Firewheel custom battery pack

[Tilmann]

Thanks, any and all help is very welcome.  I was playing around with the idea of using the old battery pack with two new packs in parallel (so the old pack could give the mainboard whatever it wants), but on second thought, it might not be the best idea to mix packs with different capacity and cells... I emailed Firewheel directly yesterday about buying a new mainboard, but haven't heard back yet, also http://scooterhelden.de/ never responded about spare parts or entire wheel without batteries (I left them a message through their contact form last thursday/friday -night, so doubt they're going to answer). Or then the contact form doesn't even work   I could still try mailing them directly...

Hmmm, they are from Berlin, but I don't recognize any of their faces. I can try calling them tomorrow if you like. Just from the looks of it, I would be more optimistic with the folks from 1Radwerkstatt. 

[esaj]

Hmmm, they are from Berlin, but I don't recognize any of their faces. I can try calling them tomorrow if you like. Just from the looks of it, I would be more optimistic with the folks from 1Radwerkstatt. 

That would be great, if you could bother   Although I think the reason they haven't responded simply is that they don't keep spare parts in stock or just don't sell them...  I would rather order the mainboard from within EU, as then I'd probably avoid the hassle with customs and it would arrive a lot faster (although would probably cost more than from China, maybe), but don't know if any store actually keeps spare parts in stock, or if they just order them from the factory for warranty repairs. 

The people from 1Radwerkstatt seem to be also the ones who build the custom packs for electro-sport.de, as electo-sports' battery-page tells to contact their email (1radwerkstatt @ g**.de), maybe for scooterhelden also... Sounds like they know a lot about batteries, so maybe they actually know what that cursed wire does.  If they've build packs for older Firewheels, they've bound to have hit it before.

Hobby16 probably could also take a pretty good guess at what it does, but he's gone into hiding. Maybe he's working on the telemetry-project, or simply otherwise busy, he did mention sometime early August that he might be offline a lot, as internet offers too many distractions  

[esaj]

It looks like my luck is finally about to change   @Tilmann got in contact with the people from 1RadWerkstatt ( https://www.facebook.com/1radwerkstatt ), acting as a translator (apparently they don't speak very much English, if at all?), and... my custom-packs work with my current mainboard. They knew what the red wire was and what to do about it. So I modified my current mainboard a bit, changed the XT60-connector to Deans I've used in my connector-adapter for the batteries, and the wheel starts up with the new packs without error messages    But, there's a catch, they asked me NOT to disclose any detailed information about it, and as they did such a huge favor to me, I feel I must honor that request. So sorry, no pictures of the modded board.

Also, it turns out they have Firewheel mainboards for sale, or even an entire wheel without batteries, should I choose so, and the prices seem very good. They also build professional quality custom-packs for specific wheels, and apparently also any other sort of custom-pack you may want (of course as long as parts for such pack exist, so you can forget that single 10kWh pack you were going to ask for your wheel ). From what I've gathered, they are highly professional people and use only high quality cells and parts, and the prices don't seem that bad either...well, of course high quality isn't exactly cheap, but I could have got 840Wh packs (3500mAh cells) with separate BMSs specifically made to fit Firewheel for about the same price I now paid for the 768Wh, and wouldn't have had to fight with the fitting.   Had I known about them when I started this project, I would have saved a lot of time and grief just ordering directly from them, as they build plug & play -packs for most common wheels. So, if you want bigger battery packs for your wheel, or need some spare parts (I don't know if they usually sell them, and for what wheels they got them, but it can't hurt to ask), I'd highly recommend them, especially if you live in EU-area, since then you don't need to pay customs or VAT. Do note though that I'm not 100% sure if they speak English at all? Also, based on their Facebook-page I think they sell entire wheels too (probably could be fitter with custom packs, if you so wish)? There's at least their "own" wheel + 14" & 18" Gotways mentioned as having arrived... Sorry if it feels like an advertisement, but I haven't seen this kind of service yet anywhere  

As for the battery-project otherwise, I've pretty much finished the left (non-mainboard) -side of the shell. The battery compartments aren't sealed with silicone yet, but otherwise I consider it "done". And now that I could stop moping about the mainboard-issue and don't need to wait for the Firewheel factory-people to answer my emails, I got started on cleaning & piping the mainboard shell-half. Still work to do, but I got the pipes in place today:

I'll have to see how much wires I can pull through the upper pipe, but that already helps a little, as the original channel to the front was really crowded. Also, the front-side (left in the picture) battery pack will have its wiring going through the pipe straight to the front, but as I had problems pushing four Andersons with 14AWG wires through similar pipe on the other side, I might have to wire the aft-side pack through the mainboard compartment. For now, as I've only got 3 packs, and two of them are already in the other shell, I can wire it through the pipe. I'll have to re-open this side anyway later on to replace the mainboard (yes, I'll still replace it, as there's been the suspicious starting problem now and then, and I suspect that the mainboard might not last very long anymore). Another issue is all the battery wiring, although single Anderson-connectors are small, 8 of them (plus and minus for each pack) does take up space... see the connector adapter in the left battery compartment? That's 8 Andersons connected to single Deans... and that's supposed to fit on the front of the wheel with all the other wiring  At least the upper pipe saves some room, but it's still going to be tight...

[esaj]

Ok, got the mainboard-side shell mostly finished last night, and then played a couple of hours of Tetris with the wiring:

Cut off & sanded (after this picture) the bumps. 

Added one more wire channel, which turned out to be a little too shallow in the end... 

Gluing the channel in place, added some more glue on the others too (they still likely need be sealed with silicone also, the glue might come off partially and allow water to enter). There's one red Anderson-connector about center and towards the top on the picture, on top of the piece of sandpaper, they're pretty small at 8x8x25mm, but there's a lot of them...

Trying to figure out how to fit all the wiring, big mess at this point, the biggest pain is going to be the wire channel from the mainboard-compartment to the front, not only does it contain the most wires (even with original batteries), it's also the shallowest of all wire channels in the wheel... And I have to run the power-button -wire (the connector won't fit through the other channels) + 6 * 2.5mm² -battery connector wires through it + 2 * 0.75mm² - charging wires for the aft-battery. It looks like they should fit, but barely   In the picture there's still all 8 discharge-adaptor wires running to the front from the mb-compartment, but later on I figured I could bend two battery discharge-connector wires from the adapter to stay within the mainboard-compartment, as the aft-battery wires will run there anyway, no point to take them to the front and back, only need the charging wires to the front. Testing that battery connectors work with Charge Doctor at the same time.

In the end, I had to cut off the battery display holders to get more space (I wasn't planning on keeping the display there any way, as it's pretty useless). That gave a lot more space for the Andersons.

I also tried one of the voltage-displays in place of the battery display, but it's kinda hard to read, as it's sideways  And takes up a lot of space.

Made the plastic foam lid-raisers to this side too, for now I've stuck the secondary PCB to the empty compartment. After the new mainboard & 4th battery arrive, I'll have to fit it into the mainboard compartment. Could have of course done it now too, but...

Display-shunt through-bolted to mainboard metal plate. Since I won't have the battery display anymore (although it wasn't very reliable anyway), I figured I better get something to at least show me a good "guess" of how much battery is left. While the display probably isn't very accurate, I'd expect it to still be more reliable than the battery display. Of course I could have used only the voltage-measurement side, which doesn't need a shunt, but figured I'd do this on the same go. The mainboards' still on the old metal plate, I'm saving the new one for the new mainboard (unless the new mainboard comes with its own plate, didn't ask...).

The shunt itself is surrounded by a piece I cut off from a DRAM-stick protection cover, should prevent any wires hitting the bare metal parts. But also takes up space depth-wise, not sure if the secondary PCB will fit under it in the compartment, that's why I stuck it into the empty battery compartment for now, but might still move it.

Calibrating the voltage-side of the display with Charge Doctor. I use the old internal charging-connector for measuring the voltage for the display, good thing I left it there (and now it really is useful that the battery has no separate diode-protected charging pad, otherwise this wouldn't work).

I was going to cut a hole for the display at right where the Charge Doctor sits, but as the width of the shell at that point isnt enough, the display would come partially outside from the sides. While I didn't mind that, I did figure out that cutting deeply there is going to do a lot of damage to the structural integrity on a critical point, where the forces from the handle will cause most strain while lifting. As I like to use the handle, I decided not to cut a hole for the display, instead, I'm running connectors to the outside of the wheel, so the display must be encased and attached on top of the wheel. There's going to be separate outside battery case for the display anyway, so might as well rebuild that with a bigger case to fit the display too.

Pretty much all that is left for now (before I have to do again a bunch of things after the last battery pack & mainboard arrive... ) is running the connector wires for the display, getting all he wires in their correct places & using some wire ties and hot-glue to make it tidier and reconnecting the mainboard to the motor & elsewhere. Then put it together for a while, to test a little that it works in general. I won't be sealing off everything with silicone yet, as it's a huge pain to get off cleanly (with this rebuild, I've spent more than a few hours just scraping off the old silicones and cleaning the surfaces), and it (hopefully) won't be long until I have to reopen it again. Maybe something like hot-glue could work for a while, or just duct tape . If possible, I'd like to do the "final" sealing with silicone only after everything (including the last battery-pack & the new mainboard) are in place and everything works.

