Intermittent no-start

[sajohnson]

I know this is a common problem. I've read through a few threads and got the impression that the usual culprit is the starter solenoid.

From what I understand, the dealers cannot order just the solenoid -- they can only get the complete starter/solenoid assembly -- but the solenoid is available on eBay and elsewhere.

Currently, the problem is intermittent but it is happening more frequently.

When the key is turned to 'start' I usually hear what sounds like a rely picking/energizing but the starter does not engage. It sounds like the relay is under the dash.

The relay seems to be on a timer -- once it picks, continuing to cycle the ignition switch has no obvious effect. Once it drops out, then turning the key to 'start' will cause it to pick again.

The chassis battery is good. It is on a Battery Tender and when the starter circuit works properly the engine cranks over normally and fires right up. Initial voltage drop on the ScanGauge II (while glow plugs are operating) is normal.

Any troubleshooting suggestions will be appreciated. I'd like to confirm that the solenoid is faulty before ordering one.

I wish the solenoid could be taken apart because it might be an easy fix, but I gather they are 'disposable'. That's too bad.

 

[Eric Experience]

Sa.
Another possibility is the crimp on the battery cable, after a non start feel the temperature of the crimp, it will be hot if it is faulty. Eric.

 

[220629]

...

The relay seems to be on a timer -- once it picks, continuing to cycle the ignition switch has no obvious effect. Once it drops out, then turning the key to 'start' will cause it to pick again.

...

Your observation is correct for both the T1N and NCV3. More accurately it times out.

Gone are the days that when you twist the key to "start" a signal immediately heads out to the starter solenoid.

When the key is rotated to start you are requesting a start. "HAL, if you check and find all systems ready please start my engine". If all systems are go the ECM aka ECU enables a crank sequence which then waits for feedback that the engine is actually running. If the engine doesn't start the cranking will time out and be disabled.

The sequence can be interrupted manually by rotating the key back to the off position. Otherwise 2001 A Space Odyssey HAL is in control.

It may help to keep the driver door open during start attempts. The small control relay will click. The start solenoid will more "clack" loudly when the parts move to engage the flywheel teeth (Bendix unit) and also to bring the heavy electrical contacts together.

If you don't hear a clack it could be the solenoid not responding at all, but my recollection is that the more common NCV3 failure mode is the contact points themselves. Bad contacts (non-conducting) should still mechanically clack.

The real test would be to wire a temporary (or permanent) visible LED or lamp to the starter solenoid control power connection.

vic


An example.

Sometimes similar to us dealing with our "computer always in control" Sprinters.

"This machine is too important for me to allow you to jeopardize it." MB Sprinter Programming Basic Premise


https://www.youtube.com/watch?v=qDrDUmuUBTo

 

[showkey]

Johnson.........you could monitor the voltage at the solenoid signal wire. If the start signal is arriving at the starter and you get the click........solenoid would be a suspect.

You are correct the solenoid is NOT serviceable.

The whole starter ( facrtory replacement part) is available at large discounts from many suppliers like europarts.

 

[caeppinger]

You can also remove the starter and bench test it, it should take less than 30 minutes to remove the starter (it is not hard, easier removal than many other vehicles).
I bought my new solenoid from summit this year www.summitracing.com/parts/bch-2339305066 it worked great.

 

[220629]

You can also remove the starter and bench test it, it should take less than 30 minutes to remove the starter (it is not hard, easier removal than many other vehicles).
I bought my new solenoid from summit this year www.summitracing.com/parts/bch-2339305066 it worked great.

A good suggestion, except bench testing can sometimes be hit or miss with an intermittent.

If I were to take the time to remove the starter, for the 50 bucks, I would replace with the part you suggest, re-install the starter, and know the solenoid is good. I suppose that still leaves question for the brushes, etc., but they have pretty reliable service history.

vic

 

[showkey]

A good suggestion, except bench testing can sometimes be hit or miss with an intermittent.

