Considerations for using two 14S packs to achieve 28S

[spinningmagnets]

I am curious about what issues would occur if I wanted to use 28S, like Zero motorcycles, but...using two common 14S packs..

If the packs are connected/separated electrically by a large DC switch(s), then I imagine it would be very easy to charge the two sub-packs with readily available 52V chargers. That being said, I'm also sure it would not be difficult to stack Meanwell power supplies to achieve 28S / 4.1V x 28 cell-strings = 115V when fully charged.

Brushless IPM motors typically can run at a wide range of voltages, and I suspect there is no problem finding a wide selection of models that would work well at 28S.

I guess the biggest question is which controllers would be appropriate for 115V and maybe 100A-200A?

Also, would I have to remove the 14S BMS's and run with no BMS's?

Also, is there anything I missed?

 

[amberwolf]

Regarding seriesing the packs, there are a number of good threads discussing the good, bad, and ugly of various methods of that, and problems that may arise. Some of them are under titles similar to "two batteries / packs / bms (etc) in series"; others may be under more difficult to find titles.

There are also threads about contactor / relay / etc BMS current path control instead of FETs. I think the most recent is by Harrisonpatm?

It's a fair bet that a 14s BMS doesn't have FETs that can handle 28s. Methods has a recent thread set about Daly BMS issues that include a search for appropriate FETs for BMS use that can help find replacement FETs you might be able to use.

For myself, on the SB Cruiser trike at some point I'll very likely be switching from 14s2p to 28s1p on my EIG packs, to run those Lebowski-brained Honda IMA controllers--once I actually finish the controllers, and make enough 1p busbars for that (or maybe I'll just series two of the 14s2p packs; the problem is they are rather large and heavy, at nearly 40lbs each). No BMS on the packs, as they stay balanced on their own (used well within their capabilities and not drained to empty, though until recently I did charge them to almost full (4.1 out of 4.15v)), so I have been able to just monitor system voltage and use system-level LVC, and periodic cell-level checks, manually.

 

[harrisonpatm]

If you ask the forum, there seem to be more negative than positive experiences with the APT96600 controller. I have it as overkill on my 72V build cuz I got a deal, and I can see why some people don't like it, but as a commuter driver, it does just fine. Plus I never push it. It's specced to 120vdc and 400-600 battery amps, so should be within your range.

I dont have experience with high-S BMS myself, but if you check out Aliexpress, JBD is a moderately-well-recieved BMS that comes up in 28S range, and there's a lot of youtube/forum use on it, if it get sticky. And if you do the relay-controlled BMS, the rated amp range of the BMS doesn't matter. Just as long as you wire it in a way that works for you.

Hope you go through with this idea, I'd look forward to seeing what you could come up with,

 

[E-HP]

I guess the biggest question is which controllers would be appropriate for 115V and maybe 100A-200A?

There's a few threads on the forum about the Fardriver controllers. Looks like they have 4 models that handle 115V or more (128V) with liquid cooling options:

https://www.far-driver.com/nd96530black-blue/

 

[docw009]

Until you upgrade the FET's, I believe there was a way to protect a series BMS with diodes, but you would have to look it up, Probably not too efficient if it involves putting them betwwwn the two series batteries.

 

[DogDipstick]

I run two 10s, 25ah packs in series, one on each side of the bike. There is a 200A MIDI fuse inline in between them to protect the wire between them. No BMS. I try to balance at least once a season but that is even an exercise in futility, often, they never go far out of balance. Might switch to balance 2x a year .

Here is a pic of a pack I built recently. These ( new) cells had been sitting since 2019; run one baseline capacity test on one of... then just hit them all with a charge current and a taper termination current of c/100 ( 0.25A... 0.3A... ) ( 2.5A termination would be c/10) as is spec in the datasheet to a happy level of 4.0v/cell until the bike is built. Point beeing, start with new cells, keep the cells all aging together. Good results. Start with bad cells, get trash results.

It has been sitting a couple months since I built it. This is a good balance for a pack that has never had a BMS on it. Lol. 0.0005v (5mV) is the greatest deviation between cells. Never been balanced.

Yeah I have had no problems with having half my pack on one side and half on another. Took about two feet of 6awg instead of the 1 foot a single pack would have needed . Fuse in between cell # 10 and #11, and fuse protecting charge input.

 

[ZeroEm]

Just curious, not good at spotting trends. Is the mind set on higher voltage changing?
When I built my trike and decided on 72v has some push back. Now it seems like higher voltage is becoming more acceptable and 72v seems low now.
Just wondering?

by spinningmagnets » Dec 15 2022 6:01pm

I am curious about what issues would occur if I wanted to use 28S, like Zero motorcycles, but...using two common 14S packs..

