2009 Zero X Upgrade/Rebuild w/ sia155-48, cl700, & 28s12p p42as

[thebes]

This thread is going to be half me documenting my thoughts/thinking out loud and half asking for advice, hopefully yall dont mind!

Picked up a 2009 Zero X (according to the vin at least) in a somewhat unknown condition with a presumably dead battery for $300 a few days ago. It seems to have been in somewhat of a rebuild before being put away in a storage unit which eventually got auctioned and the guy who got it had no idea what to do with it. As a result it has a mix of 2012 mx plastics and components with the forks appearing to be off a 2010 or 2009 mx, what I think is an upgraded rear shock, and an upgraded mars motor? I believe these came with the agni, please correct me if Im mistaken! Also came with a bin of spare plastics and other knickknacks and a spare rear wheel!

(apologies for the poor picture, we got back pretty late...)


In its current state, apart from the dead battery (which I might make a separate thread on disassembling or even post it here if anyones interested), I think the altrax controller might be dead, when powered with an external psu contactor clicks, display comes on, get a steady green flash (which I assume is good) from the speed/torque selector and lights turn but dead throttle...Ill do more in depth troubleshooting later this week (hopefully!) and hopefully get it at least driving with a spare farsis brick I have. Oh and both tires are flat and leaking, and the brakes do not brake

I am not a particular fan of the motenergy/mars motor since its brushed and not particularly high powered. Inspired by docs drag bike built with the same frame Im hoping to be able to fit a zf75-7 without modifying anything but the mounting brackets, though I might have to cut out and extend the bottom bracket :/ (bit of an issue since I dont have a welder that can do aluminum). This may be ambitious though...I havent really been able to find any higher power alternatives that would fit in that space apart from the somewhat under powered (at least when compared to the zero motors lol) qs165 50h. If anyone has any suggestions Id be very interested! The max depth of the motor appears to be ~160mm and the max diameter is ~203mm (the zero motors are all 220mm). Perhaps one of those new surron motors but theyre pretty pricey and dont really appear to be all that great :/ Though before deciding on anything I think Ill print a mockup of a 75-7 and see just how much of the frame Id need to chop out...
Also not quite sure if the drag bike build was ever finished since the thread was never updated, would be very interested in the end result if it was!

As for the rest of the powertrain, I have 540 p42as that are begging to be put into something, planning to arrange them into 28s20p with nylon printed cell holders, 0.2mm copper sheets for the connections and a 420a ""cont"" ~1ka peak 30s ant bms. Thisll be powering my cl700 whichll be the weakest link in this build it seems, if I can find one of the larger sevcons somewhere (and someone who has the dongle) for cheap I might consider switching to that...
Also if anyone knows what the other set of pins on the battery connector on the stock bike are please let me know, I know two are for a temp probe and I assume the other two are a sort of enable signal to the bms or something like that but I cant seem to find any info :/

Oh and a brake upgrade will definitely be in order haha, cant believe they shipped these with mountain bike brakes lol.

 

[amberwolf]

if I can find one of the larger sevcons somewhere

If it helps, I have a sevcon Gen4 80v 350A and the main connector with some wiring harness on it. I do not know it's functional state, it was sent to me along with some other parts; I think from one Zero bike or another, but don't know which one--some marker on the heatsink says "15fx" which if that's a 2015 ZeroFX means it should be setup for a ZF5-7 motor?

https://www.motorcycle.com/specs/zero/on-off-road/2015/fx/zf5-7/detail.html.

I'll never be able to use it because I can't program it (don't have the hardware or software or expertise, or the money for any of those) (and it is totally overkill for anything I can ever see myself needing).

If you want to try it out, then if you'll pay for whatever shipping would be from Phoenix AZ 85051 (guessing around 7-8lbs, size of a big thick hardcover book), I can send it to you, then if it does what you want you can donate whatever you feel like to my projects or dog (well, JellyBean doesn't really qualify as a dog, just an alien Schmoo). If it doesn't, then just forward it on to the next person that wants to mess with it.

 

[amberwolf]

In its current state, apart from the dead battery (which I might make a separate thread on disassembling or even post it here if anyones interested), I think the altrax controller might be dead, when powered with an external psu contactor clicks, display comes on, get a steady green flash (which I assume is good) from the speed/torque selector and lights turn but dead throttle...Ill do more in depth troubleshooting later this week (hopefully!) and hopefully get it at least driving with a spare farsis brick I have. Oh and both tires are flat and leaking, and the brakes do not brake

It would be interesting to see the battery.

