Vectrix VX-1 Leaf module conversion

[Spoonman]

Morning all - So I'm just getting set to finish up converting a 2010 registered VX-1 from its original NiMh pack to a stack of 19 Leaf modules and figured I'd throw the tale up on here for your amusement.

For the unitiated, this is a Vectrix VX-1 (stock photo, don't have a decent wide shot of her):

From VX-1

This particular VX-1 was a dead bike for the last two years or so, and belongs to a colleague for whom I'm doing the work.

The job that needs doing is to remove the 6 banks of NiMh cells from the bike, replace them with 19 Leaf modules, reprogram the charger and the motor controller, and button everything back up.

The teardown is pretty straight forward, and particuarly for this one as it would appear that this is a later 'Lithium' chassis that had NiMh installed, meaning there were a couple of tasks such as sealing chassis holes and cutting plastics, which I didn't need to perform.

Once the plastics were out of the way, the original battery was exposed and, well, I was expecting this part to be a bit of a challenge as everything I'd heard about the process suggested that these batteries were quite a chore to remove.

...they weren't... they came straight out, not mess, no fuss.

From VX-1

(You can see the leaf modules in the background of that shot in a stack of 24, exactly as I removed them from the battery housing - but more on that later.)

leaving behind a space which quite literally couldn't be any better suited to the Leaf modules

From VX-1

So no, onwards to the fitting and connection of the Leaf modules.

The fillet of the Leaf pack lies in the 24 module block to its rear - and I was lucky enough to get exactly that when the complete battery housing was purchased

From VX-1

Some delicate removal of a lot of wiring and fixings yields this from the battery housing

From VX-1

In order to fit this into the VX, it has to be shortened by 5 modules to 19, the retaining bars were cut and rethreaded to suit the new stack length, and the BMS looming and battery power connectors were reconfigured and retained. There was a lot of angle grinding of the fixtures for the pack in order to get it to fit as well as it does, but I was able to use some of the waste material to create a very effective lifting rig for the pack which made getting it into place a whole lot easier (I should have taken more pictures of that process).

Once in place, it wouldn't look much better if it came factory fitted.

From VX-1

From VX-1

The battery bay is lined with adhesive backed closed cell foam in order to cushion and snug the battery into place, as well as to block the now redundant air channels which would only serve to bring moisture into the bay.

The BMS connections can be seen - I'll be building an interface to monitor the pack in the future and it will connect directly onto those plugs.

The battery is reconnected to the motor controller using a 60w filiment bulb as an inrush current limiter to precharge the capacitors before being permanently connected.
Power now restored, and the ignition switch turned, the dash lights up and all displays function correctly.

At this point it's time to upload the revised softwares to the Motor controller and the charger - there was a bit of a song and dance to be performed in finding the correct version of the scooter diag software, and the correct system file for the CANbus interface, but once that was done I was able to connect to the scooter using my laptop (Win10 & ScooterDiag2012). It was at this point that I found out that the charger installed was not communicating with the CANbus at all - everything else was fine, just not the charger.

[Spoonman]

Now, during the removal process, I encountered a blown temperature board in the front pack - a known weakness with these bikes - but not, ordinarily, in itself enough to kill it.
Examination over the CANbus however revealed that the charger wasn't responding; and as the temperature boards operate off the CANbus interface from the charger, the likelyhood became that whatever even blew the snot out of the temp sensors, took the charger CANbus interface with it.

From VX-1

It was the Runke charger, which is purported by the community to be the better charger to have, but I was chatting to David Dugas at http://www.vectrixsupport.com/ and sent him some pictures of the insides of the charger to see if he could advise on a fix, and his response was, verbatim, "yuk...thats an old charger .... Those were junk".

Multiple failure modes apparently, which start in the signal level components and cacade up to the power elements; there being nothing to loose I powered the unit up direcly on the bench and was instantly greeted with two capacitor failures, most likely overvoltage as there wasn't any load and inrush wasn't a factor - on inspection they had also failed short circuit and taken out another IC nearby.

From VX-1

From VX-1

So given the absence, and unavailability, of a schematic, it became obvious that the old charger was goosed - luckily I was aware of a parts bike not too far from me, so I arranged with its owner to pop along a whip the charger off of it. This turned out to be an ESB charger - which the community would tell you was the older and less desirable charger; although to my eye it appears to be far better build in every regard.

From VX-1

So I got it home and installed into the bike

From VX-1

From VX-1

and it showed up perfectly on the CANbus so now all is looking well.

As of yesterday evening I have the motor controller and the charger reprogrammed with the custom code, and I've charged for 1 hour so far.
I'll charge for another few hours this evening and just keep and eye on the pack as it approaches EoC to make sure that it all stays in balance, but assuming it does we're just about done.

[DRZ400]

Cool...looks factory!

[Spoonman]

Yeah, I'm super stoked with how well it came together - everything about the Leaf-VX combination is like a dream so far!

Finished off the CC portion of the charge there this evening and watched the current drop from 9.5A down to 4A, whilst constantly checking the individual cell voltages - thing's running like a swiss watch and those Leaf modules continue to astound me; 0.02V Spread over 38 cells after 4.5 hours of charging! Utterly incredible.

So I've left it to finish itself off whilst I get some shut eye

Can't wait to button her up and ride the thing now before the week is out.

Plan is to do a couple of short trips and and top up charges, checking the cell voltages before and after in each case, and then I'll take her on a 45min motorway spin once I'm happy that everything is working as well as it currently appears to be - charging back up from that then should tell the full story.

[Spoonman]

So I took her for a maiden spin on the Leaf modules today - 70km a mixture of speeds up to 110km/h - the performance was perfect, am massively impressed by it.

Based on the pack voltage after the spin, the 70km used about 45% of the nominal capacity of the pack, so 120-130km usable range should be expected at higher speeds; significantly more than that in an urban environment.

Plan now is to recharge then do a full discharge trip in the coming days, recharge again and reassess pack balance at EoC; I'll be installing a monitoring system regardless but for now I'd just like to get an idea of the stability of the pack.

[DRZ400]

I've got over 100 cycles on my 11 modules....no BMS. Charge to 90v, discharge down to 80v. All cells within .005v

[Spoonman]

Have been hearing tales like that from all over the place - these Leaf modules really do seem to be a complete work of art!

[DRZ400]

Have been hearing tales like that from all over the place - these Leaf modules really do seem to be a complete work of art!

They are...but it really helps to not top charge to a full 4.2v/cell. Avoid the cliffs just like Nissan does.

[Spoonman]

Yeah, charger is set to terminate at 4.1v/cell; and I've the Low voltage alarm kicking in a 3.4v/cell so it's all operating well within the more linear portion of the curve.

[Spoonman]

Literally just finishing up another one of these this evening - latest one had a Zeva BMS as part of the deal and I got some learning related to that along the way.
Worth noting that I've found the vectrix parts lad in Poland to be very reliable and would have no problem buying from them again - wouldn't be keen to rely on them for support though, so you need to go in knowing what you need and how to use it.
David Dugas has made all his stuff freely available, which makes programming and configuring a breeze once you have the Peak PCAN cable.

Might throw up a few pics of the process later if I get a minute, along with a couple of the salient learnings.