If the packs are connected/separated electrically by a large DC switch(s), then I imagine it would be very easy to charge the two sub-packs with readily available 52V chargers. That being said, I'm also sure it would not be difficult to stack Meanwell power supplies to achieve 28S / 4.1V x 28 cell-strings = 115V when fully charged.

Brushless IPM motors typically can run at a wide range of voltages, and I suspect there is no problem finding a wide selection of models that would work well at 28S.

I guess the biggest question is which controllers would be appropriate for 115V and maybe 100A-200A?

Also, would I have to remove the 14S BMS's and run with no BMS's?

Also, is there anything I missed?

 

[lnanek]

> Is the mind set on higher voltage changing?

High voltage like that is used for e motorcycles more than ebikes. First sentence of the OP does say he's cloning a Zero e motorcycle, though.

 

[spinningmagnets]

Just curious, not good at spotting trends. Is the mind set on higher voltage changing?

I live in tornado country. I like my odds of avoiding major damage, but regional power outages are the norm (on occasion). The batteries that are 13S / 14S (48V / 52V) are able to run an common and affordable inverter to power household items.

Although I am happy with 52V for my EGO lawn mower and two ebikes, I have wondered what I would do if a project fell into my lap that would benefit from higher voltage.

Although 60V chargers and controllers exist, there is very little difference in price if I go to 72V (20S) which has the option of using two 36V packs. That being said, I am intrigued by the possibility of jumping straight to using two 14S packs, because I already have several uses for that.

Small EV batteries are some of the most expensive things I will buy, and I am frustrated to build or buy something so large and have it only being useful for one thing. Plus, if there is indeed a power outage that lasts longer than a few days, I can imagine me staring at a $1,000 72V motorcycle battery that cannot be used to power my home devices.

I would likely add a way to charge at 28S (in addition to 2 X 14S), but the main thing is to find out what issues occur when trying to series two packs. One nice thing is that if two 52V packs are not exactly at the same voltage (51V + 54V) then paralleling them would generate a ton of heat, but putting them in series is quite easy and safe.

For reference, my region has large quads for sale very cheap with a fried engine. They are considered serious ranch equipment, instead of only being recreational.

 

[john61ct]

Did not read the whole thread.

There are NO disadvantages to serial'ing sub-packs to build higher voltage packs, as far as the cells themselves are concerned.

Added weight/complexity/cost of wiring infrastructure, switches diodes fuses maybe sure.

The key issue arises IF you want to use a monolithic BMS for protecting the cells, rather than using separate components for the various separate functions.

Balancing should be done per sub pack using active units that work at higher balancing current at any desired point in the V/SoC curve.

The controller may handle LVC while traveling.

The charger system should reliably handle HVC.

But depending on whether you top-balance or bottom-balance, at least one of those latter two "should" be based on per-cell/group voltages.

So if "a BMS" is used for that, then it should be:

1. wired at the whole-pack level, (act as) one single unit for all the cell/groups in this example one 28S not multiple 7S or 14S in serial

2. wired so that it is easily removable when replacement is required, if not every charging cycle

3. built to withstand 20% higher than total max amps the controller will ever pull

 

[john61ct]

Note that the definition of BMS does not require the propulsion current flowing through internal FETs.

Controlling external contactors is a much more robust design, eliminating a frequent point of failure

 

[ZeroEm]

Little lost on this one? accessible balance wires to External BMS.

y john61ct » Dec 17 2022 10:36am

Note that the definition of BMS does not require the propulsion current flowing through internal FETs.

Controlling external contactors is a much more robust design, eliminating a frequent point of failure

Thank you for the reply! My interest is close to yours then. E-trikes to home backup power to small EV.

by spinningmagnets » Dec 17 2022 10:00am

Just curious, not good at spotting trends. Is the mind set on higher voltage changing?
I live in tornado country. I like my odds of avoiding major damage, but regional power outages are the norm (on occasion). The batteries that are 13S / 14S (48V / 52V) are able to run an common and affordable inverter to power household items.

Although I am happy with 52V for my EGO lawn mower and two ebikes, I have wondered what I would do if a project fell into my lap that would benefit from higher voltage.

Although 60V chargers and controllers exist, there is very little difference in price if I go to 72V (20S) which has the option of using two 36V packs. That being said, I am intrigued by the possibility of jumping straight to using two 14S packs, because I already have several uses for that.