 

 

[dmethvin]

Nice! I know what you mean about the silicone. I have been meaning to finally shunt my wheel but keep forgetting to get some new silicone. 

By the way, there are also fiberglass-reinforced cutoff wheels for the Dremel that are a lot more durable and only a little thicker than the skinny "heavy duty" ones. I can't count how many of those I have broken when they're barely used, compared to the fiberglass ones I can wear down to a nub!

Once I get it open I think I'm going to permanently set it in sport mode and  use the sport/comfort switch to control the speaker.

[esaj]

Nice! I know what you mean about the silicone. I have been meaning to finally shunt my wheel but keep forgetting to get some new silicone. 

Yeah, from the point of sealing, it's really a good thing that it sticks so well (I've wiped the surfaces with acetone before applying and it REALLY sticks), the downside is the pain of getting it off if need be, it usually comes off in little pieces and then you have to tear the rest with fingernails/wooden stick/whatever, and it really takes a long, long time...

By the way, there are also fiberglass-reinforced cutoff wheels for the Dremel that are a lot more durable and only a little thicker than the skinny "heavy duty" ones. I can't count how many of those I have broken when they're barely used, compared to the fiberglass ones I can wear down to a nub!

Thanks for the tip, have to look for some... mine is not a real Dremel, but a cheap clone from Biltema. The discs that came with it actually have too small holes for the screw to go through (so I always have  use a round small file that I push through the hole and turn a bit to make it bigger), and they shatter really, really easily. If you're careful of not bending the disc even the slightest bit (which usually occurs when the disc gets suddenly stuck in a deeper groove), it actually lasts for a good while, but even the slightest wrong move causes it to break.

Once I get it open I think I'm going to permanently set it in sport mode and  use the sport/comfort switch to control the speaker.

Oh yeah, forgot to mention that, I did this also (shorted the wires so it's always in sport-mode), the mode-switch isn't currently used for anything. Kill-switch for speaker could be a good use, at first I thought I'd use it to switch on and off the voltage display but now that I'll make it separable, no need for that.

[EUC Extreme]

I use this for sealing. It is very flexible and weather-resistant. Flexibility allows it remains even in heavy exertion closed.
However, the leaves off in large pieces, remove the hand.

This photo was the only one I found.

http://www.tori.fi/pohjois-pohjanmaa/Polymeeri_liima_ja_tiivistemassa_19658735.htm

 

 

[esaj]

I use this for sealing. It is very flexible and weather-resistant. Flexibility allows it remains even in heavy exertion closed.
However, the leaves off in large pieces, remove the hand.

This photo was the only one I found.

http://www.tori.fi/pohjois-pohjanmaa/Polymeeri_liima_ja_tiivistemassa_19658735.htm

I think I have that exact product in the garage...

Edit: Yup, had that in the garage, except its "Best before"-date has gone a year ago, plus it says to store it between +10...+25C, and it's been in the garage which has no insulation or heating, so it has probably frozen over the winters...

[EUC Extreme]

If it comes out of the tube, can be used.

[esaj]

If it comes out of the tube, can be used.

I'll try if it works... for now, I'm stuck with getting the motor cable in place, it was a pain the last time I tore the wheel down, and it's an even greater pain now

I soldered the motor power wires back to the mainboard (they didn't have any connectors before either), which of course makes the motor and the mainboard inseparable. Then I tucked all the cabling in, closed the compartment and tested that the wheel starts:

Also the display works, don't know how much off the amperage is, but at least it shows "something"   Then comes the painful part:

I had to leave the other battery compartment open at first to get the motor wire in place (the plastic parts of the lids that are supposed to cover the wire channel between the compartments come on top),, and then close the second compartment. No problems there. But actually getting the motor cable in correct position is the hard part:

There's not much room, I can barely put my hand in there and try to nudge the cable to go right, but no success yet (been at it for about an hour and a half, taking the whole thing open once, and then doing it again, no luck).  The wire gets stuck between the shell and the screw holes that are for attaching the shell-half to the pedal-frame. The mainboard-compartment is full of wires and the motor-wire is fairly rigid, so I couldn't get it to push in there either. Can't pull it from the mainboard compartment, as it must be closed before I can put the motor/tire/pedal-frame -assembly in place. I have the small metal piece seen in the earlier picture between the battery-compartments, that I made during the first tear-down, and the last time I put it back together I could get the wire to bend enough so I could get it under that and stay there, but now it just wouldn't go right... Any tips? I was thinking of cutting a small hole in the outside on the mainboard compartment (it will be covered by the black round-part later on), so I could pull the cable from inside the compartment. Actually already took it apart again just for that, but I guess I should eat something first, haven't had but coffee today yet, and it's closing 11PM  These are the kind of situations where I really wish all the compartments were accessible from outside...

