Y22 - idle speed control actuator '93 300TD

[Alex Crow]

with warm engine expect zero voltage.

 

[Number_Cruncher]

Yes, I would spend a little time checking. Check also for short circuits between the wires - insulation failure is a possibility. I don't know if the ELR controller is robust against short circuits, but, if not, a short could take out any new ELR controller you might fit.

That the tacho is working suggests that the ELR control unit is getting power, and is, fundamentally, working.

For the sake of completeness, and before spending serious money, I would also spend a little time making sure that the coolant temperature signal, as seen by the ELR controller, is correct.

 

[balge59]

Yes, I would spend a little time checking. Check also for short circuits between the wires - insulation failure is a possibility. I don't know if the ELR controller is robust against short circuits, but, if not, a short could take out any new ELR controller you might fit.
Absolutely, I don't want to get sloppy - not sure, but looks to me like a diode in the diagram so worth being careful!


That the tacho is working suggests that the ELR control unit is getting power, and is, fundamentally, working.
well it seems to be getting all the signals it should, now - crank sensor and 12V from the OVP tested ok now, temp sensor - tested ok at the sensor itself, will test from the ELR socket to be sure!

For the sake of completeness, and before spending serious money, I would also spend a little time making sure that the coolant temperature signal, as seen by the ELR controller, is correct.

cheers again, I feel more confident of getting it right first time with all this help!

 

[dieselman]

Connect a pair of wires to the ELR solenoid and just touch the +12v post for a second. The solenoid will react instantly if it's good.

 

[balge59]

Connect a pair of wires to the ELR solenoid and just touch the +12v post for a second. The solenoid will react instantly if it's good.

good one, tried it and the actuator works, kicked up the tickover instantly with battery attached.

temp sensor tested ok, test by removing and replacing connector failed eg tickover stayed put but -
Problem now - not sure how to read the manual when it says -
'engine idling - unplug coupling at actuator, measure Voltage = 12V'
I get 0V across the two terminals , but 13V between red/blue terminal and earth (and 12 without engine running..) not sure what they mean here?

Think I need another look...

 

[dieselman]

Problem now - not sure how to read the manual when it says -
'engine idling - unplug coupling at actuator, measure Voltage = 12V'
I get 0V across the two terminals , but 13V between red/blue terminal and earth (and 12 without engine running..) not sure what they mean here?

Think I need another look...

The actuator is fed with a constant +12v on one wire and the control ECU shorts the other to Gnd when the actuator is to operate.

It sounds like either the ECU isn't triggering the solenoid for whatever reason, or the return/signal wire is broken.
You can check the whole circuit by measuring for +12v at the ECU connection for the signal wire, alternatively short the signal wire to Gnd to make the solenoid operate.

If this isn't a known good ECU check for dry solder joints internally.

 

[Number_Cruncher]

One other thing may be happening - if someone has manually adjusted the tick over to be too high, the ELR may never activate the solenoid, and you would correctly measure 0v across the terminals. I would make sure the engine idle speed is correctly adjusted.

While this is for the wrong engine, the data for the OM606 with ELR shows the principle;

http://w124-zone.com/downloads/MB CD/W124/w124CD1/Program/Engine/605_606/07.1-2053.pdf

Without ELR connected, the tickover speed is set approximately 50 to 100 rpm lower than the controlled idle speed range, thus allowing ELR to raise and control the speed.

Considering the engine's history (thermostat!), the possibility of ill informed fiddling is quite high!

 

[balge59]

Temp sensor?

OK, had a chance to look at the wiring properly, everything tested ok, but while re-checking the coolant temp sensor - now I have a thermostat! - discovered its out of spec!

Test says - 290 -370 Ohm at 80C - got 290, just in spec,
BUT - at 20C it should be 2.2 - 2.8 kOhm - I only get 900 Ohm! Well short, according to the spec sheet equal to 50 C, nowhere near 'cold start temps?

Would that be enough to throw the ELR off?

So, trouble is I can't work out what the part number is, all the ones on EPC say 'for a/c' and Eurocarparts only lists petrol ones, will they work?.....or is it worth trying my 'spare' sensor of '88 230TE or is that not compatible?
Found this on fleabay http://cgi.ebay.co.uk/Mercedes-Coolant-Sensor-OEM-0055450324-190D-W124-Diesel-/110313444685?cmd=ViewItem&pt=UK_CarsParts_Vehicles_CarParts_SM&hash=item19af31954d#ht_1387wt_907 but it doesn't list OM603?

 

[Number_Cruncher]

Not ideal, but, I can't imagine those resistance readings preventing the system working at all - it isn't as though the readings are completely outside the allowable range of values - it's neither a short or open circuit.

Does the idle stop screw show signs of having being messed with?

