Hard to start. Insight needed.

[IslaTurbine]

Posted on Rennlist but figured this is a better place for this. A poster over there suggested it might be a ground issue.

 

My 996 has started to give me trouble with starting recently. I've done many forum searches but can't seem to find any posts matching my issue.

 

When I start my car, whether it's hot or cold, the starter engages but sounds like it barely has enough electrical power to start the car. One time it failed to start the car and started making clicking noises and I had to jump the car to get it going. Otherwise on the rest of the starts it simply sounds like it's struggling. 

 

At idle, the battery is getting 13.2 volts from the alternator. The battery is 1 month old and was tested yesterday to ensure that it wasn't the cause of the issue. The battery test reported that the battery was low on charge and needed to be charged up in order to be tested, which makes me suspect it's the alternator. 

 

At highway speeds, the voltmeter in the car reads around 13.5 volts. Service history on the alternator, starter, and ignition switch are unknown to me. Accessory belt is about 4 years old and appears to be in good condition.

 

Any thoughts?

[Ahsai]

13.2v at idle is a low. What's the reading when fully warmed up and with a/c and low beam on? It should be between 13.5-14v at all times even with some good loading.

You will need voltage drop tests to figure out whether the alternator is not generating enough or you lose voltage on the cable(s)

http://www.renntech.org/forums/topic/47504-troubleshooting-low-voltage-and-charging-issues/?hl=%2Bvoltage+%2Bdrop+%2Btest

[IslaTurbine]

Thanks for that link. I'll do the tests and see what comes up.

[AWDGuy]

sounds similar to what I experienced last April. the voltage regulator that bolts onto the alternator was the issue for me.

$80 at pelican and 2 hours in the garage and some pelican parts how-to tech articles, I was back to 14+ volts at idle and 0 issues starting it.

I always start with the cheapest solution.

[Paul Grainger]

Try checking your main engine earth (ground) connection. Mine became loose at the chassis connection point and I had the exact same symptoms. I can't believe its not a common fault, when I tightened up the connection the stud just sheared off. I drilled and fitted a bolted connection, instant cure and good starter motor performance restored.

 

Paul G

[AWDGuy]

yes, check the grounds 1st before anything else. nothing but time.

[IslaTurbine]

So here's the latest. Due to work commitments I haven't been able run any tests on the system. But I have tinkered with a few things during my commutes.

First, I have DriveSense from eSurance which plugs into the OBD port and monitors driving behavior for potential insurance discounts. I had noticed that it would be warm after the car had sat, even after many days. It appears this gizmo was drawing a fair amount of juice and draining the battery to some degree. So I've since unplugged it and the car seems to perform better during cold starts.

The second thing I've noticed is with the power draw by the climate control. If I start the car and turn off the climate control, the dash voltmeter instantly jumps to 14v. If I turn the climate control back on, the needle drops to about 13 to 13.5v. If I then additionally turn on the headlights (Litronics), the volts drop to about 12.5 to 13 while at idle. Is this sort of power draw, particularly from the climate control, normal?

[Ahsai]

It's normal to drop 0.2 to 0.3v when you turn on the a/c (due to the load from the compressor and blower fan). Your drop is on the high side.

With a/c and low beam on 12.5 to 13v is low for sure. The only question now is whether it's your alternator (regulator most likely) or it's one of the cables (e.g., ground, the starter "y" cable).

 

You can reproduce the above low voltage condition, then measure the voltage between the back of the alternator (the alternator output) and its casing. If it's below 13.5v, the alternator is bad. Otherwise, you will have to perform the other voltage drop tests I mentioned before to pinpoint which cable is bad.

[IslaTurbine]

Here's another update.

I simply didn't have time to do testing on my own so my local Indy did it for only $45. They concluded it was a weak alternator. I figured it could still be a bad regulator but thought it was probably worth my time to simply replace the whole thing if I'm going to the trouble of taking it out. So I ordered a remanufactured Bosch unit and installed it today.

When I got out the old alternator, I noticed it said Bosch and had the Porsche part number 997.603.012.7. Why in the world would the previous owner install a 997 alternator into a '99 996? Are they interchangeable?

Additionally, I'm not sure the new alternator is performing any better. A cold start produced 14v, auto climate control turned on dropped it to about 13.4ish, and then additionally turning on my Litronics dropped it all to about 12.5v.

I'll keep an eye on it for a week or so and see how it's starting performance is. If my problems continue, then I guess it's likely the starter/alternator cable.

[Ahsai]

The 997 part number is OK. Lots of our updated parts have 997 numbers.

 

Check with your indy to see if he measured the voltage directly at the output of the alternator (which is very difficult to access). It's possible the alternator is good but you lose voltage over that infamous "Y" starter cable.

