Honda Insight Forum banner

21 - 40 of 43 Posts

·
Premium Member
Joined
·
2,885 Posts
my main thing when we're talking that long is the number of applications goes down due to costs going down and further breakthroughs happening.
Larger batteries in the future could cost a lot less, a new tech breakthrough in batteries/panels or both could outperform today's tech of any kind by a lot, or other unforeseen stuff. The longer the timeframe the more likely vastly better stuff will be cheaper than today's stuff (not a guarantee but tends to happen especially in tech). Saving half of your money today and spending it at 20 years instead of 50 isn't necessarily a loss depends on what you can get in 20 years for those dollars.
kind of

I will 100% agree application matters enormously .. use the right tool for the job .. pros and cons for a the needs of a specific application .. devil in the details.

but .. as for the other part ..

The rate of technological improvement has been slowing down for many years now .. there is not infinite growth .. sure progress is still being made .. but it's slower and slower each new decade .. less and less improvement each new decade.

The easiest place to see this is in computers .. which used to be the poster child for exponential technological improvement .. from 1995 to 2000 a home PC got about 20x better in almost every way faster , more storage , more diversity of applications , etc .. but from 2000 to today 2020 .. no where near that .. sure they have continued to get better .. but how much it is further improved is less and less .. the rate of progress is slowing and slowing .. today's home PC is all of about ~4x better than the PC from 2000 .. which is good .. but 4x in 20 years is vastly slower rate of improvement than 20x in 5 years .. at the current curve of decreasing rate of improvement , computers can be reasonably expected to get maybe 2x better than today , over the next ~30 years or so.

But that same over all trend of .. slowing of improvement .. is seen in pretty much all technological sectors .. cell phones , batteries , material science , medical care , etc .. etc .. not all sectors are slowing at the same rate .. but the rate of improvement is slowing in all of them.
 

·
Registered
Joined
·
80 Posts
kind of

I will 100% agree application matters enormously .. use the right tool for the job .. pros and cons for a the needs of a specific application .. devil in the details.

but .. as for the other part ..

The rate of technological improvement has been slowing down for many years now .. there is not infinite growth .. sure progress is still being made .. but it's slower and slower each new decade .. less and less improvement each new decade.

The easiest place to see this is in computers .. which used to be the poster child for exponential technological improvement .. from 1995 to 2000 a home PC got about 20x better in almost every way faster , more storage , more diversity of applications , etc .. but from 2000 to today 2020 .. no where near that .. sure they have continued to get better .. but how much it is further improved is less and less .. the rate of progress is slowing and slowing .. today's home PC is all of about ~4x better than the PC from 2000 .. which is good .. but 4x in 20 years is vastly slower rate of improvement than 20x in 5 years .. at the current curve of decreasing rate of improvement , computers can be reasonably expected to get maybe 2x better than today , over the next ~30 years or so.

But that same over all trend of .. slowing of improvement .. is seen in pretty much all technological sectors .. cell phones , batteries , material science , medical care , etc .. etc .. not all sectors are slowing at the same rate .. but the rate of improvement is slowing in all of them.
In the end predicting the future is hard to do and just comes down to beliefs and guessing. 20 years is a long time though for both the tech and the cost.
 

·
Registered
Joined
·
62 Posts
Discussion Starter #23
Oh ok I've seen Daly start to make LTO BMS's. For whatever reason they've all been 80A which isn't good enough for my setup which peak draw is just over 100A. You never mentioned the specifics of what voltage you're going to run at and what your peak continuous load is. You've mentioned how many packs so I gather you're building around a 40kwh pack (must be a large house you're looking to power) but not the other parameters. Is 80A going to be good enough? If so yeah you can probably get them down to $45-$50 instead of $80.

One thing I'd take a look at is Jehu has a video that shows what happens when BMS's in series trip what happens. It certainly can be done but default behavior is not what you'd expect (nor did he expect in the video). The many BMS route raises your cost but makes using the packs straightforward (and you might be wasting some cells if you are going 11s or 22s). The fuse block route makes using the packs involve a little bit of wiring up front but you don't waste any cells and it costs less than the BMS.
I have one other question here. Wouldn't my setup be putting the BMS in parallel and not in series? Would this change the behavior?
 

·
Premium Member
2001 5S "Turbo"
Joined
·
10,579 Posts
AmpLee:
Please include your Location in your Profile. Thank You.
 

