Phev Insight

[IamIan]

Originally Posted by crx_rogus

I'm still not getting why every single balance cable would see such assist current vs. the 100A -rated outer main cables that see the full ~144V potential,

Let's take a look at what is happening... in a bit more detail.

The IMA system asks the 120 cell OEM pack for ~100 Amps of Assist.
That ~100 Amps coming out is the result of the OEM battery SoC, Ohms, and Voltage and how they react with the Voltage , Ohms , and duty cycle of the MDM.

All batteries connected in series no matter what kind of chemistry they are... will all see the same current level... but not the same voltage level.

Even though there are minor differences in those 120 cells in Ohms , Voltage , SoC... they all put out the same Amps because they are all connected in series.... + terminal of A to - terminal of B... + terminal of B to - terminal of C ... etc...

A single 6 cell series sub-pack stick in order for it to put out ~100 Amps... is the same process we see at the larger scale ... the 6 cell stick Battery's SoC, Ohms, Voltage compared to the Voltage , Ohms , Duty cycle of what is pulling that current out... in order for it to have ~100 Amps pulled out of it... those things that it is seeing at its terminals determine what it gives... which as long as there are even the tiniest variations in SoC, Voltage, Ohms in any of those 120 cells in series ... some will see a different dV across their terminals than others see... This is what drives a series connection to all have the same current of amps.... but not the same voltage levels.

Now a Parallel connection + terminal of A to + terminal of B ...etc... works a bit differently.
A Parallel connection sees the same Voltage across all terminals ... but not the same current in Amps.... again each individual battery connected in parallel will have variations in Ohms, Voltage, SoC... in order for the parallel connection to keep all things connected in parallel at the same terminal voltage the Current in Amps will vary from connected battery to connected battery.

Once you connect two 6 cell sub-pack sticks in parallel .. + to + ... - to - ... each of those 6 cell sub-packs individually still act in series ... but collectively the two act in parallel.... so the OEM pack asks for ~100 Amps ... that combination of two paralleled sub-packs will see a dV that will result in that combination giving out ~100 Amps... but sense the Voltage is the same in a parallel pack that dV is seen across both of the 6 cell sub-packs.... each of those 6 cell sub-packs might not react exactly the same to the same dV... one might provide ~40 Amps... the other ~60 Amps... etc... for the OEM series connection it doesn't matter because the net out of that parallel connection is still the same ~100 amps it is asking for from the series connection.

Originally Posted by crx_rogus

It also explains how stringing 120 cells won't fail catastrophically with failing cells getting via increased ohms most of the ~144V charge voltage all by themselves; they just short themselves out of the way of the others.

Well that depends a bit on the chemistry ... a PbA .. or Li batteries fail as an open circuit with HIGH OHMs... but the series connection still forces all parts to have the same current flow ... so if a series connection or PbA or Li has a single cell fail... the series connection forces it to still flow the same current of amps... as the failed cell's OHMs go up but the series connection forces the same AMPs of current through it the dV and heat from that failed cell rise drastically... and bad things can happen.

The expensive Li packs for EVs and such often use a BMS that will disconnect a bad cell from the series and bypass it ... so that the series current does not keep going through the rising OHMs of the bad cell... or they sound an alarm , or turn off the system.

Cheaper PbA EVs sometimes use BMS that do the same thing... but other times sense the battery investment is less they don't bother with the BMS investment.

NiMH is different ... sense it fails toward 0 Ohms a failed cell weakens the whole pack... but does not cause the major problems a failed PbA or Li cell can... but NiMH still benefits from a BMS as well.


Now a separate issue from cell failure ... is cell reversal.
Cell reversal happens when the cell has not failed ... but has reached it 0% SoC before the rest of the series connected battery pack has... because the rest of the series connected battery pack will continue to force equal amounts of AMPs through it ... you begin to charge that cell in reverse ... so the voltage of the cell reverses ... the + terminal becomes a - terminal ... etc... reversed charging of a cell has allot of OHMs so it produces allot of heat and dV... it also is very bad for the battery cell, no matter what chemistry it is... voltage reversals should be avoided ... bad things can happen.

