Honda Insight Forum banner

61 - 80 of 120 Posts

·
Registered
Joined
·
6,745 Posts
My 2 cent conclusion/opinion based on stuff written thus far:

Despite what me and Peter wrote about potential limitations of the 2.9Ah cells, I still think trying that option has the most research value. We know what the 20Ah SCiB cells do - and we already have an option for that (the FIT packs).

The 10Ah cells are the same physical size as the 20Ah, so there's really no point to those, i.e. there's a lot of finagling needed to get the 20Ah to fit, so if you've got to do that might as well pay a little extra and get larger capacity cells...

IF you have latitude in terms of what your supplier builds, at this point I think 6-cell modules would be best:

-each module would have busbars welded to the terminals, same/similar busbars as the Fit packs
-end busbars would have wire attachment points, for voltage taps to be connected to OEM
-end busbars would allow attachment to adjacent modules, to form the series string
-I think each busbar should have an attachment point for cell-level monitoring, basically a 6-cell harness with JST connector, just like the typical Lipo...
-we should be able to attach large 4-6 AWG ring terminals or similar to the module-end busbars, so we can split pack in middle and connect at ends, connecting to stock electronics board main current carrying terminals...

As Jime mentioned, the cells need to be electrically isolated from one another - hopefully you're supplier would know that much...

I'm not for the Dow Kokam LTO cells - no "soft packs." Plus, if they ARE Kokam cells, well, those are different, right? - not SCiBs...

On the electronics 'magic box' side of things, with this configuration you'd need a top voltage limiter (Peter's BCM Interceptor, I think, would do for the time being), or to manually avoid over-voltage.

A 7-cell module would have less of a need for this - the max in car is 192V, but you rarely see that; 2.7 X 70=189V, probably close enough. BUT, a 7-cell module also wouldn't charge enough within the stock management window, so you'd need additional electronics to sort that out...

IF you went with a larger string, say your 80 cell version, you'd need something more complex, other of Peter's devices, perhaps just the BCM Interceptor, BCM Fooler, and a couple resistors in a couple places. Or maybe the 'BCM Replacer'. Peter could answer this part better than me... You need an OBDIIC&C, too...

I think that's it.

The modules would fit in the stock case between the stock 'racks' that hold the NiMH sticks, they'd be accessed from the rear, where the fan shroud is (your image of the case posted earlier is a gutted case, it has no 'racks'). I think it'd be easy to make trays that hold the modules out of widely available PVC 90 degree angle stock.

edit: also, I must have tested the fitment before, because I'm pretty sure you can't fit two long strings of the 2.9Ah cells within the stock battery case. I think Jime alluded to this earlier. They need to be split into at minimum 3 or do the small module approach, the 6-cell modules, as I described. There could be alternative configurations if you gut the case, but that's just one more probably unnecessary step. Basically, if you do 6-cell or whatever smaller modules, it will be easier to finagle fitment.
 

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #62
A few years ago, I put together a basic info sheet for a client (attached below). This is not to insult anyone in this thread, it's meant for those new to the Li-ion world who are monitoring all these Li-ion related threads to make sure they have the industry-standard terminology right. Maybe we should post these basic info sheets, esp yours on electronics in another thread for the beginners to download, instead of having them trying to dig out key info pieces among the millions of posts in thousands of threads. Maybe we can call it Technology Basics. I have many many more info sheets on Li-ion cells to post in the future.

In my opinion (IMO), some Insight owners care about the original look of the car inside-n-out. For them, the battery pack (OEM or new) has to fit in the original space. For others like me (function over form), I don't mind a larger battery pack or the rear deck not looking original; as long as I have a reliable-running car with good mpg, no more battery replacements, and without spending more than US$1.5K. The right Li-ion pack can outlast the ICE and other expensive major components.

I'll be happy to test different batteries and configurations if you can supply the mating electronics. I think we should have more than one test vehicle. Maybe one of the cars have other problems that could affect the data. I'll be happy to start with my car with two 40s1p 2.9Ah LTO packs in series. If necessary, I can increase to four 40s1p. My guess is the brake regen will be sufficient to keep the pack voltage up high enough to avoid the system to throw the towel in. Average daily mileage in the U.S. is around 30 miles. Around 65% of Americans own homes. So, the homeowners can grid-charge at night if they do lots of driving each day. Many Condo owners have access to grid power where they park their cars. Apartment dwellers can do grid charging (if really necessary) at work or at public charging sources. More and more free chargers are at American shopping centers, nowadays.

