DIY Electric Car Forums banner
441 - 460 of 465 Posts

·
Super Moderator
Joined
·
1,376 Posts
I find it odd that somehow the positive plate wouldn't be isolated from the battery case. That seems like a bigger problem than a "just put an insulating plate on it". Like... why is it not isolated? Who sold you these batteries? EV West again? :p But I don't think that's it.

Back to what we diagnosed earlier in the build, you do not have a short. You almost certainly don't even have a leak, because if it was a leak it would be the tiniest, tiniest drip of a leak, too small to even measure. So, what we know is that it's charging up some amount of capacitance. You know this because a leak would have a consistent voltage. Capacitance would drop merely by the act of measuring it, which is what you observed. Capacitance would charge up, and then the current drain would stop when the voltages equalized (minus capacitive leakage). An actual leak would continuously discharge the battery.

So, that said, what is a capacitor? A capacitor is any two conductors separated by an insulator. Like, say, a conductive battery case, a conductive cooling plate, and a thin film of thermal paste. Or two wires running side-by-side. Not a dangerous amount, yeah it might be 300+v, but, rubbing your feet on the carpet charges you up to 10,000v, there's just only a minuscule amount of energy stored, and it can't sustain that pulse so when you touch a doorknob, you don't die, the discharge is only microseconds long. In your case, not enough of a leak or energy stored to even measure it without it vanishing in a couple seconds.

I think the capacitance is probably not your problem. There is capacitance, maybe more than there should be, but it's a wild goose chase to try to track every opportunity of it down. That's just going to exist in your car by virtue of dozens of parts of it mechanically having some non-zero capacitance. Again, any two conductors separated by an insulator, the voltage of which will also ebb and flow as the conductors are conducting (i.e. pre-charging, running systems, driving, etc).

But somehow this capacitance is also meanwhile getting into the control electronics. Which is a separate problem, and the real problem. That probably means your data cables aren't shielded well enough. Or, it's inside some of the electronics you bought.

...

In other news, I'm more starstruck by you getting to hang out with John Kelly than I was with you hanging out with Zach. I didn't know that Weber was also in the LDS fraternity of educational Youtubers. I have a theory on 2 cultural anchors that have led to so many good educational channels founded by LDS members. One is that the mandatory 2-year proselytizing is a brain-grinder for both effective presentation and communication of your ideas with strangers in a way that they'd understand (essentially: hosting, script-writing, content selection, succinctness, etc), and a humbling experience of near constant rejection that makes you impervious to fear of rejection or shame of attempting new outgoing things. The other is the sense of agency over your own behavior and future.
 

·
Registered
Joined
·
6,526 Posts
I find it odd that somehow the positive plate wouldn't be isolated from the battery case.
There is no single "positive plate" in a module,; although of course the most-positive cell has positive side, it is not isolated any differently from the other cells. Each of the heat transfer fins which runs between a pair of cells and turns to contact the cooling plate is exposed to a different potential - each of them is positive compared to one neighboring fin and negative compared to the neighbor on the other side, if they are not completely isolated from the cells that they are cooling. If the fins where conductively connected to the cell terminals, any conductive plate clamped to them would be a short across every second cell of the module, and across the entire module. If the plates were conductively mounted to the frame, the entire pack voltage would be shorted through the frame.

The outside of each pouch is intended to be isolated from both terminals, but that isolation is not expected to be complete - there is certainly capacitive coupling, but there is also some conductivity.

In this design, which is widely used in LG Chem and other modules, there is no case between the cooling plates and the whatever surrounds the module. The modules are not intended to be exposed, but are supposed to be (as they are in snowdog's car) enclosed in a controlled environment, in contact with a heat-transfer plate using suitable materials (and those materials are an aspect which is being worked on). The "case" of the module is really a stack of retention frames, rather than a housing.
 

·
Super Moderator
Joined
·
1,376 Posts
High voltage problems are solved (I think)
Mmmm, I hope so.

But your symptoms didn't line up with a voltage leak. You know this, because it was so slow of a trickle that you couldn't even measure it without it dropping. If your cooling plates were leaking voltage you would've been able to measure it as a steady voltage. So, isolating the batteries better is good regardless, but, if you were getting high voltage into your control electronics, that's not solved.