[esaj]

Firewheel PITA (as in Pain In The Ass) MK 1. 576Wh (3 * 192Wh) LG MH1 -batteries. Non-shunted. Non-sealed. The display fitting is just temporary. No idea if it's going to explode, catch fire or cut-out when I try to ride it. The street lights are out, as it's over midnight on a night against a normal weekday, so I can't see shit outside. If I don't edit the last part of the build process here by next night, I'm either dead or in a hospital. Laters...

[esaj]

Ok, no explosions, fire or cutouts. Nice ride otherwise, but pretty chilly, and I didn't take any gloves, so my fingers were freezing when I got back    Started with 57.2V at batteries (normally fully charged batteries original on FW were around 67.2-67.8V), ended with 54.3V, jumped back to 54.5V after a while, but looks like it's staying there. Got the battery warning twice when returning home, both times while pushing it harder uphill (don't remember what the voltage exactly showed at those points when at rest, 53.something). Not sure if the battery warning is triggered by voltage drop (didn't see how low it went there) or if there's other logic behind it, at lowest I saw the display dipping slightly below 51V earlier going up hill, and stood around 54V when riding on level or slight inclines. Not sure if the display shows it exactly correct though, it was off by 0.1V compared to Charge Doctor when I got back, the value could wander a bit with temperature, but Firewheel definitely lets the voltage drop way below the around 56V which has been where the charging has usually started. Now the voltage at end wasn't that near to 56V. Not going to leave them to charge overnight, I'll charge them tomorrow while working, as I don't want to risk anything happening while I'm asleed (I haven't charged the new batteries ever yet). Probably the batteries weren't "really" empty, just the voltage drops on the climbs caused the warnings (didn't ride much after those, but a little while).

The calibration of the current-display is probably off, I saw values of 53V + closer to 40A accelerating uphill, which would total around 2kW. While that's not entirely impossible (if comparing to the values I've got out from vee's MCM2S), I still doubt if it was that high. Since I have no trustworthy way of calibrating it now, don't know what I'll do with it, probably just let it be. Btw, I had to take off the bolts holding the shunt in place before I finally got the wheel assembled, as one of the nuts was scraping against the tire, and it's now just hot glued to the plate. If that won't hold, well, shit. It's just the negative wire feeding power from all the batteries, so not that important, all the power will cut instantly.   Although, the wires are both screwed & soldered to the shunt, so if it's dangling there, they still shouldn't come off easy... maybe?

In total I rode about 7-8km, didn't have the bike computer attached, so no idea on more exact distance, but fairly familiar short route for me. Only paved roads, no off-roading. Had to take the first and last about 2.5km long parts slower, as there were no streetlights, and while I can see some way ahead with Firewheels own headlight, it's not enough for high speed riding. Kept it around the first warning, so roughly 18-20km/h on those.

Once I got to the part which has a long straight next to a freeway, and which has street lights on, I lifted my speed above the second warning (25+km/h) and pretty much kept it there for the next few kilometers. Never took it all the way to the "Take care"-speed, as I have no idea on the possible cut-out behavior. But of course the BMSs could still cut power with low enough voltage. Had a funny thought earlier, now that the same wires actually work both ways (discharging and charging), a single shunted pack actually could "revive" the rest, if they went to latched low-voltage protection mode, if it can give them enough voltage (so if some packs went dead during acceleration, they could still be reset after stopping, if the voltage from other pack(s) rises high enough to release the protection).

So, I got no speed- or distance-data, as I didn't have the bike computer. Performance-wise, I'm pretty sure there's a difference, but needs more testing. I didn't get the "sluggish" feeling at any point that usually comes before the battery warnings, or at least I didn't notice it. Even at low voltage, taking it above 25km/h was no problem going up somewhat steeper uphill, accelerated with ease (that's where I saw the voltage dip, accelerating above the second warning on uphill). I didn't dare to really let it rip due to the combination of low battery, poor visibility, rabbits (I actually almost hit one while peeking down at the display ) and coldish weather (+8C). Can't really tell if it has faster acceleration now either, as I didn't do any really fast accelerations.