If I were to take the time to remove the starter, for the 50 bucks, I would replace with the part you suggest, re-install the starter, and know the solenoid is good. I suppose that still leaves question for the brushes, etc., but they have pretty reliable service history.

vic

Agree.....especially in this case when the problem is intermittent. I had my starter off twice before the solenoid got bad enough to become less intermittent. Mine was random intermittent.......then was perfect for months, then this spring got worse where I had to do multiple jumps of the solenoid to get it to crank. All this with 30,000 miles.

On the up side it is pretty easy to get the starter out.

 

[sajohnson]

Sa.
Another possibility is the crimp on the battery cable, after a non start feel the temperature of the crimp, it will be hot if it is faulty. Eric.

Thanks Eric.

You're definitely correct about that crimp getting hot when it is faulty.

When we still had the original 'Y cable' I suspected it was bad (low voltage on ScanGauge was one clue) so I thought I would give it a stress test. With the engine running, I plugged a toaster into the kitchen outlet, which was being powered by the inverter I installed. That caused the coach battery voltage to sag and a lot of current to flow through the Y cable.

Then, because I've put off buying an IR heat gun, I slid under the front end and felt the lug on the starter solenoid to see if it was warm. It was smokin' hot and left a blister on my finger!

IOW, the test was a success!

Since I replaced the Y cable we've had no further problems. Also, I imagine that if the crimped connection were resistive the no-start would not be so intermittent, but who knows? I guess vibration could cause the resistance to vary.

 

[sajohnson]

Your observation is correct for both the T1N and NCV3. More accurately it times out.

Gone are the days that when you twist the key to "start" a signal immediately heads out to the starter solenoid.

When the key is rotated to start you are requesting a start. "HAL, if you check and find all systems ready please start my engine". If all systems are go the ECM aka ECU enables a crank sequence which then waits for feedback that the engine is actually running. If the engine doesn't start the cranking will time out and be disabled.

The sequence can be interrupted manually by rotating the key back to the off position. Otherwise 2001 A Space Odyssey HAL is in control.

It may help to keep the driver door open during start attempts. The small control relay will click. The start solenoid will more "clack" loudly when the parts move to engage the flywheel teeth (Bendix unit) and also to bring the heavy electrical contacts together.

If you don't hear a clack it could be the solenoid not responding at all, but my recollection is that the more common NCV3 failure mode is the contact points themselves. Bad contacts (non-conducting) should still mechanically clack.

The real test would be to wire a temporary (or permanent) visible LED or lamp to the starter solenoid control power connection.

vic


An example.

Sometimes similar to us dealing with our Sprinters.

"This machine is too important for me to allow you to jeopardize it." MB Programming Basic Premise


https://www.youtube.com/watch?v=qDrDUmuUBTo

Very funny vic!

You know, I actually thought about using the phrase "timed out" (or 'ran time'). That was my instinct because I worked as an automatic train control tech for 27+ years, and a large part of the system still uses 100 year old relay logic -- large relays the size of a dictionary and (in some cases) timers. Those were phrases we used.

If the solenoid operates, I should hear the 'clack' -- especially if I have the door open as you suggested. Then I'll know it's either the starter solenoid, or the starter itself.

From what I've read -- and what you said above -- the usual failure is the solenoid contacts, so at that point I could probably go ahead and order the solenoid.

Of course there would still be a possibility that the starter itself is bad, but that does not seem to be very common. Then again, many owners just replace the entire starter assembly (or pay the dealer to do it). In those cases of course we don't know what caused the failure.

In other 'no-start' threads I saw photos of solenoids that had been cut open and the contacts had been arcing and were clearly dirty.

I don't recall reading any posts by owners who had replaced just the solenoid, only to find it was actually the starter -- but I did not read through every no-start thread.

I like your idea to wire a temporary (or permanent) visible LED or lamp to the starter solenoid control power connection. I guess I'll wait and confirm whether the solenoid is operating when there is a no-start condition. If not, then I'll wire up the LED to determine whether it's getting power from the relay under the dash.

Great post, thank you!