 

[Alex Crow]

..........BUT - at 20C it should be 2.2 - 2.8 kOhm - I only get 900 Ohm! Well short, according to the spec sheet equal to 50 C, nowhere near 'cold start temps?

Would that be enough to throw the ELR off?

there IS a problem! can you be 100% sure about those values? if so you have a hot suspect - air temp here today is +25 degC btw.

 

[balge59]

Not ideal, but, I can't imagine those resistance readings preventing the system working at all - it isn't as though the readings are completely outside the allowable range of values - it's neither a short or open circuit.

Does the idle stop screw show signs of having being messed with?

I was thinking - the ELR must be sampling the temp sensor by measuring the resistance, but because the resistance isn't going high enough when cold, the ELR thinks the engine is too warm to need the cold running on?
The resistance is in spec at high temps, so the temp gauge is still reading right when warmed up?

Haven't looked closely at the idle stop yet - tickover is about 500 rpm, bit higher when it warms up, maybe a little low but close enough?

cheers!

 

[Number_Cruncher]

>>maybe a little low

I was thinking the opposite - if, owing to the failure of the crank speed sensor, the system hadn't been working, and the idle stop had been mis-adjusted to raise the engine speed too high, the ELR would effectively switch off.

By switch off, I mean that supplying power to the ELR solenoid can only ever increase engine speed, it is the absence of power which reduces engine speed. If the engine speed is already too high => no voltage applied to the solenoid.

AP hints at this above, when he says expect to read no voltage at the solenoid when the engine is warm.

 

[balge59]

The test sheet I have says tickover with ELR uncoupled = " 570F40/min " which I assumed meant 570 + or - 40 rpm, the car is definitely ticking over at less than that - according to the tach at least! - especially when it is cold.

Quote dieselman - "It sounds like either the ECU isn't triggering the solenoid for whatever reason, or the return/signal wire is broken. You can check the whole circuit by measuring for +12v at the ECU connection for the signal wire, alternatively short the signal wire to Gnd to make the solenoid operate. "

Earthing the signal wire with the ELR out upped the tickover so that wiring is ok, but the Voltage was only about 0.1V with it back in?

I found a coolant sender cheap enough to take a flyer on so I'll try slapping that in, and check the resistance, see if that works!

 

[balge59]

there IS a problem! can you be 100% sure about those values? if so you have a hot suspect - air temp here today is +25 degC btw.

Ooops missed this one :Oops: As I said (somewhere...) I have ordered a sensor, so hopefully it will be the answer! I tested it a couple of times, first thing in the morning before I even started the car, and later when the the gauge was showing 85 (yay for thermostats!) so pretty confident ---- until I just reread the sheet where it says ' measure the resistance...' but then quotes e.g. " 20C = 2.2 - 2.8 kW " surely Watts are power not resistance? is that a misprint? I was testing with Ohmmeter - thats how to test resistance surely?

 

[dieselman]

Earthing the signal wire with the ELR out upped the tickover so that wiring is ok, but the Voltage was only about 0.1V with it back in?

Which is correct because the return line goes open circuit when the solenoid is not operating.

 

[balge59]

Which is correct because the return line goes open circuit when the solenoid is not operating.

So the next question is - does the wrong reading from the coolant temp sensor explain no output from the ELR....all the other inputs seem to be right...

- just thought, the kickdown works too, although as there's no test for it I guess its not relevant to the idle...? Its a shared feed to the diagnostic port so probably not involved?

 

[dieselman]

So the next question is - does the wrong reading from the coolant temp sensor explain no output from the ELR....all the other inputs seem to be right...

I suspect not as the ELR is meant to be in use all the time for idle stabilisation, is it not?

 

[balge59]

I suspect not as the ELR is meant to be in use all the time for idle stabilisation, is it not?

Hmm ... maybe it depends on the temp reading to know if its cold , and then cuts out when the tach signal says the revs have gone up? Like you said, it can only raise the revs when powered, when the powers off the revs drop again,... but then it must read the tach signal to know when the revs are too low as well...

How does the darn thing work anyway, I looked inside and it looks like a pretty complex PCB! And I thought it was a 'relay' lol

Would an error on the diagnostic thing turn off the ELR?... maybe I should try my DIY code reader...

 

[Number_Cruncher]

DM's right, ELR is active all the time the engine is running. The temp sensor just enables the ELR controller to provide extra correction when the engine is cold.

I wouldn't trust the dash board mounted rpm gauge to indicate how quickly the engine is running - they're fine for keeping the driver entertained, but, don't mistake them with a serious instrument.

 

[Alex Crow]

IF the NTC temp sensor does read 900 ohms @ 20 degC then it will be the problem. HOWEVER (!), these sensors are simple temp dependent resistors and if ok at, say, 80 degC will usually be ok at 20 degC. my suspicion is still the control unit.