If the batteries can handle the high inbound current, then more regen energy can be captured. Hopefully, we can come up with a power pack that will not require any grid charging to top up or balance the cells. I can get the cells to handle the current and the BMS to do the proper balancing. We will need guys/gals to help with the rest of the electronics to interface with the OEM electronics to work as originally intended (e.g. dash displays, cooling fans, brake regen, etc.). If all the new electronics can be packaged in one box, then the DIY'er just need to plug in the OEM lines and cables.

If the fear in using a small pack (e.g. 80s1p 2.9Ah) is due to low voltage cutout, then I'd be happy to carry or wire-in a 35s1p pack of NMC 18650 cells (192V) to do a manual jump-start to get the internal charger working again. My easy-style of driving should not ever deplete the pack down below the cutout threshold voltage. I'd like to test the small pack first before we all firm up on a Mammoth pack. Trying other chemistries is also OK, but I'm leaning towards LTO for long life.

While it's easy for me to put a pack together and I know how to solder, I just don't have the time to tinker with the various loose boards. Any chance I can trouble some of you experts to put together a single magic box for me to plug in all the vehicle's cables and wires? I know extension harnesses may be required, but I'll be happy to pay for all the hardware. This can be a first-run in prototyping our Translator or Magic Box for future DIY'ers. We need to save more Insights from junk yards. Please excuse my long-winded posts.
 

Attachments

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #63
Please excuse the quality of the photos below. For me, these types of photos will help a beginner DIY'er (like me) to start to sort out the connections to connect to our Magic Box. For us, it's a good way to discuss new battery case size and design. The multi-views show the available space for potential extra battery modules, BMS, the Magic Box or any other components.

I was even thinking about a fold-able solar panel array covering the rear deck to trickle charge the power pack when out in the sun (solar cells will deliver 50% power on cloudy days). If the pack is at 90-95% SOC, then the solar power will operate the fans to cool the pack off. This might be another thread later on solar cooling and pack top-up.

86039


86040


86041


86042


86043
 

·
Administrator
Joined
·
11,583 Posts
Re the photos don't forget the OEM temperature harness/and or equivalent sensors and connector including PTC fooling resistor.



There really is no single magic box that can be assembled at present.

In order for the car to function with lithium certain systems have to be fooled or modified.
They aren't necessarily adjacent to each other physically or electrically so a single magic box can't work.

For your most basic testing of the 80S1P pack you will need a BCM fooler with pre-resistor, a VPIN resistor and PTC resistor. They all go in seperate places in the harnesses, so no magic box they have to be installed into wires.

With that cheap setup you will have no automatic voltage control or monitoring capability whatsoever. I don't know how you intend to monitor to protect your pack or cells..

Of course monitoring, protection and control can be offered by things like the BCM Interceptor, OBDIIC&C, IMAC&C P&P but your 1.5K budget then probably becomes inadequate.

To automatically protect against pack over or under voltage and perhaps have some control over SOC you really need a BCM Interceptor setup specially for your pack parameters. Again that needs installing in harness wires or inside the MCM. Of course the BCM Interceptor cannot do anything about cell imbalance that's down to your BMS.
 

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #65
What will be the cost for:

  • BCM fooler with pre-resistor
  • VPIN resistor and PTC resistor
  • BCM Interceptor
  • OBDIIC&C
  • IMAC&C P&P
Including instructions on where to install these components.
 

·
Administrator
Joined
·
11,583 Posts
IF you went with a larger string, say your 80 cell version, you'd need something more complex, other of Peter's devices, perhaps just the BCM Interceptor, BCM Fooler, and a couple resistors in a couple places. Or maybe the 'BCM Replacer'. Peter could answer this part better than me... You need an OBDIIC&C, too...
For the most basic 80 string testing! i.e. Does it start and drive up/down?

Then the BCM Fooler and three resistors VPIN,PTC etc would do. $65
I'm assuming you have your own pack voltmeters etc.