 

[sajohnson]

You can also remove the starter and bench test it, it should take less than 30 minutes to remove the starter (it is not hard, easier removal than many other vehicles).
I bought my new solenoid from summit this year www.summitracing.com/parts/bch-2339305066 it worked great.

Thanks for the link!

$42 isn't bad, but I'd feel better knowing that the design of that solenoid had been improved.

I'd rather spend $70 or $80 once, and be done with it, than $42 several times.

Our View has about 68,000 miles on it, and being a RV, pretty much by definition, when it is fired up it is driven a long distance. So, the solenoid has not been cycled very much -- relatively speaking. The number of cycles is probably about the same as on the average car with 20K miles on it. The starter (& solenoid) in most vehicles lasts at least 150-200,000 miles, sometimes longer -- and most cars are started several times a day.

IIRC, I read posts by owners who have replaced the solenoid 2-3 times. It's very discouraging that a company like Bosch, with a generally good reputation, would continue to make such a lame part.

I hope I'm mistaken and they have changed the design.

 

[showkey]

I felt my Y cable was my suspect as well ( replaced) ..........but.........since the click was so intermittent it was not able to be confirmed and it passed both visual inspection and prior voltage drop tests. Until months later when the click returned and solenoid was the true faulty part.

If I am out on the road I carry a spare starter ( the old starter with new new solenoid installed), along with a MD802, spare inter cooler hoses and an orange gasket.

 

[sajohnson]

I just thought of a test that might be helpful -- measuring voltage drop across the solenoid contacts.

Checking the voltage drop on the bench with no load probably wouldn't show much, but with the starter on the engine and under load (cranking a cold, high compression diesel engine) and pulling hundreds of amps, even a small amount of resistance would show a voltage drop.

A 250A draw with just 0.01 ohms of resistance = 2.5v

Just the voltage drop across the battery cable is significant. I used this...:
http://voltagedropcalculator.net/

...and plugged in #2 gauge cable; 12V; 250A; over 5 feet and got almost 0.5V (0.48V).

That's just an example. Most of those numbers are estimates.

It would be easy enough to set it to DC volts and measure across the solenoid contacts while the starter is running. Having an assistant would help -- otherwise you'd have to use jumpers so the meter could be in the cab (or at least visible).

 

[sajohnson]

I felt my Y cable was my suspect as well ( replaced) ..........but.........since the click was so intermittent it was not able to be confirmed and it passed both visual inspection and prior voltage drop tests. Until months later when the click returned and solenoid was the true faulty part.

If I am out on the road I carry a spare starter ( the old starter with new new solenoid installed), along with a MD802, spare inter cooler hoses and an orange gasket.

You know a vehicle has design problems when people routinely carry several spare parts with them...

That's actually very common with View/Navion owners. We carry a few ourselves.

I have to admit that I don't know what the orange gasket is for, or what a MD802 is.

If I understand you correctly, you initially thought the Y cable was the culprit but it turned out to be the solenoid -- is that correct?

Was the Y cable actually bad (resistive)? It is a common failure -- another "The Best or Nothing" faulty design -- so I wouldn't be surprised.

I read several similar posts -- people who replaced parts and thought they had fixed the no-start problem, but because it was so intermittent they couldn't be sure. Then the problem would come back and they'd replace something else...

I felt really bad for one guy -- he kept taking it to the dealer and they kept shotgunning the problem. He ended up spending a lot of money for a problem caused by a ~$45 solenoid.

 

[220629]

... Having an assistant would help -- otherwise you'd have to use jumpers so the meter could be in the cab (or at least visible).

A 21 watt or even smaller watt lamp across the contacts may work too. No assistant needed.

It would only need two wires up to the visible lamp location.

Normal start = no problem.

No crank with clack, monitor lamp lit (probably bright) = solenoid contact problem.

No crank with clack, monitor lamp dark = internal starter problem (brushes, etc.?).

If the intermittent crank problem surfaces, it might be good to mount a fairly permanent monitor lamp rather than doing fancier testing.

vic

 

[sajohnson]

A 21 watt or even smaller watt lamp across the contacts may work too. No assistant needed.

