Intermittent starting issue (2013 Outback)

[Takeitoutback]

My wife and I purchased our first Subaru - a 2013 Outback (~56,000 miles) in September 2016. In December, it began to exhibit an intermittent starting issue - the engine cranks (slowly) and fails to start approximately 10% of the time. The issue does not appear to be correlated to temperature. We first took the car to two mechanics who were unable to find any problem. Finally, a Subaru dealer told us that a diagnostic test revealed the car required a long crank update and that that should fix the problem. We paid for the update, but the problem persisted. A few weeks later, we took it back to the same Subaru dealer who told us that another diagnostic test found that the body control module was problematic. We paid for a replacement body control module, but the problem still persists. At this point, we've paid $700 for two misdiagnoses. We are fed up and are wondering if any one may be familiar with this problem and how to fix it. Thanks in advance for your advice.

[GeneralDisorder]

Anyone check the brushes in the starter?

 

GD

[1 Lucky Texan]

if original, battery may be getting old. It can be tested.

[mikec03]

As already mentioned, it almost certainly must be either a problem with the starter or with the power to the starter [battery or wiring].

 

But I'm puzzled and interested because I own a 14.  If the problem is so simple, why wasn't the dealership able to diagnose it.  The only reason I can think of is maybe it wouldn't manifest itself at the dealership?

[idosubaru]

Sometimes I can't believe these stories where people pay that much money for ***absolutely nothing***.

So bizarre. 

I'd be having some long talks with the dealer, management, and finding out what their overall philosophy is on incorrectly diagnosing something and how they would feel if they were paying for a service that never happen. 

 

If they paid the cleaning company to clean all their matts and carpets and the company changed their toilet paper instead for loads of money they would not accept that deal.  There's absolutely no way they're fine with that kind of operation. 

[1 Lucky Texan]

Sometimes I can't believe these stories where people pay that much money for ***absolutely nothing***.

So bizarre. 

I'd be having some long talks with the dealer, management, and finding out what their overall philosophy is on incorrectly diagnosing something and how they would feel if they were paying for a service that never happen. 

 

If they paid the cleaning company to clean all their matts and carpets and the company changed their toilet paper instead for loads of money they would not accept that deal.  There's absolutely no way they're fine with that kind of operation. 

 

 

yep - I can 'troubleshoot by the law of averages' for a lot cheaper.

[idosubaru]

This is almost a self-defeating process you're attempting - "I'm waiting until the car starts to test for a no start condition"

The vehicle SHOULD NOT BE TESTED unless it's not starting. 

 

Once that happens - then voltages need checked to determine which of these are the culprit:

 

1.  Starter

2.  How old is the battery?

3.  If the battery is new - why is it new?

4.  Are the battery terminals TIGHT and CLEAN?

 

I'd rather not guess or assume but if the shop is guessing all these random parts that aren't normally the issue - it sure sounds like near theivery. 

 

Let's check and replace things that rarely fail?  Makes no sense at all. 

 

You need to verify that the starter is seeing a full 12 volts during the no-start condition - someone with a multi-meter could figure this out in 3 minutes.  The easiest solution is to learn how to do it yourself. 

 

Otherwise you'll have to keep guessing. 

Since they're so cheap i'd maybe plug in one of these just to see if the voltage your chassis is seeing is changing during the no-start conditions:

https://www.amazon.com/Battery-Electric-Cigarette-Lighter-Voltmeter/dp/B00AQAQIGO/ref=pd_sim_263_5?_encoding=UTF8&pd_rd_i=B00AQAQIGO&pd_rd_r=794Z1KCBEXNF6QX2TE3Z&pd_rd_w=Nycsu&pd_rd_wg=k5nzu&psc=1&refRID=794Z1KCBEXNF6QX2TE3Z

 

But truthfully that's unlikely to show you anything. 

 

Better would be simply spending 3 minutes to test things yourself when it's not starting. 

This is almost a self-defeating enterprise - you're waiting until the car starts to go test it for a no-start condition.

[idosubaru]

yep - I can 'troubleshoot by the law of averages' for a lot cheaper.

 

I'll qualify my stance - I don't want to sound like I'm advocating having an attitude and accusing them - it's possible to have a polite discussion with no expectations in order that you can make a good assessment about what to do moving forward, in the future. 

