Charging a 72v pack from a 96v pack?

Joined
Jun 12, 2019
Messages
377
What would be the best way of doing this? I'm guessing buck convertors are the correct method, if so, is there a good source readily available? The 96v pack will have a low C rating and the parallel connections can be split so multiple units of 5-10 amps for each set in series would work ok if that's a better approach than a single unit (somewhere around 50-200a total).
Thanks.
 
Do not charge from storage, huge inefficiencies.

Mains power, inverter genset, solar, wind hydro, car alternator

feed all your various packs from a real charge source

DC-DC chargers, or dumb converters with an HVC circuit can handle the changes in voltaged needed

ideally downward, buck conversion, boosting has greater inefficiencies

Give details as to your use case
 
john61ct said:
...
Give details as to your use case
Thanks.
All mobile, about 3kw/hr of lifepo4 pretty much built-in that covers 90% of use cases and power needs with plenty of headroom and then anything up to about 20kw/hr of li-ion that's easily removable and only attached when extra range is needed.

Voltages for the li-ion can be anything, 96v was just a guess. Others seem to have done it with equal pack voltages and just a diode between them and I'd considered the same along with a power resistor to make sure the li-ion can never be asked for more than about 0.5-1C but it seems like a kludge and I'd imagine a hell of a lot would be lost through the power resistors due to the sag and different discharge curves, I'd guess it's better to take a correct approach from the start.
 
I had a plan to do a similar thing, but from 12s modules to my 22s battery.
I dont know if there is something of interest for you in that thread:

https://endless-sphere.com/forums/viewtopic.php?f=14&t=109850

I havent tried it, there are very few times that I actually need more battery.
 
j bjork said:
I had a plan to do a similar thing, but from 12s modules to my 22s battery.
I dont know if there is something of interest for you in that thread:

https://endless-sphere.com/forums/viewtopic.php?f=14&t=109850

I havent tried it, there are very few times that I actually need more battery.
Thanks, I was digging into it yesterday until my eyes went square but hadn't come across that thread. It's just hypothetical for now, the 3kwh of lifepo4 will do me for everything I need short-term and there should be no problem with charging points even then but the long term target is over 100 mile range. Discharge rates work out ok, only extended motorway use should need any backing off to let the lifepo4 pack catch up.

Seemed like a trivial problem initially but the more I read, the more complicated it gets! I'd have thought a simple PWM current regulation would have been adequate, just match up the discharge voltages and I'd have thought there'd have been current-limiting (instead of tripping out on overload) BMSs available but my grasp on the finer points of electrickery is weak at best, might have taken a wrong turn with this one :/
 
So what are the energy inputs?

Be very wary of the non-LFP packs if you are talking a mobile living space for humans.

Why not just use LFP throughout?

If you do mix chemistries, and you say any voltage, then simplest is build them the same charging voltage

and charge them both the same time from the same source.
 
john61ct said:
So what are the energy inputs?

Be very wary of the non-LFP packs if you are talking a mobile living space for humans.

Why not just use LFP throughout?

If you do mix chemistries, and you say any voltage, then simplest is build them the same charging voltage

and charge them both the same time from the same source.

It's on a motorcycle so everything's open-air, the li-ion packs will be very exposed so crash protection's a biggie but it's manageable. LFP alone doesn't seem to give me anything near the density needed, enough can be built into the tighter spaces to cover the power needs (usually about 250a with anything up to about 500a intermittent) with plenty for regen/fast charging (up to about 100a) and enough range for day to day needs but LFP would take up too much space in the extended packs (in the easily accessible space) when trying to add a solid range to that.

Could be I've missed something there though, 5000mah 21700 li-ion cells look like the ideal for energy density so far but if there are LFP cells/puches that can come close then they'd solve all the issues. Discharge curves would be close enough to just link them straight up and the additional packs could just be tripped out on high currents if sag on the main pack was asking too much from them, there's enough range from the main pack to cover everything other than extended flat-out use and that's an edge case. There's also a possible scenario where a fully charged extended pack can be hooked up to an empty main pack but that's also an edge-case, easy enough to guard against with contactors between (multiple packs so only about 50a max each).

Btw, should I edit the thread title? Guessing 96v to 72v is pretty low down the list of possible solutions.

EDIT: Just a thought, could a DC (brushed) motor controller be used? Probably overkill but it looks like it could be the lowest cost off the shelf option for current limiting.
 
