something simple, like the vacuum leak, you now have issues returning multiple, fine adjustment units back to where they were originally. It's tough

350SL4spd · #4 05-08-2011, 08:12 PM

Quote:

Originally Posted by ps2cho

As I said in the previous thread though, I know you were fiddling with the fuel distributor, so if you did also the EHA, you need to return it back to its original position or direction.

This is why I don't advocate touching or adjusting anything but the duty cycle (for future readers/researchers) because if it ended up being something simple, like the vacuum leak, you now have issues returning multiple, fine adjustment units back to where they were originally. It's tough

FD or EHA were never adjusted. The FD is new (which is probably the source of the car being out of adjustment) and I have a spare EHA that was swapped out just to rule it out.

I've gone as far as 1.5 turns with the lamda tower so far for exactly the reason you mentioned: I wanted to be able to get it back to where it was if that wasn't the problem. I'll take it a bit further and see what happens.

The 1.5 turns didn't move it even a hint off of the 99% that's why I didn't push it further.

350SL4spd · #5 05-08-2011, 08:42 PM

OK. Got the car fully warmed up & very slowly turned the lamda screw in 1/2 turn at a time, every 3-5min or so. Same result every time. Brief journey to the 40% range, followed by a drop all the way down to 99% (.1v). I didn't dare go any farther after 5 full turns, lest the damn screw fall right out.

However, if I manually push the AF plate just a millimeter, the idle smooths right out & Duty Cycle shows around 45% & oscillating. Is it possible that something is wrong with the screw?

Or do I just need to adj the height of the AF Plate with the drift-pin on the top of the mixture unit?

I'm pretty surprised that moving the screw that far didn't get me anywhere...

ps2cho · #6 05-08-2011, 10:23 PM

Doubtful....

It may be time to now check the height adjustment pin. The proper way to check is by measuring the depth of the throttle plate, but you may not have the tool....so check this pin and make sure its not sticking up and is flat at the very least. Just remember if knocked in too far, it must be knocked out from the other side, so take it easy if you decide to knock it in slightly -- do it with care.

350SL4spd · #7 05-08-2011, 10:26 PM

It's flat. I've been thinking about knocking it in a bit, but: Does it only effect the static (rest) position of the air plate?

What I don't get is that it's the same mix unit as before when it was running fine, just a new FD on top. So it should be THIS far out of adjustment...

JohnM. · #8 05-09-2011, 12:05 AM

100% duty cycle indicates the following when car running, warm ,idling. This is direct from FSM. Also, in open loop your car should be at a SOLID 50% duty cycle, no matter what!

Quote:

No voltage or ground at KE control N3 (ECU) or KE control unit N3 faulty
Lambda setting too lean
Oxygen sensor faulty (short to ground)
Overvoltage protection fuse or overvoltage protection (K1) faulty

350SL4spd · #9 05-09-2011, 12:41 AM

Well, all of those are ruled out except the first, I guess. I'll have to check that tomorrow.

OVP checked out, but I do have a spare to throw in just be doubly sure.

Suppose there's any different btw 99% & 100% DC?

350SL4spd · #10 05-11-2011, 10:10 AM

So now I've:

-Swapped out OVP.
-Tested N3 (CIS Control Computer) for 12v & ground.
-Tested Air Intake Sensor.
-Tested wires from all major sensors back to N3 to make sure that there are no breaks in the harness.
-Knocked that drift pin in a couple of mm's.

Everything checked out perfectly. And the drift pin just knocked the AF Plate out of adjustment. For anyone's future reference: it only effects the rest state of the plate.

No idea where to go from here...

EDIT:

Almost forgot: I tested the ICV as well. All wires checked out as good (no breaks) but my FSM said to check for 12v btw pin #1 on the ICV connector & ground (ign. on). I was only getting a couple of volts there. Can anyone confirm that test for me?

350SL4spd · #11 05-12-2011, 11:09 AM

Interesting tid-bit: I was going through an old notebook last night looking for any notes on the last time I adjusted the DC on this car. I found some scribbled references to what my DC was back in '09 and, to my surprise, my DC back then was 90%.

At the time I was dealing with a low-idle with the A/C on problem. I replaced the EHA (leaking) and tweaked the DC, but I have no notes as to what I ended up tweaking it to, if I managed to adjust it all. (Memory is crap...) But it's run fine in the 2yrs since...

So now the question is, wtf could have been slowly degrading over a period of years, gradually causing a leaner condition??

Ring any bells for anyone??

350SL4spd · #12 05-15-2011, 10:59 AM

One thing that has be thoroughly baffled at this point is that the car shows a DC of +/-45% with the O2 disconnected and the car runs like crap @ idle. Push the AF Plate in a bit (with O2 connected) and duty cycle jumps from 99% to around 45% (depending on how far you push the plate in) and the car runs great.

So how can the same DC give me 2 vastly different idle qualities?

350SL4spd · #13 06-08-2011, 10:29 PM

After more testing and much more frustration, I have now replaced the ECM with a warrantied used one (no change) and a new O2 sensor (nearly no change).

I still am running at 99% Lean (.10v) after warm up. With the new O2 sensor it seems to run at pretty good DC for a couple minutes after warm-up, oscillating from 5.5-6.5v give or take for maybe 2min (temp is around 70deg C at this point), then it drops to the .10v when it reaches around 80deg.

The car still runs PERFECTLY if you nudge the AF Plate down a couple on millimeters with your finger.

Any new suggestions or info? I'm a couple of weeks away from a burial at sea at this point...

ps2cho · #14 06-14-2011, 10:13 PM

I have an idea....

MB makes a special tool for holding the airflow sensor plate in a specific position for fuel flow testing. How about you purchase it, attach it on and have it so it holds the throttle open those few millimeters that you need and drive with it for a bit?
Part # 201 589 00 99 00

Does the duty cycle return to a normal reading with the plate deflected

350SL4spd · #15 06-14-2011, 10:53 PM

Duty cycle is in the butter zone when the plate is deflected.

I think I've seen that tool... My big question would be if you could get the air filter on with the tool attached. Driving around w/o an air filter probably wouldn't help things... I was also hoping to sell the car later in the year to upgrade to a R129 or the like so I need a permanent solution.

The other thing that bugs me about how she's running is that the duty cycle is where I'd want it (5.5-6.5v & oscillating) after cold start before the car enters closed loop (or should) and it drops to .1v. Now, the reason that bugs me is that it still runs like crap during open loop even though it's showing a decent DC. However, if I deflect the plate even when it's cold, the idle smooths right out.

As soon as I get a chance this week, I'm going to try to adjust the AF Pot, because that is actually physically effected by the AF Plate. It tests fine, but I'll try anything right now...