Mercedes EQ All Electric Forum banner
21 - 38 of 38 Posts

·
Registered
Joined
·
81 Posts
Discussion Starter · #21 ·
I should have added that, after a journey of about 110 km yesterday, I checked the Service page in Mercedes me and the 12v battery 'monitor' now says "Charged" (previously, "partially charged"). I look forward to monitoring the 12v battery charge more accurately once I've fitted the Cetek monitor.
 

·
Registered
Joined
·
121 Posts
Did you have your car max charge set to 50%?
Yep.
If my Ctek monitor has a fault it’s that the history graph for voltage isn’t nearly as detailed as your monitor,
My no-name monitor is great bang for the buck. It's spectacularly accurate: when my calibrated Fluke multimeter claims 12.52V it claims 12.53V. With a bit of hacking you can get your hands on the raw data which will give you a timestamped voltage reading to 2 decimals places every 2 minutes, like:

12.53 Fri Jul 22 08:17:00 2022
12.53 Fri Jul 22 08:19:00 2022
12.52 Fri Jul 22 08:21:00 2022
12.53 Fri Jul 22 08:23:00 2022

If you're that way inclined, you can then drag that into a spreadsheet and zoom in on as much detail as you want. Here's a plot of a 600km roundtrip motorway trip (Gold Coast to Coffs Harbour and back). I did more DC charging than I needed to on the way home because I was testing out various charging networks.

Colorfulness Rectangle Slope Font Plot


The onboard DC/DC determined that the 12V was fully charged throughout so didn't do any charging. It just held the voltage at 12.8V for the duration thereby supplying all the loads (aircon fans, seat chillers etc.). Once at the charging stations, it steps it up to ~13.9V for the duration of the charge. It tries hard to ensure it doesn't have to charge the 12V battery once underway.

Somewhere I've got a pic of what happens when you set off with a partially discharged 12V battery. In that case the DC/DC does a full charging cycle on the 12V battery once underway. If you check the Consumption screen in that case, you can see a much higher %-age of energy go into the 12V stuff and efficiency/range suffers a bit as a result. If I can find it, I'll post it here.
 

·
Registered
EQA 300
Joined
·
74 Posts
Yep.

My no-name monitor is great bang for the buck. It's spectacularly accurate: when my calibrated Fluke multimeter claims 12.52V it claims 12.53V. With a bit of hacking you can get your hands on the raw data which will give you a timestamped voltage reading to 2 decimals places every 2 minutes, like:

12.53 Fri Jul 22 08:17:00 2022
12.53 Fri Jul 22 08:19:00 2022
12.52 Fri Jul 22 08:21:00 2022
12.53 Fri Jul 22 08:23:00 2022

If you're that way inclined, you can then drag that into a spreadsheet and zoom in on as much detail as you want. Here's a plot of a 600km roundtrip motorway trip (Gold Coast to Coffs Harbour and back). I did more DC charging than I needed to on the way home because I was testing out various charging networks.

View attachment 1418

The onboard DC/DC determined that the 12V was fully charged throughout so didn't do any charging. It just held the voltage at 12.8V for the duration thereby supplying all the loads (aircon fans, seat chillers etc.). Once at the charging stations, it steps it up to ~13.9V for the duration of the charge. It tries hard to ensure it doesn't have to charge the 12V battery once underway.

Somewhere I've got a pic of what happens when you set off with a partially discharged 12V battery. In that case the DC/DC does a full charging cycle on the 12V battery once underway. If you check the Consumption screen in that case, you can see a much higher %-age of energy go into the 12V stuff and efficiency/range suffers a bit as a result. If I can find it, I'll post it here.
Wow your monitor is at the opposite end of the scale from the Ctek as far as detail goes, Ctek is very accurate in real time (confirmed by my battery tester), it’s just the historical display in the app that is too vague. It does what I need in letting me know if I have to charge the battery when the car is parked but yours would be a better option if you needed to solve some intermittent problem during driving for example.
One thing these don’t show of course, which would be useful, is what amps are going to the battery. I sometimes found the ammeter fitted in older vehicles gave a better view of what was happening in the charging system than a voltmeter.
 

·
Registered
Joined
·
121 Posts
One thing these don’t show of course, which would be useful, is what amps are going to the battery. I sometimes found the ammeter fitted in older vehicles gave a better view of what was happening in the charging system than a voltmeter.
Yes, with just a voltmeter you kinda' have to infer what's going on. My brother uses one of these in his camping rig and it makes a half decent attempt at coulomb counting to determine the SoC. The shunt would be a pretty intrusive installation into an EQ though.

