A2B Metro (Gen 1) Restoration and Upgrade

[chuyskywalker]

Was searching for a donor bike to electrify and serendipity struck, this A2B showed up in the local listings missing all the batteries, but otherwise in good shaped. I love the styling on this and it was "new" to me. Picked it up for $340.



Trawling these here forums has been very helpful in figuring out where to take this bike. (Frustrating that so many old photos have been lost to time.)

First, tore it down and cleaned it up:



Since the bike is something like 15 years old, and the design has passed between a few companies there's absolutely no parts around anymore -- gonna be a very DIY project. Good thing I own a CNC machine and 3d printer.

First I'm going to tackle the internal battery to build a replacement. I'll be going "easy mode" -- instead of trying to stuff a 60 cell 18650 setup in there (which is what it originally shipped with), I'm going to go with a 20s2p 72v 21700 setup. While that'll only net me 10Ah of capacity, I plan to supplement that later on with a rear battery.

Here's what I've got so far:

Mapped out some 2170 cells along a path that I think should fit -- really wish I had an original battery to take some detailed measurements from -- or, better, a CAD model of the original. Anybody, uh, happen to know some of the OG A2B engineers that could accidentally email that to me?



Add a box and subtract the shape out to get a cell holder:


Since I'm 3d printing this, I decided to lop off the top/bottom so it'll print flat:


Sliced into 3 parts to fit on the printer and setup for going:


----

I'd like to be able to jam along at a comfortable 35mph, which is why I'm going 72v -- plus I have a few other 72v PEV's around, so it's nice to have less "what charger do I use?" hassles

If you have any good recommendations for what motor to pair this with, I'm all ears. I'm currently leaning towards a "GMAC 8T" paired with a Phaserunner and that should net me 35+. (Yeah, it says 32, but the simulator doesn't account for FOC which should get me that extra bit.)

 

[chuyskywalker]

The print came out out ok. I've loaded it up with my dummy 3d printed 21700 "cells" to see how it fits.


All wrapped in tape and it fits in and along the curve just fine.


---

At this point, the "internal spacer" was really just that and had a few issues. First off, it was really hard to figure out how to align each cell vertically. Second, this setup provides no protection to the pack at all. The original was "encased" in material, which was pretty smart.

I decided to try and machine it from HDPE, but quickly found that the arched nature of the cell spots was driving my CAM software up the wall:



So, instead of trying to keep all the cells arched and pointed at a far away center point, I instead just lined them up on the curve, but perfectly horizontal. This is WAY easier to machine, at least I hope so

I also moved to a 21 cell holder setup so that everything is symmetrical from the center out so that I can machine the same thing twice and just sandwich them back-to-back.



Here's the design (half) with cutouts for nickel strip and some vent holes, and it laid out for machining

 

[fechter]

Nice work on the battery. I pretty much gave up on the internal battery and went with an external rear battery mainly so I could run a higher voltage (52v).

Here's mine from years ago:
https://endless-sphere.com/forums/viewtopic.php?f=6&t=50590&hilit=a2b+metro

Replacing the tires was one of the more worthwhile mods. The stock tires handle very poorly on corners and go flat easily.
The stock internal controller is prone to overheating after a long ride. Once it cuts out on over temp, it takes about 30 minutes to cool off enough to reset. An external controller is a lot of work but you can increase the power significantly.

 

[chuyskywalker]

Nice work on the battery. I pretty much gave up on the internal battery and went with an external rear battery mainly so I could run a higher voltage (52v).

Thanks; I also considered just bolting some shark packs on to the rear rack sideways. If you pull the bar stock and rails off, they would mount right to where the rails used to be. May come back to that if I don't come up with something nicer, but it would mean losing the storage space which I'm remiss to do.

Here's mine from years ago: viewtopic.php?f=6&t=50590&hilit=a2b+metro

For sure; I have that thread book marked for the fork-flip

Replacing the tires was one of the more worthwhile mods. The stock tires handle very poorly on corners and go flat easily.