[Dj7291993]

If it's a slow start, the 3 things I would check are the battery condition, voltage drop (battery positive to starter positive & battery negative to starter case), and the starter's current draw.  If you've got a multi-meter, you can do the voltage drop yourself.  I usually put my meter on record mode for max so I don't have to watch it while I crank.  Just have to get your leads the right way round, lest your readings go negative.  Shouldn't be more than about half a volt on either side, ideally less.  They should have tested the battery, but pretty much any parts store will test them for free.  Current draw is going to be a little harder without a high current probe, but you can get an idea if you know the battery is good by watching how low voltage drops during cranking.  One other thing to try is pulling off the serpentine belt and see if it cranks faster.  It's not common, but I've seen accessories lock up that made it act like a seized motor.  You can try putting a socket on the crankshaft pulley and turning it by hand.  You should feel the resistance as it hits each cylinder's compression stroke, but it should be possible to turn by hand.

 

Basically, if it's cranking slow, it's not getting the power it needs.  That could be an issue in the supply (battery, cables, connections), or an issue in the demand (starter is worn causing it to need more current than it should, or something is causing the motor to be harder to turn).  One other thing I always test is that the solenoid is getting good voltage during cranking; however, I've never seen low voltage there cause a slow crank.  Instead, it usually causes an intermittent no crank.

[GeneralDisorder]

We have seen a rash of late model cars (in general) with cracked, worn, chewed up brushes in the starter leading to intermittent slow cranking. I think the last two were Hondas - one with 65k on it. Starter made by Mitsuba.

 

On a car that new it is highly unlikely it has bad cables, terminals, battery, etc. It's almost certainly a faulty part and if it really is, as he says, a "slow crank" then it's probably starter brushes. It's just too new to be anything else.

 

But who knows - maybe he means it's slow to *start cranking. Maybe that's why the dealer thinks it's immobilizer related.

GD

[1 Lucky Texan]

^^^ interesting.  Other than visual inspection, is there a test for that? does whacking the starter alter the brush's contact or ????

[GeneralDisorder]

^^^ interesting.  Other than visual inspection, is there a test for that? does whacking the starter alter the brush's contact or ????

 

Yes - the last Honda that came in on a tow truck we diagnosed by smacking the starter with a hammer and it started. But the starter wasn't accessible so the car had to be put on a lift to do this. Replacement involved removing most of the exhaust system from the header back. Pretty stupid design really. 

 

I disassembled the starter because I'm curious like that and the brushes were worn down to nothing and had turned to mostly dust and chunks. And this at ~60k miles. The one before that had 110k. Seems a problem with at least these Mitsuba starters anyway. 

 

GD

Edited March 9, 2017 by GeneralDisorder

[1 Lucky Texan]

thanx, I was googling a little and saw where there's a lot of dust and chunks inside some of those Mitsuba.

 

are they 2 or 4 brush? is there an upgraded brush plate or brushes?

 

and thanx for letting folks know about this.

Edited March 9, 2017 by 1 Lucky Texan

[GeneralDisorder]

It's a 4 brush arrangement. 

 

As far as parts I don't know - we just get starters with warranties - our suppliers don't carry parts for them and it's a lot easier to lean on their warranty than provide my own. 

 

We only buy the Denso and Mitsubishi remans typically or the ones from the dealer. 

 

GD

[gbrand]

if battery is OE would get it tested-my 12 forester had some slow cranks and then died stone dead 2 weekends ago-would not even take a jump-was 400 miles from home but had a friend take me to buy a new battery and got on the road-you may want to jump with another battry to see if the problem improves-or replace with another battery if you have one to swap.

[Dj7291993]

2013 is a five year old battery, plenty old to be a problem. Cables are less likely on a newer one, but considering the size of cables the OEs are running to cut weight, it doesn't take much to drop them below their threshold.

Hitting the starter is always a pretty good one if it won't go at all, but I got the impression that it was cranking slowly, ie. low rpm. Guess it depends on what they're actually experiencing.

[GreaseMonkey03]

10% of the time is interesting. This is a push button starter? Anyone know off the top of their head what the manual override is for these? If the module stops cranking due to slow crank maybe you can override and see if it starts fine right after.

 

2013 it could just be a battery going bad, which you can have tested.

 

Or

 

we have had subarus go dead apparently just from leaving the key in the ignition or the headlights not going off when the car is off. I wouldn't eliminate the possibility that something might be getting left on once in a while if it's only 10% of the time.

 

I would check the battery first though. Place to start anyway.

[Dj7291993]

10% of the time is interesting. This is a push button starter? Anyone know off the top of their head what the manual override is for these? If the module stops cranking due to slow crank maybe you can override and see if it starts fine right after.

Don't think they were using push-button yet, at least not in the US.  We have some customers bring in 2014 and 2015 models, and they still have boring old keys.

[DaveT]

5 years is end of life for lead acid batteries, under ideal conditions.

 

Sometimes they will continue to function beyond that, but a full test of capacity, and cca would show the degradation.