Using the Megger... I dunno that I would've done that. You're not just "testing" them to 1000v. I'm pretty sure the test is that you're applying 1000v to those wires. You shouldn't be doing that to anything that doesn't have 1000v+ volt rating, and being really sure anything rated less is nowhere near enough that it might be getting tested accidentally. Suppose you had insulation rated for 80v, or 400v... by feeding it 1000v to "test", you've certainly punctured through that insulation. Many micro-punctures burned through, probably not visible to the eye. Think of it like 'testing" bulletproof armor plates. Suppose it's supposed to be impervious to .22 pistol rounds, and you start shooting 7.62 rifle rounds at it and say "Aha, I can see where the holes are". What you've actually done is created new leaks, gone and punctured and bunch of holes in otherwise perfectly good small-caliber plates. Some in places you expected, and some in places you probably didn't expect. You might not even notice yet that you made all those punctures.

Especially your low voltage wiring, those wires are not going to have much for insulation. Especially if you've already been having control-electronics problems and isolation problems in and around the delicate low-voltage electronics making their way to high voltage, waving 1000v around near them and stress testing them to that is certainly going to make your problems worse. Let alone if it gets into the chips (which behave erratically after HV/static damage, which is why you use a grounding strap when installing RAM for example, it's not instant failure, it's more likely reduced life expectancy and erratic behavior) the wires will now have micro-punctures everywhere in them. So, depends what you were doing with it and where, and what might have already had weak wires.

BMS fault... who sold you the BMS? EV West again or is this already a replacement to that one?
 

·
Registered
Joined
·
242 Posts
Discussion Starter · #446 ·
Just following the steps that John (WeberAuto) and I did at his shop. I texted and emailed him throughout the troubleshooting process. I figure he knows best. He mentioned that this is commonly done in hybrid and electric cars for troubleshooting these types of problems. This is 1000V but extremely low amperage. Like if you walked through the carpet in your socks and got in your car, you are not likely to damage your electronics. This is also done with the 12V system disconnected so it is just testing the high voltage to the chassis. And all my high voltage wires are rated to over 1000V. If I were blasting new holes through insulation the meter and data would definitely confirm that.

What I learned from John, is it is not so much a "leak" as a loss of isolation. This loss of isolation was due to the battery module cooling elements. Hard to argue with John or data
 

·
Super Moderator
Joined
·
1,376 Posts
This is 1000V but extremely low amperage. Like if you walked through the carpet in your socks and got in your car, you are not likely to damage your electronics.
Hmm, that's not really equivalent.

The amperage doesn't matter, in terms of whether you punctured the insulation or not. Think of the insulation like a rubber plate filled with water, and then you punch "only" a pinhole through it. You now have a hole in it. It's not a fist-sized hole, but it's still a hole that wasn't there before. The purpose of the insulation is to prevent a certain voltage from being able to jump that gap. Any amount of hole in it is like a break in a dam. Or, think of "only" poking a balloon with a pin, not a bullet.

The reason you're not likely to damage your car from static is because you're not touching any of the parts that are sensitive to static. Their enclosures are either isolated, or grounded, to avoid damage. Otherwise yeah, you'd fry just about anything you touched. That's why you wear a grounding strap when handling unshielded electronics.

In your case, you were not certain how or where voltage was getting into places it shouldn't be, so you were not aware of whether anything sensitive was properly isolated or grounded. In fact, your symptoms show that this was particularly an issue, you were getting voltage in places you were not supposed to. Shoving 1000v at that (versus 400v or however much your pack is), to me, would be the worst thing you could do to those electronics.

That's kind of like using a hammer to find out what types of glasses are breakable.

This is also done with the 12V system disconnected so it is just testing the high voltage to the chassis.
Yes, and that would be the key to only testing the thing you wanted to test. However, you were already getting erratic behavior in your electronics. The unknown unknowns. So you already had some kind of path between your HV and your electronics that you didn't know where it was connected.

And all my high voltage wires are rated to over 1000V.
Yes, that's fine. That's like a bridge being rated for 100 tons and then driving 100 tons across it. It better survive, and if it doesn't, it'll probably be obvious where, and you need to rebuild the part that failed. That's how the megger is supposed to be used. It tests the insulating ability of the thing you're testing, below or at its maximum rating. So you did that, and you found a weakpoint in the insulation, then you fixed it, that's good so far.