By comparison to the above plot where it effectively did no charging during the drive because it knew the battery was full, here's what it does on a 1 hour drive when you launch with a depleted battery:
Colorfulness Rectangle Font Parallel Slope

The car hadn't been used for couple of days, and I ran the LH clearance lights for a few hours to further deplete it. That got it down to 12.4V when resting (and 12.0V with the lights on). That was enough for the app to declare it partially charged (and fully charged by journey's end). Conditions were cold, dark and wet, so plenty of 12V loads on (seat heaters, wipers, lights etc.)
 

·
Registered
Joined
·
81 Posts
Discussion Starter · #25 ·
Having enquired about the merit of driving in D- in the first post of this thread, with the benefit of information gleaned from ensuing posts, I now drive in D and use D- intermittently, when there's a need to slow down more quickly than by simply decelerating. In my motoring environment (mainly urban, up to 60 kph), I estimate that by driving in D my EQA250 consumes about 10% less energy than when leaving the setting on D-.

Since my last post, I have fitted a Cetek battery monitor to the 12 volt battery and several times after checking that battery's state of charge I've also checked its status in the Service section of the App. On two consecutive occasions, the App described a state of charge in the low 70% range as, respectively, Charged and Partially Charged. I don't know whether the App has a lower battery status category than Partially Charged, but in the absence of more accurate information, I suspect Partially Charged and Charged are somewhat relative, Charged indicating a satisfactory state of charge and one relatively higher than a previous reference point.

Maybe someone has a better understand of how the Mercedes me 12 volt battery state of charge indicator works. Given that it's in the Service section of the App, perhaps its intended to be used as a prompt for a non-mechanically-minded owner to visit their dealer, rather than as a monitoring device.
 

·
Registered
EQA 300
Joined
·
74 Posts
Having enquired about the merit of driving in D- in the first post of this thread, with the benefit of information gleaned from ensuing posts, I now drive in D and use D- intermittently, when there's a need to slow down more quickly than by simply decelerating. In my motoring environment (mainly urban, up to 60 kph), I estimate that by driving in D my EQA250 consumes about 10% less energy than when leaving the setting on D-.

Since my last post, I have fitted a Cetek battery monitor to the 12 volt battery and several times after checking that battery's state of charge I've also checked its status in the Service section of the App. On two consecutive occasions, the App described a state of charge in the low 70% range as, respectively, Charged and Partially Charged. I don't know whether the App has a lower battery status category than Partially Charged, but in the absence of more accurate information, I suspect Partially Charged and Charged are somewhat relative, Charged indicating a satisfactory state of charge and one relatively higher than a previous reference point.

Maybe someone has a better understand of how the Mercedes me 12 volt battery state of charge indicator works. Given that it's in the Service section of the App, perhaps its intended to be used as a prompt for a non-mechanically-minded owner to visit their dealer, rather than as a monitoring device.
I think you are correct in your interpretation of the Mercedes me app battery charge indicator function, there is a lower status than Partially charged which does tell you to go to your dealer to get the battery charged. I’ve only seen that warning once, when the battery was down to 12volts before I fitted the monitor and never seen the Partially charged status but I don’t check the app very often. If you relied on the app to know when to charge the 12v battery it would probably not have as long a life as if you kept it at a higher SOC. Plus I have a suspicion that, when parked for a while in an area of poor internet reception, if the 12v battery gets low the app doesn’t update sometimes so could be showing you a SOC 2 or 3 days out of date.
 

·
Registered
Joined
·
81 Posts
Discussion Starter · #27 ·
Thanks for your comments, FDR.

I find the behaviour of the 12 volt system in my EQA250 quite interesting. When I fitted the CTEK battery monitor on the 30th of July, the SoC was a bit over 60%. It then hovered around the 80% mark for a few days, then on the 3rd of August it increased to about 100% before discharging to a little over 80% (on the same day). On the next day it increased to the high 90% range, falling to the high 70% range on the 5th, remaining in the 70-80% range until yesterday, when it increased to 80%. Today, I drove about 30km (city running) and it has remained in the 75-80% range. I drove about 10-20km each day between the 3rd and 7th of August.

The above pattern is reasonably typical week on week, although sometimes I don't drive the car for a few days - and I've never charged the 12 volt battery with my battery charger.
 

·
Registered
Joined
·
121 Posts
when parked for a while in an area of poor internet reception, if the 12v battery gets low the app doesn’t update sometimes so could be showing you a SOC 2 or 3 days out of date.
I think that's correct, and you don't even need to be in a poor reception area. After all the pumps and servo motors eventually calm themselves, the car puts itself into a very deep sleep and doesn't communicate with the cloud for days. Most pages on the app include a "Updated from the vehicle x hours/days ago" but unfortunately the Service page doesn't so there's really no way of knowing the age of that 12V battery state display, but in my experience it can be days old if the car has been left completely idle (even with good reception) so comparing it with your live battery monitor is problematic.