·
Registered
Joined
·
80 Posts
I have one other question here. Wouldn't my setup be putting the BMS in parallel and not in series? Would this change the behavior?
Yes later posts I understood more what you wanted for your setup with bms's in parallel. It would change the behavior on that post but not the later ones.
 

·
Registered
Joined
·
62 Posts
Discussion Starter #26
Yes later posts I understood more what you wanted for your setup with bms's in parallel. It would change the behavior on that post but not the later ones.
Looking at the Daly BMS spec sheet for my custom order, I'm confused by some of the things I see here. Specifically, it looks like the voltage numbers are too high for the battery chemistry. Am I correct, or am I not understanding something? To me it looks like the voltage cut-off is above what the battery can handle at 2.7v per cell (64.8 in 24s configuration). Also, is this saying the BMS can only balance at the cell's top voltage of 2.7?

85870
 

·
Registered
Joined
·
80 Posts
Looking at the Daly BMS spec sheet for my custom order, I'm confused by some of the things I see here. Specifically, it looks like the voltage numbers are too high for the battery chemistry. Am I correct, or am I not understanding something? To me it looks like the voltage cut-off is above what the battery can handle at 2.7v per cell (64.8 in 24s configuration). Also, is this saying the BMS can only balance at the cell's top voltage of 2.7?

View attachment 85870
Slightly yes, though not likely by enough to matter (perhaps someone else has data pushing the cells to 2.8V). LTO is typically fully charged at 2.7V. The BMS you ordered looks like it starts balancing at 2.7v considers 2.8v fully charged and will shutoff the battery at 2.85v. So basically at 2.7v it starts bleeding off the high cells with a 20 mA current.

Obviously your particular application matters, what your typical charging rate and load is, and also what your inverter can handle. Also pay attention to the low end it looks like it will also shut off the battery at 1.7V. The balance voltage is slightly high but it might not matter, but the balance current might be an issue if you typically charge the batteries pretty quickly with your solar panels. You'd have to do the math if a 20 mA bleed is going to be enough to bring up the lowest cell into balance between 2.7V and 2.8V at the top.

Do you know many panels you're going to use and what the typical load is when at peak sun?
 

·
Super Moderator
Joined
·
8,148 Posts
I think you will be OK. The overcharge protection acts on the upper shoulder of the charge curve where voltage changes very rapidly. It is only .15V over the standard operating range, so should be ok - at least it seems reasonable to me. It is probably just a trigger for the highest cell before total battery balance is achieved. <edit> On my 5 cell starting battery which should be 13.5V max(2.7V x 5), I frequently see 14V in the car, and the voltage changes pretty rapidly between parked and running. It is operating right at the upper shoulder which is steep. I don't honestly know if this might shorten the life of the cells. I do plan to change to a 6 cell, just to gather data and get on the linear part of the curve.

It makes no sense to top balance and bottom balance at the same time. They have chosen a top balance approach which is, if not mistaken, pretty much standard.

I see that greggtwep16 provided a better answer on the upper limit while I was typing
 

·
Registered
Joined
·
62 Posts
Discussion Starter #29
Slightly yes, though not likely by enough to matter (perhaps someone else has data pushing the cells to 2.8V). LTO is typically fully charged at 2.7V. The BMS you ordered looks like it starts balancing at 2.7v considers 2.8v fully charged and will shutoff the battery at 2.85v. So basically at 2.7v it starts bleeding off the high cells with a 20 mA current.

Obviously your particular application matters, what your typical charging rate and load is, and also what your inverter can handle. Also pay attention to the low end it looks like it will also shut off the battery at 1.7V. The balance voltage is slightly high but it might not matter, but the balance current might be an issue if you typically charge the batteries pretty quickly with your solar panels. You'd have to do the math if a 20 mA bleed is going to be enough to bring up the lowest cell into balance between 2.7V and 2.8V at the top.

Do you know many panels you're going to use and what the typical load is when at peak sun?
I'm looking at 12+kw of panels facing east. I guess I'm unsure about the balance current and how I need to do the math.

My intention was to use a SolArk 12k, but it's voltage range tops out at 63v so it would not be able to allow the bms to balance. If I don't go with the SolArk, I could still go with the MagnaSine which covers the range, but then I'm stuck fully charging the cells which will likely lower the lifespan. I think I jumped the gun too early and will need to re-order with Daly so that I can operate within the 1.9-2.5v voltage range where most of the energy exists. Dang, it's a pricey mistake, but maybe I'll be able to sell the BMSs to recoup.
 