For more extensive testing and to automatically prevent pack over/under V and have some SOC control add the BCM Interceptor. $250 (It will have to be programmed specifically for your pack parameters)

To monitor what is going on/ fault find/clear codes etc add an OBDIIC&C display. $350

Manual IMA control IMAC&C P&P $250

The BCM Replacer is not ready for use in such a new setup.
 

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #67
For more extensive testing and to automatically prevent pack over/under V and have some SOC control add the BCM Interceptor. $250 (It will have to be programmed specifically for your pack parameters)
I watched the YouTube segment on the BCM Interceptor, not clear if it is necessary if I already have a BMS. What are the other functions?

You suggested two 40s1p modules using the 2.9Ah cell to be connected in series. Will the power in and out go through both boards in series? Or, will I need a BMS for each module?
 

·
Administrator
Joined
·
11,583 Posts
How will your BMS control power?

I have already stated you should not use a BMS that tries to cut current using power MOSFETs itself. If you do that will certainly generate ima fault codes and may damage itself or the car.

You need a simple BMS with logic high and low outputs than can interface with the BCM Interceptor.
 

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #69
The simplest BMS that we normally use only allows the pack to be charged up to a certain voltage level or discharge down to a certain voltage level, but that also uses MOFSETs for switching. It's the more precise way to do this function. I'll keep looking. I'm surprised that the FIT LTO pack BMS is compatible. If your BCM Interceptor can also act as a BMS, then I won't look for 1-2 BMSs for this 2X40s1p pack.

Another one of my suppliers just told me that the Toshiba 10Ah cell is no longer available to them either.

The other concern for a 2.9Ah pack is its capacity for vehicle acceleration. Using the maximum numbers provide, at maximum acceleration of 85A for 3-sec in each of the 5-gears, this comes out to be around 354 Wh of expended energy. I normally don't accelerate very hard, so if decreased by 50% in power output, then energy consumption will be around 177 Wh to 90 kmph. The 80s pack is only around 556 Wh, so one can't be too liberal with burning tires.
 

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #70
I might have found one BMS supplier that can supply a BMS without MOFSETs, but the two 40s need to have a safety disconnect (MSD) and fuse between them.
 

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #71
According to my BMS buddy, he said, you can only limit charge without MOSFETs if you have a smart charger which will receive charge data from the BMS and not charge by itself. The BMS sends the charge parameters, including current, to the charger. During balancing, the charger has to be able to provide very low current or it will overrun the balancing circuitry. The BMS would control two main contactors to cutoff power to the vehicle, but it can't control CHG/DIS separately this way.

So, let me know the issues you have with BMSs with MOFSETs, and see what he says.
 

·
Administrator
Joined
·
11,583 Posts
If you connect your two 40S packs to the switchboard connection points we have specified there will be a safety fuse and switch/breaker between them. It's part of the OEM system.

Re your BMS. I'm not really discussing grid/mains charging your pack here, that is another level of complication.

I'm talking about you braking and the car trying to shove 50A into your cells when one or more may be full.
Or you accelerating and the car trying to draw 80A from your cells when one or more might be empty.

Your BMS needs to signal (via 5v logic level or opto isolation) the fact a cell is empty or full, and then the BCM Interceptor can turn off regen or assist correctly to protect the pack/cells without throwing IMA codes.

If you have a BMS with its own power controlling mosfets then the car will throw IMA error codes whenever that operates and tries to cut power. It's as if you suddenly threw the breaker switch when regening down the hill at 50A. The regen voltage will suddenly spike very high, maybe 300V+ plus depending on rpm, throw an IMA code and possibly damage the BMS or the IMA system.

I suggest you organise the zoom meeting and have your BMS buddy online then we can all chat together.
Passing messages back and forth isn't going to work.
 

·
Administrator
Joined
·
11,583 Posts
To be clear the BCM Interceptor when programmed with the correct voltages can offer protection against pack over or under voltage on it's own.

I.e. Pack over 200V disable regen or Pack under 140V disable assist.

However the BCM Interceptor cannot protect against cell level events unless an external BMS asks it to do so.

The BCM Interceptor is not a BMS.

It's an interface that allows you to correctly and safely enable/disable regen or assist and manipulate the reported SOC, battery temperature, current and voltage.
 