Got any recommendations for what to toss on there? The previous thread ended up with a lot of recommendations. I probably won't do a ton of trail riding, so better speed handling is probably weighted higher on my todo list.

The stock internal controller is prone to overheating after a long ride. Once it cuts out on over temp, it takes about 30 minutes to cool off enough to reset. An external controller is a lot of work but you can increase the power significantly.

Yeah; I'm gonna just immediately jump in to a new motor/controller combo; I'm certain I want more power, but specifically, more control than what the stock system allows.

 

[fechter]

I ended up using some 16" moped tires. 16" moped tires will fit on a 20" bicycle rim. Hard to mount, but very tough. If you get new tires with a round profile, you probably don't need to flip the fork.

Rewiring the stock motor for external controller was a LOT of work. It would be significantly easier to just get a new motor that is already wired for an external controller. Having regen braking is a very nice feature you can get with an external controller.

 

[chuyskywalker]

I ended up using some 16" moped tires. 16" moped tires will fit on a 20" bicycle rim. Hard to mount, but very tough. If you get new tires with a round profile, you probably don't need to flip the fork.

Based on that, and previous threads, I went ahead and ordered some "Shinko SR714 2.25-16" tires. I kept thinking about the SR 241/244, but have to be honest that I'm not really going to be doing enough off-roading on much more than pre-packed, dry dirt trails rarely, so the knobiness is simply not justified for me.

I'll see how they do before the fork flip.

Rewiring the stock motor for external controller was a LOT of work. It would be significantly easier to just get a new motor that is already wired for an external controller. Having regen braking is a very nice feature you can get with an external controller.

Yeah, and rewiring is definitely beyond my technical reach at the moment too. Regen braking is absolutely critical to me; I can't stand systems that don't have it. So, no free-wheel geared motors for me.

 

[Audisport09]

Very impressive work,I gave up trying to put batteries in there and used the space for a SPY Motorcycle Alarm , a horn and 12v step-down. Keep us updated.

 

[chuyskywalker]

Machined the new holder from HDPE and it's a perfect fit. I might even have enough room to hide the Phaserunner inside the tube. Not sure how I'd mount it, but there seems like there could be room.


The inner half cutout. Lots of learnings on this piece, even though it did sorta work. My tabs (to keep the thing in while doing the outer cutout) were much to short and too few.


You can see the damage right at the end from the piece jumping up into the cutter before I slammed the power off. Need to invest in some straight blade plastic cutting bits...


None the less, my mock-21700 cells fit in nicely and stagger to the shape.


The outer shell piece, with some revised paths based on the first run, came out much nicer. (Here you can see how I added quite a few more tabs for this setup)


Finally, the cells all sandwiched up. You can get an impression of how the parallel and series nickel will run from this view.


Will probably have to run the balance leads along the tops after a layer or two of kapton. Or maybe down one side and drill holes for wires through. Dunno yet.

 

[chuyskywalker]

Now we're talking.



Need that spot welder to show up

 

[Skaiwerd]

The smaller bikes have their advantages, I know this is smaller than yours 12” rims.

 

[chuyskywalker]

The smaller bikes have their advantages, I know this is smaller than yours 12” rims.

I have an EUC for that

 

[chuyskywalker]

If you have any good recommendations for what motor to pair this with, I'm all ears. I'm currently leaning towards a "GMAC 8T" paired ...

To my great dismay, the GMAC motor is not compatible with 150mm dropouts. Seems like the only hub motor Grin sells is that DD45.

Or maybe get a QSMotor 205 50H V3 4T. At 820RPM (from 72v) on a 20" tire, that should be about 48mph top speed; the 3t winding @650 rpm drops that to 38mph. I suppose that would likely be good enough, but...

(Using this calculator.)

 

[amberwolf]

The All Axle motor would be compatible with any dropouts, depending on the axle you choose to install in it (though you might have to make the axle first, if one isn't already available).

Yes, it is a "front" motor, but you can use it in the rear if you don't need a freewheel (use it with the disc mount on the chain side, and use a sprocket-to-rotor adapter for a singlespeed in the rear). Then use a freewheeling crankset instead. That's assuming you are planning on pedalling at all.

