Honda Insight Forum banner

1 - 20 of 104 Posts

·
Registered
Joined
·
1,100 Posts
Discussion Starter #1 (Edited)
Hi, as many of you know by now I've been brooding on a portable parallel booster pack for the 2nd gen Insight.
Note: this is no replacement for the IMA battery but an add-on.

The idea is to have a 1 kWh pack with a voltage slightly over the average voltage of the OEM pack, so it will deplete itself into that pack and force the Insight (which senses a higher than normal voltage on its pack) to use more EV support and less (like no) forced regen.
It would save fuel and give a peppier feel to the car.

I had already started a thread about it on Ecomodder (this one) but thought I'd better not make too much noise before I knew I could actually do it, so I did not start that (same?) build thread here. Yet.
After all, I cannot do this without my wife's approval.

I just told my wife I want to spend 1K on it and she basically agreed!
So, I can put the wheels in motion in earnest now. Any kind of advice is more than welcome. :)
See the link on ecomodder for more details.
Once things come to fruition I will definitely post about it here too though.
 

·
Premium Member
Joined
·
211 Posts
there does not seem to be anyone with much experience in battery building on your thread over there.

if you build a 36S pack then it should be charged to 129.6V at a minimum with a 3.60V balance voltage on the BMS.

your statement about how a BMS will damage your battery is entirely wrong and misinformed.

if you use a 10Ah prismatic pack like the GWL then you can only expect to draw 10-20A max from it but your car will demand more when it is under load.

since the lithium will show less internal resistance to discharge, then it will provide the most current when the two parallel packs are being discharged and the nimh will recharge the lithium in between your power cycles.

but you will get more power on demand when needed. until the voltage has dropped low enuff to increase the internal resistance of the lifepo4 significantly. at that point the lifepo4 will show more internal heating. this is what shortens the life span of the lifepo4 cells.

if you use a 36S then your lifepo4 should be able to discharge down to around 80-85V if it is properly and continuously balanced by the BMS. your nimh will not go this low so much of the power stored in the lifepo4 is never useful in that case.

if you use a BMS then there is no need to use a contactor and instead you can just turn off the BMS output mosfets to shut off the pack from charge or discharge. a mosfet is a much better high voltage switch than a relay or contactor.

the headway cans will provide up to 5C of current with some significant sag, but not as bad as the winstons which really are only 2C.
 

·
Premium Member
Joined
·
9,983 Posts
The BMS thing is my hold up too as I have 4 kilowatts in 64 x 20 amp lifepo4 cells from my old enginer kit.

Plan B was to buy a prius pack that is nickel metal and assemble a pack one stick more than the insight pack so its capacity will be used in the gen 2. Then it could be charged at home and left in the car near the oem pack.

I would use a grid charger set for a lower amp to balance the prius pack so no bms would be needed every so often.
 

·
Registered
Joined
·
1,100 Posts
Discussion Starter #4 (Edited)
My reason to post it on EcoModder first was the vast experience of people there with EV builds and conversions.

Limiting the current to safe levels has always been the object, and just to be sure I will start with a big resistor in the circuit. A current limiting step-down DC converter would be great, but even if such a thing exists it will probably be too expensive to meet its ROI point ever.

I will monitor pack temperature (I have a thermometer for it ready and waiting)

My choice for LiFePO4 is no flight of fancy.
I like their relative safety; they may overheat when severely maltreated, but won't catch fire and burn.
They can be recharged over 2000 times without losing much of their capacity, even on near full cycles.

Unlike other lithium battery chemistries LiFePO4 cells keep the voltage quite steady throughout almost their entire range. For example:

This 12S pack start out at 39 Volt right from the start and keeps the voltage above 38 Volt until it has lost 80% of its charge.
The Insight IMA battery voltage hovers around 107 Volt, 1/3 of that is 35.33 Volt. This pack reaches that at about 95% of its capacity spent.
I would not want to discharge it deeper than that so it seems spot on.

I do not expect high currents to and from the parallel pack, at least not for long. The OEM pack buffers those.
Once the OEM packs voltage has been raised the current flowing from the booster pack will dwindle to less than 1C. When the Insight uses EV mode the voltage will drop, but not by as much as now because of the higher SOC and being fed from the parallel pack.
And if I can recharge the parallel pack before it is fully discharged, the OEM battery will still be in higher SOC and will not draw much current from it even from the start.

So I see no reason to use anything else than LiFePO4.

Using another NiMH hybrid pack is out of the question. That would be too heavy to be portable and takes up too much space.
Portability is a must, I will not charge the pack while it is in the car and at times I do need the space the pack would take.
The OEM NiMH pack is capable of handling the big discharge on acceleration (even better so when constantly trickle fed from its booster pack).
It does not need more NiMH.

Things to do on short notice:
Search out the right cabling, batteries, frame, docking connector (portable, right?), BMS or cell monitoring stuff, etc.

While I'm at it and to test the waters with one supplier of batteries I think about replacing my 12V battery with a 20 Ah 4S LiFePO4 block like this.

Any thoughts? I mean, "What could possibly go wrong"? ;)
 

·
Premium Member
Joined
·
211 Posts
those fake motorcycle packs have no useful BMS.

the reason you need a BMS is so that the pack can be regularly balanced by charging to full voltage.

3.65V is considered full charge, but charging over 3.60V causes the lithium to plate out away from the electrode so the lithium is depleted over time and you lose capacity.

so you want a BMS that balances at 3.60V which is where almost all BMS are designed for now that balance lifepo4.

i have plans to build a 48S lifepo4 pack from a buncha headways i still have laying around to use just like you as a parallel pack to the nimh pack. full charge would be 172.8V so i would charge to 173-174V to force the pack to balance faster when fully charged.

eventually i expect i can afford to build a C&C board using peter's stuff to be able to force the lifepo4 pack to provide power at speed. so it would b similar to a PHEV then.

i currently have a 24S 87Ah lifepo4 pack in my ZENN car that i use in parallel with a 21S 36Ah lipo pack and the 100Ah 72V SLA pack that the car has for a stock battery, except it has degraded to about 40Ah but i was able to restore some capacity of the SLA pack back up to about 55Ah.

i bot some D131 BMS with a balancing voltage of 4.18V for the lipo so i can charge both the lifepo4 and the lipo together to 87.6V which overcharges the lifepo4 to 3.65V/cell and leaves the lipo around 4.17+V on average.

i use the charge immediately so that scrubs off the higher voltage but it forces my big lifepo4 pack to balance since it is made of of recycled ping pack pouches and they need regular supervision imo.

wish i could post up pictures but they don't let me post pictures here anymore but if you wanna talk about your build over on the endless-sphere then i can post up pictures over there.
 

·
Registered
Joined
·
1,100 Posts
Discussion Starter #6
I was looking at the bicycle packs as they are extremely portable :) and come with a charger; a threesome would be manageable.
But indeed I worry about the specs and quality, and they are not that cheap.

My hopes were to build a discharge cutoff when any cell should drop below say 2.8 Volt and a charge cutoff when any cell goes above 3.55 Volt.
As the pack is removable it should not be too hard to balance the cells every once in a while, should that be needed. Or I can do with charging at leisure (like 1.5 Amps) and use these to prevent overcharging the cells.

Nowadays I use Imgur for pictures and paste their ready-made BBCode in my posts.
I did use ImageShack before but somehow I could no longer access my account there (???) and Imgur turns out to be more user friendly, so no regrets.

I will look up Endless-sphere.com ? Index page soon, thanks for mentioning.
 

·
Premium Member
Joined
·
211 Posts
for lithium cells you should always try to use the power from the battery at full charge. if you use the battery at less than full charge the internal resistance eats into useful power and it causes heating of the electrolyte at the SEI/electode interface with the electrolyte. this is where the gas is formed that causes the pouches to puff up.

internal resistance is a function of the population of ions to carry the current. the cell's transconductance is a function of the total number of ions that can move the current through the cell and the most free ions are available at full charge. many people think they are protecting their lithium battery by undercharging it but in fact it is just the opposite. the key is to not charge up and leave at full charge. charge up only when the charge is needed and can be discharged within a short period of time.

this is all separate from the plating out of the lithium metal away from the electrode above 3.60V for lifepo4.

so you should charge to 3.60V at least. you can discharge lifepo4 down into the 2.2V range and if you reduce the power and current flowing through the cell at these lower voltages it reduces the impact. 2.1V is generally considered to be full discharge but the cell will hit the 2V level when it is under load and then after the BMS cuts off for LVC, the cell will rebound to about 2.7-2.8V with no load present.

i never use those external image hosting websites and i never open them either because of the risk of viruses being planted in the images. google is the only hosting i would consider safe.

it is just too bad they don't allow posting pictures here.
 

·
Premium Member
Joined
·
9,983 Posts
I guess Red the real question is what output are you expecting? Are you expecting an hour worth or extra power? That would help to size your pack. I can tell you 32 20 amp lifepo4 pouches with wires and a metal box is about 60-80 lbs. My pack if I did it would go behind the IMA box and not removed unless the car or pack is being serviced. With my enginer kit I got 3-4 hours out of 4 kilowatts. The car could use more if the converter was built heavier. So with that weight is less of an issue and with no clear cheap reliable solution for BMS as to why I would go the used prius pack route. If you can find a good bms thats not made in china that balances the cells you are good to go.
 

·
Registered
Joined
·
1,100 Posts
Discussion Starter #9
I want a small pack, so I'd go no bigger than 10Ah. Looking for relative gains, not absolute. And I can scale up later if I wish to.
I would love to make an exception for the 20 Ah A123 pouch cells, as they are relatively light so a pack of 36 pouch cells can still be lifted out. But they are beyond my budget; they'd work great but be slow to meet their ROI.

I honestly don't expect to draw more than 5-10 Ampere on average, that should reduce my (moderately driven) fuel consumption by about 15%.
As my commute takes about 35 minutes I would not be able to deplete the cells fully in one go, and as I can charge the pack at work I will probably never have to worry about depleting cells completely. So a top balancing scheme seems best. Nonetheless, if I can have a cheap and reliable low cell voltage protection system I won't skip that.

I have not found such yet, I'll have to step up my research.
 

·
Premium Member
Joined
·
9,983 Posts
I was using 17 amps and 117 volts on my enginer converter. It could take more, but the magic smoke would escape the converter. On average unless you try to engage EV mode I think you are fine.

I tried a few China made gizmos for BMS and the cell log 8s seemed most accurate, but has no balancing capacity. You could get 4-5 of those and program the cut off voltages and daisy chain the connector to a relay to stop charging or stop discharging when in the car in use. I just do not know the MTBF for those.

Top balancing I believe is charging each cell? For my enginer pack that has the real force cells 8 cells each of 4 20 amp cell groupings for 24 volt output. I tried a few balancing gizmos and they never got them balanced and it would quit after a while. I tried moving loads and single cell chargers around to manual balance. No luck.

It wasnt til I purchased 8 1 amp lifepo4 3.6 volt wall warts and plugging in for a week to each pack did I get all cells to the same voltage. :) Having 36 of those or meanwell power supplies would make for a neat display and likely cause concern from security o ra second look from someone who works with electricity.
 

·
Premium Member
Joined
·
211 Posts
there is really no need to have use single cell chargers or cellogs. just build the battery with cells that are balanced initially and the BMS will balance them after that each time you charge.

but for the BMS to work, you have to charge to full charge. it does not help your battery to partially charge it unless you regularly charge it to full charge so that it can balance.

if you are trying to extend cycle life the best procedure to follow is never charge the pack until you need to use it. leave it discharged and then charge to full charge immediately before use.

you should never leave a cellogg attached to your battery for more than a short period of time in order to monitor it specifically. the cell log will create cell imbalances because it does not take power from the entire 8S.

there is a hack for the cellogs over on the sphere that richard started and otmar finished with his excellent idea of using a depletion mode mosfet as the active device to drop the voltage to the voltage regulator from the top cell.
 

·
Premium Member
Joined
·
9,983 Posts
Id considered just that very thought. Just balance the pack then use it by the seat of my pants and once a week or month plug in the cell log 8s to see how things are going or just plug in the single cell chargers to top things off.

Of course I would check thing after the first commute to see how they faired and again once they are fully charged from a mean well power supply set slightly below 3.6 volts.
 

·
Premium Member
Joined
·
211 Posts
Id considered just that very thought. Just balance the pack then use it by the seat of my pants and once a week or month plug in the cell log 8s to see how things are going or just plug in the single cell chargers to top things off.

Of course I would check thing after the first commute to see how they faired and again once they are fully charged from a mean well power supply set slightly below 3.6 volts.
but why would you do that? seems just backwards and has nothing relevant to how batteries are balanced with a BMS and by using the cellogs you force the battery into an imbalanced state.

by continuing to promote bad practices it legitimizes poor design in the minds of those who have no knowledge or experience in building batteries.
 

·
Premium Member
Joined
·
9,983 Posts
dnmun, post some links to where we can buy the devices you speak and praise. Ive tried a few China made devices and they are all flawed. Either the voltage readings is off, the BMS part uses the first few cells for power resulting in creating an imbalance while trying to fix one or just gives up after a while.

Hell, Peter has basically showed everyone how to copy he design and I am only aware of 1 America or is he Canadian who tried to copy it, but it does not work.

I think my plan was not too out there. Basically balance the pack with single cell chargers, then use the cell log 8s to check by plugging it in, noting readings then unplugging it and moving to another 8 cells. Do this after a charge to verify every cell is up there, do this after a use to see how they fared under discharge, then once more before and after recharging. In the mean time use the over all voltage to tell if they should be full or empty.
 

·
Premium Member
Joined
·
211 Posts
i use several D131 from bestechpower on my lithium packs. the lipo is 21S36Ah wide and the lifepo4 is 24S and 87Ah wide.

i use 2 of the D131 on the big lifepo4 pack and i use just one of the D131 on the lipo pack.

i am balancing it right now after having to take it apart and repair it and i am charging it at 87.6V right now and my 6A charger is still running the fan and the lifepo4 BMS are hot from shunting alotta current right now but i have to raise the voltage some more to force the lipo to balance finally.

i cannot imagine having to do this manually and with such precision that the BMS handles reliably.

the circuit current for the BMSs i use is just few micro amps. that does not cause the pack to go outa balance because the charge is replaced every cycle when you balance it at the end of charge.
 

·
Registered
Joined
·
1,100 Posts
Discussion Starter #16
My plan (if I go for 36 single cells) is to initially charge them in series until one cell reaches 3.6 Volt, then rearrange them in a parallel configuration so all cells take on the same voltage, then charge again in series and repeat until they all reach 3.6 Volt at the same time.

I would then use the pack for a while, no balancing just checking for high and low voltage to drop of the charge/discharge; if I find they get imbalanced I'd repeat the initial charge regime.
No need for single cell chargers that way, although a single cell charger may be handy to trickle charge the cells up to 3.6 volt when they are in parallel to avoid repeat rearrangements.
They'd be nearly full then anyway so it should not take that long even with a single charger on 36 parallel cells.
 

·
Premium Member
Joined
·
211 Posts
why? just so many things to go wrong and the wasted time. i just don't get it. we have cheap high power BMSs available and it seems you are going backwards. this is why i hate people posting up silly schemes because others think it is feasible to try this silly time wasting stuff rather than building the battery properly from the beginning. it is almost impossible to educate people when there is so much bad information about design of batteries posted up everywhere.
 

·
Registered
Joined
·
1,100 Posts
Discussion Starter #18 (Edited)
This calling me silly does not help me. I'll stop discussing the pack for a while. I'll be back when I have something to show.
 

·
Premium Member
Joined
·
9,983 Posts
Red, the cell that is close to a full charge will hog the current and easily exceed 4 volts + with that method. Plan B would be to hook them in parallel and get a 3 volt mean well power supply that can do a few amps so you hold them all at 3.55 volts or so tops. Id get a few extra cells as you are going to get some bad ones. Id let them rest a week and measure the resting voltage after the charge.

My idea to wire was to use some insulated or plastic coated terminal strips and just use the wires already attached to the cells. Id also get a few sets of 8s leads and wire those in so it would be easy to plug in a cell log 8s to take measurements.

This almost makes a 60 lb prius pack of 15 modules and a grid charger look better and better all the time? :evil:
 

·
Registered
Joined
·
1,100 Posts
Discussion Starter #20
I think you got me wrong. I'd monitor all cells and cut off the charge when the first reaches 3.6 Volt; none would exceed 3.6 Volt.

I could make my life simple by adding a couple of PCM boards like these:

Sadly, I've read about mixed results with these too.
At least they are cheap and I think I understand how they work, so I can and will verify it does not short cells and does indeed protect the cells properly.
 
1 - 20 of 104 Posts
Top