It would only need two wires up to the visible lamp location.

Normal start = no problem.

No crank with clack, monitor lamp lit (probably bright) = solenoid contact problem.

No crank with clack, monitor lamp dark = internal starter problem (brushes, etc.?).

If the intermittent crank problem surfaces, it might be good to mount a fairly permanent monitor lamp rather than doing fancier testing.

vic

I like the idea of using a lamp instead of a meter. There are a couple things I'm not sure of:

* How much voltage would be required to illuminate the lamp to a degree where it could be seen?

My concern is that the solenoid contact resistance could be high (out of spec) but the voltage drop may not be enough to visibly light the lamp. I don't know, it's not something I have experience with.

If the resistance (and voltage drop) must be very high for the starter not to work then it shouldn't be a problem.

* The other thing is that the lamp would be lit at full intensity whenever the starter was not working -- all the time, except when the solenoid contacts are closed (assuming they are clean and working properly).

Perhaps the lamp could be switched. That way it could be left off except when starting the engine.

 

[showkey]

I used the light in the solenoid signal wire to monitor the start signal.

My starter click at this time was very intermittent. My Y cable passed all tests except the voltage at the scan gauge had dropped a couple of tenths. Not bad drop but because of the history of the Y cable and the multiple changes I replaced it. At the times wrote I had no confidence it was bad after inspecting it.
I was not able to get a voltage drop test across the solenoid when the click actually happened. My starter never ever ran slow. It was either crank full speed or click......nothing in between.

But......when the click got bad, I was on to the solenoid as the problem, and jumped the solenoid signal wire to get the starter to run. After it was jumped and the starter ran the truck would start with the key. When it got real bad it would take multiple solenoid "jumps" to get the solenoid to close and run the starter.


Md802 is an affordable MB friendly scan tool, orange gasket is the gasket in the nose of the turbo inlet and the air cleaner hose attaches with a clamp and the orange gasket. If the orange gasket fails it can get sucked into the turbo. $$$$ There is a sticky note on this section that orange gasket is NOT suppose to be reused.

 

[sajohnson]

I used the light in the solenoid signal wire to monitor the start signal.

My starter click at this time was very intermittent. My Y cable passed all tests except the voltage at the scan gauge had dropped a couple of tenths. Not bad drop but because of the history of the Y cable and the multiple changes I replaced it. At the times wrote I had no confidence it was bad after inspecting it.
I was not able to get a voltage drop test across the solenoid when the click actually happened. My starter never ever ran slow. It was either crank full speed or click......nothing in between.

But......when the click got bad, I was on to the solenoid as the problem, and jumped the solenoid signal wire to get the starter to run. After it was jumped and the starter ran the truck would start with the key. When it got real bad it would take multiple solenoid "jumps" to get the solenoid to close and run the starter.


Md802 is an affordable MB friendly scan tool, orange gasket is the gasket in the nose of the turbo inlet and the air cleaner hose attaches with a clamp and the orange gasket. If the orange gasket fails it can get sucked into the turbo. $$$$ There is a sticky note on this section that orange gasket is NOT suppose to be reused.

That's the first I've heard of the orange gasket. I don't recall seeing it but I will keep that in mind.

I was also unaware of the MD802. Last I heard, there was no reasonably priced scanner that would read all of MB's top-secret, proprietary codes. All they would read were the codes that are federally mandated -- mostly emissions related. If the MD802 can read more than the ScanGauge II and the flash tool that came with the Green Diesel ECO tune then I'll definitely pick one up!

If I understand correctly, you connected a lamp to the wire that feeds the solenoid coil, so you could see when the coil had 12V on it -- is that right?

I plan to do that if I do not hear the solenoid operate (the "clack"). We've only had the no-start happen a few times, and I believe the solenoid did operate each time. If so, then of course the problem is with the solenoid contacts or the starter motor. What makes it tricky is if you turn the key to start while the control relay is still picked, nothing will happen -- in contrast with most other vehicles, in which the solenoid will operate every time the key is turned to 'start' (assuming the solenoid is working and getting power to the coil).