 

[chuyskywalker]

The All Axle motor would be compatible with any dropouts, depending on the axle you choose to install in it (though you might have to make the axle first, if one isn't already available).

That's an option, one I've not much looked into or have done before. How would that be sourced? I suppose you'd use the adapters they offer to get the right width, but I have no idea where to source the axle itself, or how to pick one that would wok.

Even then, I feel like all the windings Grin offers, now that I'm looking closer, tend to really peak out (even at 72) around the mid/upper-30's. I think, philosophically, they just don't sell motors that'll make the speed I'm really after. I know I said 35, but if I'm being honest, I probably want to be able to peak up to 40 once in a while.

This is where the qs motor is a bit more plug-and-play (fits the 150mm natively) and doesn't skimp out on winding options.

 

[amberwolf]

It's just REALLY heavy, at basically 30lbs. (three times the GAAM)

https://endless-sphere.com/forums/viewtopic.php?f=2&t=67833&p=1526261#p1526261

But it will handle a lot more power, both continuously and peak, than the GAAM.


I don't know what you would need to use for the axle, but presumably you would simply use the actual axle that fits the frame. Meaning, a regular rear axle with nuts/washers or a long QR axle, to pass thru the hollow adapters. You can look up the dual-grin-all-axle trike thread (cant' remember who it's by) for how they had hardware worked up to use them on the front of a delta trike. EDIT: https://endless-sphere.com/forums/viewtopic.php?t=88266

Keep in mind also with the speed you want, that besides being wound to be able to reach the speed, the motor needs to be able to handle the power needed to reach those speeds under the riding conditions you have (terrain, wind, etc). For many motors, the Grin Simulator will show you the "overheat in" time under the conditions you setup for a particular system and throttle value/speed.

 

[chuyskywalker]

It's just REALLY heavy, at basically 30lbs. (three times the GAAM)

Well, it'll be a "bike" with pedals, but we all know how one intends to ride it

I'll check out that thread; thanks for the link.

And yeah, I've spent an enormous amount of time with the simulator; Justin's live-stream on hub motors is invaluable in understanding it better too.

 

[amberwolf]

It's just REALLY heavy, at basically 30lbs. (three times the GAAM)

Well, it'll be a "bike" with pedals, but we all know how one intends to ride it

It's not really about whether it's a bicycle or a "motorcycle" but that the suspension on the A2B was designed around a certain amount of wheel mass. I can't remmeber the weight of the ultramotor, but it's way less than the QS205, so the suspension wont' work nearly as well with the QS as with the UM. You may be able to retune it if the shock is adjustable (enough), but I don't know how to do that.

 

[chuyskywalker]

It's not really about whether it's a bicycle or a "motorcycle" but that the suspension on the A2B was designed around a certain amount of wheel mass.

Would that matter? The motor/wheel are the unsprung weight, being that they're on the ground -- isn't it the rest of the bike/ride weight the suspension has to deal with?

 

[Audisport09]

Nice, make me one, make me one! I’m guessing this is not your first gig at bikes. Whats the amp rating? You still have room for a bms? I think the stock motor is like 9-10kg so it would only be a 3kg weight increase. QS205 for the win!
4T At 820RPM and the 3t winding @650 rpm…wrong way round possibly…? 3T should be faster (top speed) than 4T. Although I feel you may be focusing too much on top speed. That’s the second most asked question I get, how fast does it go? Doesn’t matter! Torque is where the fun is at on ebikes and the QS has plenty.

 

[chuyskywalker]

Nice, make me one, make me one! I’m guessing this is not your first gig at bikes. Whats the amp rating? You still have room for a bms?

We'll see if I make any others. At the least, if it works out, I'm happy to publish the model. First ebike, but have done up a scooter previously though and a lot of the transfers over.

I haven't fit tested it yet, but I think it will be thin enough to slap the BMS on the side actually. If not, it's much shorter than the original holder, so I think I can stuff it in the tube either way.

