How to Recondition/Service Your 12V Battery

[Mach-Lee]

Ran another discharge test at 5A today (click on graph for larger version), and let it run for two charging cycles. Power off, MME at rest, load at 12V jump points (as you noted 12V gnd sensor in circuit), used OBD "power cord" for 12V measurement (5A is not flowing through any of the voltage sense path). Horizontal axis is hours.

Not sure if 11.9V is optimal for the low end voltage, however what is clear is that MME is reliably identifying the need to charge, then charging the LVB. Unclear why some are still have 12V LVB low voltage problems. I am running the latest MME (FDRS) software as of November. Turned off the 5A active load at the very end right side.

Yeah 11.9V is a little low. I have a theory that it only checks the battery voltage every so often (once per hour?) so it might be a little late kicking in. Try discharging at something like 1-2A and see if the voltage threshold changes, that might answer that question.

 

[louibluey]

Yeah 11.9V is a little low. I have a theory that it only checks the battery voltage every so often (once per hour?) so it might be a little late kicking in. Try discharging at something like 1-2A and see if the voltage threshold changes, that might answer that question.

There may be different MME versions of the turn on strategy, but as best as I could see during the 5A test, voltage was the trigger criteria, both times at about 11.9V.

I did run a first run earlier (last evening to this morning) at 1A which showed more of those -0.3V spikes which also seem to correspond to a wake up. One of them happened around time I was looking at the DMM through the driver's window, and I wondered if it was the motion sensor, or maybe my Apple watch. This one (only one partial cycle) did come on closer to 12.0V, all of the previous converter starts were around 11.9V. Interesting that there were none of the odd 0.3V marks on the 5A chart (except at DC/DC converter turn on and turn off). I might have caused these by some kind of bluetooth signal or other activity. Right side as before, 1A active load turned off. MME was not plugged in during this test.

 

[louibluey]

These last two discharge tests at 1A and 5A were only power off (rest), and unplugged. Maybe something different happens when plugged in, or when MME goes into sleep mode. I have lost track, is there just sleep mode, or another one, deep sleep?

When I restarted FordPass, I got these warnings, the 12V service warning possibly based on the relatively short time between DC/DC converter cycles or with the battery sensor in the circuit, because of the constant loads (1A, then 5A).

I had received that 12V service soon message several times in the past before this LVB testing, just happened on its own. That is why I became interested in reconditioning and battery testing from the original post in this thread. Since then, from the battery testing discussed above, we know my 12V LVB is fine, although it is less clear what part of LVB fine is the reconditioning as per the first post. Also, the last 5A test ended with the LVB at 13V, so it is unclear whey MME thinks my 12V battery has a problem.

 

[louibluey]

FWIW, I went back to trying to understand the resistance of the ground cable and battery sensor between the negative terminal of the 12V battery (LVB) and the negative ground chassis jump post (~0.55 milliohm +/- 0.05) the resistance of the battery sensor alone (~0.17 milliohms +/- ~0.03) and double checked the entire ground leg including the battery sensor by voltage and current measurements (~0.55 milliohms +/- 0.05). While I was at it, the resistance plus jump terminal to the accessible edge of the 12V battery clamp is also ~0.5 milliohm.

BTW, if anyone knows anything more about the battery sensor, I would really like to know what type it is and how it works. There are only two wires at the connector. It could be a simple passive current shunt (resistor, maybe 100 or 150 micro ohms?), I guess that is mostly likely, but the value seems low for that. I would think an active sensor like a hall effect current sensor would have at least three wires, but who knows.

The reason all this is relevant to the battery (sort of) is that I am trying to understand 12V bus voltage droop with various LVB load currents. I was wondering how much of that is battery electrochemistry (or something else) and how much was simple voltage drop (loss) across the battery cables and battery sensor (turns out not much, under .02V for the ground cable and battery sensor combined).

 

[Mach-Lee]

There may be different MME versions of the turn on strategy, but as best as I could see during the 5A test, voltage was the trigger criteria, both times at about 11.9V.

I did run a first run earlier (last evening to this morning) at 1A which showed more of those -0.3V spikes which also seem to correspond to a wake up. One of them happened around time I was looking at the DMM through the driver's window, and I wondered if it was the motion sensor, or maybe my Apple watch. This one (only one partial cycle) did come on closer to 12.0V, all of the previous converter starts were around 11.9V. Interesting that there were none of the odd 0.3V marks on the 5A chart (except at DC/DC converter turn on and turn off). I might have caused these by some kind of bluetooth signal or other activity. Right side as before, 1A active load turned off. MME was not plugged in during this test.

I see this as an issue with the programming, recharge should be triggered before it drops below 12V. Yes I agree those dips are probably the car waking up and drawing a lot of current for a short time.

I had received that 12V service soon message several times in the past before this LVB testing, just happened on its own. That is why I became interested in reconditioning and battery testing from the original post in this thread. Since then, from the battery testing discussed above, we know my 12V LVB is fine, although it is less clear what part of LVB fine is the reconditioning as per the first post. Also, the last 5A test ended with the LVB at 13V, so it is unclear whey MME thinks my 12V battery has a problem.