Some chemistries are more or less tollerant of cell reversal than others... I have pulled NiMH cells into voltage reversal several times... it damages them and reduces their capacity ... but it isn't an instant death sentence ... voltage reversal to a Li battery will kill it very quickly... PbA will also not tolerate much voltage reversal ... but from what I've read they fall in the middle ... not as tolerant as NiMH ... but still better than Li.

Originally Posted by crx_rogus

Thanks for the very clear explanation/analogy. To avoid possibly driving you bonkers I've ordered Batteries in a Portable World to help me have a hint of a clue regarding best use and maintenance of common battery types (I use NiMH and deep-cycle PbA every day already and LiPo occasionally).

No problem at all... Batteries are one of my hobbies... a NiMH is still my personal favorite for most applications... some of the newer Li have been able to match or beat NiMH for cranking power... and Li has better energy per L or energy per kg ... but the tolerance , durability, and cost of NiMH still have many benefits over Li.... that might change eventually... but for now... what battery type best meets an applications need still depends on what those needs for the application are... PbA for instance ... with it's low $ per Wh ... beats the crap out of Li for stationary applications where weight or volume don't matter much.

There is tons you can find out there on batteries ... so enjoy.

[crx_rogus]

Originally Posted by IamIan

... each of those 6 cell sub-packs might not react exactly the same to the same dV... one might provide ~40 Amps... the other ~60 Amps... etc... for the OEM series connection it doesn't matter because the net out of that parallel connection is still the same ~100 amps it is asking for from the series connection.

I understand about series vs. parallel (I've monkeyed with electronics since I was wee), but I didn't think there would be that big a dV across a pair of subpacks that had just spent at least the last several hours or so balancing each other in a no-load situation. Now I know.

Originally Posted by IamIan

... as the failed cell's OHMs go up but the series connection forces the same AMPs of current through it the dV and heat from that failed cell rise drastically... and bad things can happen.

That's exactly why I assumed long strings of battery cells of any type are a bad idea, not realizing NiMHs ever so conveniently short themselves out of the way.

Originally Posted by IamIan

... The expensive Li packs for EVs and such often use a BMS that will disconnect a bad cell from the series and bypass it ... so that the series current does not keep going through the rising OHMs of the bad cell... or they sound an alarm , or turn off the system.

and thus probably avoid having a $40k+ car do the same thing Sony and Dell laptop computers did awhile back... have their lithium-based battery packs burst into bright, smoky "thermal runaway" states taking the value of whatever they're in with them.

Originally Posted by IamIan

... NiMH is different ... sense it fails toward 0 Ohms a failed cell weakens the whole pack... but does not cause the major problems a failed PbA or Li cell can... but NiMH still benefits from a BMS as well.

It would be nice for the Insight to have something besides a way to gradually de-EV upon noticing subpack thermal issues, something to help counteract any developing out-of-balance issues.

Originally Posted by IamIan

Now a separate issue from cell failure ... is cell reversal.
Cell reversal happens when the cell has not failed ... but has reached it 0% SoC before the rest of the series connected battery pack has... because the rest of the series connected battery pack will continue to force equal amounts of AMPs through it ... you begin to charge that cell in reverse ... so the voltage of the cell reverses ... the + terminal becomes a - terminal ... etc... reversed charging of a cell has allot of OHMs so it produces allot of heat and dV... it also is very bad for the battery cell, no matter what chemistry it is... voltage reversals should be avoided ... bad things can happen.

Some chemistries are more or less tollerant of cell reversal than others... I have pulled NiMH cells into voltage reversal several times... it damages them and reduces their capacity ... but it isn't an instant death sentence ...

Hmmm... NiMH cell failure = shorting itself out of the way, yet cell reversal can lead can eventually lead to exciting things possibly involving release of a cell's magic smoke (does it eventually short itself, or keep its radical overvolt condition?)... One issue with Insight 120S packs then is, I'm guessing, overcharging leads to relatively benign cell failure while chronic discharging getting out of hand can lead to catastrophic cell reversal.

I'm also getting the impression that the Tesla Roadster among approaching others must have some astoundingly sophisticated BMS hardware and software, with a significant cost of the vehicle development just going into BMS development. A Roadster stranded with a flaming battery pack would produce unproductive PR, especially in the courts. My multicharger alone has all sorts of warnings about choosing the correct Lix setting stemming from the 0.1V cell potential difference between LiIo and LiPo.

