V8 2nd battery vs upgrade?

[alcatraz]

The battery pack has its own bms but I don't trust it to balance the cells. Should I observe an imbalance I have balance leads to balance the battery. 

It's not done on every charge. It's a bit troublesome like you noted as I'm doing it through 4S balance leads.

I'm not an RC enthusiast with a garage full of equipment/transformers/chargers. I just have an apartment and I want from time to time (yearly?) to balance the battery. I bought one high precision hobby charger that can arrive at 4.2v with 10mv precision. What I do is I charge the battery with my wheel charger to about 95% lets say, then using an egg timer I charge the five 4S ports with my hobby charger for 10 minutes each and thus I get all my groups at 4.200V. 

Important to note when using multiple small balance leads is that you can't balance them all at once. You have to do them individually or you'll short the battery. I played with the idea to create a setup that can do them all at once but it's not worth it. Their main use is to monitor the voltages, and occasionally balance. The better the pack the less you need to balance.

Edited August 28, 2019 by alcatraz

[alcatraz]

On 8/27/2019 at 4:29 AM, leadfeathers said:

Is the original pack in 20s2p?  I'm confused about the 4s balancing part.  In 4s, your voltage is 14.4V/16.8V (nominal/peak).  I thought that for 84V, you'd be looking at a string of 20 or 24 batteries (not sure which..).  Wouldn't you need a BMS that is 20s to handle balancing along the entire string?

I'd love to use a 20s hobby charger or anything larger than 4s. The problem is that 4s is the largest that is also cheap,tiny and really high precision. I got it for like 12usd.

It needs to charge through the balance leads only. That requirement excludes a majority of hobby chargers.

Also because a charger that uses the balance lead only needs to have separate charge circuits for each cell it supports, it becomes either less precise or more expensive as it grows. I got a 3s one I intended to use first but that one had different end voltages by a lot. Like one 4.23v and one 4.21 and one 4.19. It's useless. Precision is first priority. Luckily there is a very good 4s one. I think the seller tested it to arrive at 4.201V on all groups. Sweet.

Edited August 28, 2019 by alcatraz

[uekarashi]

On 8/3/2019 at 7:19 PM, Rywokast said:

that sounds like hell tbh.. the batteries aren't meant to be swappable and opening up the wheel while out on a ride?? it sucks enough when you're at home can't imagine doing that outside all the time.. 

@hyperair has literally done this on a V8. It was not a big deal, at all. We stopped for about 15 minutes and it took all of one Philips screwdriver in terms of tools.

[uekarashi]

On 8/3/2019 at 7:07 PM, Mono said:

The vid only shows how to remove the side cover which is IMHO the smaller part of the work to remove the battery.

 

 

It's actually most of the work.  After which it's a couple of screws, and a couple of connectors.

[hyperair]

1 minute ago, uekarashi said:

It's actually most of the work.  After which it's a couple of screws, and a couple of connectors.

Yeah, the full steps to remove the battery are:

• 2 screws on the outer shell near the right pedal
• Twist the cover off
• Remove 2-4 retaining screws from the battery (4 in the old model, 2 in the new model)
• Disconnect 2-3 connectors (XT60 + 7-pin JST connector in the new model, XT60 + 5-pin JST connector + 2-pin JST connector)
• Slide the battery out

Then just do the thing in reverse to reassemble. Among all the wheels I've seen, only the Inmotion V8 had "swappable battery" on their marketing material, so don't judge the V8 by your experience with other wheels in this regard.

[hyperair]

On 8/21/2019 at 12:15 AM, h3X said:

Why can't you use a normal supower or similar bms?

You can, but you lose the safety feature where the BMS gracefully warns the wheel to tilt the rider off and turn on warning indicators (as opposed to a sudden power-cut) when a cell has dropped below critical voltage. It's a very nice feature, and I'm glad that my V5F and V8 didn't drop me on my face when the batteries fell severely out of balance.

[hyperair]

On 8/27/2019 at 5:41 AM, atdlzpae said:

Yes, you either need a 20s BMS or multiple BMS'es connected in series. :-)

Before anyone tries connecting multiple BMSes in series, you should be careful about this, as the charge/discharge cutoff MOSFETs are sometimes not rated for the stacked voltage of the battery pack.

https://us.reddit.com/r/AskElectronics/wiki/batteries#wiki_increase_the_voltage_with_batteries_in_series

[atdlzpae]

3 hours ago, hyperair said:

Before anyone tries connecting multiple BMSes in series, you should be careful about this, as the charge/discharge cutoff MOSFETs are sometimes not rated for the stacked voltage of the battery pack.