Will have to do more testing tomorrow evening, when I've got fully charged batteries... also should get a somewhat realistic total capacity value now that they should be closer to empty.

 

[esaj]

Charge over 5hrs 45minutes, voltage measured from batteries before starting was 54.7V (although it seems the charger still starts at 56V):

Left axis: voltage (volts) and charging current (amps, 100mA, so 20 on y-axis = 2A). Don't know what the drop around 220 minutes is, maybe I kicked the stool on top of which the Charge Doctor sits while passing by. Right axis is the watt hours charged into the batteries, the final value was 579.7Wh (I didn't wait the charge current to drop all the way to 0, but it was somewhere in 70-80mA range when I stopped it). So barely over nominal capacity, though the batteries weren't all the way empty... the wheel did give me the battery warning twice accelerating up hill, but on both occasions, I could just restarted the wheel and continue, when it's REALLY empty, the warning starts right away after restart and it's impossible to ride anymore.

Final voltage was 67.8V (4,2375V / cell). This was tested on the original Firewheel 2A -charger, I do have a 4A charger from a Gotway, but the Charge Doctor can't handle current above 3A.

[Jason McNeil]

Really cool!  So that 570Wh is just the three packs right? Do you know how low the Firewheel allows the voltage get before forcing you off? Will you be doing limit testing (with safety gear  ) with the new packs?  Using the services of https://www.facebook.com/1radwerkstatt, I wonder how many mediocre eWheels can be transformed into super-Wheels by simply upgrading the batteries with the best available. As always, you don't do things by half-measures, nice job!! 

[esaj]

Really cool!  So that 570Wh is just the three packs right?

Yes, 3 * 192Wh nominal, so 576Wh nominal. Probably I could get it closer to 600Wh charged (as the cells are brand new), if I really rode it to the battery-warning so many times, that it starts right away when the wheel is restarted, and waited until the charging current drops all the way to 0mA. EDIT: The original 2 packs (264Wh nominal) have about 400+ cycles and still charged to 268.3Wh the last time I checked a few weeks ago after riding them all the way to empty (so empty that it starts the warning right away when powered).

 

Do you know how low the Firewheel allows the voltage get before forcing you off? Will you be doing limit testing (with safety gear  ) with the new packs?  

Last night I did that quick 7-8km lap, and the lowest I saw while climbing a hill at around 25km/h was around 51V. I didn't check the voltage (which I of course should have) when the battery warnings occurred, but I have a feeling that it probably happens when the voltage drops to 50V. Don't know if I will ride them all the way to empty tonight (I'll just attach the bike computer, do a couple of lift-tests to see no-weight max speed and then I'm off), as it will get dark in an hour or two, plus it's chilly, and probably won't test the max speeds yet, as I don't know about the high speed cut-off -behavior (although, there is a possibility that it's gone even with non-shunted packs, as I suspect the now-bypassed BMS-input wire going to the mainboard had something to do with it). The packs have a BMS with over discharge protection voltage of 2.9V +- 0.05V, so if it's the highest (2.95V), I better not let it drop below 2.95V * 16 = 47.2V

 

Using the services of https://www.facebook.com/1radwerkstatt, I wonder how many mediocre eWheels can be transformed into super-Wheels by simply upgrading the batteries with the best available. As always, you don't do things by half-measures, nice job!! 

As long as the mainboard can take it, I'd expect most wheels to benefit from multiple high-discharge packs, not only range-wise but for power too. Still wish I had found out about 1Rad Werkstatt earlier, then I'd probably have had the packs already by the start of my vacation and would have avoided a lot of trouble with installing them...  

[dmethvin]

I think I have that exact product in the garage...

Edit: Yup, had that in the garage, except its "Best before"-date has gone a year ago, plus it says to store it between +10...+25C, and it's been in the garage which has no insulation or heating, so it has probably frozen over the winters...

For a project a few years back I had some old silicone that was about 3 years past its expiration and tried to use it. It came out of the tube but was a little grainy rather than smooth. Also it didn't seem to "set" correctly, I came back the next day and some of it was still runny. I had kept it in a cool basement so it hadn't even been exposed to any temperature extremes. I had to scrape it all off and clean it, what a mess. So I wouldn't take any chances.

[esaj]

Well, that didn't go according to plan... I did a couple of quick lift-tests (34.something was the highest I could get, it's difficult to accelerate the tire slowly so it doesn't go over the cut-out speed before the bike computer can record the max speed), and then headed out. Full gear, wheel on the ground and turned on, bike computer works, ride lights on, display on... 67.0V. Guess it must have settled a bit since it was about an hour after the charging. I get on, roll up the street, and then turn right to go down the hill.