Originally Posted by IamIan

No problem at all... Batteries are one of my hobbies... a NiMH is still my personal favorite for most applications... some of the newer Li have been able to match or beat NiMH for cranking power... and Li has better energy per L or energy per kg ... but the tolerance , durability, and cost of NiMH still have many benefits over Li.... that might change eventually... but for now... what battery type best meets an applications need still depends on what those needs for the application are... PbA for instance ... with it's low $ per Wh ... beats the crap out of Li for stationary applications where weight or volume don't matter much.

PbA works well for my home's hobby-grade solar-recharged 12V subsystem (powering the 3 D->C cycle slow refreshing of the spare pack's subpacks effortlessly), and has reliably and safely and without stink (except near the recharging racks) moved countless tons of paper, ink and anything else imaginable for at least a couple decades at work, but getting folks to remember to recharge the battery, not the drive terminals, on some of the smaller equipment is always a challenge.

Originally Posted by IamIan

There is tons you can find out there on batteries ... so enjoy.

I've been gradually reading V2G-101, learning how V2G-enabled versions of cars like the (apparently truly heading towards real production) Chevy Volt (and current Tesla Roadster) will (/can) earn their owners / leasees up to about half their new cars' monthly payments back via buffer service micro-contracts with the local grid utility. But first, we need a smart grid. Flywheel, hydraulic and pneumatic -based hybrids also appear viable. There is indeed alot of really fascinating stuff going on out there, mostly off-topic for this thread.

Regards,
Roger

[IamIan]

Originally Posted by crx_rogus

I understand about series vs. parallel (I've monkeyed with electronics since I was wee), but I didn't think there would be that big a dV across a pair of subpacks that had just spent at least the last several hours or so balancing each other in a no-load situation. Now I know.

As long as the two packs are always connected to each other then sure they should have the same Voltage as any 2 parallel batteries would.... but not the same current.

If you charge up the secondary battery and then connect it to the OEM battery now you have a difference and will have a surge of amps... and like I showed ... the worst case for the two batteries by themselves would be a surge of ~66 Amps... probably wouldn't last very long ... but it is a worst case to plan for.

------------

If they stay connected when you start up the car and drive off... they are still connected when the IMA system asks for ~100 Amps of assist... so like any set of batteries in parallel the current might not be equally distributed... between the two... and your secondary battery could easily end up being asked for ~50+ Amps.

Just think about it like this... your secondary battery that is connected in parallel to the OEM battery will see the same dV that the OEM battery sees... and that causes the OEM battery to put out up to ~100 Amps when it is by itself ... if your secondary battery has equal Ohms and SoC as the OEM battery than it will see 1/2 of the ~100 Amps ... or ~50 Amps come from the secondary battery and the other ~50 Amps come from the OEM battery... if the secondary battery has less Ohms or a Different SoC... it might end up sending more than ~50 Amps... If the secondary battery has more Ohms or a different SoC it might end up sending less than ~50 Amps.

Unless you have some type of device to limit the current that will flow between the secondary and the OEM battery.

Originally Posted by crx_rogus

That's exactly why I assumed long strings of battery cells of any type are a bad idea, not realizing NiMHs ever so conveniently short themselves out of the way.

Series connections are needed for the higher voltages ... you get the same amount of Watts from 200V at 1 Amp... as you do at 20V at 10 Amps... but losses and heat from resistance are not driven by the voltage ... Resistance losses come from R*I^2 ... so while 200V & 1 Amp = 200 Watts = 20V & 10 Amps... for 1 Ohm the 1 Amp produces 1 Watt of waste heat due to resistance ... but the same 1 Ohm for the 10 Amps produces 100 Watts of waste heat due to resistance.... so the same amount of power to do work 200Watts is made either way... but by having higher voltages you greatly reduce the losses due to resistance.

In order to increase the voltage from a battery pack above the 2V per cell PbA ... or the 3.6V per cell of Li ... you have to connect them in series.

You can try to use a DC-DC device ... but even at 95% efficiency a DC-DC device would loose ~10 Watts from a 200 Watt load... plus any of the resistance losses in order to get the amps and volts into the DC-DC in the first place ... so it still ends up being much more efficient to run multiple cells in series to get higher voltages.

