RagingGrandpa's charge-and-ride Greenworks backpack

[RagingGrandpa]

My charge-and-ride backpack:

 

 

Conclusion:
 

This was a pain to build.

It's useful now that it's complete.

A 58.8V (14s) pack would have been better...

But I should have just purchased the ~$800 KS-Power device.

 

 

The good:
 

It adds about 20 miles of range, in my typical on-road riding.

It also lets me ride at ~30mph for 60 minutes straight, something that I couldn't do previously because of the slowly decreasing EUC voltage.

Efficient: Since it's the same pack voltage as my 84V EUC, the boost converter doesn't have to work very hard, doesn't get hot (no fan needed), and doesn't waste much power. Also, the backpack can be depleted to a much lower voltage than the EUC pack would allow (I prefer ~2.8V/cell), meaning you can make full use of its capacity.

Flexible: The backpack will deplete itself in about 90 minutes of riding. After it's empty, the EUC will still have ~50% SOC, which gives me the option to stow the backpack (in a car, etc) and ride a final hour or so without wearing a big lump on my back. If I don't need those last few miles of range, I can stow the backpack earlier in the ride.

Affordable: the Greenworks backpack is available new for about $500, then add a DC-DC, charger for the backpack, and supplies: this is about a $600 kit.

Generic: no modification to the EUC was necessary. (84V MSX with 4-pin charging.)

Relatively safe: (...for my 84V MSX; other wheels not analyzed.) Single-point failures during use cannot cause the EUC to cut-out, nor the EUC cells to burn.

 

 

The bad:

 

Can't ride-then-charge:

There is no buck converter, and the backpack is the same voltage as the EUC, so I must connect and use the backpack at the beginning of a ride. If I use the EUC without the backpack, I cannot connect the backpack later in the ride (current would be unregulated). This is one reason a 14s pack would have been better.

Of course, I can "take breaks" and stop riding whenever I please, as long as the backpack remains connected.

Mfg labor:

IF you knew what to do and had all the parts ready-to-assemble, this is easily 4 hours of labor. Add "learning about it" and some time for shopping, fine-tuning, etc and a week's worth of free time will disappear quickly.

15lb weight.

Riding "unencumbered" is more fun...

 

 

Details:

Pack:

Greenworks GL900 "82V"

LG 18650 HE4 2500mah cells, 20s 5p

Remanufacturing it with modern 3500mah cells sounds awesome for performance, but is practically impossible due to the very secure welded-and-glued design.

The backpack itself is built to fall down the stairs, with lots of attention paid to mechanical durability: cells mounted in frames, crush space built-in, etc.

It was splash-resistant but not fully gasketed. I added more holes for wire passages and my converter is exposed, so rain, puddles, and tire spray must be avoided.

 

DC-DC converter:

Yeeco BT900W 900W boost converter, 0-60V input.

^ modified to isolate its logic power source from the input voltage: logic is now powered by a separate connection to cell 6 (25V) thru a 5 ohm resistor to avoid inrush damage. (Learning this lesson took blowing up two converters...) With this new current limiting resistor in the logic supply, I might be able to power the whole thing from 84... potentially with more component changes to keep things within their ratings at 84V... but since it's working now, I don't feel compelled to dig deeper.

I added RTV to secure the jiggly wire-wound inductor.

 

I was unable to find an off-the-shelf DC-DC rated for 84V input. There is a variant of this Yeeco model that is sold promising a 12-90V input, which I ordered. It was identical to the "8-60V" model though (errors / lies in the online description). The input capacitor is rated at 100V, so the high-current path should be fine at 84.
More boost converters are available for <60V input- a 14s pack would have made this a lot easier... but they still may blow up during power-up with the huge current that our big lithium packs can provide.

 

Almost all of the ebay/amazon boost converters 400W-1800W are hobby-grade devices. Mine had sloppy manufacturing: dented capacitors, wobbly unsupported wire-wound inductor, cold through-hole solderjoints. The SMT soldering seemed fine. If you use one, be prepared to inspect it thoroughly, retouch solderjoints, and test it with current-limited devices (not battery packs!) before use. Plan for failures and troubleshooting regardless of which model you choose.

 

Backpack BMS:

The backpack has a microprocessor-controlled BMS, which includes cell temperature sensors and is capable of engaging balancing resistors, interrupting the pack output (except that I bypassed this), and interrupting the charging input (if using a Greenworks charger).

The BMS is not 'active' unless woken by its 'command' wire. Pulling this wire to +VBatt will wake it up, but more complex signaling is needed for it to become fully functional. Balancing does not occur when the command wire is pulled up this way.

I could not use the BMS main (negative) output with the boost converter- the BMS would set an error code "erratic tool voltage" and interrupt the output. So, I bypassed the BMS for the main output to the boost converter.
 