I'm going down the hill, keeping the speed fairly steady, below the first warning so the sound won't annoy the neighbours, decelerating just a bit before a larger bump... whoa! I just feel the wheel go limp, land on my feet and take a few running steps down the street, seeing that the wheel is rolling ahead of me, all lights out. Shit, don't hit a car, don't hit a car... after about 10 meters it tumbles into some bushes. At this point I was already pretty sure of what had happened: over-charge protection due to full battery & regenerative braking. If only I had thought about that roughly 30 seconds earlier...  

The display battery case is in the middle of the sidewalk, and I fish out the wheel from the bushes. A couple of neighbours come out to their balcony, as I'm in the front of their house and ask what's the matter and what that noise was. "I fell on my wheel" (well, technically I didn't fall, but I wasn't going to start explaining it to them). "What wheel, where is it?" "Right out there, leaning on your fence" "I can't see it, is it a motorcycle or a bicycle?" Btw, the Finnish word "pyörä" is used as a short form for "polkupyörä" (bicycle) and "moottoripyörä" (motorcycle), but it also means a tire or a wheel. "No, it's a unicycle" "Oh, that thing"... then they just asked if I'm alright, and I told them I was fine and said I was sorry for the noise. So no biggie. Time to inspect the damages.

The bike computer on top of the handle was smashed, the display's filled with the black liquid. The voltage/current display connector was still intact and in place, and the battery case had ripped a bit from one side, but seemed otherwise ok. No idea where the display itself was. No visual damage on the wheel. Tried turning it on. Everything works, motor runs and balancing works, front & tail light work. I tried to ride it up the hill a bit. No problem, but didn't want to try downhill again. I looked for the display for a while, but didn't find it until a bit later, after I went home to get a flashlight. It was in the middle of the bushes where the wheel had landed, the connectors had just released and it still works just fine. After I got it to wheel, and put it on, it showed 67.3V. So even more sure that it was the over charge protection.

So, no speed or mileage data. After I rode some circles around the street to get the voltage lower, I just picked the wheel up and carried it downhill. In the end I did go ride around only for about an hour, didn't try any really fast accelerations and was paranoidly alternating watching in front of me and at the voltage display every time I descended any hills (and there are a lot of them here)  I did go off-roading at the start and the end too (that was interesting in pitch black, potholed and bumpy paths). In total I rode maybe about 15-20km, mostly testing uphill climbing (yes, it definitely is more powerful at hill climbing now), but was pretty careful for the first half an hour or so, as I was constantly wary should the wheel shut down on me again. Never happened, and coming back I was going uphills above the second warning speed and riding on the flats a little below it. The voltage as already low enough for me to not worry about it that much, but I still did check it going down any hills.

Batteries were at 64.4V at the end, but no idea on the mileage. I'm not going to charge it up right now (that way I would know how much Wh I spent, but not much use without knowing the distance), as I'm probably not going for another trip tonight, and even if I were, looks like charging it to full voltage is a no-no, unless I figure out a way to take the voltage down a notch or always carry it downhill    As I can use the charge-port for discharge too, maybe I could put something there to "eat up" the excess charge...

[esaj]

Here's where it happened, one of the cars in the yard was actually closer to the street, and the wheel missed it by about a meter or a meter and a half. Phew.

 

[Tilmann]

Glad to hear, you didn't harm yourself and nothing essential was broken. Looking at the scene, that could have gotten garstly expensive if the wheel hit something more valuable than the bushes. Do you have insurance? I am presently trying to buy liability protection for my wheels.

So, I guess you discovered another deficiency separating us from the perfect EUC: BMS overcharge protection can bring you down without warning. It's probably not trivial to solve. Naive idea: When battery voltage is to high to consume the energy generated by braking, a switch should send it to a power resistor and immediately alarm you in an unmistakable way (as the resistor will overheat quickly if you carry on). At first I was thinking of a 1000 Watt halogen bulb (which would be unlikely to go unnoticed ), but those would probably be destroyed by the vibration really fast...

Any more realistic suggestions?

[esaj]

Glad to hear, you didn't harm yourself and nothing essential was broken. Looking at the scene, that could have gotten garstly expensive if the wheel hit something more valuable than the bushes. Do you have insurance? I am presently trying to buy liability protection for my wheels.