OK then forget the LFP

I assumed like a House solar / storage use bank for a camper, not propulsion

Go li-ion all packs the same chemistry

ideally all the same production run.
 
stan.distortion said:
Could be I've missed something there though, 5000mah 21700 li-ion cells look like the ideal for energy density so far but if there are LFP cells/puches that can come close then they'd solve all the issues.
There's nothing even remotely close. The 21700s are 230Wh/kg; automotive LFP stalled with the AMP20M1HD-A at 130Wh/kg.

This is a pretty common design requirement, and the standard solution is to spec the auxiliary pack(s) to the same nominal voltage as the primary pack and then just parallel them. Especially with high series counts, the chemistry doesn't have to match -- just adjust the series count to match the nominal voltages, and ensure that paralleling them doesn't exceed the maximum voltage of either.

So a 24s LFP primary at 76.8V nominal could be supplemented with 20s Li-Ion auxiliary at 74V nominal, 84V hot. Just be careful to undercharge or draw down the LFP pack from 87.6V hot down to 84V (3.5V/c).
 
If you are just looking for a reserve and your controller can handle the voltage, you could just add a 12 or 24V boost pack in series. That way the battery chemistry won't matter.

I did that, adding a 6s boost pack to a 18s main only I started out on 24s and unplugged the 6s after my CA told me I had about used up the boost pack. Did all the switching with plugs.
 
fatty said:
...
There's nothing even remotely close. The 21700s are 230Wh/kg; automotive LFP stalled with the AMP20M1HD-A at 130Wh/kg.

This is a pretty common design requirement, and the standard solution is to spec the auxiliary pack(s) to the same nominal voltage as the primary pack and then just parallel them. Especially with high series counts, the chemistry doesn't have to match -- just adjust the series count to match the nominal voltages, and ensure that paralleling them doesn't exceed the maximum voltage of either.

So a 24s LFP primary at 76.8V nominal could be supplemented with 20s Li-Ion auxiliary at 74V nominal, 84V hot. Just be careful to undercharge or draw down the LFP pack from 87.6V hot down to 84V (3.5V/c).
Probably best to just give that a try and see how it pans out then, probably just a small auxiliary pack of some of the cheaper 18650s initially to see how it works out (as I can still use them in cordless powertool packs afterwards) and it will have it's own BMS anyway so should be well enough protected. It would certainly make it a hell of a lot simpler! I've got it in my head that the discharge curves will be fighting each other but it doesn't look that way when sag at different currents is taken into account, might be an issue if the LFP cells had a lot of sag but seems to work out well on paper with the ones I'm using, will be interesting to see how that compares with real world testing.

john61ct said:
OK then forget the LFP

I assumed like a House solar / storage use bank for a camper, not propulsion

Go li-ion all packs the same chemistry

ideally all the same production run.

Already ordered and I'm fairly determined to get this worked out. The advice almost always seems to be "don't waste your time with that, more of one type works out better" and that's certainly been true for a long time but I doubt it's going to stay that way forever. Things like power density and energy density are moving so far apart with the different technologies that mixing them to exploit each of their strengths will be essential at some point imo.
 
stan.distortion said:
Probably best to just give that a try and see how it pans out then, probably just a small auxiliary pack of some of the cheaper 18650s initially to see how it works out (as I can still use them in cordless powertool packs afterwards) and it will have it's own BMS anyway so should be well enough protected.
Don't rely on the BMS, but that does add another layer of protection. It'll be fine.

stan.distortion said:
I've got it in my head that the discharge curves will be fighting each other but it doesn't look that way when sag at different currents is taken into account, might be an issue if the LFP cells had a lot of sag but seems to work out well on paper with the ones I'm using, will be interesting to see how that compares with real world testing.
Once they're paralleled, they electrically become one, their characteristics (discharge curves) are aggregated -- they won't be fighting each other.
 
stan.distortion said:
Already ordered and I'm fairly determined to get this worked out. The advice almost always seems to be "don't waste your time with that, more of one type works out better" and that's certainly been true for a long time but I doubt it's going to stay that way forever.
It's certainly better if you haven't ordered yet. But there are certainly edge cases like yours, where a small primary pack uses a power density chemistry to meet discharge requirements, but auxiliary packs can use energy density chemistry.

stan.distortion said:
Things like power density and energy density are moving so far apart with the different technologies that mixing them to exploit each of their strengths will be essential at some point imo.
Outside of edge cases as above, this does not seem to be the current trend in the industry. Rather, the industry is trending towards application-specific cells that are tuned to balance the power- and energy-density requirements of that specific application.
 
Yes, all the same chemistry selected for a given use case is an immutable principle not going to change with "trends".