·
Registered
Joined
·
6,745 Posts
The other concern for a 2.9Ah pack is its capacity for vehicle acceleration. Using the maximum numbers provided, at maximum acceleration of 85A for 3-sec in each of the 5-gears, this comes out to be around 354 Wh of expended energy.
I think your calculations must be wrong. I calculate ~13.6Wh for 3 seconds, and then however you want to multiply that for "each gear." But you only get that assist level in 2nd and 3rd, so it'd only be X2 at max. Still, only about 27Wh.

The concern when it comes to max assist and the 2.9Ah cells is how large (small) a charge state range the cells will be able to deliver it. For example, it's possible the cells won't deliver it under around 20% charge state. If that were the case and getting full assist across a large range were critical, then you've shrunk the usable capacity window by 20 points. Something similar happens on the top end with regen. So once you clip the top and bottom to accommodate full assist and full regen, the usable capacity range shrinks from 3000 to 0.6 X 3000=1800mAh...

The actual numbers may vary as we can't be sure how the stock BCM will handle the cells.
 

·
Registered
2000 Honda Insight, manual transmission
Joined
·
77 Posts
Discussion Starter #77
My BMS buddy passes on these thoughts:

"I hear what he's saying, and he's right. You can't cut FETs during regen, but you also can't OV the cells. Here's what I suggest: BMS will signal no regen capability when the cells are near top-of-charge voltage. For LFP, you need to signal this near top of SOC (not voltage, voltage is not reliable for LFP). For that, you need a smart BMS. I'm not sure how the original pack does it; it sounds like this guy made a BCM interceptor that handles this already, but it needs some 5v/opto-isolated signal from your BMS to tell it when to disable gen/regen. You just need a programmable BMS which knows SoC and can message that to the interceptor."

Your thoughts? Can the FIT BMS handle 40s?
 

·
Administrator
Joined
·
11,583 Posts
The FIT BMS from the 20ah LTO packs is not practicable for other packs for several reasons.

1) It doesn't have a working balancing function as yet. You probably need that with such small capacity cells.

2) It is only available in very limited number if salvaged from LTO FIT packs. (Most LTO FIT pack owners will want to keep the BMS on those packs so they can read cell voltages)

3) It is 12 cells per module and each module has a unique CAN ID, you would need 7 modules all with unique ID's, that not going to be cheap.

4) The physical layout of the FIT BMS is designed specifically for the 20ah LTO FIT blocks.
It would be very difficult and complicated to mount and wire these to any other sort of pack.
 

·
Administrator
Joined
·
11,583 Posts
For LFP, you need to signal this near top of SOC (not voltage, voltage is not reliable for LFP).
I don't know if this is correct? Links/sources?

It sounds like this guy made a BCM interceptor that handles this already, but it needs some 5v/opto-isolated signal from your BMS to tell it when to disable gen/regen.

You just need a programmable BMS which knows SoC and can message that to the interceptor."
So we are back to the 'Orion2' for your 80 cell LTO/LFP which is very sophisticated and probably counts as a smart BMS. That can certainly talk to the BCM Interceptor.

'Simple drop in Lithium' for the Insight for a reliable saleable commercial product isn't simple or cheap.
 

·
Administrator
Joined
·
11,583 Posts
For now probably just stick to a simple high/low cell voltage indication BMS.

All these cheap BMS modules have current control power MOSFETS on them which cost money.

You/we don't want the mosfets bit.

We just want the detection and signalling part of the BMS boards that tells the mosfets when to turn on/off etc.

We can probably do the rest.

It might be possible to just remove the mosfet section or bypass it one of these cheap mass produced boards.

Buy the lowest current one you can. A 20S costs about $25 from China..

You might still have to top balance from time to time with very low current charging that doesn't overwhelm the cell bypass resistors. That sort of charging can't be done by the car in normal use so something else for consideration.


Or maybe don't bother with a conventional BMS, but use an autonomous active balancing module.

e-bay active balancing module

Fit them and forget.

These transfer energy around the pack once a cell goes more than 100mv off.
They stop balancing when all cell get back to within 30mv.

If you knew all the cells were within 100mv of each other then you could just use the BCM interceptor and pack level voltages to protect the pack.
 
61 - 80 of 120 Posts
Top