It may not trigger the warning message right away. I'm not really clear on the criteria. I should emphasize that the car shouldn't have to do the "wakeup recharge" thing if everything was working correctly. A recharge being triggered while off is a cause of minor concern, so that may trigger a warning. Perhaps it has to happen 2-3x in a certain period before you get notified.

BTW, if anyone knows anything more about the battery sensor, I would really like to know what type it is and how it works. There are only two wires at the connector. It could be a simple passive current shunt (resistor, maybe 100 or 150 micro ohms?), I guess that is mostly likely, but the value seems low for that. I would think an active sensor like a hall effect current sensor would have at least three wires, but who knows.

Looks like the sensor is "smart shunt" type:

• It measures current, voltage, and temperature of the battery and has a microprocessor that is capable of highly accurate calculations of battery state variables that are sent to the BCM via LIN communication line.

One wire is + from the other terminal, and the other wire is the LIN bus (single wire data bus) that goes back to the BCM.

https://www.hella.com/mining/en/Intelligent-12-V-battery-sensor-IBS-2652.html

I found a spec for the shunt, it's 68 µΩ.

The internal resistance of the battery is by far the largest contributor to voltage drop. I think we said it was like 6 mΩ, vs. fractions of a milliohm for the cables.

 

[louibluey]

Excellent, thanks. I found a couple of more references too (below). Interesting, so they get the 12V bus measurement from the same small wire that powers the unit (so it can do current by shunt, voltage, and temperature).

I watched the 12V bus for some 12 hours or so while on L2. There were a few dips, roughly corresponding to 15 mΩ (probably battery electro-chemistry). A remote unlock caused a couple of dips about a second wide several seconds apart an some other activity which dropped the battery voltage 0.2V which seemed high compared to the energy used.

Otherwise, I have seen one more battery saver message on Sync4, and one more sleep on FordSync. Curious.

The bms being a micro with LIN reporting, and the corresponding software processes on the other side, whatever modules are reacting to it, certainly adds another dimension to whatever is going on with MME 12V maintenance. OTOH, apparently Ford has been using this LVB negative terminal BMS for many years, so it shouldn't be new tech on their side.

Bosch Mobility Solutions

Ford Battery Management System

How Long Does a Battery Temperature Sensor Last

Update: fixed the line above. Also, interesting, so the battery sensor really is a three wire device (makes sense), +12V, one wire data (LIN bus) on the two wires, and of course, it has one of the best ground connections in the house (installed directly on the 12V battery terminal).

I got another FordPass notification this morning, your 12V battery needs to be replaced soon! (I did the bms reset when I installed the new one, but did not change the bms as suggested).

I am looking into monitoring the BMS LIN line. I read somewhere the Ford wants under 50 mA parasitic draw when MME is completely asleep. So, I have a bad feeling BMS also goes to sleep with some MME sleep mode.

 

[Kevin P]

I read through this and didn't see an answer... Is it a good idea, or bad, to have the HVB charger plugged in while also charging the 12V battery with an external charger?

 

[ProximusAl]

I read through this and didn't see an answer... Is it a good idea, or bad, to have the HVB charger plugged in while also charging the 12V battery with an external charger?

I've just charged my LVB before the winter cold comes, and I did not have my HVB plugged in.
I plugged it in after I finished charging my LVB.

The simple reason for me is certainly when programming using FDRS it explicitly says NOT to be plugged in to an HVB charger whilst doing updates.

One probably has nothing to do with the other, but I apply it everywhere. When charging LVB, or programming with FDRS, I do not have HVB plugged in.

 

[Mach-Lee]

I read through this and didn't see an answer... Is it a good idea, or bad, to have the HVB charger plugged in while also charging the 12V battery with an external charger?

It mostly doesn’t matter, which is why it wasn’t specified. 12V will get charged either way, unless the car’s charging of the battery during a HV charging cycle shuts down your battery charger early. If using a power supply it wouldn’t matter.

 

[Kevin P]

Interesting findings here... I pulled the panel off in the frunk and charged the LVB for a good while. After that, I am not having the (so far) issue where the car ignores the HVB parameters and tries to charge to 100%. I think what was POSSIBLY happening is the car was detecting the LVB having low voltage and "waking up" and using the HVB to charge it, which then started the external charging too. Last night it stopped at 90% and didn't restart heading to 100% for the first time. More testing needed...

I probably didn't charge it long enough to really desulfate completely, but I was at 12.4V no-load when I started and over 13V no-load when it was done. My charger would kick off when the amp draw got low enough (below 2 amps).

 

[louibluey]

@Mach-Lee was thinking maybe under some circumstances MME waits too long to check LVB voltage, so if I understand correctly, lets LVB discharge too much before taking action (sill a possibility).

My limited testing so far on my FE showed that unplugged (no charge cord), by about 12 to 11.9V, at least twice during the 5A discharge test (no idea if there is a limit), the DC/DC converter did come on and charge the LVB back to 13V. At least in these very limited examples, something seemed to be sampling at least on the order of minutes or seconds.