NiMH only short themselves toward 0 Ohms at cell failure ... not cell reversal.

Originally Posted by crx_rogus

and thus probably avoid having a $40k+ car do the same thing Sony and Dell laptop computers did awhile back... have their lithium-based battery packs burst into bright, smoky "thermal runaway" states taking the value of whatever they're in with them.

yup.

Originally Posted by crx_rogus

It would be nice for the Insight to have something besides a way to gradually de-EV upon noticing subpack thermal issues, something to help counteract any developing out-of-balance issues.

People are working on it...
See thread: grid charger/balancer

Originally Posted by crx_rogus

Hmmm... NiMH cell failure = shorting itself out of the way, yet cell reversal can lead can eventually lead to exciting things possibly involving release of a cell's magic smoke (does it eventually short itself, or keep its radical overvolt condition?)...

Over voltage comes from over charging ... and the BCM is supposed to prevent that.... if it fails to do so... NiMH will tolerate some over charging ... better than Li ... but still limited.

Eventually if continued to be over charged further and further ... just like PbA ... NiMH will get very very hot ... then start to vent electrolyte as a gas that contains a mix of Hydrogen and Oxygen gas.

Venting Hydrogen and Oxygen gas into a hot area containing electrical current flow ... is asking for problems.... just like it is with PbA.

but all of that is just from over charging...

Over discharging is where you get voltage reversal.... this is kind of like putting a 0% SoC battery in a battery charger backwards ( reversing the polarity and forcing the charge anyway ) ... in a cell reversal you are forcing the + button side terminal of the battery to be a - terminal ... and you are forcing the - Flat side terminal of the battery to be a + terminal... you are just like the term sounds like ... you are reversing the voltage of the cell.

Cell Reversal will expedite a cells death into cell failure ... but NiMH is tollerant enough of cell reversal that it won't die fast enough ... the cell in reversal will end up being a liability event long before it ends up failing.

Originally Posted by crx_rogus

One issue with Insight 120S packs then is, I'm guessing, overcharging leads to relatively benign cell failure while chronic discharging getting out of hand can lead to catastrophic cell reversal.

Actually the BCM is a bit smarter than that... in order to prolong NiMH service life the OEM battery pack is not charged up to 100% SoC ... and it is not discharged down to 0% SoC... instead the BCM avoids the ends... and keeps the SoC closer to the middle... in this way it extends a NiMH battery cycle life from a few hundred cycles ... to many thousands of cycles.... which is needed for vehicular service considerations.

If the BCM sees / calculates that one of the cells is getting very near 100% SoC... it stops charging the whole pack before it gets there... if the BCM sees / calculates that one of the cells is getting very near 0% SoC ... it stops discharging the whole pack.

In order to try and avoid over charging or over discharging it instead reduces the effective usable capacity of the battery pack.

If the battery pack gets to low in BCM usable capacity it lights up the IMA error code light ... to indicate a problem ... Honda will then recommend you replace the whole battery pack.

Because the Ohms vary with temperature as shown before ... and because not all of the 120 cells are the same temperature under use... eventually some of the cells reach a different SoC than other cells ... even if they have not effectively lost any capacity at all.... but the OEM BCM was not built to rebalance the SoC... instead it just keeps on restricting the usable capacity.

Which is what lead to the IMA battery rebuilding service:
Hybrid-Battery-Repair

Originally Posted by crx_rogus

I'm also getting the impression that the Tesla Roadster among approaching others must have some astoundingly sophisticated BMS hardware and software, with a significant cost of the vehicle development just going into BMS development. A Roadster stranded with a flaming battery pack would produce unproductive PR, especially in the courts. My multicharger alone has all sorts of warnings about choosing the correct Lix setting stemming from the 0.1V cell potential difference between LiIo and LiPo.

Yup... lots of time , effort , and $.

But the #1 issue most people end up having with EVs is the range per charge / charge times... both of which are things that current modern Li have significant advantages in ... as long as you include a proper BMS.... and have a large enough budget for them.

Originally Posted by crx_rogus

PbA works well for my home's hobby-grade solar-recharged 12V subsystem (powering the 3 D->C cycle slow refreshing of the spare pack's subpacks effortlessly), and has reliably and safely and without stink (except near the recharging racks) moved countless tons of paper, ink and anything else imaginable for at least a couple decades at work, but getting folks to remember to recharge the battery, not the drive terminals, on some of the smaller equipment is always a challenge.