So, although my wife told me there was no solenoid 'clack', she may have been thinking of the times she turned the key to start while the control relay was still energized.

Unless it is very resistive, troubleshooting the Y cable is tricky. Like the starter solenoid contacts (or the 'boost' solenoid contacts in RVs), the only way to test the Y cable is to pass a huge amount of current through it and measure the voltage drop. With just a small amount of current flow, there may not be much voltage drop at all and the cable or contacts will appear good.

Interesting that your starter either ran full speed or not at all. Come to think of it, the couple times I experienced the no-start with our View it seemed to be that way also. There was no in-between, no slow cranking -- just go/no go. I wonder why that is? I'd think that dirty contacts would -- at least sometimes -- create a resistive connection, but allow enough current to pass to run the starter motor slowly.

I'm not sure I understand the following:

"...when the click got bad, I was on to the solenoid as the problem, and jumped the solenoid signal wire to get the starter to run. After it was jumped and the starter ran the truck would start with the key. When it got real bad it would take multiple solenoid "jumps" to get the solenoid to close and run the starter."

When you say, "...jumped the solenoid signal wire to get the starter to run", it sounds like you were jumping across the solenoid contacts -- putting 12V from the battery cable directly on the starter motor. Is that correct?

I'm not sure why doing that would allow the truck to start with the key.

What do you mean by, "when the click got bad"? When the no-start condition became more frequent?

Just curious, thanks.

 

[220629]

...

If the resistance (and voltage drop) must be very high for the starter not to work then it shouldn't be a problem.

Yes. It was intended as a gross test.

...
* The other thing is that the lamp would be lit at full intensity whenever the starter was not working -- all the time, except when the solenoid contacts are closed (assuming they are clean and working properly).

You are exactly correct.

...Perhaps the lamp could be switched. That way it could be left off except when starting the engine.

That's a solution. It adds complexity in that the switch will need to be remembered.
Remembering to switch off won't be a problem, especially at night.

Maybe the light across the contacts isn't such a great monitor.

Some good test ideas are coming out in this thread.

vic

 

[showkey]

The truck would start after multiple "jumps" of the solenoid because the contacts would shift or align or make better contact. The contacts are burnt ( I cut mine apart) and the pull in of the contacts is likely weak. I think they are burnt enough ( poor contact) to make the starter a go no go failure situation.
This jumping of the solenoid was my go to method of getting the truck start when the click occurred. I always thought I could get it cranking using this method. Being in remote camp ground and hearing click was always a concern ........in my case I felt I could get it going on my own.

Published this photo prior in "my turn with the click" but here it is again for a reference to the solenoid internals:

[/URL][/IMG]






There are multiple posts on the merits of the AUTEL MD802. It does a much better job than the scangauge at code reading and live data stream for troubleshooting. The scan gauge is very good driving monitor.

 

[220629]

...

Published this photo prior in "my turn with the click" but here it is again for a reference to the solenoid internals:

...

The picture didn't display for me?

Was it from here?

Update:

New starter arrived, installed .......cranks and starts normal.

Bench tested old starter.........FAILED..........solenoid mechanical closes but failed to make contact to run the motor. Loud click and pinion moves. Multiple jumping (7-10) gets starter to run on the bench just like in the truck.

Cut open solenoid ( it is not a serviceable part).
Found top contact burnt and black. If this was a serviceable part confident it could be cleaned and returned to service. Think this is similar to what Froggy found on his multiple solenoid failures. All my issues are a low mileage seasonal use RV with + 30k miles.

My issues were very intermittent last year and went away for months and intermittent but more repeatable this spring.
My confidence on some of my other repairs was low.........I do feel this is a true smoking gun ( solenoid) and would be very surprised if the click returns any time in the near future






Ordered a new solenoid from eBay looks to be the correct part ? Won't be here for a week. It was $42 delivered. As others have posted solenoids are not available at the dealer and can be difficult to find in the US market.

vic