Given it's only 2p, it's a pretty low 20amps continuous, so not gonna be winning any drag races and DEFINITELY not going to be taking full advantage of just about any motor I put on it, especially not a 205, lol. But since I plan to parallel more power into the system, I think I'll get somewhere more appropriate eventually -- at least up to the 50ish amps that the PhaseRunner can handle. That'll be enough for me, pretty sure.

I think the stock motor is like 9-10kg so it would only be a 3kg weight increase. QS205 for the win!

I'll have to weigh it, but that's a good point.

4T At 820RPM and the 3t winding @650 rpm…wrong way round possibly…? 3T should be faster (top speed) than 4T. Although I feel you may be focusing too much on top speed. That’s the second most asked question I get, how fast does it go? Doesn’t matter! Torque is where the fun is at on ebikes and the QS has plenty.

Yeah, you're right I got the windings/rpms mixed up. In any case the up/down step on winding falls outside my top speed range desires. 4T gets the closest. And, yes, torque is a lot of fun, but I'm more of a destination cruiser type, so I don't need to smoke everybody off the line WHILE doing a wheely, but I do need to be able to keep up

 

[amberwolf]

Would that matter? The motor/wheel are the unsprung weight, being that they're on the ground -- isn't it the rest of the bike/ride weight the suspension has to deal with?

Yes, the unsprung weight is what makes or breaks a suspension's performance. If it's tuned for a different amount of mass there, it won't operate as intended.

There's a few threads about suspension (some of htem are about a bike or questions about a potential build, but have suspension discussions in them). The ones that might help the most should have "unsprung" in them (they might use weight or mass or other terms, but "unsprung" should be the same in most cases; "unsuspended" might be used as well). I expect MadRhino has contributed to most of those discussions; there's others here that also know a good bit about suspensions.


But basically, there's two things.

The suspension for any wheel has a loaded mass on it that is everything not on the swingarm, or the suspension on other wheels, and that mass is different for different riders and uses (like if you intend to carry a backpack or extra battery or whatever, all the time). That's something you can setup the preload on the suspension to take to minimize sag, and get closer to the suspension response you want out of it vs what it would be with a different load on it.

Then the suspension has a "response time" that depends on inertia of all the stuff that's between the suspension and the ground--meaning, how fast that stuff will start moving, vs the pressure of the suspension on it, vs whatever it hits on the road that pushes it up or whatever hole it falls into that lets it fall down. The more mass, the slower this response is, so if the suspension is set for less mass than you have, it'll be even slower. To some degree, some suspensions can be retuned for a different mass than they're built for, though I don't know how to do it.

The cheaper the suspension design is, the less likely you'll have much if any control over it's response, beyond perhaps preload.

 

[chuyskywalker]

Seems kinda hard to find the QS205 in 4T winding with a 16" rim premade. Gonna keep shopping around and see if anything pops up. Not super excited about the idea of lacing a wheel myself.

 

[amberwolf]

You could contact QSMotors directly, via the info they provide in their sales thread over in items-for-sale-new (since I don't know which website out there with their name being used on it is actually theirs). It will probably take more time and might cost more shipping, but you can get exactly what you want from them.

 

[chuyskywalker]

I did; they only directly lace down to 17" rims.

Also shot out an email to this guy offering a 16" laced 205 @ 5t which is almost exactly what I want, except for the slightly off winding. Hoping either I can get one through them or have them do the lacing (they seem to be local, ca bay area, to me). The hunt continues

 

[chuyskywalker]

No replies, and nothing else pre-made showing up.

Having gone through several other threads here where folks just open up and remove the internal controller, this motor seems to be pretty good. I don't really feel like burning a month or two waiting on a motor from China (nor paying 2x the price of the motor in shipping, yikes pandemic...) Plus then getting a new rim, and figuring out spoke lengths, and doing the lacing. And, really, I wouldn't be using a 205 to it's real potential anyway.

I'm more certain I can handle pulling this motor apart, yanking out the controller, and rewiring the phase/hall wires. And, aside from waiting on the gear puller to get here in a day or two, I'll have everything to get a working motor up and running.