[eq1]

Rebound in general is normal...

Looking at this table and the others, I'm asking myself, 'What the hell's going on with taps that you're NOT shorting?' That seems really weird to me. I wouldn't expect voltage to have dropped at all over just a few days. But here you've got -110mV, -110mV, -70mV, -80mV, -100mV on 2,5,6,8,9 respectively...

In my experience you'd have to have a drain on the pack in order to end up with voltage drops like that...

Later... Re-read most of this and the other, short related thread, trying to get a handle on things. One thing I caught: there was driving/IMA usage after the stick-level load tests. After those you put the pack back in the car and said it worked fine. Looks like you got at least a week of usage:

"Reinstalled the battery pack and drove the car for about a week without any codes or issues. About a week ago, a P1449 was set."

So, when I re-jiggered your loaded stick voltage measurements into taps, and compared to your most recent measurements, the difference reflects IMA usage and potential drain/self-discharge. So at time 2, the lower voltages could simply reflect whatever usage the pack saw during that short interval of IMA usage, not self-discharge and/or a potential HV short/drain alone...

* * *

What is the current state of your pack? It's in the car, fully installed, but you've got the BCM unplugged and the pack switch OFF. Is that right?

If this is the case, then I think you need to do a few HV measurements again to try to figure out if you've still got some short or what. Basically, I'm back at post #9.

...I wrote a bunch of stuff, some ideas, but I decided they don't make sense - so I deleted it. I'm not sure where to go from here.

 

[blackcloud77]

After those you put the pack back in the car and said it worked fine. Looks like you got at least a week of usage:

"Reinstalled the battery pack and drove the car for about a week without any codes or issues. About a week ago, a P1449 was set."

That‘s correct. That was a while ago. Since then, the IMA has been self-disabled but I had not killed the breaker or disconnected the BCM until last week when we discussed it. If there is some kind of drain, I would think knowing the breaker HAS been off and the BCM disconnected during the last two measurements should narrow it down.

I would not compare my stick voltages with the recent tap measurements because you are correct, those would reflect any discharge from the point I removed the pack and about a week of IMA use. There has been zero IMA use after the P1449 was set.

But we are seeing some voltage drop in 48 hour time spans. I don’t know what to look for in self discharge or much else, that’s why I’m posting the tables.

What is the current state of your pack? It's in the car, fully installed, but you've got the BCM unplugged and the pack switch OFF. Is that right?

Correct.

If this is the case, then I think you need to do a few HV measurements again to try to figure out if you've still got some short or what. Basically, I'm back at post #9.

I remember getting some guidance on that and posting the results, I will have to back up in the thread. There either wasn’t a response to it or there was and I didnt understand it enough to move on any recommendation. I do recall finding some things I thought were strange, but that doesn’t mean they weren’t normal.

We haven’t seen a change in the plan that I am aware of so despite these anomalies in measurements I think I am going to install the new harness today and start a full grid charge, unless advised otherwise. Good thing I have this week off.

 

[blackcloud77]

Installed the harness and started grid charging at about 8:30 this morning, starting at 148v. Camera only captured the “8” due to the shutter speed Being faster than the LCD’s scan rate.

 

[eq1]

...But we are seeing some voltage drop in 48 hour time spans. I don’t know what to look for in self discharge or much else, that’s why I’m posting the tables.

Right, those drops don't look normal to me. I can't see how you'd get drops that big and even, unless there were a drain on the pack. It'd be a very tiny drain, smaller than the ~380ohm tap shorts for instance, maybe 1/100 that load?...

Here's a measurement you can do while grid charging, to see if there's a short to ground: positive probe of DMM to red wire terminal on MCM rear connector - that two-wire connector on the right hand computer, as seen from back of hatch area. That connects to the positive end of the pack. Negative probe to ground, like over on the left, on top of the DCDC, where that orange-clad wire crosses from IMA pack to 'stuff on left'. Is there voltage? There shouldn't be. Make sure your grid charger isn't resting on any of the metal chassis parts, isolate it on something, like a blanket or pillow...

If there is voltage, disconnect that orange clad wire from the top of the DCDC and measure again.

Here's an image I posted earlier, it shows that 'orange-clad wire' (orange tape) I'm talking about:

 

[blackcloud77]

OK, so I put the grid charger on a piece of styrofoam and measured from the red terminal on the MCM to the ground going to the DC/DC. Got about 162v steady. Checked the same to the aluminum IMA battery enclosure, and I assume the chassis as well...same 162v. Then...I got a weird idea, tagging on the reason why one would isolate the grid charger. The grid charger goes to ground/earth via the household AC connection...but I didn't flesh that idea out completely, it was faster just to test. So I disconnected the grid charger from the household AC and got what I now vividly recall seeing before when we were primarily chasing the P1444. I got about 60v between the first suggested test locations and it immediately and gradually fell to zero over a maybe 10 seconds or so.