[IslaTurbine]

I believed that he did but now I'm certainly questioning that. I'll call them on Monday to check. They brag about their ability to diagnose electrical issues because they have a side business which deals solely with hybrid/electric cars....I'll be annoyed if they didn't test this properly.

[IslaTurbine]

The saga continues......

 

I spoke to my indy again and they said they did specifically test the alternator. They said it wasn't producing linear, or stable, power and was giving fluctuating indications. FWIW, when I removed the old unit, it did make a slight ticking noise when I rotated the pulley wheel with my hands. The new unit made no such noise.

 

I went out and used my multimeter on the battery and checked it during cold idle and warm idle (190ish degrees) with various amounts of load (high beams, A/C, radio, hazard flashers) on the system. The readings at the battery lined up with what is being indicated on the dash voltmeter, so I think I can safely rule out having a faulty dash gauge. The readings were quite low....12.0 volts with warm idle, A/C running, high beams, flashers, radio on (with M490 amp), 1 door and the hood open and their associated lights.

 

So I guess now I'm down to either a ground issue, though they both appear to look ok. Or it's the Y-cable from the alternator/starter/+ post.

[Ahsai]

First don't trust the look. The starter cable can look shinny and still fails. Ask me how I know.

 

You can use a stiff insulated wire like this to measure the alternator output http://www.renntech.org/forums/topic/44652-alternator-terminal-b-access/?hl=weapon#entry241245 Connect one test probe to this wire and the other to the alternator casing. It should read 14-14.5v with a/c and low beam ON.

 

Then with a/c and low beam ON, connect one test probe to the back of the alternator (using the same wire) and the other probe to the engine jump start post. If you see 0.2v or higher, your starter "Y" cable is bad. If not, connect the test probes to the alternator casing and the engine brace or any exhaust component. This tests your ground strap and it should be very small ~0.2v or less.

Edited April 10, 2015 by Ahsai

[IslaTurbine]

Yeah I meant the ground in the right rear wheel well and the one in the battery box both appear ok. But you're absolutely right regardless, looks don't mean anything with this type of stuff.

[Ahsai]

Take your time to measure them. That's the only way to be 100% sure. A bad connection/wire will show up like a sore thumb with voltage drop test.

[mcbit500]

Before my engine was rebuilt I had the low voltage issue, so while the engine was out I had the voltage regulator and earth strap as a preemptive attempt at a cheap cure. When the engine went back in I still had the issue, so one after another every HD cable in the starter / charging circuit was changed including the front to rear ones; no result, voltage was still low. Completely frustrated I decided to take a look myself. Taking the easy route first, with the engine hot and on load I measured the voltage across the battery directly and then from the +ve terminal to the chassis, both were exactly the same. Moving to the engine bay I measured from the +ve jumper terminal to the engine (13.6V) and then from the jumper terminal to chassis (12.3V) and concluded that the earth strap was faulty. Took the car back to the shop and demonstrated my findings and they said that they had actually checked the earth connections and found that the connections were good. They agreed to test the car using the earth strap from another car in the shop which worked just fine; the shiny new Porsche original earth strap on my car was faulty.

 

A temporary fix was made using a (not very well made up) earth strap from a local electrical shop; I ordered a heavier duty, marine quality, ground strap online and this is what is now fitted to the car.

 

As Ahsai says above, just because it is new or looks OK, don't ignore it when checking.

[IslaTurbine]

Before my engine was rebuilt I had the low voltage issue, so while the engine was out I had the voltage regulator and earth strap as a preemptive attempt at a cheap cure. When the engine went back in I still had the issue, so one after another every HD cable in the starter / charging circuit was changed including the front to rear ones; no result, voltage was still low. Completely frustrated I decided to take a look myself. Taking the easy route first, with the engine hot and on load I measured the voltage across the battery directly and then from the +ve terminal to the chassis, both were exactly the same. Moving to the engine bay I measured from the +ve jumper terminal to the engine (13.6V) and then from the jumper terminal to chassis (12.3V) and concluded that the earth strap was faulty. Took the car back to the shop and demonstrated my findings and they said that they had actually checked the earth connections and found that the connections were good. They agreed to test the car using the earth strap from another car in the shop which worked just fine; the shiny new Porsche original earth strap on my car was faulty.

A temporary fix was made using a (not very well made up) earth strap from a local electrical shop; I ordered a heavier duty, marine quality, ground strap online and this is what is now fitted to the car.

As Ahsai says above, just because it is new or looks OK, don't ignore it when checking.

Finally got around to testing the ground. Positive post to chassis gave me 12.5 volts with a hot car, AC off, radio on, and cooling fan on. Positive post to the engine gave me the same reading, or within .1 depending on where I was touching on the engine.

