12v battery time (sigh)

[gregd]

Are you sure that the old, dead, battery still being in the circuit is not affecting things? 16A for a couple of door solenoids seems ridiculous.

Oh, no, it's worse than that. No door solenoids, just the unlock button on the key fob. Beep Beep and a couple of flashes of the headlights, even if the car is already awake. I tried turning on the headlights with the car asleep (so, sudden wake-up and lights on at the same time), and no such behavior. The tug, however, lasts only about a second, and it's specifically triggered by the unlock command from the key fob (or OVMS).

But, no, I don't think the existing battery is affecting things. After the voltage dip, it recovers very quickly, so I think it's just drying out, not shorting, and still has some small capacity left. When the car is asleep, it appears that the charging current is dropped to just a trickle (under 100ma). It's enough to keep the battery at 13.77v (which is where it's been for the past 5 years), and will very slowly charge another battery when I parallel them up. When the car is awake, the charge to the second battery jumps up to several amps, which is effective. Actually driving the car to charge the 12v battery (which has been reported in other threads) doesn't appear to be required. The other threads are correct that there's not much charging taking place when the car is asleep.

You certainly would not want a heavy object in your car that is not securely mounted.

Exactly. It would be right above the brake assist vacuum hose, and some of the hard brake lines; not something I'd want several lbs of battery to be bouncing on top of on rough roads. It appears one could build a platform on top of some of the metal structure in there, but at this point I don't think it would be worth the effort given I don't have the proper tools. But something one might consider as an alternative to the Gruber trunk solution, if you could do it right.

I might run a wire temporarily by the passenger door jam for a few days, until I get the "real" battery installed, just to avoid the alerts.

 

[Hebbe213]

Find someone off the roadster group in your area that does garage calls or knows how to service lotus elise; Luckily someone found me on the group and it was the right person to have working on my car here in Florida.

He repainted my Battery bracket from corrosion in which Service center will not do things like these.

View attachment 609446 View attachment 609447

Good info and pictures thank you

 

[gregd]

The tug, however, lasts only about a second, and it's specifically triggered by the unlock command from the key fob (or OVMS).

Oops, change that. I wasn't watching the meter carefully enough. There's a spike to about 20 amps, then it drops back to 17-18 amps for as long as the headlights are on, then drops off and starts charging coincident with the headlights turning off. So it seems like the battery is running the lights during that time, or perhaps there's simply a battery load test that runs while the lights are powered from the APS. Hard to tell what. Total is several seconds.

I've got a wire snaked from the Power Pole under the hood into the passenger side of the cabin, with the 9AH UPS battery hooked up there. Not by any means a permanent solution, but it's enough for me to drive around a bit without worrying about hassles getting into the car, or annoying beeps from the system. No point adding to the log file. More work after Thanksgiving.

 

[ML Auto]

It takes a few seconds for the APS to power up. Without the aux battery connected, activating the door opening switch does nothing until the APS turns on (it does start that process, however). The lights, alarm and door locks are all powered by the 12V battery until then.

 

[gregd]

It takes a few seconds for the APS to power up. Without the aux battery connected, activating the door opening switch does nothing until the APS turns on (it does start that process, however). The lights, alarm and door locks are all powered by the 12V battery until then.

Yeah, perhaps... Interesting that all this heartache with the 12v battery is just to save, what, 3 seconds delay between pushing the unlock button and the ability to open the door?

But the observed behavior occurs even if the car is already awake, with the APS at full steam. Seems like there's an opportunity to optimize the car firmware here.

 

[slcasner]

Seems like there's an opportunity to optimize the car firmware here.

One of terribly many! Source code, please!

 

[markwj]

Yeah, perhaps... Interesting that all this heartache with the 12v battery is just to save, what, 3 seconds delay between pushing the unlock button and the ability to open the door?

The 12v battery is there to keep emergency systems (lights, hazard warnings, etc) operating in the event the main pack is shutdown in an accident.

 

[gregd]

The 12v battery is there to keep emergency systems (lights, hazard warnings, etc) operating in the event the main pack is shutdown in an accident.

Sure, but why is it pulled into action when the unlock button is pressed, even with the APS running?

This is why I'm wondering if the current draw is simply a battery load test, so that when the car is driven (after being unlocked), they're sure the battery can perform that emergency function. The alert may be safety-by-annoyance.

 

[markwj]

Sure, but why is it pulled into action when the unlock button is pressed, even with the APS running?

This is why I'm wondering if the current draw is simply a battery load test, so that when the car is driven (after being unlocked), they're sure the battery can perform that emergency function. The alert may be safety-by-annoyance.

I think 'door locks' are part of the 12v emergency system to allow them to operate with the main pack down.

 

[gregd]

I think 'door locks' are part of the 12v emergency system to allow them to operate with the main pack down.

Oh, very likely they are. Been thinking more about this... (yeah, dangerous). Logic follows:

What is puzzling is how the software unlock (as opposed to physical the door opening) function works. If the car is awake, doors open, everything humming along just normal, and one presses the unlock button on the key fob, or issues an unlock command via OVMS (*), why does the car start pulling 17+ amps from the battery for several seconds? Is it actually disconnecting the APS from the system and running the car from just the 12v battery, or is it simply a load test to be sure the battery is healthy? There's certainly no "need" to reroute power, when under those conditions it's essentially just a software state flag in the VMS that marks the car locked or unlocked.