And LFP has no advantage wrt power density for this use case, its advantages are cost, longevity and thermal safety.

The lower energy density is a deal breaker for most in the motorcycle domain.

Decently large cars, design sophistication is overcoming that, driven by price and supply advantages long term, e.g. Chinese-market Tesla as opposed to the first-world models
 
john61ct said:
And LFP has no advantage wrt power density for this use case, its advantages are cost, longevity and thermal safety.
This is incorrect. The AMP20M1HD-A still has higher specific power than competing automotive cells, especially contemporary cells like the EiG C020 or SPIM08, or modern energy-density cells like the LG N2.1, and is competitive with the latest A123 NMC cells.
 
I did not mean in theory.

Since the EV use case usually requires an average burn rate well under 1C. . .
 
john61ct said:
I did not mean in theory.

Since the EV use case usually requires an average burn rate well under 1C. . .
Ah, I see you mean current production EVs (which do not use LFP). My bad -- I thought you meant the OP's use case.

I am curious if LFP (or LTO?) will be (re-)researched once EVs saturate the current in-town vehicle market, and are legislated to replace long-haul vehicles as well. Energy density is still an order of magnitude or two lower than gas, so simply adding capacity to extend range as current EVs do doesn't help if charge rates remain low. Seems like the only way to make long-haul practical is to recharge as fast as stopping for gas.
 
fatty said:
Ah, I see you mean current production EVs (which do not use LFP).
Wut? hundreds do if not thousands. Did you miss Tesla's deal with CATL, and very recently the plan to go global with LFP?

Huge buses, cargo vehicles, tons of applications where maximum density is not the top priority.

Yes consumers taking long road trips will need lots of changes including realistic expectations.

Easily accomplished by just banning FF vehicles sooner rather than later.

Not as if CC isn't an emergency from us waiting too many decades doing squat about it.

Public transport, reducing private vehicles weighing tons used by single humans by 80-90% also part of the solution.
 
john61ct said:
Wut? hundreds do if not thousands. Did you miss Tesla's deal with CATL, and very recently the plan to go global with LFP?
Seems like you're going back and forth on this. The trend for Western consumer markets is definitively, if not exclusively Li-Ion, not LFP. And I think the CATL deal is for Li-Ion, not LFP.

john61ct said:
Yes consumers taking long road trips will need lots of changes including realistic expectations.

Easily accomplished by just banning FF vehicles sooner rather than later.

Public transport, reducing private vehicles weighing tons used by single humans by 80-90% also part of the solution.
This is not even remotely "easily" accomplished.
Gas is 13,000Wh/kg, Li-Ion is 230Wh/kg, 56x lower.
10gal of gas fills in 1 minute = 28.4kg/min = 369.2kWh, 85kWh Tesla battery fill takes 75min on a 150kW Supercharger, 326x slower.

They're simply not suitable for the same use cases. That's not to say that use cases won't change, but "easily accomplished" is absurd.
 
Wrong, on CATL, spend ten seconds googling.

And easily as in relatively, over time, compared to trying to make it happen using the "free market", make it the law as many countries have, then the market will respond.
 
john61ct said:
Wrong, on CATL, spend ten seconds googling.
Can't follow everything. In any case, sounds like Tesla-CATL LFP is China domestic market only.

john61ct said:
And easily as in relatively, over time, compared to trying to make it happen using the "free market", make it the law as many countries have, then the market will respond.
US is not remotely comparable to Europe.
Guess we'll see what happens..
 
LFP powered Teslas are being exported to both Oz and Europe, Elon has stated in many tweets and interviews this will be expanded worldwide, otherwise Nickel will become a huge bottleneck.

And I'm not saying the US **will** get smart enough fast enough on EVs, green tech, leaving FF in the ground

just that we should.
 
Once EVs are as readily available as FF I think we'll see countries steadily hike up fuel taxes to unaffordable levels, maybe with exemptions in places where FF is the only viable option unless battery tech takes a big jump in capacity. LFP and LTO is possibly one area where Tesla can stay relevant, I know at least one BMW EV owner who thought it was the best thing ever for the first couple of years but was totally turned off by the sting in the tail when the batteries started to give up the ghost, he'll never buy an EV again if he has a choice.

That's the main reason LFP was a no-brainer for me, 10+ years of daily use if they live up to the datasheet specs. I've no problem with easily replaceable packs that might only last a few hundred charges for extended range use as that's pretty much the same 10 year time frame for me and if that ever changed, longer range was needed on a daily basis, then lower capacity long life packs are still an easy option.
 
Back
Top