Well another issue and sorry for the long post.

I posted about the battery issue in my "Plugs Away" post. After replacing the battery, and changing the plugs/wires a few days ago, he goes out to run some errands and his battery light comes on again. It lights for a minute or so and goes back out. Double check the battery connections and triple check the Battery voltage under load and get the same results as before. High 13's at idle, no load. Mid 12s at idle with everything one. Rev to 2K and it's back up to high 13's even just over 14v.

So check the belts and notice the alt belt is a bit frayed and does feel a bit loose. So I replaced the the belts (pwr steering/ac as well, it was pretty bad). I also pulled the connector off the alt to spray some cleaner in it. Get the belts tensioned. Start it up and check for charging voltages, all looks good.

He goes off to run some errands and about 40 min later, tells me the battery light is on and this time won't go out. He makes it home and I put the voltmeter on and with the engine running it reads just over 12v Rev the engine and no change. The new belt also feels pretty loose and even squealed a bit when I rev'ed it (I had intended to retighten the belts after he drove it a bit anyway). Scanned for codes and got a P1631. While tightening the belt, I pulled the alt connector back off, sprayed some more cleaner in them, use my airgun to blow them out. Reconnect everything and it's charging again. I was pretty sure I fully seated the connector but perhaps i didn't and it came loose. I didn't see any signs of oxidation or other corrosion on either side of the connectors.

So it's definitely charging. I used some long alligator jumpers to extend my voltmeter leads into the cabin so I could monitor the battery voltage while driving. I also turned on combinations of head lights (with and without fogs) and AC (with different fan speeds) to see what happens.

In summary, driving with no lights or AC, voltage remains mid-high 13's even crossing 14v and coming back down but mostly high 13s. Come to a full stop, trans in drive, voltage drops a bit a first but back up to mid-high 13s. While driving (we're talking under 40 mph), voltage remains high 13s. Turning something on drops the voltage a bit but climbs back up. However with lights and AC on max, when coming to a stop, the voltage slowly drops to mid 12s. Accelerate (or just put in neutral and give it some gas), voltage climbs back up again into the 13s. My scanner is still unable to reliably test the battery (it's new), but the alt test still passes.

So while dropping to 12s concerns me, it seems to be fine once I get going. The FSM details how to check the charging system but not by measuring battery voltages (at least I couldn't find it). So before I start pulling up carpet and going through the detailed steps, checking voltage drops, etc., does this sound normal? I don't want to fire the parts cannon (again) and just swap in a new alt unless I have to do so. I certainly don't want to start tracing wires and probing connectors but will if I have to.

Thanks for any help.

 

Well it hard to diagnosis this old a vehicle with a charging problem that is still OBDII on a forum... but I suspect a failing generator? That is a PURE GUESS because tha e other would be a failing ECU/PCM and again this is prure guess and I did consult with my .................eight ball.
You need to use your scanner and test the ECU/PCM charging cycles on the live scope screen of your scanner. If you don't have one or an oscilloscope your trouble shooting this charging system my present a DIY challenge and a half for you. Sorry.

 

I used to see that light activity on Bosch alternators when the voltage regulator was beginning to fail.

 

I believe if you look his year the generator voltage is ECU/PCM regulated.

 

Well that now gets bad. What I know of ECU controlled alternators is they reduce and increase when the demand is needed to improve things like idle speed regulation. I have seen people replace alternators when they see a low charge voltage when it's now a normal situation.

 

Hey I just did a search and the voltage regulator is in the alternator and my diagnosis aligns with what happens when it begins to fail. The ECU has intervention but the main regulation is done by the internal voltage regulator. They are about 40 bucks. Do a google.

 

Right, Regulator is internal to the alt. The connector is only 2 pins. My "magic 8 ball" said answer unclear. So I asked Chat GPT. It said the "under full load" at idle results needs to be above 13 and it's not. All the other testing scenarios (idle no load, or all tests at 2k rpm) check out ok. So I think I'm going replace the alt. I did find the regulator on Amazon for about $40 but, frankly, I don't trust the parts.

First, I gonna check the belt tension again. The Alt belt was a Gates, the PS a Duralast. The alt belt needed tightening yesterday. The PS seemed OK

 

The ECU/PCM is programed with a DTC that will activate the check engine light if the voltage regulator is failing. The exception is the ECU/PCM programing will be place an anomalous or incomplete constant issue DTC in the permanent file of the ECU/PCM during engine cycles and if the occurrence is variable or inconsistent it will continue to keep it in the permanent file area until the sporadic issue has been resolved. The programing at some point may if the problem is not corrected after several complete engine cycles, then activate the check engine light.
One would need to check 2 area's of the ECU/PCM to confirm this, the permanent DTC file and the ECU/PCM counters.
Disclaimer; A 2 pin alternator with an internal voltage regulator can be ECU/PCM controlled ....however I did not look up the exact year and Model Mazda to confirm that. My responses are universally learned or trained basic diagnostics
OBDII and CANbus (since 1995) ECU and PCMs are a great diagnostic aide when learns how to use them... lol
AI like Chat GPT to which I use one program on another vehicle forum daily is always making little or not exactly accurate information. Mostly because it gets it much of its information based on the number of results posted not only by empirical sources but also by the number of answers from the anyone posting them. What is comical is that occasionally I will directly respond to the AI posted information and the response to me directly in a follow up response " you made a good point"..... lol

We are not yet at the level of DATA on Star Trex regarding AI.

 

The P1631 code did not trigger the CEL, although the battery light did come on. I was able to clear it when I re-tensioned the belts and cleaned the alt connector. Although we haven't driven it far (only about 20 miles), the code has not returned nor has the battery light come on.

I have an Innova 6310 handheld. While it has a lot of functions, not all are supported by the P5. For example, I can do the alternator test (basically what I did with the voltmeter), but not get alternator live data. The scanner has the test function but reports the P5 doesn't support it.

I'm certainly no Data. I'm more like B4 on this one!

 

Auh a fellow Trexer!!


Keep in mind that if the check engine light is activated on any OBDII or CANbus clearing the DTC only removes it from the viewing caution of the check engine light it does not remove it from the ECU/PCM and may return after 2-5 complete engine cycles if the issue causing the DTC(s) (check engine light activation) was not corrected. This is one reason I often respond that DIY should also check and note the current conditions of the I/M monitors when scanning the ECU/PCM if applicable for the year regarding emission standards that were implemented during the manufacturing year.