·
Registered
Joined
·
80 Posts
I'm looking at 12+kw of panels facing east. I guess I'm unsure about the balance current and how I need to do the math.

My intention was to use a SolArk 12k, but it's voltage range tops out at 63v so it would not be able to allow the bms to balance. If I don't go with the SolArk, I could still go with the MagnaSine which covers the range, but then I'm stuck fully charging the cells which will likely lower the lifespan. I think I jumped the gun too early and will need to re-order with Daly so that I can operate within the 1.9-2.5v voltage range where most of the energy exists. Dang, it's a pricey mistake, but maybe I'll be able to sell the BMSs to recoup.
Yeah it will likely only be the more expensive inverters that will play nice with the balancing. Basically balancing starts roughly at 64.8V and then it feels it's fully charged at 67.2V. The inverter/charge controller/grid charger that I like would never balance as well as it stops at 64.0V. Other than the magnasine and a few other high end inverters 24s wouldn't really balance and you'd have to wire to a 23s or get a different BMS. You could inquire with them if they have one that starts balancing at 2.6V and then is fully charged at 2.7V.

Personally as far as BMS's go I prefer the bluetooth ones instead of the black box ones. Both because the bluetooth ones you set your own voltages to use and because you can in real-time view the voltages of each cell which verifies the BMS is doing what it should on your phone without getting a multimeter. That being said they cost at least double what you paid and you do need 36 of them so I understand why you can't go that route.

Daly has been covered many times on most of the youtube channels and as far as wiring, mosfets used, etc. they are fine and use decent components. They do tend to reach a bit and hurry some things though like when they first added low temp cutoff that didn't work. This is likely their first attempt at LTO and are more used to Li ion so for playing nice with most charge controllers and inverters their current voltages don't seem quite right. The price is right but you'll make up the price on the inverter side using those. In 6 months I'm sure they'll learn and have one that would work with the cheaper inverters.

As for the math, assuming everything was perfectly balanced you have 12000W going into 48V batteries so if it was 1 pack it's 250 amps. You have 36 in parallel so if perfectly wired that would be 7 amps per pack. Each pack has 24 cells so it would be .28 A per cell. So under that scenario a .02 A bleed would likely not be sufficient. This is if things were perfect but also assumes no load at the time which also likely doesn't happen but you get the picture. I'd probably want a .5A (500 mA) or a 1A balance current rating.
 

·
Registered
Joined
·
62 Posts
Discussion Starter #31
Yeah it will likely only be the more expensive inverters that will play nice with the balancing. Basically balancing starts roughly at 64.8V and then it feels it's fully charged at 67.2V. The inverter/charge controller/grid charger that I like would never balance as well as it stops at 64.0V. Other than the magnasine and a few other high end inverters 24s wouldn't really balance and you'd have to wire to a 23s or get a different BMS. You could inquire with them if they have one that starts balancing at 2.6V and then is fully charged at 2.7V.

Personally as far as BMS's go I prefer the bluetooth ones instead of the black box ones. Both because the bluetooth ones you set your own voltages to use and because you can in real-time view the voltages of each cell which verifies the BMS is doing what it should on your phone without getting a multimeter. That being said they cost at least double what you paid and you do need 36 of them so I understand why you can't go that route.

Daly has been covered many times on most of the youtube channels and as far as wiring, mosfets used, etc. they are fine and use decent components. They do tend to reach a bit and hurry some things though like when they first added low temp cutoff that didn't work. This is likely their first attempt at LTO and are more used to Li ion so for playing nice with most charge controllers and inverters their current voltages don't seem quite right. The price is right but you'll make up the price on the inverter side using those. In 6 months I'm sure they'll learn and have one that would work with the cheaper inverters.

As for the math, assuming everything was perfectly balanced you have 12000W going into 48V batteries so if it was 1 pack it's 250 amps. You have 36 in parallel so if perfectly wired that would be 7 amps per pack. Each pack has 24 cells so it would be .28 A per cell. So under that scenario a .02 A bleed would likely not be sufficient. This is if things were perfect but also assumes no load at the time which also likely doesn't happen but you get the picture. I'd probably want a .5A (500 mA) or a 1A balance current rating.
Thanks for helping me understand things. Since the nominal voltage is 60V, would that be the number to divide by and not 48V which would be closer to the low voltage cut-off? This doesn't really change the end result, just curious.
 

