Honda Insight Forum banner

1 - 20 of 42 Posts

·
Administrator
Joined
·
11,603 Posts
Discussion Starter #1 (Edited)
We have had successful LTO 48, 60, 72, 84 cell conversions. :)

But can we go for higher voltage, ultimate power potential and fit one with 96 cells?

Can it be done reasonably easily and how? OK some random ramblings..

8 blocks of 12 would be very difficult to fit in the OEM pack space.
I'm not going to even attempt to fit all the packs in there.

I'll use one of my 48 cell setups as it and just reconfigure the high power wiring.
I'll but the other 48 cells in a block in the spare wheel well pushed as far towards the front of the car as possible.
I'll mount them on a dexion or aluminium tray bolted into the wheel well like I did with my first Lithium conversion in 2008!
This will still give some crush area behind them in the event of a rear end accident.

85847


85848

The standard car detects voltage at three points..

1) The 10x BCM voltage taps. (Pack voltage / 10)
2) The MCM high voltage input. (Full pack voltage)
3) The MCM 0-5V VPIN input. (0-5V)

1) We can use the standard 10k BCM Fooler resistor matrix to simulate a perfectly balanced pack.
We can use an extra say 20-40k resistor in the + feed to the BCM Fooler to drop the pack voltage down to the acceptable system range.

2) This is covered by 1 above and is fed from the same BCM fooler so not a concern.

3) This as we know can be fooled by the insertion of a simple resistor into the wire, equivalent to the one used in the pre- BCM fooler + input, so maybe 20-40k and adjusted as necessary.

So fooling the IMA electronics that the pack voltage is in the acceptable range is very easy with simple analog techniques..

Other considerations.

How high can the stock DC-DC actually operate?
IIRC it is upto about 220V before it shuts down, so that's probably going to be problematic.

The IMA pack actual resting voltage might easily be 230-240V when the pack is fully charged.
That's too high for the stock DC-DC which will shut down.
So it's almost certain a Meanwell PSU will have to be used in it's place.
This will easily accept and operate at this higher range, plus it's smaller and lighter than the stock unit.
No pre 3 phase to 1 phase rectifier will be required as the IMA system IGBT performs this function.

Specs.

A 96 cell pack would have about 5kwh of capacity

240V would be 96 x 2.5V and fully charged.
182 would be 96 x 1.9V and empty.

So if you're interested in the results and dyno testing later bookmark this thread.

PS

Anyone got 3 or 4 LTO Fit block busbars and plastic clip on covers they can pop in the post to me in the UK please? Let me know price inc postage.
 

·
Registered
Joined
·
793 Posts
@retepsnikrep I have the same question.

I am wondering if there is room for a fourth LTO pack outside of the IMA compartment, under the Tom Mix bar and behind the driver (which side depends on which country.)

From a handling perspective, getting the weight forward may be of greatest benefit.

The electronics mods are not my greatest concern. What I wonder is whether the IMA motor is capable of producing sufficiently high voltage and enough current at different RPMs to keep the pack charged at the highest voltages.

If it can produce the voltage but not the current, it may be time to look at the CV shaft and motor mount mods needed to add a second IMA motor (which would kind of be the point of putting in so many packs - you don't need as much current to drive two motors if you have more voltage).

It would be valuable to have a three dimensional chart of voltage, current and RPM produced by the IMA, as well as motor winding temperature at each of the data points. This would require running the car on a dyno, with the IMA motor connected to a rectifier and a variable resistive load (that would be a LOT of heat generated) to find not only whether enough voltage can be generated, but how hot the motor gets when producing different amounts of current at different RPMs.
 

·
Administrator
Joined
·
11,603 Posts
Discussion Starter #3
I am wondering if there is room for a fourth LTO pack outside of the IMA compartment, under the Tom Mix bar and behind the driver (which side depends on which country.)
Sorry not going to even look at that for this proof of concept testing.

What I wonder is whether the IMA motor is capable of producing sufficiently high voltage and enough current at different RPMs to keep the pack charged at the highest voltages.
I'll find out in actual use and road testing.
No point IMO rigging up a complicated test setup the car already has one built in.
If the packs slowly discharge on a journey then some simple tricks are possible to encourage charge more and assist slightly less.

It would be valuable to have a three dimensional chart of voltage, current and RPM produced by the IMA, as well as motor winding temperature at each of the data points. This would require running the car on a dyno, with the IMA motor connected to a rectifier and a variable resistive load (that would be a LOT of heat generated) to find not only whether enough voltage can be generated, but how hot the motor gets when producing different amounts of current at different RPMs.
I def won't be doing that unless I win the lottery.
 

·
Super Moderator
Joined
·
8,156 Posts
It will be an interesting experiment. I'm wondering if there is an actual a way to get the IMA motor/gen to regenerate at voltage that high, but perhaps this is not what you have in mind:)

I'll check my busbar supply. I know I have a couple for sure. You might need to hit up Atheos for one. He has one bad block, so by definition he has one busbar.
 

·
Premium Member
Chicago & Detroit
Joined
·
847 Posts
Interesting experiment but I'm not sure of the end game other than "conquer the beast".
 

