XDmToter,
Your question about why the failure doesn't show at 10°F or higher, but has a 50% probability
of showing at 0°F or lower is a good one. Mechanical items with greased pivot points are known
to stiffen up as the temperatures drop, but even so, plenty of starters are reliable at subzero temps.
On the other hand, we know that batteries generate electricity via chemistry, and further the
rate of chemical reactions is dependent upon ambient temps. Check out this chart, taken from
the side of a carbon pile battery load tester, and see how the 'test pass' voltages drop as the
temps drop:
You must be registered for see images attach
NOTE: A battery load test is simply pulling 1/2 of the battery's CCA (Cold Cranking Amps)
rating for 15 seconds, and then verifying that the battery voltage hasn't dropped below
the pass voltage listed at the ambient temp that the test was taken. For example,
if you have a 600 CCA battery, you would connect & then adjust the tester to pull 300 amps,
and when the beeper goes off 15 seconds later you note the voltage and then remove the load.
60°F out? The voltage needs to be 9.6V or above.
0°F out? The voltage now needs to be 8.6V or above...a full volt lower. (!)
As we can see, a battery's ability to generate current without excessive voltage drop is
directly related to ambient temperatures. Does this mean that I am going to blame your
battery via long-distance troubleshooting? Not yet, for there's more to consider.
****
Now, we have 4 possibilities as to what's causing us to end up with insufficient power for reliable starter
operation once we get to 0°F. They are as follows:
* Battery
* Starter
* Total Circuit Resistance of the purple solenoid wire.
* The Total Circuit Resistance of the large diameter positive wire between the battery and the Starter Solenoid.
Weak Battery that shows full on a battery charger
Battery is weak & is primarily surface charging. This will cause the battery charger to see a 'full' voltage level prematurely,
and (incorrectly) indicate a fully charged battery, only for the battery to subsequently fail to deliver the necessary
current when called upon by the starter. So of course we need to test the battery's current delivery capacity against
the specifications, and act accordingly.
And in my humble opinion, all the fancy handheld testers that measure internal battery resistance via different AC test
frequencies are too clever by half, and have lied to me. In my humble opinion, the only foolproof test is to make the
battery perform real work into a real load, requiring the use of a carbon pile load tester. (When I first started fixing in
a shop, I was raised on a Snap-On VAT-40, which took ALL of the guesswork out of whether or not a battery had enough
reserve capacity for the upcoming winter season. (See attached for a picture of the gold standard of battery testing.)
Of course a new commercial-strength battery load tester similar to the VAT-40 will set you back a nontrivial amount of money.
In order to have the same testing capability that would fit into my finite budget, a dozen years ago I purchased
a light duty 500 Amp carbon pile tester, and it has since paid for itself many times over, both in avoiding chucking a battery
still capable of deep current delivery, as well as more importantly stopping the troubleshooting madness where a weak/marginal
'Bad Actor' battery would act fine on a battery charger, always make a voltmeter read correctly...but at the same time cause
the car owner to lose confidence in the starting reliability of their DD by letting them down when it got really cold.
Note: This tester is identical to mine except for the color: (Budget 500 Amp battery load tester that doesn't lie to me)
****
The Starter
What a lot of people don't realize is that the starter operation can be broken down into 2 distinct steps:
1) Starter solenoid windings are powered by a human turning a key to the Start position. During this phase,
*all* power supplied to the starter is through the Purple wire only. (!) In order to understand this 1st step
in the starting process, take a close look at this image:
You must be registered for see images attach
Alright. The voltage/current delivered to the solenoid winding by the purple wire is going to make the
solenoid kick the starter's pinion gear out towards the flywheel ring gear thanks to the
pivot point. In other words, the solenoid is actually moving from Right to Left, while the pinion gear
is being forced Left to Right.
Why is this important? Because the BIG power being fed via the larger diameter positive cable down
to the starter motor doesn't occur until the big brass contacts are made at the connector end of the
starter solenoid. And keep in mind that the starter solenoid windings are like the voice coil in a
woofer -- the higher the voltage, the farther/harder the assembly is going to move from the at-rest
position. EDIT: In the case of the starter, the higher the voltage applied the further the solenoid will be
moved against the spring trying to hold the solenoid in the unpowered/at-rest position.