Skeeters were out in force so I didn't take a video with the grid charger disconnected but I'm pretty sure I have one from when I was testing months ago. I don't know if seeing voltage and then ramping down is normal. Should I be more alarmed by the 164v steady between the pack and the chassis, albeit only with the grid charger connected to AC?

Am I correct in believing the grid charger being physically connected to residential ground via its AC input would nullify a test between pack positive and the chassis? Has the presence of voltage (and its consequent fall to zero) without the grid charger connected to AC told us anything?

Edit: Re-reading your response, I forgot to remove that orange ground wire and test again. Skeeters are bad out there, so I'll do that in the morning.

 

[blackcloud77]

OK so the voltage ramp-down seems to have something to do with the grid charger, voltage is measurable and ramps down after the grid charger has been operating.

So I went out there this morning, disconnected the ground (since I didn't do that last night) and this is what I got: Seems since the grid charger was operating, (disconnected from AC when measuring), I can measure voltage to chassis ground and it ramps down gradually.

Just to be sure, I connected the ground back up and got this: No voltage to chassis, after having bled it off with the first test.

To show what I'm seeing between the grid charger connected and disconnected to AC, here's this: After the voltage was bled off, measured to ground again, then measured after turning the grid charger on. Testing after the grid charger is turned off, voltage is measurable and it ramps down again. Don't know if this would happen in normal operating conditions, but at least in testing the grid charger seems to produce this voltage presence and it bleeds off when grounded to chassis.

Bottom line...these observations suggest the effects of the grid charger notwithstanding, no current is measurable between the HV+ and chassis ground. Do you agree?

 

[eq1]

Note that we're getting into things I don't understand very well, so whatever I'm saying here can't be more than...above average guesses...

The first issue is the grid charger: When I asked you to measure between pack positive (red MCM wire) and chassis ground, ideally we would have been doing that without the charger connected. But I didn't think it would matter too much. It might matter more than I thought, but I'm not sure why or how. My thinking when I said to 'isolate it' was that it was possible the charger itself wouldn't have proper isolation between its AC input and DC output, and often these units have the metal enclosure grounded (like with the 'green' or bare AC grounding wire, which is also bonded to the white neutral back at the service entrance panel). So, if that were the case, and you had the unit resting on chassis ground, there could be leakage that way...

Now, even with the case isolated from chassis, I'm thinking it still could pose issues with this measurement - if the AC and DC in the charger aren't properly isolated. I don't really understand this well. In the past, when talking about finding proper power supplies for grid chargers, 'proper isolation' was often an issue. In general I think the AC needs to be 'properly isolated' from the DC output, like they can't share a ground or something like that. Without proper isolation I guess you can end up having the pack DC potential pushing current through the AC ground, or something.

That's not good, we don't want that, and I think it can pose safety issues. But I don't know enough about it to say one way or the other, and it's not the issue we're trying to get a handle on... Bottom line here is, I think, that if you're really measuring the full DC voltage between the pack and chassis ground when grid charging, you should be extra careful around the car, touching things.

So I went out there this morning, disconnected the ground ['snubber' center tap] and this is what I got: Seems since the grid charger was operating, (disconnected from AC when measuring), I can measure voltage to chassis ground and it ramps down gradually...

So, your grid charger is NOT hooked-up in this video, correct? If that's the case, then what you're measuring here seems like an issue. It looks like you're getting a 63V value at first stab on that center tap. You're actually doing this backwards, though: I wanted you to measure to chassis first, with that center tap wire connected, and then remove the wire and measure to chassis again.

I actually made some measurements myself on these things - I did see some voltage, but it was in the 0-350mV range and fluctuating. The only time I saw zero voltage was when I disconnected the orange-clad snubber center tap.

Just to be sure, I connected the ground back up and got this: No voltage to chassis, after having bled it off with the first test.

OK, that's closer to what we should expect, as far as I can tell... I measured some voltage, but never anything high...

To show what I'm seeing grid charger connected vs. disconnected from AC, here's this: After the voltage was bled off, measured to ground again, then measured after turning the grid charger on. Testing after the grid charger is turned off, voltage is measurable and it ramps down again. Don't know if this would happen in normal operating conditions, but at least in testing the grid charger seems to produce this voltage presence and it bleeds off when grounded to chassis.

hmm... If you go back up to image of electronics board, you can see that 'snubber' unit has direct connections to grid charger... Grid charger positive goes directly to one side of snubber, grid charger negative goes directly to the other side. The problem is I don't know or understand the configuration of the snubber. Are those polarized or unpolarized capacitors? What's the center tap connected to? Why the two rows? *

If indeed you're measuring zero volts between pack positive and chassis, with charger disconnected - but then full (or is it ~1/2?) pack/grid charger output voltage between pack positive and chassis with charger connected -- Why? It seems like the charger can charge up those caps, but the pack can't or isn't...