I do have liability insurance, but I think the insurance company would squirm away from paying anything based on the fact that I was using an illegal vehicle  

So, I guess you discovered another deficiency separating us from the perfect EUC: BMS overcharge protection can bring you down without warning. It's probably not trivial to solve. Naive idea: When battery voltage is to high to consume the energy generated by braking, a switch should send it to a power resistor and immediately alarm you in an unmistakable way (as the resistor will overheat quickly if you carry on). At first I was thinking of a 1000 Watt halogen bulb (which would be unlikely to go unnoticed ), but those would probably be destroyed by the vibration really fast...

Any more realistic suggestions?

Well, it never did that with the original batteries, I don't know where the extra energy then went (I've gone down that hill as the first thing on my trips with batteries filled to the brim pretty much every time I've ridden), but I suspect that there must be something in the mainboard that handles it. Now that the BMS-input is bypassed, it doesn't know the situation, so it probably just tries to charge the batteries in every situation   And yes, I do need something there that can burn off the excess and prevent the shutdown, or otherwise I can never charge the batteries except to about half full to keep the voltage down... Or ride lots and lots of circles before going forward, or always carry the wheel downhill. I just mailed my battery-guy for the datasheets of the BMSs, don't know if those help much, but at least they should list the over charge voltage (which probably is 4.25V, given that the voltage went to 4.2375V per cell while charging, and it didn't cut the charging).

[Tilmann]

I do have liability insurance, but I think the insurance company would squirm away from paying anything based on the fact that I was using an illegal vehicle  

yep, that's what I meant. I am also afraid, that personal liability insurance will bail out if my EUC causes a damage. In Germany, small motorbikes and scooters (max. 40 km/h I believe), have to carry a small license plate issued by the insurance companies and sold in combination with vehicle liability coverage. I thought, you can buy these only for vehicles approved for street use, but that does not seem to be the case. I ordered one for my Msuper, but don't have it yet. Cross fingers...

BTW: the guy, who found out about this option is Alex Segmüller (Berlin). This guy :

 

[esaj]

Looks like the BMS-cutoff issue is pretty much confirmed now, I charged the wheel to full (20mA charging current when I disconnected the charger), so it was "hot off the charger", at 67.8V. By the time I took it out and walked to the top of the hill, the voltage had dropped to 67.3V. I then pushed the wheel down the hill with my hand, and slightly tilted it back at the same time... pretty much instant shutdown. The wheel goes limp, the LED on the power-button stays on, but there's no usual "Please restart the unicycle"-message playing, so my best guess it that the firmware crashes/gets stuck due to momentary power cut. Restarting it, it starts up just normally. Tried it a couple of times and shot a short video of it happening (the voltage/current display looks funny, because in the video you see how it actually refreshes itself one number at a time ):

EDIT: In the BMS specs, it's stated that the overcharge protection triggers at 4.2 +- 0.05V (so 66.4 - 68V). As the charging voltage goes all the way to 67.8V, and there's no cut-off, I think it's somewhere near 68V. As the voltage raise is very momentary, I expect that it releases pretty much immediately after the power is cut, and starts giving power again, but the wheel has already shutdown/crashed at this point... 

[Chriull]

...

Well, it never did that with the original batteries, I don't know where the extra energy then went (I've gone down that hill as the first thing on my trips with batteries filled to the brim pretty much every time I've ridden), but I suspect that there must be something in the mainboard that handles it. Now that the BMS-input is bypassed, it doesn't know the situation, so it probably just tries to charge the batteries in every situation   

This was the extra cable from the original BMS to the mainboard, you mentioned sometimes somewhere?

Could be, that this cable gave an alarm, that the BMS is about to shut down due to overvoltage (or the batteries are full) so that the mainboard could switch from regenerative breaking to active breaking (consuming accu capacity)?

 

And yes, I do need something there that can burn off the excess and prevent the shutdown, or otherwise I can never charge the batteries except to about half full to keep the voltage down... Or ride lots and lots of circles before going forward, or always carry the wheel downhill. I just mailed my battery-guy for the datasheets of the BMSs, don't know if those help much, but at least they should list the over charge voltage (which probably is 4.25V, given that the voltage went to 4.2375V per cell while charging, and it didn't cut the charging).

Burning of the excess power from regenerative breaking does not sound like an easy solution. You'd need to sensor the direction of the current (charging or discarging) and according to that burn the power from the motor without discharging the accus at the same time. Could be quite some cooling needed, too.

If i remember right with the extra cable from the original BMS and this is really a signal for the mainboard to choose between regenerative/active breaking this could be your way to go? So as first step try to set the mainboard permanent to active breaking and if this works design a small circuit which gives the right signal to the mainboard.