At least one other zero here has posts about alltrax problems; I think it was a throttle thing but was a while back.

 

[thebes]

If it helps, I have a sevcon Gen4 80v 350A and the main connector with some wiring harness on it. I do not know it's functional state, it was sent to me along with some other parts; I think from one Zero bike or another, but don't know which one--some marker on the heatsink says "15fx" which if that's a 2015 ZeroFX means it should be setup for a ZF5-7 motor?

Yep its most likely from a 2015 fx, which ran with a zf75-5. Sadly though, Ill have to decline your offer for now since Im still not a 100% sure what motor Ill end up going with. If I do find a zf75-5 somewhere Ill reconsider but without a way to reprogram it for regen :/ Again though, thank you a lot for your offer!

It would be interesting to see the battery.

Alright! Ill do a writeup.

At least one other zero here has posts about alltrax problems; I think it was a throttle thing but was a while back.

I think Ive seen that thread a while ago...Ive also had another alltrax die on me after sitting for about ~10 years, it was an axe4834 I believe, its just completely dead, no error codes, no lights, no nothing, Ive also got another one sitting in a 2010 zero s which Im hoping is functional but I dont have particularly high hopes...

 

[amberwolf]

a controller that just dies sitting ther eproblaby has bad capacitors. they are chemically operated inside so they have a limited lifespan. some are made poortly so they go bad far soon er than tehy should. happens a lot in power supplies, which is basically what controllers are inside.


it's likely i'll have the sevcon around for the foresceeeable future so if you end up needing to try it out, it'll probalby bve here. or if you know anyone else that wants to try it.

if i knew it's condition and programming setup i could just sell it (cuz they';re apparently worht quite a bit) but not knowing that i wouldn't feel very good if it ended up being a deader or soemthing... given who it came from i expect it probably workse perfectly and is still setup exactly as it had been for the bike it probably came from, but i can't know without testing, wich i don't ahve the motor or other bike hardware to do. :/

 

[thebes]

On to the battery disassembly, I regrettably did not take a before picture but after standing it up on its side and pulling out the ~30 odd m3 or m4 screws out of the outer casing and the 3 holding in the black battery connector, pulling the side plates off was fairly easy. Here is the side with the bms which appears to have a 3d printed (I think in tpu?) cover

Heres the cover:

Also heres a better look a the text on the cell

And here is the other side with the main discharge leads and balance leads

Heres a better look at the discharge connector and the current meter, I believe its a 500a unit, forgot to take a picture of the model number :/


At this point it was possible to pull the cells out of the aluminum case with a fair bit of hassle due to not really have anything to grab on but they did come out! It is a 7s1p arrangement of these modules which each are 2s12p of molicel imr26700as which are 2.9ah and rated at 40 (or 60 depending on where you look) amps discharge. Total pack arrangement is 14s12p for a total of ~1.7kwhs capable of ~500 amps or ~25kws.

Heres just another angle

And a better look at the connectors which are mirrored on either side (the marker was already there haha)

Better look at all the balance connectors and discharge leads, the balance connectors are all connected with those crappy shovel connectors and fused with 4a micro fuses. The main leads are 8 gauge fused with a 300a anl fuse.

Here is that side after pulling everything off

Now time to pull these apart! All the modules are vhb taped together which I had to use a 3d printer scraper to slice through as I couldnt quite get my razor in there


All the modules apart:

Im not surprised they had so many issues with these packs, dont know how this was ever greenlit haha. The cells are sandwhiched together in a cardboard tube and the bus bars (which are more of a braided copper rope type thing) are held on with rubber bands that remind me of sliced up inner tubes.


Here are the cells:

Made in canada is interesting, I thought molicels factories were in asia...Sadly though all of these cells are completely dead, read no voltage both with my crappy amazon meter and my friends' nicer rack one :/


Better look:
(I sliced my thumb pretty badly with the part scraper)


Did later try force charging one but this was probably a waste of time, these have been dead for a longgggg time and it appears water got into the pack at some point, or the cells leaked? or both? There was some residual substance that indicated water in the aluminum casing but this looks more like leaky cells to me...


Basically all the modules had this sort of damage on them, sort of a shame that I cant reuse these for some other project but it was to be expected with a ~16 year old bike. Overall the pack construction is as abysmal as everyone says it was, again I cant imagine shipping this to customers, but it was a different time I suppose...If anyone is interested in the bms let me know! I dont really have any use for it as I dont believe its needed to run the rest of the bike with? And even if it was, the bike will be getting an entirely new powertrain haha...