Bottom line...these observations suggest the effects of the grid charger notwithstanding, no current is measurable between the HV+ and chassis ground. Do you agree?

That's what it looks like in your second video. I wish I understood why the grid charger can charge those caps, but the pack can't. I'm not convinced there isn't an issue here, like with the 'snubber', but I don't see what else you can do but button stuff up and wait to see if you get another P1444... Maybe there's some fluke happening, where you got the code because the grid charger was in play at some point - and that's what caused it. On the other hand, anything having to do with the grid charger doesn't explain why you got those large voltage drops on your taps after sitting only a few days...

* * *

* I'm taking a bit of time to look at this. Looking at a circuit schematic, the cap symbols show polarized caps, the two rows of caps are in series, the positive side ('left-hand' wire) goes to pack negative, negative side ('right-hand' wire) goes to pack positive, center tap splits the two rows, where it connects to positive of left row and negative of right row. [note: I'm looking at Mudder's corrected merged IMA wiring diagram.]

So what does it all mean? [edit: I got most of this backwards, have crossed out backwards stuff and replaced, and generally deleted most of the rest as it doesn't make sense.]

Well, if you wanted to charge the caps, you'd have to push current through the 'left side', 'right side' (from neg side of caps to pos side), potential at left right side would have to be higher than potential at right left side. Since the center tap is a ground - zero volts - any voltage at far left right would charge the first, left right row of caps - but the ground would have to be shared with the source, and that's not the case with the pack, at least not normally...

 

[blackcloud77]

Without being too knowledgeable, it seems normal. I don't know what happens in normal driving, if those caps would charge and you'd see a similar bleeding off if you did the same test after driving. I think the important thing is that we are not seeing high voltage to the chassis when the grid charger is disconnected (normal use). The P1444 hasn't come back, so I think we can dismiss that for now, unless of course it comes back or someone else has something to say about it.

Grid charged until about noon today, when I had to steal the cord to use some power tools. That was about 28hrs of charging. Display showed 180v when I pulled the plug. Just hooked it back up about 5:00pm, it showed 160v when I plugged it back in. I'm gonna let it run overnight I think. I checked the taps real quick, too.

Tap	11-18-2020	@48hrs Disch.	Diff	@48hrs Disch.	Diff	[email protected]
1	15.42	14.57	-0.85	14.73	+0.16	17.38
2	14.77	14.79	-0.02	14.68	-0.11	17.19
3	15.34	13.08	-2.26	14.18	+1.10	17.38
4	15.38	15.38	0.00	13.23	-2.15	17.32
5	15.12	14.79	-0.33	14.68	-0.11	17.14
6	14.79	14.70	-0.09	14.63	-0.07	17.14
7	15.28	13.08	-2.20	13.88	+0.80	17.39
8	14.98	14.60	-0.38	14.52	-0.08	17.22
9	14.58	14.51	-0.07	14.41	-0.10	17.22
10	15.27	12.87	-2.40	13.68	-0.81	17.39

 

[eq1]

^ Is that last column the final loaded voltage, or resting? You say the charger meter read 180V when you 'pulled the plug' - and then 160V when you plugged it back in - what? - some hours later? That's not right. Even with relatively fast self discharge cells you would not see a voltage drop like that, that's huge...

If your tap voltages were the far column values, resting, end of charge about 17.2V to 17.4V - and they dropped to about 16V in hours? Not right. Maybe you'd see a drop to 16.8V within a few hours, but not to 16V...

I wouldn't grid charge over night. If those 180V and 160V measurements are more or less correct (i.e. there's probably a little error), you have to have a problem. Even if you don't have a high voltage leak/drain, cells that self-discharge that fast would be worthless. But I don't think you can have self discharge like that...

I think you need to verify the voltage values with your meter, get that grid charger out of the picture. Use it to charge, but take measurements with the meter.

I'd charge for an hour, take tap voltage measurements loaded, remove grid charger completely (disconnect from harness), measure resting voltages a few minutes later, then measure tap voltages again in some number of hours, perhaps the same number as before where you saw the drop from 180V to 160V. That will be the baseline. What you do after that depends on what you see at this baseline...

 

[blackcloud77]

^ Is that last column the final loaded voltage, or resting?

I'm not sure what "loaded" voltage could be, would that be with the grid charger connected? I never checked taps while charging, I didn't think that would be a useful figure at all. I was just probing the taps at the BCM connector, no charger connected.

I can't explain the strange readings, I think we just move on for now. It's documented in case it becomes relevant later.