How do other wheels solve this? With some ninebots this was reported to happen - with some not, imho one got the mainboard/whole wheel exchanged and then there was no shutoff going downhill with a full accu pack anymore? And the ninebot has no extra wire from the BMS to the mainboard. So they sensor the accu state directly on the mainboard and decide to charge the accu or consume for breaking?

ps.: i hope there is something like active breaking for ewheels - if not the whole post makes no sense

.. But on the other side - if active breaking does not exist every wheel would have to shut down due to the overvoltage protection from the BMS going downhill once the accus are full?

[esaj]

This was the extra cable from the original BMS to the mainboard, you mentioned sometimes somewhere?

Could be, that this cable gave an alarm, that the BMS is about to shut down due to overvoltage (or the batteries are full) so that the mainboard could switch from regenerative breaking to active breaking (consuming accu capacity)?

That's my suspicion, of course I don't know for sure. I've got a new mainboard on the way, the newer models apparently no longer have the BMS-input wire at all, so hopefully this has been somehow remedied in those, otherwise I'm either stuck with never charging them all the way to full, or hoping that someone knows how to build a circuitry to either limit the voltage going back to the battery or otherwise prevent it...

 

Burning of the excess power from regenerative breaking does not sound like an easy solution. You'd need to sensor the direction of the current (charging or discarging) and according to that burn the power from the motor without discharging the accus at the same time. Could be quite some cooling needed, too.

If i remember right with the extra cable from the original BMS and this is really a signal for the mainboard to choose between regenerative/active breaking this could be your way to go? So as first step try to set the mainboard permanent to active breaking and if this works design a small circuit which gives the right signal to the mainboard.

Yeah, I wouldn't even mind if the whole regenerative braking was entirely gone, of course that would probably drop the maximum range, but it's a better option than always having to monitor the charging to stop it before the voltage gets "too high"

 

How do other wheels solve this? With some ninebots this was reported to happen - with some not, imho one got the mainboard/whole wheel exchanged and then there was no shutoff going downhill with a full accu pack anymore? And the ninebot has no extra wire from the BMS to the mainboard. So they sensor the accu state directly on the mainboard and decide to charge the accu or consume for breaking?

Yeah, I don't know if any other wheel has the input-line from BMS to the mainboard, so they must have some current-sensing scheme going on and then revert to other type of braking or burn off the excess energy somewhere.

ps.: i hope there is something like active breaking for ewheels - if not the whole post makes no sense 

.. But on the other side - if active breaking does not exist every wheel would have to shut down due to the overvoltage protection from the BMS going downhill once the accus are full?

I do think there is some sort of "active braking", or at least regenerative & plugging-type & dynamic braking, see here:  http://www.electrical4u.com/what-is-braking-types-of-braking-regenerative-plugging-dynamic-braking/  I'm not 100% sure if I understand it correctly though   With my 14" generic, the power-braking is much less effective when the batteries are more depleted, so there must be a way to use the battery power for braking too. I once asked about this from hobby16, and he then had asked some DC-motor driving specialist at work, who had said that the braking can be done by using external power, not just regenerative. Maybe they reverse the switching of the poles, so that they start to attract/repulse in "backwards", goes way over my head though  

Other sources I've been going through related to this (but not much help yet, at least partially because my understanding of electronics isn't that good):

http://www.societyofrobots.com/robotforum/index.php?topic=15105.0  (This actually talks about limiting voltage during regenerative braking, 22V motor can generate 300V during regen?! Some images are missing though... )

http://electronics.stackexchange.com/questions/56186/how-can-i-implement-regenerative-braking-of-a-dc-motor

But these mostly just speak about how to make the braking regenerative, not how to stop it  

Edit: Now that I'm fairly sure that the overcharge is the issue, I think I'm going to bother the 1Rad Werkstatt guys again... Hopefully they would have some ideas, although not so sure how keen they are to keep helping me and answering questions, as all I've done is order a mainboard from them

 

[Chriull]

That's my suspicion, of course I don't know for sure. I've got a new mainboard on the way, the newer models apparently no longer have the BMS-input wire at all, so hopefully this has been somehow remedied in those, otherwise I'm either stuck with never charging them all the way to full, or hoping that someone knows how to build a circuitry to either limit the voltage going back to the battery or otherwise prevent it...

...

What have you done with the BMS input cable? Connected it to ground ?

The BMS and the Mainboard imho should only share the ground line. So the way to "communicate" for them would be that the cable has a pull up resistor on the mainboard and the BMS shortens it to ground or leaves the cable "open" (unconnected) to signal the two states (accu can be recharged/accu is full)

Or did you mention in your post that the wheel did not start at all when you left the cable unconnected?