 

[thebes]

a controller that just dies sitting ther eproblaby has bad capacitors. they are chemically operated inside so they have a limited lifespan. some are made poortly so they go bad far soon er than tehy should. happens a lot in power supplies, which is basically what controllers are inside.

This was my line of thinking as well, I originally intended to try and diagnose/repair the boards but they turned out to be potted, I dont have any pictures of this but its about an inch of black epoxy that as poured into the casing, the db9 connector for programming them sits just above it.

it's likely i'll have the sevcon around for the foresceeeable future so if you end up needing to try it out, it'll probalby bve here. or if you know anyone else that wants to try it.

Ill reach out to some of my buddies who might potentially be interested!

if i knew it's condition and programming setup i could just sell it (cuz they';re apparently worht quite a bit) but not knowing that i wouldn't feel very good if it ended up being a deader or soemthing... given who it came from i expect it probably workse perfectly and is still setup exactly as it had been for the bike it probably came from, but i can't know without testing, wich i don't ahve the motor or other bike hardware to do. :/

Your caution is quite understandable!

 

[j bjork]

The battery construction looks exactly like on my 2011 DS, but in a different shape. I was a little shocked when I took it apart too

 

[jkrienert]

I'm really surprised, but also curious why they used rubber (?) as the fastener for the HV bus. Wouldn't any heat generated compromise the tension applied by the rubber? Maybe its not rubber in the vulcanized elastic sense, but it sure looks like the innertube material as noted. Maybe the rubber just a cushion for the heat shrink acting as the 'fastener'? Perhaps it was an effort to minimize mechanical metal bits that could become work loosened and present a fire hazard?

Thanks for this dissection of the pack. It gives some effect of, "oh wow, my build isn't that bad after all...". Were these packs UL approved?

 

[thebes]

I'm really surprised, but also curious why they used rubber (?) as the fastener for the HV bus. Wouldn't any heat generated compromise the tension applied by the rubber?

I do not think these cells got particularly hot as they were being run at about half their rated discharge rate but it is still a valid concern, I think Ill put a couple of the stripes under a hairdryer later today to see how it reacts, if it reacts at all.

Maybe the rubber just a cushion for the heat shrink acting as the 'fastener'? Perhaps it was an effort to minimize mechanical metal bits that could become work loosened and present a fire hazard?

They are both working together, I dont think either would alone have been enough to keep the braided strips in place (not that they are enough together either but its better at least).

Thanks for this dissection of the pack. It gives some effect of, "oh wow, my build isn't that bad after all...". Were these packs UL approved?

Thanks! And I dont believe so, I couldnt find any mention of any certification...The only thing they mention at all is that it was "patent pending" which doesnt really mean much...

 

[jkrienert]

Honestly, it leaves me questioning (in a good way) how I am making the series connection in the pack I'm creating. Maybe not the intent of the builders or your reveal of such, but certainly a worthwhile moment of pause to reflect on a design part, that cascades effect to all others.

They were not using this system in any way to establish quasi cell level fusing right? Since, the braided copper would still have intermittent contact if say, one of the rubber belts broke or if in addition the shrink wrap failed?

 

[PaulM]

Interesting thread, I’ll be watching how this progresses. I have a 2010 mx that I bought new in 2011, all original except a cycle analyst and replacement controller. Mine has the perm132 motor. The original npx4834 controller would quit intermittently. Usually it would start working again after a bit but I did push the bike home once. I put an axe4834 on that I had bought used for another project, no issues with it so far. I’m sure my battery has lost capacity, but I just use the bike for short rides on local trails so it’s good enough for now.

 

[PaulM]

 

[thebes]

Having some trouble sourcing zero motors/me1507s, most forum listings have either been sold or the seller is not responding to my inquiries, there are a couple on ebay but they are all zf75-5s and not the 75-7s Im after...plus theyre a bit more than I was hoping to pay. Seems the cheapest one is ~$650 (before shipping) but I was in talks with a couple local folks to pay that much (if not a bit less) for 75-7s...they probably wont working out though, one guy might end up keeping two for a dual drive gokart project whilst the other is not really sure if he wants to sell and is trying to find a use for it . And of course no one has an me1507 haha...