****
Pulling this all together, we can see it's possible that a marginal battery can generate enough current
(while maintaining enough voltage) to make the starter solenoid kick the full travel with enough authority
to ensure that the high current contacts make and allow the starter motor to begin spinning at 40° F.
But when we get closer to 0°, the voltage starts to fall off, and you finally reach a point where the
starter solenoid kicks most of the way, but not quite far enough to make the all-important high
current contacts for the starter motor to start spinning. But substituting a tested-good battery
now makes it happen 100% of the time at whatever ambient temps are encountered.
****
But let's say that a tested-good battery is in place, and the problem still occurs? Is it
possible for a starter to wear in such a way that it requires too much from the battery
in order to work reliably? Absolutely. In the old days a starter solenoid that started
acting flaky would be changed out, and the original starter with a new starter solenoid
would soldier on. But we don't do that anymore. Instead, the whole starter/starter solenoid
is changed out as a warranteed assembly.
But what about the wiring to the starter? Of course the grounds must be in new condition,
no surprises here.
And lots of attention is lavished on the 2 gauge positive cable between battery and starter,
for oftentimes this is the root cause of the problem.
But what a lot of people don't realize is that the entire circuit between the 'CRANK' fuse, the
Ignition Switch, the Neutral Safety Switch, and finally the purple wire/connection to the
starter solenoid can, with enough resistance at any point in the path, make a good starter
or battery seem bad, because voltage lost in the circuit is not available to make the
solenoid throw far enough to get the big power contacts engaged.
How do you troubleshoot this? A: When it's cold enough that your vehicle only works 5 times
out of 10 tries, time to disconnect the purple wire and substitute a remote starter
switch in it's place. (Remote starter switch)
IF the starter now operates 10 times out of 10 tries, you have discovered the root cause
of your cold weather intermittent.
IMPORTANT NOTE: You have to keep your wits about you, for if the truck is a stick and
in gear, or it's an automatic and NOT in Park or Neutral, hot-wiring the starter solenoid
for troubleshooting purposes will get you into hot water immediately if the engine
starts. (!) I'm thinking chocks, parking brake/helper on the brake pedal, clear
communication, etc.
Where do we go from here?
I propose that we test our way out of here. Here's the order I would do if this was
my GMT400:
1) Reverify that the grounds, main starter positive lead, and purple wire connections
are all clean/tight/new condition.
2) Pull the battery (or just drive the truck) to a place where it will actually be load tested,
so that we no-kidding know what it is capable of delivering. Don't be surprised if your
new battery is marginal, for I have found that the less I drive a modern vehicle, the
more trouble I have with the battery. IF the battery is good, then on to step #3.
3) If it's cold enough for the problem to present itself, disconnect the purple
wire and try starting using the Remote Starter Switch. If the purple wire
gives you a 5 times out of 10 success rate, but the remote starter switch
gives you 10x out of 10, then you know the electrical path that will need
attention. But let's say that the starter is STILL flaky with the remote
starter switch? Proceed to step #4.
4) The next step is to have the starter tested. Only problem is, the parts store
really needs to have the tester located inside a walk-in freezer in order to
duplicate the failure scenario...and allow you to leave the starter in this
freezer overnight. Seriously, you need to consider putting the suspect starter
into your deep freezer prior to going to have it tested. Bonus points for showing
up at the store with the starter inside a cooler. :0)
If you luck out, the chilled starter will be balky on the test stand, whereas the
new starter with warranty works with authority.
****
Best case scenario: You drive the truck somewhere to get the battery load
tested. Own, buy, or borrow the remote starter switch in order to prove/disprove
that the control wire feed circuit is like new & good at 0° F.
If a load-tested good battery + good cabling + remote starter switch don't
change the bad behavior, then your next step is to test/replace the starter.
No parts cannon, no guessing.
Apologies for the length, but wanted to share a no-nonsense test strategy
that has worked for me over the years. (When in the military you simply didn't
have the luxury of calling into the shop saying that your car wouldn't start. Even
in the hills of VT. :0)
Best of luck. Let us know what your discover.
Cheers --
Attachments
Last edited:
