alternator gauge

[VolsFan]

Hi, my dad has an 03 3500 6 spd truck. He was leaving my house and got about 2 miles down the road and a light came on that said check gauges. He said his alternator gauge was flat. He came back and we checked everything out, looked fine, turned it back on and the gauge started to go up to normal again. Is this just a problem with the gauge? I dont know a whole lot about 3rd gen+ trucks. Also, how do you check the codes on it, my scanner only does 98. 5-02, I imagine it threw a code out when it did this, but the check gauge light went off. He said it never said 'check engine' but "check gauges". I want to get this taken care of before he is broke down somewhere.

thanks for any info

Rick

 

[Prairie Dog]

It's actually a volt meter. Have your dad cycle the ignition 3 times within 5 seconds, leaving it on the last cycle, and if any codes are present they will display where the odometer is. I'm sure you'll see a low PCM voltage code but hopefully there will be more.

 

[cerberusiam]

The gauge is actually a fake and set by what the ECU thinks it should read. The problem could be in the ECU, the charging system, the grounds and battery connections, or in the cluster itself. Use the key trick to pull stored codes and go from there.

 

[rbattelle]

The gauge is actually a fake and set by what the ECU thinks it should read.

I respectfully disagree. In my opinion, the voltmeter is an accurate representation of the charging voltage based on measurements taken by the ECM except when the grid heaters are cycling.



It would be convenient if you could get a voltmeter on the batteries when (if) it happens again. That would help eliminate the gauge itself as the culprit.



How old are the batteries? How many miles are on the truck?



-Ryan

 

[cerberusiam]

I respectfully disagree. In my opinion, the voltmeter is an accurate representation of the charging voltage based on measurements taken by the ECM except when the grid heaters are cycling.

Well, if its not accurate when the grid heaters are cycling why would it be accurate when they are not? How do you know its not a switch that comes on at 12 volts and off under that?



Irregardless, it is not in any way shape or form hooked up to the batteries to read the actual charge. It is a fake set by the cluster controller, from a data stream pulled off the bus to the cluster controller, fed by the ECU, sensed from wherever the ECU reads voltage.



A real voltage meter would bounce high and hard every time the grid heaters engaged\disenaged, thats a given. The nice slow measured swing of the needle is obviously a manufactured signal.



As I pointed out, there are a LOT of pieces between the batteries and that needle position on the dash and a LOT of potential for errors. As you said, its representative, not actual.

 

[monkeyman]

To check gauges operation, with key in ignition but off, push odometer trip meter, hold it down for five seconds, while holding it turn ignition on,not start so just lights come on. Gauges will start to sweep, release odometer button and gauges will sweep back and forth. Gauge that does not do full right to left or left to right sweep is bad or going bad. Sometimes you have to repeat twice
Rick

 

[Prairie Dog]

I'm going to have to agree with Ryan. I realize the gage is heavily dampened but it does reflect when ever the batts in my truck have been low or when I had one battery that was defective. The battery would intermittently drop it's voltage as it was starting to short between plates. Then there's the times I've used the batts during camping trips and the meter showed their low state of charge. It also reflects the high charge level after starting and the slight drop as the batts reach full SOC. Checking it against my Fluke DVM shows the meter to be quite accurate.

 

[TCDiesel]

Ok . . I am going to upset everyone with this post. . 3gen charging is base on battery temp and the sensor is on under the drivers side battery , the system charges the rider side battery 1st . The charge is control By the PCM on 03/04 Autos 03/04 manuals and All later models control the by internal Module installed in the ECM. The system looks for signals created by the bus system . disconnect the rider or drive side Neg cable... you will see the gage showing just above 12 volts. . BUT YOU ARE CHARGING MUCH HIGHER THEN THAT... IF YOU DO THIS FOR TOO LONG YOU WILL BLOW UP THE RIDER SIDE BATTERY,ALTHOUGH YOU ARE SEEING JUST ABOVE 12 VOLTS ON THE GAGE. the solution to this is to add another charging cable form the lead pole on the alt to the positive drivers side post . The system reads the Bus signal created and the temp sensor under the driver side battery then commands accordingly. You can have the gage showing 12 Volts but may be charging 15+...

 

[VolsFan]

awesome, thanks for all the responses. I went turkey hunting down by my dads this morning so I'm waiting for him to get in so we can pull up some codes and troubleshoot the gauges. I dont know how many miles are on the truck, I would say 200k+ and they are the original batteries. Will post back with the results

 

[VolsFan]

just checked the codes. P2502- charging system output voltage P0483-cooling fan speed P0622-Generator Field Control Unit P0633- SKIM key not programmed in PCM. I think a couple of these are old (p0483 and p0633). So where do I go from here? Thanks

 

[VolsFan]

oh, and all of the gauges are working as they should, did the test on them. Thanks again

 

[cerberusiam]

Start by cleaning up all the battery connections and grounds. One bad connection will cause the ECU to sense a problem. Load test the batteries and see if there is an obvious weak point. It may be an intermittent thing and the batteries will need replacing.

 

[rscurtis]

200K + and original batteries? After you replace them, buy some lottery tickets.

 

[rbattelle]

Ok . . I am going to upset everyone with this post. . 3gen charging is base on battery temp and the sensor is on under the drivers side battery , the system charges the rider side battery 1st ... disconnect the rider or drive side Neg cable... you will see the gage showing just above 12 volts. . BUT YOU ARE CHARGING MUCH HIGHER THEN THAT... You can have the gage showing 12 Volts but may be charging 15+...

You're right, of course, the battery temp sensor is the key to the charge controller. But I didn't know the gauge works this way (I don't dispute you). Disappointing.

200K + and original batteries? After you replace them, buy some lottery tickets.

I agree. In fact, I wonder if the alternator might also be due for replacement after all those miles.

Well, if its not accurate when the grid heaters are cycling why would it be accurate when they are not? How do you know its not a switch that comes on at 12 volts and off under that?

Well, there's nothing in the service manual to suggest it's just a switch.

A battery temperature sensor, located in the battery
tray housing, is used to sense battery temperature.
This temperature data, along with data from
monitored line voltage, is used by the PCM (ECM
Diesel) to vary the battery charging rate.

Obviously, line voltage is trivial to monitor, so there would be little reason to use a simple binary switch.

System Voltage Message - Each time the cluster
receives a system voltage message from the PCM
or ECM indicating the system voltage is between
about 9. 5 volts and about 15 volts, the gauge needle
is moved to the relative voltage position on the gauge
scale.

I read that to indicate the instrument cluster receives a voltage reading that is normally between 9. 5 and 15 volts. If it were a switch, I'd expect it to say something like, "... indicating the system voltage is above about 9. 5 volts... "

Here's some interesting things from the rest of the voltage gauge section of the manual:

System Voltage Low (Charge Fail) Message
- Each time the cluster receives three consecutive
messages from the PCM or ECM indicating the electrical
system voltage is less than about 9 volts
(charge fail condition), the gauge needle is moved to
the graduation on the far left end of the gauge scale
and the check gauges indicator is illuminated.

System Voltage High Message - Each time
the cluster receives three consecutive messages from
the PCM or ECM indicating the electrical system
voltage is greater than about 15. 5 volts, the gauge
needle is moved to the graduation on the far right
end of the gauge scale and the check gauges indicator
is illuminated.

On vehicles with a diesel engine,
the ECM continually monitors the system voltage to
control the generator output. The PCM or ECM then
sends the proper system voltage messages to the
instrument cluster.

It is possible to interpret this as the voltage signal generated by the ECM is only based on whether the charging voltage is >9. 5v, but it would be a stretch, in my opinion, to say that >12v is the determining factor.

So I could be entirely wrong about the voltage gauge giving an accurate reading of charging voltage. It certainly won't if 9. 5v>Vcharge>15. 5v! But for 9. 5v < Vcharge < 15. 5v, I guess I can't say for certain that the needle position is accurate. It's always matched what I get on an independent voltmeter, but I suppose that could be coincidence.