I wouldn't grid charge over night. If those 180V and 160V measurements are more or less correct (i.e. there's probably a little error), you have to have a problem. Even if you don't have a high voltage leak/drain, cells that self-discharge that fast would be worthless. But I don't think you can have self discharge like that...

After reading your post, I disconnected the grid charger at that point and did not grid charge overnight.

The car has sat since then, I did not work Thanksgiving week so I didn't drive it at all. Back to work this morning, I disconnected the grid charger and reconnected the BCM. IMA light is off, I did not reset it...went off by itself. I scanned the codes, the P1449 was still set and another one related to the BCM...I assume that's due to it being disconnected. I cleared the codes and they have not come back. I got a few bars of charge for a bit, then I think a positive recal (SOC bars went to full?). Assist, regen and auto stop all functioning normally. Regular commute this morning, everything seems to be functioning back to normal.

I think it's prudent to ascertain whether or not there's a significant problem, or I have a faulty DVM, or I'm less competent than I thought at making simple voltage measurements. What would be a reasonable strategy? Check the tap voltages after some amount of normal usage? Maybe hit it with another grid charge so we start from a known point? Or return to the last suggested strategy:

I'd charge for an hour, take tap voltage measurements loaded, remove grid charger completely (disconnect from harness), measure resting voltages a few minutes later, then measure tap voltages again in some number of hours, perhaps the same number as before where you saw the drop from 180V to 160V. That will be the baseline. What you do after that depends on what you see at this baseline...

I also have my HA automatic discharger back, so pack level discharge cycles become an option again.

 

[eq1]

I'm not sure what "loaded" voltage could be, would that be with the grid charger connected?

Loaded voltages=grid charger still connected, charging.

Back to work this morning, I disconnected the grid charger and reconnected the BCM. IMA light is off, I did not reset it...went off by itself. I scanned the codes, the P1449 was still set and another one related to the BCM...I assume that's due to it being disconnected. I cleared the codes and they have not come back. I got a few bars of charge for a bit, then I think a positive recal (SOC bars went to full?). Assist, regen and auto stop all functioning normally.

Your timeline seems a little muddled to me. For example, here you say you disconnected the grid charger, but you had said you disconnected the grid charger after your last charge some days ago, so either you disconnected it some days ago or today, it can't be both...

Not sure how the IMA light can be off if you then scan for codes and read a P1449 code and another. In general, yes, you'd have something like a P1568 having the BCM disconnected. Few bars of charge and then bars to full=pos recal...

I think it's prudent to ascertain whether or not there's a significant problem, or I have a faulty DVM, or I'm less competent than I thought at making simple voltage measurements. What would be a reasonable strategy? Check the tap voltages after some amount of normal usage? Maybe hit it with another grid charge so we start from a known point? Or return to the last suggested strategy. I also have my HA automatic discharger back, so pack level discharge cycles become an option again.

If there's a 'significant' problem you should get a trouble code at some point. I think it'd be best to just drive until you run into problems again, in part because I've run out of gas trying to troubleshoot. It becomes too difficult trying to keep everything straight, methodical, etc... You could return to the last strategy I outlined. If that 180V to 160V drop over some hours is true, I can't see how you don't have an issue, either fast self discharge or a short and drain somewhere, or perhaps a faulty voltage meter on the charger.

 

[blackcloud77]

Your timeline seems a little muddled to me. For example, here you say you disconnected the grid charger, but you had said you disconnected the grid charger after your last charge some days ago, so either you disconnected it some days ago or today, it can't be both...

I get that it can be confusing, there's certainly a lot going on here. I can't remember exactly but without reviewing too much, I probably meant I unplugged the charger from power, to imply the stopping of charging (when you said don't charge it overnight), this morning I mentioned I disconnected it just because the harness was still plugged in...knowing that keeping it connected would keep the IMA light on.

Not sure how the IMA light can be off if you then scan for codes and read a P1449 code and another. In general, yes, you'd have something like a P1568 having the BCM disconnected.

I can't explain most of what I'm seeing, that's why I need help from the forum! To be sure, when I started the car this morning (got an IMA start too, which was good), the IMA light was OFF but the CEL was still on, presumably because the P1449 was still stored, as was the (you're correct) the P1548. Maybe the P1449 was stored, and just needed cleared? I don't recall under what conditions that code clears itself, might have been a few driving cycles? Not sure. When stopped at a light, I cleared the codes to see if either would come back. Neither have. Yet, fingers crossed.

I agree, all the details are running together so at this point, let's just drive it and see what it does. I think if there's a serious problem and I'm not completely inept with a DVM, then something is likely to resurface soon. We'll see how it goes. Maybe I'll put the discharger on it the next time I have a week off.