I've got a 1993 MK III Golf, and have been having problems starting the car when it is hot, or the weather is hot.
What happens is that when I turn the ignition key, the lights come up on the dashboard, but the starter refuses to turn.
I know that the starter is not faulty, because when I bridge the solenoid the car will start with no problems.
When I leave the car to cool right down, or if the weather is not really hot, the car will start fine.
I've checked the wiring, and it all seems fine. I have been told that this does happen to some Golfs but I don't know why. I was thinking of putting in a relay to correct the problem, but I'm not sure how many amps the relay should be.
Any help would be much appreciated.
Josh
|
|
It sounds like the solenoid is sticking, you'll need to remove the starter then the solenoid from it. You'll need a good quality torx T25 driver to get the solenoid off the starter, if it's anything like a mk4 one anyway. Probably just dirt or rust in there. It will be very difficult and expensive to source a relay big enough to carry the starter current, definatly more than a whole new starter.
|
Thanks HarlequinVW for your reply. But I dont know why the temperature affects wheather the solenoid engages or not?
Just to clarify when I bridge the starter I mean, between the battery and the solnoid energising terminal, and not the actual starter motor terminal, (i.e the terminal which the ignition switch plugs into).
Everything works fine when I power the solenoid directly when the engine or weather is hot.
Thats why I was thinking of installing a relay at this point, but not sure how many amps the solenoid itself draws.
|
Josh, you are correct in your diagnosis, I have met this problem before with this equipment. Any electromagnet (eg a starter solenoid) will produce a force when energised which is a function of the number of ampere turns (current x turns) in the coil. The number of turns is fixed, but the current is goverened by the resistance of the coil. Copper wire has a positive temperature coefficient of resistance, hence the coil resistance will increase with temperature, and the current will reduce for a fixed applied voltage. Soak back heat in a starter mounted near an exhaust is a well known problem, and some Bosch starters are very marginal in this respect.
So much for the theory. The easiest fix is exactly as you suggest - fit a relay, using heavy wire in the contacts circuit, preferably sourcing the starter solenoid current from a fused 60A supply. Your relay coil is of course energised via the existing wire which originally went to the starter solenoid (terminal 50). Mount the relay in a sensible place on the bulkhead, terminals down.
659.
|
|
It could be the ignition switch then, which is the ONLY thing you are substituting when using the wire straight from the battery.
|
In the VW (and some other) applications, the wires to and from the switch, together with the contact resistance of the switch itself are high enough to tip the balance into a non hot start due to the marginality of some of these starters. Remember that the solenoid is actually dual wound, with a pull-in and a hold-in winding, and the pull-in current is about 30A.
I had exactly the same problem with a SAAB (ignition switch between the seats, so longer wires) and cured it with a relay. I was gratified to see that my next SAAB had wires of twice the cross section to and from the switch. The second one gave no trouble. The Swedish SAABs had a canted ehgine, with the starter under the exhaust manifold, so the early ones were a prime candidate for this (although it probably never showed up in Sweden!).
In any event, a relay is cheaper than a new switch.
659.
|
Thats Brilliant!! Thanks!! I'm glad to get an answer to this problem. Would you say a 30A or 40A relay would be enough to energise the solenoid?
Many thanks,
Josh
|
One possibility is to use a solenoid from any old inertia type starter motor system. There is no doubt about such a relay's current capacity, or its suitability for the engine bay environment.
These solenoids are dual wound, similarly to the one on your pre-engaged starter, but don't need to move the pinion, they only need to make the main contact.
If you can get hold of one, some of them even have a button to allow you to make the contact manually!
Number_Cruncher
|
Solenoid switches for inertia starters did not have dual coils as there was no requirement to move the pinion. A relay of such capacity is an overkill for this job - a 30A "matchbox" type relay will do fine, but don't forget to use a source fused at about 60A as the pull-in current when very cold is high.
659.
|
I agree entirely about such a relay being overkill - but robust.
Number_Cruncher
|
Just checked, and, 659 is, of course, correct about those solenoids not being double wound.
I was sure that they had a pull in coil - but I was wrong!