If you have 2 batteries in series, when one battery's BMS opens its power switch during discharge, it sees the entire battery voltage (twice the voltage of each battery) across the that power switch, but in the negative direction. This video explains why.

Very good point! I never thought about it that way. Do NOT connect multiple BMS'es in series unless you read it's MOSFET's datasheet.

Edited August 29, 2019 by atdlzpae

[Mono]

On 8/29/2019 at 9:36 AM, uekarashi said:

It's actually most of the work.  After which it's a couple of screws, and a couple of connectors.

Not for me. Removing the side cover takes (much) less time for me than taking out the battery afterwards. Maybe that's because I have done it so many times more. The battery I only removed three or four times. One interesting challenge is to find the right screws which hold the battery.

[alcatraz]

This might be offtopic but another useful thing to do when building a pack from used cells, is to "calibrate" the 18650 charger that is being used to determine the capacity of the cells. The purpose is not to determine precise absolute values but to make the slots comparable to eachother which is what you really need when testing many cells and grouping them together.

Mark four 18650 cells with individual numbers. Do a capacity test in each slot of the charger. (4 cells x 4 times). You now have 4 different capacities for each cell. 

Calculate an average for each cell and calculate the deviation to that average by a percentage coefficient for each charge slot.

example result:

slot 1: +0.015

slot 2: -0.005

slot 3: +0.010

slot 4: -0.032

(You'll actually get four deviations for each slot, because you did the test with 4 cells instead of 1. You can average the deviations out for a bit better results.)

Then you just test all the cells and before writing down their capacity you just normalize it by multiplying the coefficient. That way the capacity differences can be better determined and thus arrange cell group better.

Edited September 4, 2019 by alcatraz

[hyperair]

On 8/30/2019 at 9:05 PM, Mono said:

Not for me. Removing the side cover takes (much) less time for me than taking out the battery afterwards. Maybe that's because I have done it so many times more. The battery I only removed three or four times. One interesting challenge is to find the right screws which hold the battery.

There's an easy trick for this:

• Screws that have their heads near the top surface of the battery pack hold the battery pack in the wheel. Unscrew these. On new Inmotion V8s, there are only two of these.
• Screws that have their heads at about half-depth hold the battery pack casing together. Don't touch these.

1 hour ago, alcatraz said:

Then you just test all the cells and before writing down their capacity you just normalize it by multiplying the coefficient. That way the capacity differences can be better determined and thus arrange cell group better.

I've wondered about which way is better for capacity matching -- do you put the weakest cells (lowest capacity) in parallel with the strongest cells (highest capacity), or do you match cells with the closest capacities together in parallel? I can think of pros and cons for both sides, but I'm not sure which is ultimately better/safer.

[alcatraz]

I'm fully aware that I make compromises when mixing cells of different age.

From my very limited observation it seems that the cells of same model retain a similar discharge characteristic as they age, but at a lower capacity. It means that as long as they don't self discharge at an accelerated rate they discharge in groups similarly to as they do individually. 

I essentially group the weakest strong cell I have with the strongest weak cell. Assuming I have two sets of cells that I want to group together. Check the last pic here: 

 

I don't recommend it to anyone that aims to have a maintenance free battery pack. My pack is 99.99% maintenance free but I didn't expect it to be. That's why I added balance wires. I didn't know if I needed biweekly or annual balancing. It did come out better than expected considering I had 10% capacity differences between strongest and weakest cell.

Edited September 4, 2019 by alcatraz

[hyperair]

21 hours ago, alcatraz said:

I essentially group the weakest strong cell I have with the strongest weak cell. Assuming I have two sets of cells that I want to group together. Check the last pic here: 

My original cells are falling out of balance, and I was thinking of rearranging them by pairing the strongest cell to the weakest cell to even out the capacities of each pair. If the discharge characteristics don't change, this should in theory make the pack less likely to fall out of balance, but I'm worried that I've overlooked something that could cause it to fail catastrophically instead of safely when it finally reaches end-of-life.