I figure the most innocuous cloud transaction you can do is to modify the high voltage battery's Maximum state of charge setting, so I changed mine from 80% to 60%. You really wouldn't expect that to use much 12V energy at all; it just has to wake up when the message comes in, modify a stored setting and then go back to sleep. But even that is enough to show up on the battery monitor:
Rectangle Slope Font Parallel Pattern


But when I leave the car idle for 2 days, I see no such blips - just the tiniest of very gradual decays, which leads me to believe there aren't regular updates being sent to the cloud. I suspect all such messages are event driven - something has to happen to make it wake up and report in. FDR's experience suggests that one such event is the 12V battery becoming critically low - that inspires it to wake up and send out a call for help.
 

·
Registered
Joined
·
81 Posts
Discussion Starter · #29 ·
Thank you, dBC. It's helpful to know the work-arounds.

I don't usually lock my car, because it's kept in a locked, alarm-monitored garage, but soon after I open it for some reason (while in the garage), the 'your car is unlocked' nanny sends me a reminder message on my phone. I guess that notification qualifies as a "wake up and send out a call for help", albeit less critical than a low 12 volt battery SoC?
 

·
Registered
Joined
·
121 Posts
Are you saying above that at the end of a trip the battery voltage was lower than at the beginning? I mostly find each drive leaves the 12V battery in better shape than it found it. On this particular day there were 4 trips. The battery started the day at about 12.6V and ended it at 12.7V. I guess it was slightly depleted during the 4th trip (the shortest) but only just. I often see those 3 or 4 disturbances about an hour after parking - I think they correspond to some whirring servo-motor noises, maybe it's changing the position of some coolant valves or something. If you book that last disturbance at around 18:00 to the last trip, then you could say that trip depleted the battery from about 12.9V down to about 12.7V.

You probably know this already, but in case not, this whole technique of using the battery voltage to estimate its charge state relies on a rested battery. You really want to see a nice flat line for an hour or two before you jump to too many conclusions. If you were take a voltage reading immediately after a trip ended you'd risk picking up that divot that invariably happens at trip end. For example, using trip 4 below, I'd say the resting voltage at the end of the trip was 12.8V, but if you happened to take a one-shot reading at an unlucky time (16:59) you might think it was 12.5V.

Colorfulness Slope Rectangle Font Terrestrial plant
 

·
Registered
Joined
·
81 Posts
Discussion Starter · #31 ·
Wow your monitor is at the opposite end of the scale from the Ctek as far as detail goes, Ctek is very accurate in real time (confirmed by my battery tester), it’s just the historical display in the app that is too vague. It does what I need in letting me know if I have to charge the battery when the car is parked but yours would be a better option if you needed to solve some intermittent problem during driving for example.
dBC, I quote (in part) FDR's earlier post in which he described the detail available from the CTEK monitor, which I also have fitted to my EQA. When one views the App, it displays the most recent reading of Voltage, State of Charge and Temperature. Mine, for instance, currently displays the following values for my EQA: 73%, 12.48 v and 16 degrees C. The historical graphs are far less precise, Voltage, for instance, hovering most of the time around the 12 v mark (the graduation is in 5 v increments - 5, 10, 15, 20, 25 v - so even that's a bit of an estimate), with occasional spikes to about 16 v (certainly, over 15 v).

Those spikes seem to correspond with discrete journeys, but the fact that the display indicates peak voltages in the order of 16 v perturbs me somewhat. I would expect peak voltage to be in the order of 14.9 v, as shown in your graph above. I therefore query whether my car's voltage regulator - or its equivalent - is allowing the 12 v battery to be charged at higher voltages than the typical 'high 14 v', or whether the CTEK monitor is somewhat inaccurate.

To answer your question, therefore, I'm not sure about voltages at the beginning and the end of a journey; the historical display from the CTEK battery monitor is too vague to show other than an (almost) flat line, with occasional spikes and dips. I tend to monitor the State of Charge value (and historical graph), much as I do with the high voltage battery.

Question (for dBC): what is the source of the images you've posted (as in your previous post); i.e. is it a screen shot from your mobile phone, or some other device?
 

·
Registered
Joined
·
121 Posts
Those simple battery monitors like yours and mine are basically voltmeters, so I tend to focus on the Voltage reading. They really have no idea what the actual SoC is. They guestimate that with a simple lookup table indexed by the measured voltage. There are two problems with that: a) there are lots of different views on what values should go in that table and b) it only ever makes sense to do that Voltage -> SoC % mapping when the battery has been at rest for an hour or two. That means no charging, and no discharging. No charging is pretty easy to arrange: just don't start the car. No discharging is harder to achieve. There'll always be a slight background discharge both from the car and the battery monitor itself, but in the EQs that's very very small and can be ignored. But if you trigger anything like central locking, interior "EQ display", interior lights etc., or the car decides to make some of its autonomous whirring servo-motor noises, then you've done some discharging and need to wait another hour or two before you attempt to convert the Voltage to a SoC %. In reality, with the battery in my EQC, waiting about 30 minutes is generally sufficient. That's long enough for it to settle back down to the resting voltage where it will stay indefinitely. Only then does it make sense to use that voltage reading to look up the table and produce a SoC % guestimate.

