Leaf charging vs dual string pack

[CasaDelGato]

So, my truck that I plan on using the Zombieverter/Leaf motor combo in, has an unusual battery pack setup.
It has 2, 350v (identical) packs in parallel, each with it's own Orion BMS.
To make sure they play together safely, they are diode isolated from each other.
I have a pair of whopping huge diodes used when driving, or pulling power from the pack.
There is a separate, smaller, pair of diodes used when charging the pack (as the current flow is in the opposite direction).

This was working quite well in my old DC setup, as the charger was a completely separate unit from the controller.
With the leaf however... there is only 1 set of wires going from the pack to the electronics. Charger doesn't have separate pack wires.
I can't just have a relay switch which diode set is being used - as I'd bet that the Leaf charger won't start if there isn't pack voltage. If the diodes are not switched - there will be pack voltage, but no current can flow from the charger, which is likely to cause a huge voltage spike and damage things.

This is looking like it's going to be a PITA - OR I switch back to an external charger.
and I just realized that this will also block regen. AARRGGHHH

I may have to UNisolate them when driving? End up with just have separate contactors for each string?

Other suggestions?

[tom91]

End up with just have separate contactors for each string?

You need to design a controller that manages to gather data from both packs/bmses then determines if they can be combined together on the HV bus safely. It will need to then need to control the precharging of one pack and then the second one if the stack voltages are close enough.

[CasaDelGato]

Well, that is a non-answer.
For the Leaf drive to USE both packs, simultaneously, they will need to be connected in parallel while driving, and charging.
This means that the measured pack voltage is guaranteed to be the same on both - as they are tied directly together.
About the only thing I can see to try as a safety measure, is to have something monitoring the current from each pack, and shut things down if it's significantly unbalanced.
I could separate them while parked and not charging, but it's far less likely to have a cell failure that will cause a problem at that time. It would possibly work if one side had a higher rate of self-discharge - but that isn't likely to be significant amount.
The most likely failure will be during driving/charging - when they HAVE to be connected together.

Looks like about the only thing I can sanely do is to fuse both packs separately - so at least if something does cause a major unwanted current flow - the fuse will blow.

[tom91]

Well, that is a non-answer.

With your limited information on your system layout or wiring you will get high level answers.

Yes ofcourse you should fuse the strings separately along with fusing any cable running between battery pack enclosures.

[jrbe]

Well, that is a non-answer.

That was a crap response to the right answer.

So you think it needs a controller that manages to gather data from both packs/bmses then determines if they can be combined together? And drop one out if necessary?

[johu]

I think it is now established that the diode solution is no longer appropriate with a Leaf setup because it would also limit regen power.

Now there are two solutions:
a) bring both packs to the same voltage and then PERMANENTLY parallel them
b) equip each pack with its own contactor. Contactor A can always come on. Contactor B will only come on if the voltage of pack B is within, say, 1V of pack A

Please correct me if I'm wrong but I think most people here who use 2 parallel packs are going for a) because KISS

Well, that is a non-answer.

viewtopic.php?t=5714

[CasaDelGato]

A) has the issue that if something does fail - you are likely to cause a fire.
B) is probably what I'm going to go with - in combination with my Orion BMS. If the BMS detects one or more cells that are WAY too low, I'll have it drop the contactor for that pack. Not an optimal solution, but at least it has a better chance of preventing a fire.
I don't like depending on a computer for simple safety conditions.
yes, each pack will be separately fused, but it's quite possible for a bad cell to not cause currents that exceed the fuse rating.

[arber333]

Hm... what you need is one DC contactor between both packs and a large diode across contactor in reverse.
Contactor is there to connect both packs and diode to provide path for charging...
Here is the link to similar setup.
viewtopic.php?t=301

[johu]

you are likely to cause a fire

That seems far fetched. I'm writing this not so much to convince you but to replace fears by my humble knowledge - as other people are going to stumble across it.
So, what needs to happen for a battery to catch on fire?
a) thermal runaway by overpowering it (e.g. fast charging) and lack of temp monitoring and derating
b) overcharging it

b) must be avoided by the BMS communicating with the charger. In the best case this is a current limit via CAN but even a simple "stop charging now" signal does the job.
a) is somewhat similar to b). The BMS must always have (indirect) control over currents going in and out of the pack. It's how all EVs work. The BMS broadcasts limits and downstream ECUs obey it.

Now, in your special case there is another possibility for overpowering and that would be compensating current flowing from one pack to another. Obviously when you parallel an empty and a full pack these are going to be huge, cause a lot of heat and with a bit of luck blow one or both pack fuses. But you're not going to do that.

When the packs are permanently paralleled there can only be minuscule compensating current if one pack has a higher self discharge than the other. We are talking a few mA. One pack is not suddenly going to drop in voltage when treated right as in b).

A remaining fault case is a cell developing an internal short circuit. That is an issue whether or not you run packs in parallel. Such an event caused by fabrication faults will cause thermal runaway by itself. In parallel packs you just have the additional booster of feeding extra current into the fault. That is also the case with battery modules that internally have cells in parallel - which is super common.

This last fault caused a recall for the Opel Ampera (and probably Chevy Bolt). It's best avoided by using reputable sources

[tom91]

I don't like depending on a computer for simple safety conditions.

Yet you are depending on a BMS to tell you when to drop the contactor. But what will allow closing of the contactor for the two packs? Or are you looking to have contactors always on?

Again without any information on your intended system layout it is very hard to provide any recommendations or consider any of the safety implications.