Edited September 5, 2019 by hyperair

[alcatraz]

Some (many?) panasonic cells retain over 70% capacity for thousands of cycles. They seem to last forever.

It's not the reduced capacity you need to worry about fore and foremost. I'd say it's the internal resistance that's going to make the cell (and maybe the pack) useless. So don't skip on the long term self discharge test. This is what's going to keep that pack alive.

If the cells have a low capacity, sure thats not a good sign. But I found newer cells to have accelerated discharge too. I had to reject a cell with the highest capacity and three with low capacity, after the self discharge test.

Had that one cell been put in a pack it would have constantly pulled that one out of 20 cell groups down.

I recommend you do the self discharge test (leave cells fully charged for 1-2 months and measure their voltages afterwards). Anything that deviates is showing signs of degradation and is essentially unfit to be paired with the rest.

And I recommend to attach balance wires even if you don't plan to use them. They can be used to keep a pack within spec for a while longer should you notices some bad cells. You can also sleep better at night knowing your voltages are fine and that you won't burn your house down next time you charge your wheel unattended.

[Tadas]

5 hours ago, hyperair said:

My original cells are falling out of balance, and I was thinking of rearranging them by pairing the strongest cell to the weakest cell to even out the capacities of each pair. If the discharge characteristics don't change, this should in theory make the pack less likely to fall out of balance, but I'm worried that I've overlooked something that could cause it to fail catastrophically instead of safely when it finally reaches end-of-life.

What about self discharge? If some cells have higher self discharge, they will drain the good cells that are connected in parallel with the bad cells.

[hyperair]

@alcatraz These are great tips, thanks! I was a little worried about finding space to tuck the balance wires away after reinstalling the battery pack though. Do you route them out of the hard case, or tuck them inside?

[alcatraz]

There are some small channels in the battery cavity that can be used for the wires. 

Those channels are for the screws that stick out externally from the original battery pack. Those won't be used anymore anyway.

Just solder on the balance leads and before heatshrinking the battery do a testfit to see where they should go. I think my balance wires were 25cm long. It's about the right length. I didn't solder them to the nickel strips but to the terminals on the bms. I used other cables between cells and bms.

I have one cell group that is discharging faster than the others. I will need to balance monthly. Darn it. Ah well, at least I have the balance leads to do it.

I think I might have damaged one of the cells a bit when I tried to fit the side cover on the V8. Inside the side cover are some bracing pieces that kind of cut into the battery if you don't trim them a bit. Don't rush after the pack is finished. My doing so might have influenced the maintenance interval.

Edited September 11, 2019 by alcatraz

[alcatraz]

I have some good news and it's that it _seems_ charging stops before overcharging the healthy cell groups. It means the pack could end up at less than 84V with a green light on the charger.

If this is true then the bms cuts charging when the highest group has reached 4.2v. A nice feature. I can't 100% confirm it as I've only had a small imbalance for now. I do know the charger outputs 84.5V and that my pack never goes over 84V so something is going on with the bms. I think the lowest I've gotten a green light at is 83.5V thereabout. The cells normally reach 4.19v thereabouts. With my precision charger I get 84.0v on the pack.

Edited September 11, 2019 by alcatraz

[Chriull]

2 hours ago, alcatraz said:

I have some good news and it's that it _seems_ charging stops before overcharging the healthy cell groups. It means the pack could end up at less than 84V with a green light on the charger.

If this is true then the bms cuts charging when the highest group has reached 4.2v. A nice feature. I can't 100% confirm it as I've only had a small imbalance for now.

The normally used BMS for EUCs have a charge cutoff for single cell overvoltage. It"s common at 4.28V. At 4.2V the balancing resistor gets active.

But of course these value could be adopted by inmotion.

2 hours ago, alcatraz said:

I do know the charger outputs 84.5V and that my pack never goes over 84V so something is going on with the bms.

This difference could be the reverse polarity protection at the charger input.

[hyperair]

On 9/11/2019 at 12:55 PM, alcatraz said:

I have some good news and it's that it _seems_ charging stops before overcharging the healthy cell groups. It means the pack could end up at less than 84V with a green light on the charger.