Small EV batteries are some of the most expensive things I will buy, and I am frustrated to build or buy something so large and have it only being useful for one thing. Plus, if there is indeed a power outage that lasts longer than a few days, I can imagine me staring at a $1,000 72V motorcycle battery that cannot be used to power my home devices.

I would likely add a way to charge at 28S (in addition to 2 X 14S), but the main thing is to find out what issues occur when trying to series two packs. One nice thing is that if two 52V packs are not exactly at the same voltage (51V + 54V) then paralleling them would generate a ton of heat, but putting them in series is quite easy and safe.

For reference, my region has large quads for sale very cheap with a fried engine. They are considered serious ranch equipment, instead of only being recreational.

 

[lnanek]

but putting them in series is quite easy and safe

Not at all. Putting them in series means the second battery needs to handle double the voltage it was designed for. Paralleling two identical packs is considered much safer.

Even when building a battery, people don't make the parallel connections between cells very thick because only slight balancing current occurs there, but the serial connections between different p-groups need a lot of metal. Also the discharge wires at the end often need to be soldered to multiple spots to handle all the current. It's those series connections that are suddenly going to have to handle twice as much in the second battery when in series.

power outage that lasts longer than a few days, I can imagine me staring at a $1,000 72V motorcycle battery that cannot be used to power my home devices

Oh, you are doing all this to get close to 120V AC? I use a 12V DC to 120V AC inverter in my car all the time. It's completely unnecessary to input 120V DC. Even the standard American household inputs 240V DC and creates 120V AC through a center tapped transformer. So 120V DC isn't what is going into the wiring cabinet anyway.

Even when running household lights on DC off an ebike battery instead of AC in order to drain the battery for storage, it's annoying to feed the light directly off the battery because it dims as the battery discharges. So I use a cheap DC-DC boost converter that keeps the voltage steady at whatever I set the potentiometer to:
https://a.aliexpress.com/_msjXTIw

So there isn't any reason to complicate your DC battery to try to get closer to your AC voltage when there are much safer pre-made converters for that purpose.

 

[ZeroEm]

The train has jump the tracks.

 

[spinningmagnets]

There may be others who have series'ed two packs that may benefit from a discussion of this. 2X 36V = 72V, etc...which was common about 12 years ago, when high amp cells were hard to source.

I am thinking I might run without a pack BMS, and I would attach a wiring harness that would allow each 14S sub-pack to be balanced by an external BMS when charging, on about every tenth charge. Standard charging and running would be with no BMS, on a set of 14S battery packs I make myself.

I have two significant inverters that can accept 13S / 14S inputs to provide 120V AC when I might want that. https://endless-sphere.com/forums/viewtopic.php?t=100355

If I was to convert a near-free quad-runner to electric, what voltage should I use? If I build a large 52V pack, I might find the amps to be a limiting factor. If the system costs are near the same for 2X 14S to make 28S, or comparing it to a large 14S pack...I suspect that the controller would have to provide a huge amount of amps for the 14S version. I have found that if you compare 28S / 100A to 14S / 200A (with both providing 11.5-kW in theory), the total watts are about the same, BUT...the size and cost of the controller is likely to be much more for the 200A.

One problem is simply fitting a huge controller onto the frame when space is at a premium. Of course the cost is always an issue. The less expensive the controller, then, the more money I have to direct to other parts of the system.

The battery pack would be a physically large block. It will be either a huge 14S pack, or two smaller 14S packs in series to produce 28S / 115V

 

[amberwolf]

If I was to convert a near-free quad-runner to electric, what voltage should I use?

What speed do you want it to go, and what is the motor kV (and it's gearing to the ground, which includes the tire diameter)? Assuming you can't change the gearing between motor and ground, those are the two things that primarily determine what voltage you'll need to run, at minimum, down to LVC, to reach and maintain that speed.


(other things also determine speed under various conditions, but those determine the base motor speed).

 

[spinningmagnets]

I just saw a youtube with a dirt bike conversion using 28S, and he used a VESC from 3Shul.

The builder stated that it was not difficult to program, which is something I've been searching for. If it was just for me, I would pay an expert to program a controller to match a certain motor, but I want to gain hands-on experience with something that I can recommend to others.

I was a bit surprised at how compact the controller was, and as a comparison, Kelly controllers are known to be very bulky per a given wattage (specifically scaling up for the amps).

https://3shulmotors.com/shop-page/?v=7516fd43adaa

 

[amberwolf]

3Shul started out here
https://endless-sphere.com/forums/viewtopic.php?f=30&t=110371