·
Registered
Joined
·
80 Posts
Thanks for helping me understand things. Since the nominal voltage is 60V, would that be the number to divide by and not 48V which would be closer to the low voltage cut-off? This doesn't really change the end result, just curious.
There are a ton of assumptions in the math already both for the batteries as well as the charge controller (and no inverter efficiency is ever 100%). In reality there will be different efficiencies based on the voltage of the panels, batteries, wire loses, and how the inverter is in sync with this. For 24 LTO batteries their nominal voltage would be 55.2V (2.3V per cell).

The biggest factor in a massive parallel setup like that is probably going to be the wiring anyways. With 36 in parallel I strongly suggest option 3 below. If not balanced you need to account for the parallel pack that is doing the most work.

 

·
Premium Member
Joined
·
2,885 Posts
I think I jumped the gun too early and will need to re-order with Daly so that I can operate within the 1.9-2.5v voltage range where most of the energy exists. Dang, it's a pricey mistake, but maybe I'll be able to sell the BMSs to recoup.
If you just got it .. less than 30 days .. you might be able to return / exchange it for the 1.9-2.5v you want .. can't hurt to ask.
 

·
Registered
Joined
·
62 Posts
Discussion Starter #34
There are a ton of assumptions in the math already both for the batteries as well as the charge controller (and no inverter efficiency is ever 100%). In reality there will be different efficiencies based on the voltage of the panels, batteries, wire loses, and how the inverter is in sync with this. For 24 LTO batteries their nominal voltage would be 55.2V (2.3V per cell).

The biggest factor in a massive parallel setup like that is probably going to be the wiring anyways. With 36 in parallel I strongly suggest option 3 below. If not balanced you need to account for the parallel pack that is doing the most work.

In that third option, where and how are the extra terminal posts installed?

Or is there a name for this type of setup that I can look-up to find examples?
 

·
Premium Member
Joined
·
2,885 Posts
In that third option, where and how are the extra terminal posts installed?

Or is there a name for this type of setup that I can look-up to find examples?
That is re-wiring the cells .. always be careful and avoid any shorts .. check twice.

It might be best to test cells before actually doing that .. the test results could effect which cells are put where.

Mine will be a little different from yours .. but I plan to re-wire each 12s subpack from Honda into a 12p subpack .. crude concept pic attached.
85875
 

·
Registered
Joined
·
62 Posts
Discussion Starter #36
That is re-wiring the cells .. always be careful and avoid any shorts .. check twice.

It might be best to test cells before actually doing that .. the test results could effect which cells are put where.

Mine will be a little different from yours .. but I plan to re-wire each 12s subpack from Honda into a 12p subpack .. crude concept pic attached.
View attachment 85875
Is this for a vehicle install or something else?
 

·
Registered
Joined
·
62 Posts
Discussion Starter #37
Just wanted to express my gratitude for how helpful you all have been. I can't offer much in return on this topic but if anyone ever needs any advice on anything to do with homesteading and self-sufficiency, I'm happy to return the favor. Or I'll just pay the kindnesses forward.
 

·
Registered
Joined
·
80 Posts
That is re-wiring the cells .. always be careful and avoid any shorts .. check twice.

It might be best to test cells before actually doing that .. the test results could effect which cells are put where.

Mine will be a little different from yours .. but I plan to re-wire each 12s subpack from Honda into a 12p subpack .. crude concept pic attached.
View attachment 85875
He has 36 24-cell 1.1kwh packs in parallel so no this is for the inter pack connections not rewiring cells within the pack.
 

·
Registered
Joined
·
80 Posts
In that third option, where and how are the extra terminal posts installed?

Or is there a name for this type of setup that I can look-up to find examples?
I'm not sure if there is a name for it but it's the easiest to do since you have way more in parallel than normal. It makes verifying the balance easy as well with a clamp meter.

It's been awhile but I believe David Poz had a video on this and the fuse layout he had on his original Chevy volt battery bank if memory serves me right.
 

·
Registered
Joined
·
62 Posts
Discussion Starter #40
I'm not sure if there is a name for it but it's the easiest to do since you have way more in parallel than normal. It makes verifying the balance easy as well with a clamp meter.

It's been awhile but I believe David Poz had a video on this and the fuse layout he had on his original Chevy volt battery bank if memory serves me right.
Okay, I think I understand. He's cutting the wires to the same length and wiring them to a common bus bar with fuses attached to the positive wires. In my setup, I'd be doing something similar, except wiring each battery to two square D QO breaker boxes with the 30amp 2pole breakers for each battery. Same concept though.
 
21 - 40 of 43 Posts
Top