Yup... there is no one battery... they all including PbA continue to have pros and cons ... and each application will have needs that determine which is the best fit.

Originally Posted by crx_rogus

I've been gradually reading V2G-101, learning how V2G-enabled versions of cars like the (apparently truly heading towards real production) Chevy Volt (and current Tesla Roadster) will (/can) earn their owners / leasees up to about half their new cars' monthly payments back via buffer service micro-contracts with the local grid utility. But first, we need a smart grid. Flywheel, hydraulic and pneumatic -based hybrids also appear viable. There is indeed alot of really fascinating stuff going on out there, mostly off-topic for this thread.

Agreed.

[retepsnikrep]

I've got my car back in the workshop now for the final push and moving of cells to within the ipu compartment. Got to re-arrange quite a bit of stuff. Got all the gubbins out and on bench.

Going to relocate BCM, MCM behind seat panel.

Ditto Pack relay board.

Charger will fit down left or right hand side of IPU comp.

Lots to do.

[Mike Dabrowski 2000]

I got some information from the only Japanese MIMA owner.
He is driving a 1999 CVT, and is part of a group of Insight owners. It seems that they are also playing with batterys and BMS mods. Wish I could read the language, looks interesting.
on-bbs

[IamIan]

Originally Posted by Mike Dabrowski 2000

I got some information from the only Japanese MIMA owner.
He is driving a 1999 CVT, and is part of a group of Insight owners. It seems that they are also playing with batterys and BMS mods. Wish I could read the language, looks interesting.
on-bbs

Welcome to Google Translate:

Google Translate

Main Google Translate page is at:
Google Translate

From there you can enter any web site url and the language to translate from and to.

not 100% perfect... but better than nothing

[retepsnikrep]

Well I stood looking at the ipu compartment for hours, offering stuff up and moving things about and the cells just will not fit in without hacking everything about, so they will have to remain where they are in the wheel well. My 40ah ones are a great fit in the wheel well, but just too big to get all 50 in the IPU comp.

I've moved on to re-assembly and general tidy up + programming cells for Pulse width instead of serial data on Master Bus. Got to move Master and re do buttons at front etc. Still quite a lot to do. However I am determined to get BMS fully working during this overhaul.

[Mike Dabrowski 2000]

Sorry to hear that there is not enough space, that would have been a very elegant PHEV install.
Have you considered splitting the pack with some cells under the small cubby on the passenger side, and the rest in the IPU box?

[boelle]

could be fun if someone could trick the car to display correct SOC... ie know that the pack is 10x bigger

EDIT: how many cells could you fit inside the org. pack "cage"?

[retepsnikrep]

We can already make the car display the correct SOC using our knowledge of the RS485 METSCI bus but this involves another chip/board which I haven't had time to implement yet.

Also the ultimate plan is to remove the BCM, but that will have to wait until I get to MK2 of my BMS Master board, as it will have to have RS485 in and out to replace BCM. I still haven't had enough time to fully study the BCM analysis either. Sadly I have to work for a living. Perhaps If I won the lottery I would get on a bit quicker.

If I had gone for the 30ah cells they probbaly would have fitted OK.

I looked at lots of options for mounting, and removed the ipu compartment, the mounting beam, and moved the power module over to the left.

The critical thing IMO is physical security of the cells in event of an accident. (Can they be restrained in situ adequately) In the spare wheel well they are secure (Each block of 25 cells has six bolts) and in two blocks of 25.

It becomes a wiring and hardware mounting nightmare if you move them about and split them into smaller units/chunks. Yes it's a bit tail heavy in wheel well, but CG is very low thought which helps. It could be done perhaps if you were prepared to sacrifice/hack about the car, and make changes that meant you could never go back to the stock setup. I would have needed to remove the fuel tank as access to floorpan would be reqd to allow bolting of cell trays throught floor to secure them.

You also need a solution which enables them to be removed fairly easily for servicing etc.

I have 3 Anderson connectors, 4 D connectors and 12 bolts to remove and I can get all 50 cells out in about 15 minutes.