Irregardless, it is not in any way shape or form hooked up to the batteries to read the actual charge. It is a fake set by the cluster controller, from a data stream pulled off the bus to the cluster controller, fed by the ECU, sensed from wherever the ECU reads voltage.

I agree that you couldn't find a wire going directly from the batteries to the voltage gauge in the instrument cluster. But it seems practical to use the ECM charge control circuit to generate the voltage signal for the cluster at the same time.

-Ryan

 

[VolsFan]

well i cleaned all the posts and traced down all of the grounds and cleaned them all, and it worked great! Or so we thought. Had the batteries load tested at advanced auto and they were about half. He drove it home from my house, and it started acting up again. What I dont understand is how it would just go flat after being at full charge. The truck does have 205k on it with the original batteries, but I figured it would be a gradual drain rather than all at once. I swear, that is a great truck, he does not maintain a thing on it, and it performs flawlessly, until now that is. I changed the oil on that poor thing, I dont think he changed it for atleast 20k, being generous at that. I'm rambling, but it sounds like batteries I guess, any suggestions? Thanks

Rick

 

[cerberusiam]

I agree that you couldn't find a wire going directly from the batteries to the voltage gauge in the instrument cluster. But it seems practical to use the ECM charge control circuit to generate the voltage signal for the cluster at the same time.

Oh, the ECU uses the charge control circuit to generate the signal, converts it to digital, sends down the bus to the controller, the controller sets the needle based on a predetermined set of values. I see 4 set points: 0, 13 , 15, 18 volts DEPENDING on what the ECU senses. I KNOW the range is a bit broader than that and the moves more rapid from testing, dash gauge does show that.



Its representative of the last sampled condition based the smoothing algorithm but that is all. Accurate is not a word I can associate with displayed readings, anymore than I can the temp or oil pressure gauge.



The point is there is a LOT going on from actual voltage reading to display on the dash. Its no more accurate then the over head mpg display given the known issues with sensors.

I'm rambling, but it sounds like batteries I guess, any suggestions? Thanks

Rick

Could be a floating plate in the batteries is causing the issue and thats tough to track. The 03's also have known issues with the ECU connectors corroding and causing a myriad of problems. It could be a slightly correded pin on the connectors causing the issue.

 

[rbattelle]

but I figured it would be a gradual drain rather than all at once.

I'm not too surprised. Seems like modern batteries are often this way - fine one day, dead the next.



-Ryan

 

[Kirwin]

So what your saying cerberusiam...

Let me make sure I understand what your saying: the gauge is using a algorithm to display a voltage on a dash gauge. Seems entirely logical to me. My Fluke 88 DMM does the same thing. No, it's not as highly dampened as the dash gauge, but its still a computer taking an analog input and converting it to a reading on a display just like the PCM does. I fail to see the difference. Perhaps you would like an ociliscope so that you could see the peak values from the full wave bridge rectifier that the alternator is using to covert the AC that the alternator puts out to a very lumpy DC voltage. Using a DMM you can even see the spikes in voltage if your meter is good enough. Heavily dampend readings are the norm in this type of application.



Thanks rbattelle for researching what the service manual actually says.



I also would like a real oil pressure gauge but since I've never heard of a Cummins engine failing due to a lack of oil pressure I guess they figured it just wasn't worth the cost. So again they used a algorithm to get the gauge on the dash to display a value base on inputs including actual oil pressure to lie to you. Again just what a real oil pressure gauge would do unless you actually calibrated it every so often. My $0. 02 Ken Irwin

 

[steved]

At 200k, and the fact the needle flatlines, I would suspect the brushes in the alternator are shot... mine were barely touching at 245k, and I'm surprised it was still working (my thread is fairly recent, and has pictures if you want to search). Think about this, what in a battery would cause voltage to suddenly drop to zero, let alone with two batteries in parallel.

 

[VolsFan]

Thanks for all the great information. I think he is going to get the batteries swapped, probably with stock ones, if the price is right. It seems like we always go aftermarket on things, but for 200k on the original mopar batteries, can't really knock that.