What I'm concerned about is that you didn't know how voltage was getting into the delicate parts of your electronics, and I don't think you removed them from the vehicle. You only disconnected them from the places they were supposed to be connected. You then tested it for leaks by applying 1000v shocks to everything. And then listened for where you could hear it puncturing through and conducting where it shouldn't be. How many other places was this happening that didn't make a noise?

If I were blasting new holes through insulation the meter and data would definitely confirm that.
Yes and no. Maybe. You know that voltage was getting to places it shouldn't have been before already, and interfering with your electronics. That's why you're troubleshooting. So, if it was already getting through there, then you give it 1000v, that's definitely enough to punch through that same area, harder. Maybe it arced and then the arcing created oxidized buildup or fried something, which then insulated it. This can happen almost instantly and imperceptibly. That's fairly normal, which leads to inconsistent behavior in the future.

Considering the nature of your glitching, I don't think I would've felt safe using a megger without removing all the sensitive electronics first (which, I don't even know is possible in your case without gutting it entirely).

What I learned from John, is it is not so much a "leak" as a loss of isolation. This loss of isolation was due to the battery module cooling elements.
Hmm. How are you defining a "loss of isolation"?

If something isn't isolated, then voltage (of whatever relevant level) will cause current to flow.

I agree that you did a thing, and it discovered a thing, and then you fixed a thing. That's all good. That was worth doing for its own sake.

But, you know you didn't have current flowing. And you knew you were still getting static buildup, and that seemed to be the best guess for the symptoms you were having. However voltage was interfering with your electronics, that's still an issue, no? Like, they're still not isolated or grounded properly, in addition to what you've discovered about the battery trays.

I hope that what you did fixed things, seems to be only half the solution though.

shrugs
 

·
Super Moderator
Joined
·
1,376 Posts
"Looks like we've got a couple different styles of ground clamp"

Nope. This was probably obvious to you later, but..

The first one, the red one, is called a "stinger", it's the stick welding electrode holder. It is not an alternative to the ground clamp, the ground clamp is still used to complete the circuit. The stinger is what you'd use in place of the torch. You bought a multi-function machine and stick welding is the easiest. Any welding machine that isn't stick, is a multi-function machine that includes stick. You just need a voltage, the electrode does all the rest. You'll probably never use it when you've got MIG and TIG on the same machine. Almost zero use case for it except maybe if you're outdoors and it's too windy for shield gas.

123015


I was a bit skeptical of the trunk looking decent if you cut the triangle off, but, it looks like it was designed that way. Lines work perfectly fine that way.

If you look at how much the nose dove into the bodywork, (10-11"?), that is how much your scissoring mechanism would have had to either lift, or pull it rearward, which is way beyond what any trunk mechanism I've seen could accomplish. This is something you can easily eyeball (on any hinge, not just the trunk), by guessing your pivot line and then opening the mechanism and seeing how much lift and pull you get compared to what the angle of the trunk tip would have to do. Your Mazdas were able to fudge an inch or so at most forward, and maybe 4 or 5 inches of upward scissoring. Any solution would have had to put the hinge pivot line at least in the same ballpark of the forward-most part of the trunk lid. Then, to address clearance around the trunk opening, you may or may not have to also have some hook to the hinge (though trunks usually have one or the other, a hook or a scissor).

The farthest back you chose to fit your pivot point (else you get too pidgeontoed like your previous attempt) was like, a foot away from the rear of the trunk. That means that if you opened the trunk completely vertical, the frontward nose of the trunk lid would protrude downward that same foot. (You didn't quite open it 90 degrees, more like, 60, but, it's almost the same measurement at that point in the arc). It's just geometry, you can't escape it. I don't think there's a scissor out there that has that significant of a lift.

123016


Just by the limitations of your trunk bay, nothing practical was going to work. The scissor mechanism accomplished like, 1/3 of what it needed to based on the place you chose for your pivot point. If you'd have been able to put it within 4" of the back of the trunk, the 4" scissor would have worked, but the triangle you removed was itself at least that big.

So, I agree, if you couldn't fit your pivor point significantly farther forward, changing the shape of the trunk was the only reasonable option. And it looks great anyway.
 

·
Registered
Joined
·
242 Posts
Discussion Starter · #453 ·
I got the rear clamshell fastened with some supports for the access panels. Also fixed the BMS. Driving again to test things out.

 
441 - 460 of 465 Posts
Top