Number_Cruncher
|
N C - a true gentleman.
659.
|
With car problems, it is ALWAYS better to fix the original fault, rather than perform a workaround. But in your situation this is especially true, as the mk3 golfs are known to have ignition switch problems. These generally effect the running of the car rather than the starting. With this in mind (although bearing in mind it's a different contact within the ignition switch) there is more value in a new ignition switch than has been given credit for in this thread.
Lot's of mk3 problems have been resolved here:
www.clubgti.com/forum/forum_topics.asp?FID=24
and here: www.clubgti.com/forum/forum_topics.asp?FID=13
including some that have been caused by the ignition switch.
The search is not always performing that well on these ^ forums.
|
A work around is valid where there is a design problem, as in this case. The value of forums such as this is that specialised knowledge and specific experience can be shared. Replacing the ignition switch in this instance would be a "dealer" solution which will not correct a design fault. In this case, there have to be several factors which are adverse in order for the fault to show up. The chances of a dealer finding this, admitting to a design fault, or even being interested are in my experience, remote.
659.
|
|
I'm with HarlequinVW here, your not fixing the problem. This car's starter relies on the current through the solenoid to pass through the starter to iniate rotar rotation to ensure pre-engagement of the toothed gears. I suspect your starter is not all you think it is or you have a connection/ignition switch problem. Establish what voltage you are getting at the starter solenoid terminal when attempting to start. (If you have 10 to 12 volts and the engine to body/battery earths are good but the starter does not spin up then the starter is unwell, probabley worn brushes, it only takes a couple of ohms to stop this from spinning up and engaging. Good Luck. Regarda Peter
|
The question is: how long will it be before the resistance of the cranking contact gets high enough that it will not even pull in the retro-fit relay anymore?
The ignition contacts are frequently become such high resistance that they will not power the fuel pump and ECU anymore! (although i am aware that the ecu and fuel pump relays are compounding this problem)
|
The extra reistance of a 2 ohm @ 20 oC solenoid, will be 0.63 ohms @ 100 oC (copper has a thermal resistance coefficient of +0.393% per oC), see how that figure compares to the actual cranking switch resistance on your car.
In all probability the max solenoid temperature will be 60 oC.
|
I used to design starter systems for a (now non existant) UK manufacturer. The maximum soak back temperature I have measured on an unfavourable starter installation was 195 degC. It is of course, very easy indeed to make such a measurement by looking at the change in coil resistance, ie. using the coil itself as a resistance thermometer.
As far as starter (ignition) switches are concerned, any unassisted switch has to deal with the 30A pull-in current, together with the inductive effects of the solenoid which will give rise to a high back emf when the circuit is broken. This causes the contacts to arc and erode. By fitting a relay, the key switch only has to deal with about 200mA, so even a switch in poor shape will do this without difficulty for a long time - a cheap fix.
659.
|
659:
where do you get the 30A figure from? at 2 ohm DCR the mk4 golf starter i have would have trouble reaching that.
You could indeed attempt to measure the temperature of a starter motor be measuring the coil resitance. However this would only be a very crude indication; the non-linearity of the copper +Temperature Coefficient of Restance, the thermocoule/peltier effect of the differing metals on the starter contact (@circa 195 oC) and also the unknown extra resistance of the starter contact (@circa 195 oC) would all be sources of major measurment error.
For the car in question, a golf mk3, the engine block/cooling system is regulated at about 95 oC. The starter is conected via the thermal mass of the aluminium bell-housing in a well ventilated area of the engine bay. The solenoid is then relatively loosly mechanically coupled to the starter motor, although from above. I can't see it reaching anywhere near my earlier predicted 100 oC worst case. This is backed by the fact that i can put my hand on it always, my touch gets too hot at 55 oC.
From the transition from the mk2 to mk3 golf, VW themselves have fitted interposing relays for the fuel pump and ecu:
Digifant FUEL PUMP relay 191 906 383C £17.83 + VAT
Digifant ECU relay 165 906 381 £8.81 + VAT
Which are both a major source of unreliability.
If there was a known problem for the starter solenoid, do you not think they would have taken this opportunity to fit one there too?
|