Recharging the backpack:
This was the saddest part of the story... I tried three different Greenworks chargers (including the rare GC900 charger which is only intended for use with this backpack) but could never get the recharging session to begin. All I can conclude here is that my backpack's BMS communication is damaged somehow. Perhaps I've burned out something by connecting the command wire directly to 84V with my SW2; or perhaps it was damaged by the previous owner (my backpack was sold as used with no guarantees).
Because I can't get the charging mode to activate, the BMS won't perform balancing. Ugh. To remedy this, I retrofit some active balancers to the pack, which are nice because they will balance the cells regardless of voltage, making balancing possible without need to charge all the way to 4.2V/cell.
And so I bypass the BMS to recharge, using a typical 'dumb' 84V CV-CC Gotway charger. This method is risky because faults in balancing or faults in the charger could allow cells to go over-voltage. Leaving BMS monitoring active using SW2 in my harness enables a beeping alert for this condition, but will not interrupt the charging. So I only recharge while I am within earshot. Fussy.

 

Short-circuit:

I was thrilled to confirm that shorting my 84V MSX charging port while riding the wheel is safe. My Gotway BMS will 'trip' and interrupt the charging input for excessive current flow, and the EUC continues operating normally. (My MSX packs were manufactured in 2019; other models may not be protected, so don't take my word for it.) Short-circuits in the boost converter or cables will open fuses in the backpack. At last resort, the backpack cells are protected by a 40A 250V soldered ceramic fuse at their only positive output.

 

EUC overvoltage:

This is prevented by Gotway BMS- if any cell exceeds 4.25V, the charging input is interrupted. It could happen if I were to sustain EUC braking while the EUC pack was already full- it would interrupt the backpack, which is good, and wouldn't let the backpack flow current again until the charging plug is cycled. I don't think cells runaway until after 5V, so this seemed safe to me.
 

Backpack undervoltage:

The boost converter is adjusted to shutoff when the backpack falls below 56V.

The backpack's BMS can be 'active' (using my SW2) during the ride- it will start beeping if any single cell goes undervoltage... but the beeper isn't loud enough to hear while riding fast (it's more quiet than the EUC beeper). This opens the possibility that if the converter became faulted or misadjusted, and the beeper was ignored, it would overdischarge the backpack and damage its cells. I plan to inspect voltage each time the backpack is depleted to verify the shutoff is working correctly; and the BMS should alert me during recharging if cells are truly damaged. "Backpack fire while riding" doesn't sound disastrous- just pull over and take off the backpack; fires indoors during recharging concern me more.

 

Backpack overcurrent:

Normally, current is limited by the boost converter's own logic. I set mine to output 4A (when the backpack is depleted) thru 8A (when the backpack is fully charged).
But because the boost converter's output voltage can never be lower than its input, user errors like connecting the backpack when the EUC is at a lower voltage will result in unregulated charging of the EUC. In this case, I expect the 15A replaceable fuse in the boost converter to melt.

 

Alternatives:

The elegance of a 100V backpack used with an unmodified 900W AC charger intrigues me... because it avoids the custom converter hassle and lets you carry one charger to serve 3 roles (AC recharge backpack or EUC; DC charge-and-ride). Maybe an AC charger rated for 100VAC would still function at the 72VDC 'brownout' condition (24s @ 3.0V/cell)? If I was using a 100V wheel, I'd investigate, especially as the KS-Power is limited to 84V.

 

Reference materials:

https://photos.app.goo.gl/HB9NXyW5P2EdLDso7

https://drive.google.com/drive/folders/1QuAEVpmo42wd5oesPe8hk3nEp1Djvz8Z?usp=sharing

https://euc.world/tour/590376164096239

 

The internet helped me; hope this helps you.

 

Edited October 7, 2020 by RagingGrandpa

[meepmeepmayer]

Wow interesting!

[Murdomeek]

6 hours ago, RagingGrandpa said:

But because the boost converter's output voltage can never be lower than its input, user errors like connecting the backpack when the EUC is at a lower voltage will result in unregulated charging of the EUC.

What?  This seems like a pretty big drawback.

Are you able to get another converter that supports this so you can hook on/off the backpack at any time regardless of the voltage on the backpack or EUC

[VolcanoSurfer]

On 9/21/2020 at 3:18 AM, RagingGrandpa said:

All I can conclude here is that my backpack's BMS communication is damaged somehow. Perhaps I've burned out something by connecting the command wire directly to 84V with my SW2; or perhaps it was damaged by the previous owner (my backpack was sold as used with no guarantees).

Hi RagingGrandpa!  I was stoked to find your post while doing some research about this battery pack, especially in an EUC forum!  Too bad to hear about your issues with the onboard bms balancing and charging though…

I recently got my hands on one of these BMS units that was pulled from the same type of backpack battery.  I got it from a guy who had a few of these packs that he tore apart to salvage the cells, and I’m supposed to report back if I have any luck waking my board up with a new set of cells.  

Im wondering if you might have any tips as to what is worth trying and what I should avoid while I attempt to wake my BMS up.

Mahalo   🤙

Eric

[RagingGrandpa]

@Azzias and I had a nice correspondence on it in the earlier thread: https://forum.electricunicycle.org/topic/19450-do-you-use-a-backpack-battery/?do=findComment&comment=328379

But neither of us have found a way to make the Greenworks BMS perform balancing, other than connecting it to a real Greenworks charger