When the car is asleep, APS off, the 12v system appears to be given enough power from the DC-DC inverter in the ESS to run the VMS in its sleep state, the alarm and other base systems, and about 100ma extra to keep the 12v battery at its float charge level. So, an unlock at that point means the VMS is coming up to full speed, battery pump is turning on, the lights are flashing, etc., all while the APS is in the process of turning on. In that case, sure, the behavior could make sense, with the battery making up the difference. But when the car is already awake? Is it just an un-optimized process where it disconnects stuff as if the car was asleep? Doesn't seem right.

Rather, I suspect that the fundamental design of the car's 12v system didn't change that much between the 1.5 and 2.0. Specifically, the 2.x cars are not "running from the 12v battery" while asleep. An 8ah battery wouldn't last an hour, if even that long. They are still getting 12v power from the ESS, but instead of a 2-sheets tap it's now a full-pack DC-DC inverter, still with only enough capacity to run the base stuff. A Mini-APS, if you will. I think the 12v battery is just there for the emergency situation, where the ESS might be damaged or disabled, a case which was apparently not adequately covered in the 1.5 design. To me, this suggests that the unlock behavior of the 12v system is not necessary for the transfer of power from mini-APS to the main APS during the unlock (especially considering that the 1.5's unlock just fine without it), but rather just a health check on the emergency 12v supply. Also supporting this is that a 2.x car that's missing its 12v battery doesn't crash and completely reboot during an unlock. That battery cannot be critical to the base operation.


(*) With the car awake, doors open, and obviously unlocked, if you tell the car to Lock by OVMS, the icon on the Android app changes to Locked even though it's clearly not. A second unlock command moves it back to Unlocked, and triggers the 12v system dance. I didn't try messing with the door latches or such while in that first pseudo-locked state, as it was late at night and didn't want to tempt setting off the alarm...

 

[markwj]

You are correct that _most_ of the car doesn't run off that 12V, even when asleep. The APS in the main pack runs things like VMS. I suggest you check the wiring diagram. This all runs through the switchpack, but it does show the connection paths, and idea behind it.

Note that the immobilizer, for example, definitely runs from 12V battery in 2.x cars. You can see this because with a dead 12V, and asleep car, pressing the key fob won't unlock the car. It first needs to be woken up (charge port, door handle, etc), and then a few second later the key fob unlock starts to work. That is a classic symptom of a dead 12V battery. You can see that as '12V Redundant' from the switchpack.

 

[ML Auto]

There is indeed a small voltage converter that is inside the ESS that supplies a little current when the APS is off (the standby line). In an accident situation where the ESS power is cut, the 12v battery still allows you to unlock the doors, trunk and glove box (which they apparently forgot about when they made the Model 3). That circuit is not directly connected to the APS output, so most of the current would come from the 12V battery, regardless of the APS status.

 

[gregd]

There is indeed a small voltage converter that is inside the ESS that supplies a little current when the APS is off (the standby line). In an accident situation where the ESS power is cut, the 12v battery still allows you to unlock the doors, trunk and glove box (which they apparently forgot about when they made the Model 3). That circuit is not directly connected to the APS output, so most of the current would come from the 12V battery, regardless of the APS status.

Right. But the 12v battery is held at the 13.77v float charge at all times, other than the 3 seconds immediately following the Unlock key fob press. What changes at that instant?

Suddenly turning on the headlights when the car is asleep (to surprise the 12v circuit while APS is off) doesn't even twitch the 12v battery voltage.

 

[jason.tv]

Right. But the 12v battery is held at the 13.77v float charge at all times, other than the 3 seconds immediately following the Unlock key fob press. What changes at that instant?

Suddenly turning on the headlights when the car is asleep (to surprise the 12v circuit while APS is off) doesn't even twitch the 12v battery voltage.

So don't put a battery in. Just leave it

 

[ML Auto]

Right. But the 12v battery is held at the 13.77v float charge at all times, other than the 3 seconds immediately following the Unlock key fob press. What changes at that instant?

Suddenly turning on the headlights when the car is asleep (to surprise the 12v circuit while APS is off) doesn't even twitch the 12v battery voltage.

The headlights are connected tp the switchpack, along with the current to charge the 12v battery when the APS is on. My guess is that the unlock/ headlight flash draws current from the 12V battery (and the float charge from the tiny standby converter) until the APS and switchpack are up and running. After just checking my headlight on response time, there is a slight delay when the car is asleep but none once awake. So the APS takes only a fraction of a second to turn on, but the switchpack, VMS, VDS, etc. take several seconds to become fully operational.

 

[gregd]

The headlights are connected tp the switchpack, along with the current to charge the 12v battery when the APS is on. My guess is that the unlock/ headlight flash draws current from the 12V battery (and the float charge from the tiny standby converter) until the APS and switchpack are up and running. After just checking my headlight on response time, there is a slight delay when the car is asleep but none once awake. So the APS takes only a fraction of a second to turn on, but the switchpack, VMS, VDS, etc. take several seconds to become fully operational.