That left me looking at the qs165 50h since the 60h has a gearbox which adds too much length to the motor and wont fit as a result. Qs rates the 50h at 10kws "nominal" and mentions it being able to do 25kws "sometimes" which is about ~350 battery amps at 72v. This sould translate into about the power Im looking for at 28s, though how much of this translates into power actually at the wheel, and how much of this power can it take continuously? No idea! What isnt particularly inspiring though is that it weighs about half as much as a 75-5 which theoretically is about the same power input...The upside is that it should be able to fit into the frame without modification of either the motor nor the frame as its quite small, and that it costs a ~550 to my door...Oh and also no one seems to have these in the wild yet, or at least anyone that can provide actual numbers and performance data, unlike those surron kids on youtube haha....

Hell maybe Im overthinking all of this and I should just pick up a mag rimmed qs273...

I couldnt really find any other viable high power options that would fit into the given space mostly because it seems manufacturers tend to lengthen motors when trying to create something that can take more power rather than widening/increasing the diameter. The form factor is sort of like a pancake motor, max width from the mounting holes is 160mm (maybe 161mm) and max diameter is ~205mm...I suppose Ill keep looking

Oh also dont think Ive mentioned this before, I can pretty easily fit 28s12p into the stock battery box which translates to about ~5kwhs with p42as. I would like to be able to cram more into the frame so I will probably not end up using it, though Im not fully decided on this.

They were not using this system in any way to establish quasi cell level fusing right? Since, the braided copper would still have intermittent contact if say, one of the rubber belts broke or if in addition the shrink wrap failed?

It would have full contact just possibly higher resistence due to slightly less pressure keeping the braid down. I think the main force holding the braid down is the shrinkwrap, feels as if the bands are more there to just keep the braid in place? But it does add a bit I suppose...Also I blasted it with a heatgun, it did not deform or soften to any noticable extent at the temps that a battery would regularly experience. It only started melting way after it was way too hot to touch, cant really put a number on it but most likely somewhere above a 150c.

Mine has the perm132 motor. The original npx4834 controller would quit intermittently. Usually it would start working again after a bit but I did push the bike home once. I put an axe4834 on that I had bought used for another project, no issues with it so far. I’m sure my battery has lost capacity, but I just use the bike for short rides on local trails so it’s good enough for now.

Oh interesting, I thought they were running an angi r95 with an axe7234, my bike has that controller but a presumably upgraded motor, one of the mars ones, dont remember the model off the top of my head. Also neat to see someone else still running around/doing something with one of these!

 

[thebes]

Bit of a minor update, been working on some cad (can post later!) for the cell holders, seems I should be able to fit 28s14p in the stock battery box which will give me roughly 6kwhs and in the ballpark of 50kws of discharge using p42as, (the battery box can actually accommodate 420 21700s but this leads to an awkward arrangement and I would prefer to stick to 28s). Ive already purchased one of the 30s 400a ""cont"" 1ka peak ant bmses, have used these before and theyve been fairly solid. Will be printing the cell holders out of nylon, it seems to be about the best commonly available and fairly cheap material, also had a couple rolls laying around from another project haha .

I also made an impulse purchase and am now expecting a sia155 48 at my door in a week or two, got ~15% off because of the aliexpress sale (and having spent over a thousand dollars in one order )!

Shouldnt be quite as powerful as a 75-7r and it is not much cheaper but it should be a lot easier to mount which is a major benefit as I dont have the equipment to weld aluminum...

For anyone following at home, I took more precise measurements, the max motor length you can fit is 163.3mm (would be a very very tight fit though), max diameter is 202mm, mounting holes are ~149.5mm away from the center of the shaft in a cross pattern, and the width between the mounting plates is about 70mm.

 

[thebes]

Heres an image dump of the cad models, turns out I measured one of the dimensions incorrectly, seems absolute max is 30s12p, (two cells short of 14p, 390 vs 392) Ill be going for 28s12p as I want to have full regen capabilities and again 13p is awkward...

Cell layout


How itll look in the case


And heres what 30s12p would look like


In the case


Didnt model it here but the bms will be placed in the same place as the stock one, with the box will acting as one large heatsink for it, should reduce the chance of any temp issues...
Oh just to have it mentioned somewhere, will be doing a couple layers of copper/nickle per cell connection so it should be plenty for the planned peak of ~30a per cell.

Have also started a couple test prints in pla to well, test fitment, we'll see how it goes!

Also apologies for the somewhat lack of substance so far, its been mostly just talk haha, will change pretty soon when I start getting parts in and figure out exactly what I want to do!