If this is true then the bms cuts charging when the highest group has reached 4.2v. A nice feature. I can't 100% confirm it as I've only had a small imbalance for now. I do know the charger outputs 84.5V and that my pack never goes over 84V so something is going on with the bms. I think the lowest I've gotten a green light at is 83.5V thereabout. The cells normally reach 4.19v thereabouts. With my precision charger I get 84.0v on the pack.

Yeah, I can confirm the same behaviour -- my battery pack has not charged up to 84V in a while despite being left plugged into the stock charger for extended periods in the hopes that it would balance itself. In the end, I did have to pop it open for manual balancing from time to time. I believe all multi-S BMSes need to do the same, or you will encounter overcharge-induced thermal runaway as the pack falls out of balance towards end-of-life.

[Nic]

On 8/20/2019 at 8:50 AM, alcatraz said:

You need a few things, and patience.

1. Multimeter - 10usd?

2. 18650 charger with capacity measurement - 10usd? Liitokala 500

3. Spotwelder - cheapo ~ 40usd

4. Nickel strips, I used 6x0.15mm and ran it side by side (better welding with gap between spotweld electrodes). I reckon limit 6 amps per strip so I ran at least 4 strips ~ 24 amps for short connections and 5 strips for longer ~30 amps. Heat = bad. A big roll was like 10usd. Enough for several packs.

5. Balance wires. 5x4S for external balancing. Didn't really need this but I like my pack balanced. I got a cheap 4s high voltage precision hobby charger (balance lead charging only). Wires ~ 2usd? Charger ~15usd

6. Soldering iron ~ 5 usd?

7. 20-25 extra LG MH1 cells. I got 30 of them used but in near new condition for ~40usd. The rest of the 35-40 cells you can harvest from the old pack. Ideally to save time just order like 63-65 cells to ensure you have 60 matched ones. Then you save time.

8. Some blue shrinkwrap, tape, protective isolation paper wrap, adhesive paper washers to put on top of the cells. ~10usd?

9. Glue gun. 5 usd?

10. Patience

Check cells for damage/capacity/self discharge. Repair/arrange them. Use the glue gun to fix the cells together in a configuration that fits the cavity. Arrange them in groups of 3s. After spotwelding together the terminals protect them from accidental contact by covering them with tape. Solder balance leads to bms. Shrinkwrap. Enjoy. This took me several days. I had a spare V8 battery pack to play with so I didn't have to rush.

@alcatrazWhat size heatshrink did you use and/or can you provide a link to the one you purchased? Thanks.

That cavity is really tight and has caused me some problems, but I am almost there now.

[alcatraz]

160mm and 130mm.

【130MM18650锂电池收缩套膜 PVC热收缩管 配件电池套管 蓝色收缩膜】https://m.tb.cn/h.eNDdA9s?sm=a9403f 点击链接，再选择浏览器咑閞；或椱ァ製这段描述$DtSOYng15D5$后到👉◇綯℡寳👈

[Nic]

40 minutes ago, alcatraz said:

160mm and 130mm.

【130MM18650锂电池收缩套膜 PVC热收缩管 配件电池套管 蓝色收缩膜】https://m.tb.cn/h.eNDdA9s?sm=a9403f 点击链接，再选择浏览器咑閞；或椱ァ製这段描述$DtSOYng15D5$后到👉◇綯℡寳👈

Thank you.

130mm seems quite small ... is the BMS also inside the heatshrink?

[alcatraz]

130 is for the sides. 160 for the middle.

You can go larger but you risk getting a bad wrap around the ends of the pack. If your configuration looks similar to mine I'd stick with 130-140 and 160-180.

The larger it is the harder it is to get the waves out.

Yes the bms is inside.

After having my pack for a few months I'm noticing I need to balance monthly. If you don't like this I suggest you get new cells or cells that come from the same batch and that have no damage.

I'll continue ride the pack as it is for the time being because I'm lazy to break it apart. 

I recommend you to heatshrink/protect the balance wires inside the wrap if they go over an edge like the bms board.

[leadfeathers]

Hey @alcatraz, after your battery upgrade, are you observing that the battery state on the wheel correlates to the new battery capacity?  In other words, is your battery meter depleting more slowly than before?

I ask because it dawned on me that if the control board was measuring capacity based on a coulomb counter, then it wouldn't ever indicate state of charge properly because the coulomb counter would be calibrated to the mAh of the original battery.   However if the board is measuring battery state based on some moving average of voltage, then this is a non-issue.