That makes some sense, though for the behavior I am seeing to be explained, the car must be resetting the switchpack (or something similar) when the Unlock button is pressed. One can imagine the code, using a big hammer approach instead of handling the car-asleep and car-awake cases separately.

Anyway, it was a busy afternoon. And an annoying one. I decided that turning the wheels to the full-right position should give me enough access, and to not mess with taking off the wheel. Scotty's instructions on the process are probably the best: Replacing the 12V aux Battery except that he shows a nice flexible "felt liner" that needs to be pulled back to get the access panel off. Mine is a rather stiff plastic. Managed to get the panel off without breaking anything. Once I had access... lookie what I found:



Little problem there on the positive...! I considered just cleaning it up, hoping a possible bad connection was at fault, but in spite of the corrosion, it seemed mechanically and electrically tight. It's more likely that the leak is what has stunted the battery's capacity, and a clean-up isn't going to have any effect. The white wires, btw, are the ones that go up to my Power Pole connector up top.

Unfortunately, the local car parts store doesn't carry the FAYTX-9 battery. Hunting through their catalog, we came up with an ETX-9 battery, which has the same exact dimensions and 8 AH capacity. But it's CCA rating is only 120 amps, vs the OEM battery's 135 amps. Price is roughly the same. There is a much cheaper version, but you have to "activate" it by filling it up with acid, and they say it should be checked on occasion. No can do. The other alternative is to drive down the hill into an area with a 2x higher Covid rate than here, so I opted for the ETX9. We'll see how long it lasts. The 20+ amp "starting" current surely isn't going to be a short term stress; it's the longer term that I'm unsure about. I also note that the Gruber battery is intended for UPS applications, and they generally don't even have a CCA rating.

I also opted to attach the wires to the side instead of the top, for easier access next time.



Done! I hope... The battery held its own with an unlock test (no alerts posted), but seems kind of slow to recover. There was still a big dip during the 3 seconds of whatever-it-is, down to 11.89 volts, ramping back up to 13.75v with the 5 minute "awake" time. But when that charging ended, it dropped way down with only the float current available. It's sitting at 13.26v right now, very slowly ramping up; it was 13.20v about an hour ago. I tried hooking a 1A "wall wart" that I use as a battery charger to it, and it zoomed up past 14.5v so I stopped it. Will let it sit until morning, and see if it gets up to the full 13.77v float by then.


P.s. What is that extra connector above the battery for?

 

[jason.tv]

Did you use any Corrosion gel; def needed as the battery is in open space; I would of cut the tips off and sliced new ones.

NOCO Brush-On NCP-2 Battery Corrosion Preventative, 4 oz.

 

[jason.tv]

I believe the extra connector is for the fog lights which are not required in USA. Could be for the HID headlights if being upgraded as well but from the smaller pins looks likes it's for the Fog lights.

You might want to redo and clip the corrosion tips and start fresh; the corroded bundled up wire when it's like that means it loses conductive properties. If you couldn't break apart and or loosen the wires free then you need to snip them off and hook up new connector or tip.

Snip; new tip ; squeeze with Plyer tool, it's that easy.

 

[gregd]

Thanks, both. Unfortunately, no, I didn't put any goop on either of the wires, but did wipe them with an alcohol wipe and all the stuff came right off. So, it didn't seem to have been there very long; no real corrosion. My reference to not being able to wiggle stuff was that the connection to the battery was tight, and not a loose connection that would have been the cause of the voltage drop under load.

That said, looking back at the pictures I see there's a little bit of the white stuff up in the connector cover that I missed. Dang. But at this point, everything is put back together, and too annoying to take apart again. The wires don't have enough slack to them to simply be trimmed and re-connectored without adding a short pigtail. If I have trouble down the road, I'll know what to do.

Jason, I already have the HID upgrade, so the extra connector's use for fog lights is probably correct. Is there a kit for adding them? That could be cool. My 2.0 has the 2.5 front and rear fascia, an upgrade done by the original owner. Not sure where the lamps would go.

 

[jason.tv]

Thanks, both. Unfortunately, no, I didn't put any goop on either of the wires, but did wipe them with an alcohol wipe and all the stuff came right off. So, it didn't seem to have been there very long; no real corrosion. My reference to not being able to wiggle stuff was that the connection to the battery was tight, and not a loose connection that would have been the cause of the voltage drop under load.

That said, looking back at the pictures I see there's a little bit of the white stuff up in the connector cover that I missed. Dang. But at this point, everything is put back together, and too annoying to take apart again. The wires don't have enough slack to them to simply be trimmed and re-connectored without adding a short pigtail. If I have trouble down the road, I'll know what to do.

Jason, I already have the HID upgrade, so the extra connector's use for fog lights is probably correct. Is there a kit for adding them? That could be cool. My 2.0 has the 2.5 front and rear fascia, an upgrade done by the original owner. Not sure where the lamps would go.

Yep the goop is the secret stuff lol; You can clip and extend with a new wire; which we did.

What is a Front a rear end fascia?