Voltage drop causing power loss is M278 >: |

[chevota]

Took a while to figure this one out but low voltage is hurting my power. My voltage varies from 14.5 to 12.5V all the time, which I know is an issue but maybe it's one others also have? The problem, on top of that, is under full load it drops ~1V. Not sure if this is normal or not but if it's ~14.5 and drops to 13.5 everything (boost) seems fine. If at 12.5 and drops to 11.5, I lose a good chunk of boost. It seems obvious to me the solenoid can't hold position with that voltage and bleeds vacuum... It's a pattern I see when I look at old data where the turbo DC is good but boost varies, and the only thing that coincides is voltage.

I have the default voltage compensation selected on the turbo page but my assumption is it's dropping lower than it can comp for. I assumed the voltage was 5V, but a big assumption on my part. The next most obvious voltage choice would be 12V, and clearly it can't comp for 11.5. Either way I bet $ it comps by adjusting DC, not voltage, and since it cannot adjust DC higher than the max DC, any V loss will hurt power. I adjusted the max DC from 90 to 95 and it seems to have helped, on avg, but very difficult to tell since the voltage is different each time, as are IAT etc. But on avg I think it helped. I also tried 100% DC which again seems to help a bit more. It's certainly no fix, just a clue that voltage is the issue.

Example for today: With 11.69V @ 5800 I netted 8.8psi >: |
Minutes later with 13.16V @ 5800 I net 13.3psi.
Both cases the turbo duty cycle is the same, throttle fully open and IAT was ~4F lower with the low boost.
Of course I can't say the rpm is accurate, per my post a few weeks ago on that subject; https://forum.hptuners.com/showthread.php?102508
or the boost for that matter, but the boost to voltage issue is a repeatable pattern. When everything is happy and V is good I get ~13.5 @ 5800, 12.5 @ 6k. So regardless of rpm error, 8.8 is simply not possible if the DC is maxed and throttle open etc.

It also makes me wonder about others with an M278 who have low boost @ 6k, which we discussed Aug '22 starting at post 57:
https://forum.hptuners.com/showthread.php?95688

So if you guys keep tabs on voltage, does it drop under load? If not monitoring it, please do and post what you see!

[TheMechanic]

Need a full charging system test. Battery and all. Also possible the battery is failing and after a new is installed you need to perform the battery replacement. Just like BMW.
My first thought if those test ok is to check the serp belt for slipping. Do you get any chirping noise when it shifts from under the hood?

[chevota]

I suppose the belt could be slipping but it's wrapped more than 3/4 ways around the pulley, same with the crank. It's never made a peep and voltage does not seem related to rpm, just load.
Aside from eng load the only other predictable thing is elect load. If I turn on the seat heater, AC etc the voltage goes up. Maybe tomorrow I'll leave the seat heaters on and see if voltage and boost stay high.

Rekon I'll get a new batt this week and see if it changes anything. Unfortunately I haven't been monitoring voltage the whole time so I don't know what it did when the batt was new, 18 months ago. I only added voltage to the list after the batt died a couple times because the car won't sleep. It's active 24/7 unless I force it off by killing power to a relay that all the power is going through. I didn't want to replace the batt until I fixed that, but I doubt it can be fixed.
I don't believe there is a batt reset on mine. I looked for it last year and found nothing in the car menus, Autel or Xentry. Maybe I overlooked it?

[TheMechanic]

Possible it does not have a reset. Many EU vehicles have it. It will change the charging strategy over time as the batteries capacity goes down. You could put a socket on the end of the pulley and give it a turn with a 3/8" ratchet. You shouldn't be able to turn it at all. If it slips even a little it is possible your tensioner is bad and or belt is worn/slipping. Regardless I really suggest getting a proper system diagnostic and includes a loaded alternator test. Not just a voltage output but an actual carbon pile load test at about 1,800 or so rpm. along with of course a battery capacity test. A good functional battery should keep the voltage in the 12.0 range even under extreme load with a failed alternator. A good battery is usually tested to 50% CCA or typically upwards of 400 amps. Unlikely that with every accessory on and every single thing loaded in your vehicle it wouldn't take more than 175 amps and that's momentary.