If you have a significant background load (perhaps like FDR's always-on dashcam) then you can't determine the SoC by looking at the voltage. In that case your CTEK (and my equivalent) will report a SoC % lower than reality. You'd need to disconnect the load.. wait an hour or two (or 30 mins) and then inspect the voltage reading (and corresponding SoC %).

So when your CTEK tells you: "73%, 12.48 v" you can trust the 12.48V as an instantaneous voltage reading (hopefully) but you should take the 73% with a grain of salt unless you know the battery hasn't been disturbed for at least 30 mins. One sure sign that it is recovering from a disturbance is that the 12.48V won't be constant.. it'll be climbing even though nobody is doing any charging. In that case, wait until it settles. Then you can have some faith in the SoC %age.

Hopefully your 16V reading is just a rounding issue with your course grained display.

Those images of mine above are graphs out of my spreadsheet. With a bit of hacking, I can get the 2-min samples out of my battery monitor into a CSV file and then graph as I see fit. So for example those disturbances around 18:00 in the image above are just a plot of this data:

Font Material property Pattern Rectangle Number


And what follows is an example of a Voltage -> SoC % table. This is just one I found from a google search, but if you look you'll find many, all with slightly different values. CTEK will have picked one and baked it into their app (as has my monitor). The resultant SoC % is a useful approximation, but only after the battery has rested. This table recommends a 30 minute resting time.

Font Rectangle Line Parallel Slope
 

·
Registered
Joined
·
81 Posts
Discussion Starter · #33 ·
Thank you, dBC; your understanding of technology is far greater than mine!

Is your data sourced via the Jaycar / Powertech QP2265 monitor? If so, can you download historic data to a PC, or do you have to do that manually?
 

·
Registered
Joined
·
121 Posts
Yes, that's the battery monitor I use. The 2-minute historic data is stored in the phone app. Each time you connect to the battery monitor via bluetooth you'll see "Synchronising Data" as it sucks the latest readings off the battery monitor and into the phone via BT. It grabs what's accumulated on the device since the last fetch and appends it to the database on the phone app, which grows indefinitely.

To get it from there onto your PC involves a bit of hackery and I only know how to do it on Android phones and Linux PCs. The basic steps are:

. use adb to backup the app onto your PC
. untar the resultant backup file to reveal all the apps files on your PC
. use sqlitebrowser to open up the database file (apps/com.dc.battery.monitor2/db/bm_db_helper)
. find the new_his_volt_table and export it as a JSON file
. extract the timestamped voltage data from the JSON file into a CSV file and then open that in your favourite spreadsheet

I wrote a little C program to do that last step and you're welcome to the source code.
 

·
Registered
Joined
·
121 Posts
If all that skulduggery is too much to bear the app also comes with a handy cursor that lets you view the detail on any data point. In this screenshot I've scrolled back to 5/8/22 and picked out the voltage at 18:03 to confirm an entry in the spreadsheet above. So you could use that to verify that your 16V really isn't 16V.
Font Slope Parallel Audio equipment Pattern
 

·
Registered
Joined
·
81 Posts
Discussion Starter · #36 ·
Given that I have an iphone (8) and a PC with Windows 10 operating system, the "skulduggery" might stretch me, dBC.

The Powertech monitor would be worthwhile, if only to (manually) check that maximum voltage reading - which appears consistently. I can't, on my CETEK monitor, mimic the historical data set shown in your image above.
 

·
Registered
EQA 300
Joined
·
74 Posts
My Ctek monitor also, at first glance, appears to show some voltage peaks of over 15volts but if I zoom in on the peak to show more detail the peaks are shown to only be 15 volts max.
I don’t pay much attention to the SOC figures, I just read the present time voltage 20 minutes or so after the car goes quiet then decide if it needs a charge. The present time voltage figures are accurate as I’ve checked them with a meter when I first got the Ctek to see if it was any use.
 

·
Registered
Joined
·
81 Posts
Discussion Starter · #38 ·
Good point about expanding the voltage display on the Cetek monitor, FDR. I agree, when the display is stretched, those apparent 'over 15 v' peaks do recede to 15 v or perhaps even slightly under.
 
21 - 38 of 38 Posts
Top