I wonder which LVB voltage MME is looking at when it decides whether to charge the LVB? We now understand that that BMS (the microcomputer in that module on the negative terminal) sends 12V voltage, current, and temperature data to the body control module (BCM) where it is sent on to other modules by CAN bus. BMS talks to BCM by a "1 wire" automotive bus called "LIN".

In FDRS data logger, there is at least one other LVB value called something like module voltage, and certainly other modules have access to 12V voltage without needing the LVB data from BMS.

My limited unplugged testing was during "rest", I do not yet understand if that is the same as "sleep" and/or if there some further "deep sleep mode".

Short term, I am curious to watch the BMS data directly on LIN bus #4. My lab scope can decode LIN to hex with some optional software, looking to get a 30 day demo to try it out. otherwise there are at least two good options for sniffing a LIN bus (1 wire referenced to ground) such as: Part Number: APGDT001 LIN SERIAL ANALYZER or, CANedge2, by CSS electronics, CANedge2 is really interesting where it can log data to a SD card, or send data to a server by WIFI ($500). They have at least one slightly less costly model that only logs to SD card and other CAN products. CANedge2 does both LIN (up to two LIN bus, and CAN). CSS has good tutorials on CAN and LIN, such as this LIN tutorial. If I keep going with this (not sure), something like the rugged CANedge2 is better than leaving my lab scope out in the unheated, sometime high humidity, garage for any length of time. Note that none of this equipment gives more than raw hex values, just like CAN and PIDs, the device data frames, offset and scale factors need to be determined. Also, the LIN slave devices, like BMS, get polled by a master, so unfortunately, you cannot just setup a BMS on a 12V battery have it broadcast alone in the lab.

Of course beyond learning about LIN, I am not sure if monitoring BMS will shed any light on why LVB is going low sometimes. BMS also has a sleep mode, so I am curious when and if they put BMS to sleep. OTOH, if LVB charging is determined by some other 12V monitor by some other module, it may not matter what BMS is doing.

 

[generaltso]

I noticed these values in the BCM. I wonder if increasing the Battery Target SOC% will change the DC/DC charging behavior. It's interesting that the car seems to think the battery is 60Ah.

 

[Mach-Lee]

I noticed these values in the BCM. I wonder if increasing the Battery Target SOC% will change the DC/DC charging behavior. It's interesting that the car seems to think the battery is 60Ah.

Hmm, the 90% is expected, that’s similar to other Ford vehicles. The 60ah is interesting though, what scan tool are you using? I’m curious is that’s something the scan tool is interpreting itself or if that text is coming from the car. Increasing target SoC may or may not help, it should work just fine with a 90% SoC target.

 

[generaltso]

Hmm, the 90% is expected, that’s similar to other Ford vehicles. The 60ah is interesting though, what scan tool are you using? I’m curious is that’s something the scan tool is interpreting itself or if that text is coming from the car. Increasing target SoC may or may not help, it should work just fine with a 90% SoC target.

This is with FORScan, so I assume it’s coming from the car.

 

[louibluey]

Tapped into BMS LIN! Okay, got to make a simple harness to tap LIN bus #4 today at the BMS sensor. Got the LIN decode software for the scope. As soon as I turned MME on, the frames were continuous (before it was quiet at rest / off, possibly a frame here or there). Not sure how many other devices share LIN bus #4 with the BMS.

Ordered a CANedge2, by CSS electronics, probably the easiest way to sit on LIN #4 bus and to log and/or chart the data from the LVB (voltage, current, and temperature, possibly more). The scope is a good first start to measure the LIN parameters. It only started to decode when I set 10.4 kbs baud rate, so odds are that's it for us (TBD). There is another parameter Sync Break to figure out, apparently 11, 12, or 13. Also, LIN bus standard LIN 1.3 or LIN 2.x?

Also, LIN seems to be Vbat (+12V) referenced, which makes sense, because the two wires into BMS are +12 and LIN signal (white violet wire). My first look was LIN to gnd because of the grounded scope input. So, I need to hook up that extra thin black wire in the pair after all (just taps in to the second green blue wire). Okay, so the little C2 box can run on the same Vbat that runs the BMS, so just three wires, Vbat, LIN, and ground. On to learning about LIN frames.

So, this is related to the LVB topic (although maybe I should start a BMS/LIN thread) because it should be interesting to see when BMS sends (on request from BCM) 12V battery information out. For example, when MME is off, is BMS sending out LVB information less often, periodically, or not at all?

Started a new thread, as I think this is going too far off topic from this thread: Looking at MME LIN BUS - 12V BMS Intelligent Battery Sensor

Here is a summary of where I am - wiring harness:

Lin tap:

chassis ground (not available at the connector):

Tapped! (I've done better work, but it will do for now):

10.4 kbs baud rate:

sample decode (no idea if it is correct yet, just starting today, with a lot to learn about LIN bus. [The red checksum means a bad decode, this was just initial looking for the signal work]. Seems like too much information for that frame anyway, oh well, today was just to access and see the frame signals):