Initially it seemed my issue was present regardless of whether or not the car was hot or cold. It now appears that the volts only take a dive as the engine warms up. I'll be replacing the Y-cable soon.

[Ahsai]

To measure voltage drop across a cable, the best is to probe the two end points in question and read the relative voltage drop directly in a single measurement, as opposed to taking two separate absolute measurements from a third common point and then do the subtraction. The reason is each absolute measurement is affected by the load (current) so if the load changes between the two measurements, the subtraction will create errors.

 

To test the ground strap, all you need to do is to probe the alternator casing and the chassis (e.g., the brass bolt on the airbox) and the measurement should read 0.1v or less.

[IslaTurbine]

Did as you suggested and the reading was zero.

[Ahsai]

Your ground strap is fine then. However, that still leaves two suspects, the alternator and the y cable. I know your alternator is new but you'll be surprised how often a new one fails.

[mcbit500]

To measure voltage drop across a cable, the best is to probe the two end points in question and read the relative voltage drop directly in a single measurement, as opposed to taking two separate absolute measurements from a third common point and then do the subtraction. The reason is each absolute measurement is affected by the load (current) so if the load changes between the two measurements, the subtraction will create errors.

 

To test the ground strap, all you need to do is to probe the alternator casing and the chassis (e.g., the brass bolt on the airbox) and the measurement should read 0.1v or less.

 

There is more than one way to skin a cat. :) 

 

I would never advocate basing a diagnosis on a single reading; you can repeat the same test about 5 times in a minute and if the values measured are consistent then the test is reliable. The results from the OPs test on the jump terminal, tell him that the earth is OK and that the voltage is either low at the alternator or that the voltage drop is in the cable between the alternator and the jump terminal.

[Ahsai]

 

To measure voltage drop across a cable, the best is to probe the two end points in question and read the relative voltage drop directly in a single measurement, as opposed to taking two separate absolute measurements from a third common point and then do the subtraction. The reason is each absolute measurement is affected by the load (current) so if the load changes between the two measurements, the subtraction will create errors.

 

To test the ground strap, all you need to do is to probe the alternator casing and the chassis (e.g., the brass bolt on the airbox) and the measurement should read 0.1v or less.

 

There is more than one way to skin a cat. :)

 

I would never advocate basing a diagnosis on a single reading; you can repeat the same test about 5 times in a minute and if the values measured are consistent then the test is reliable. The results from the OPs test on the jump terminal, tell him that the earth is OK and that the voltage is either low at the alternator or that the voltage drop is in the cable between the alternator and the jump terminal.

 

 

I should have been clearer. I didn't mean to say you should rely on a single measurement to make your diagnosis. You can measure multiple times and average like you said as long as you measure what I prescribed. My point was do not do subtractions between two measurements taken at two different times (via a common reference point) to get voltage drop. You can get away with that practice ONLY if the load is constant and the alternator voltage is constant. Neither is true on an idling hot engine.

 

Fans kick in, low setting, high setting, alternator changes its output voltage to compensate for the load. So the voltage you measure at one point in time is only valid at that instant. That's why it's important to measure the voltage drop directly between two points at a single instant. You can repeat the measurements to get more confidence, just no subtractions should be used.

 

Hope this clarifies my comment.

[mcbit500]

I wasn't questioning your methodology, I was merely stating that I would not base a diagnosis on a single reading which is why i have 100% confidence in the results I received. Fact is that during the test, hot engine at idle, the readings were consistently 13.6V and 12.3V and the alternator will maintain a constant voltage for minutes at a time. This a simple test from an easily accessible point which has three fundamental outcomes which can immediately point to the area which needs further investigation.

[farmer boy]

I had EXACTLY the same issues and after a lot of messing about a new starter motor solved the problem.  I know how embarrasing it is to push start a 911! 

 

.......just to clarify - the starter was working but only turned the engine over slowly - a bit like having a flat battery.....  I fitted new alternator, new battery and checked all of the connections.  Replacing the starter solved the problem - it apparently had a short "inside" and was "working" but very inneficiently.  Tricky devil to change!

Edited April 17, 2015 by farmer boy

[Ahsai]

I wasn't questioning your methodology, I was merely stating that I would not base a diagnosis on a single reading which is why i have 100% confidence in the results I received. Fact is that during the test, hot engine at idle, the readings were consistently 13.6V and 12.3V and the alternator will maintain a constant voltage for minutes at a time. This a simple test from an easily accessible point which has three fundamental outcomes which can immediately point to the area which needs further investigation.

Yes, I'm sure in your case your diagnosis was correct because your voltages happened to be consistent a the time of the measurements so subtractions did work out. I also explained the same.

 

However, this may not be true in general. If you probe the output of the alternator, you will see that the alternator does change the output when the load changes.