[Schnell]

Interesting...will check my logs

[chevota]

I looked at some M278 logs from others here and they drop voltage under load too. The same .5 to 1v but their normal voltage is higher so with the drop they're still at least 13v. Not many logs available to view but I'll assume that's normal and dropping below 13 is abby normal.

I bumped my turbo DC to 100 and it helped a lot, almost as good as 90% DC and 13v. In the past, before my low voltage, anything >90% DC was pointless. So I guess 90% and 13 is just barely enough and less of either will cost you. But also note I have another turbo solenoid that doesn't work nearly as good. At 90% DC the vacuum at the turbos is much less. I forget the # but it was a lot and it hurts power a lot to match. At oem boost it was not a big deal, but above that you're screwed. So don't assume your solenoid is good simply because it seems to work. I had to put a gauge on the turbo line to see it, and the brand new solenoid I bought is the bad one. From what I recall the good oem solenoid was 25" vac at 90% DC and the other was something <20.
I think I'll take the bad turbo solenoid apart and mod it to work at the lower voltage, to both fix it and I don't like the idea that even a good one is borderline enough. And get off my butt and get a new batt.

[TheMechanic]

Might want to find out if there is a higher amp rating alternator out there for your application or have a good re builder wind it for a higher current. It would be interesting to see the current load at its peek and of course the current capability of your installed alternator.

[clifftoo]

On tunedmerceds.com I wrote this up as one of my first mods. Coming from Range Rover I saw first hand how the drop can affect the performance of sensitive electronics which may not result in optimal performance.

[TheMechanic]

On tunedmerceds.com I wrote this up as one of my first mods. Coming from Range Rover I saw first hand how the drop can affect the performance of sensitive electronics which may not result in optimal performance.

you wrote what up? The original issue? The diagnosis? The fix?

[chevota]

I bought a new batt and voltage was 14+ on my drive home. Car sat a couple days and batt was dead. Swapped in my old batt, voltage was at 14 for my drive to work. On the way back it started to fluctuate on down to 12. Exchanged new/bad batt for another installed and voltage back to 14+. Next day fluctuating to 12 again. So it seems when I disconnect the batt it resets whatever is causing the low voltage issue, but after xx time it reverts to its old ways. Wth could cause that?
If it were the belt or Alt I don't see how it would change, so maybe the little BS things on the Pos and Neg batt cables? They look like they were designed to cause people problems because they are not needed.

[TheMechanic]

If I am not mistaken this alternator is controlled by the PCM via a LIN bus connection. Simple 2 wire alternator. Three possible ways of testing. One is full fielding the alt via a PCM/Scan tool and measuring current and voltage. "O" scope the LIN bus line to see if you have communication stoppage. GM is easy peazzy with the Data Bus Diagnostic Tool. Can't believe MB doesn't have the same. Last but not least I believe that MB does have a stand alone full fielder for the alternators. I have one for those pesky early model GM full fielders when they started controlling ALT's with PCM. Lots came in to my dealership with new alternators only to find out it was a bad PCM. That was back when a PCM was $500-$600. Thats early 2000 dollars. That would be like $50 million now.

[chevota]

By PCM I assume you mean ECU? Because I have a spare ECU I could swap in.
I don't think the Alt has an output issue considering I can load it down and the more I load it the better the voltage (on avg). Then there's the whole works fine after I disconnect the batt.

I bought a little scope last year to check the car but have yet to do it. It means I have to read and learn stuff I have zero interest in, hence why I have not. Getting a headache thinking about it. I was also unaware there was yet another bus aside from CAN. Now I do have a headache. I do appreciate your help because when it comes to mystery electronic issues in something I have no schematic for and the mfg is of zero help, it's frustrating.

I don't suppose you know what those little thingies are on the batt cables? I think the pos side is just a fuse but the neg side I'm baffled, and very suspicious:

batt thingie.jpg

[TheMechanic]

Yes ECM. Engine Control Module. Not PCM powertrain control module.
As for that part it is a shunt. MB calls them "Intelligent Battery Sensor". Not really "intelligent". All it really is, is a low ohm resistor. The plastic plug on the side are wiring on each side of the sensor. It uses voltage drop across that resistor to determine current. Ohms law if you know the resistance in a circuit and you know the voltage than you can calculate the amps through the circuit. If you have another ECM then that would be a great cost free test.

[chevota]

No change after ECU swap.