·
Registered
Joined
·
793 Posts
I love a good engineering challenge - it's too bad I'm still focused on more mundane mechanical stuff and haven't shifted to IMA stuff, so I will have to live vicariously through you and others! :cool:

I am wondering if there is room for a fourth LTO pack outside of the IMA compartment, under the Tom Mix bar and behind the driver (which side depends on which country.)
Sorry not going to even look at that for this proof of concept testing.
Just looked, but can't measure at the moment. It might work. It would certainly be better for handling and general ride to move the 35 pounds from behind the rear axles.

What I wonder is whether the IMA motor is capable of producing sufficiently high voltage and enough current at different RPMs to keep the pack charged at the highest voltages.
I'll find out in actual use and road testing.
No point IMO rigging up a complicated test setup ...
I wonder if you can find this out without making any mods - disconnect the motor and measure the raw 3 phase voltage at different RPMs?

If the voltage is found to be sufficient, the next step is to see if the current is sufficient. I wonder if hooking up a Saleae and driving the car while measuring OBD and xSCI data would unlock the IGBT interface. Or maybe @mudder has done this?

<EDIT> Of course he has, long ago. His full IMA system schematic is worth reposting. (Peter, while searching I found you posted that you had this laminated. What a great idea!)

Mudder's IGBT module diagram aligns with the diagram for what is likely the very similar PM200CVA060 or PM300CVA060 on Mitsubishi's site (the connectors and package differ, at least).</EDIT>

Then you could control a spare IGBT with a PIC or Arduino, connect it to the motor, and configure it to dump current from the P/N terminals into a low resistance shunt. Use spare current sensors to measure the current?

I have a feeling this isn't difficult and would leave you with a nice PIC- or Arduino-controlled IGBT module that could be employed for all sorts of cool experiments/tests/projects.

It would be valuable to have a three dimensional chart of voltage, current and RPM produced by the IMA...
I def won't be doing that unless I win the lottery.
You mentioned dyno testing in the earlier thread, which would be the perfect place to do this kind of measurement. But then again, maybe no dyno is needed with aforementioned IGBT load tester.

I'm really looking forward to seeing what kind of hard data comes from this experiment! (y)
 

·
Registered
Joined
·
793 Posts
@minor4326 thinks aloud:
Interesting experiment but I'm not sure of the end game other than "conquer the beast".
I was thinking about this for several applications:
  • Power source for an RV towing the Insight. Moar power is better...
  • Driving out west: allows more use of the IMA on a grade when there is a lot of regen opportunity on the back side
  • Second IMA motor mod would need more voltage to reduce the current demands
  • Model 3 rear end?
yes, this is dreaming, but that last one probably would probably conquer any beast hiding under the bed...
 

·
Registered
Joined
·
793 Posts
I realized that I have all the parts I need to do this experiment (measure voltage and current at different RPMs) sometime in the next few weeks or month.
 

·
Administrator
Joined
·
11,603 Posts

·
Administrator
Joined
·
11,603 Posts

·
Administrator
Joined
·
11,603 Posts

·
Premium Member
Chicago & Detroit
Joined
·
847 Posts
Your back must be better than mine, hefting those backs into the two different locations. 😢
 

·
Administrator
Joined
·
11,603 Posts
Discussion Starter #15
96 cells is actually a bit of a PITA because the voltage when full at 2.5V cell (240V) is far too high for the stock DC-DC.
That will only tolerate upto about 220V before shutting down.

(It really might be useful to disable the high voltage checks in that, but I have looked and could not find the relevant circuit.)

It will have to be replaced in this 96 cell setup with a Meanwell unit.
That's why 84 cells is a more realistic maximum. (Being 30V less) 210 v 240V

That's probably the job for later today.

Now I may simply disconnect the stock DC-DC but leave it in situ
and just mount the meanwell somewhere handy as a temporary measure.

This is a maximum power experiment after all, and if I don't like the results (like the 48 cell version), then It will be easier to revert without having to pull out the MDM to re-install the DC-DC etc.

For power testing my intention (if it doesn't blow up) is to visit a local rolling road.

I will do two power runs.

One with IMA disabled to get a baseline for my gutless 320,000 mile engine.

One with batteries warm, full and everything on and +40% current, imac&c etc.

I would expect the pack too hold up at about 215V (2.2V Cell) at 130A load or so.

That equates to about 28kw. Well we shall see in due course. :)
 

·
Administrator
Joined
·
11,603 Posts
Discussion Starter #16
Ready for install. CAN bus tested and working. Cell deviation 20mv.

Wheel well pack assembled, holes drilled etc, just need a lift with it all.

 

·
Administrator
Joined
·
11,603 Posts
Discussion Starter #17 (Edited)
96 Cell LTO in and running. :) First test report!

Well as per title. It's all installed securely and safely. Started first time with the IMA. :)


Note. The video was shot before I plugged in the dc-dc and went for the test drive!

As the pack was under 220V I plugged in the stock DC-DC for a quick ten mile road test.

The stock DC-DC cuts off at ~225V approx 2.3V per cell so it will have to be changed for the Meanwell so I can charge the pack to a higher voltage and nearer 100% SOC. That's probably tomorrows job.

I will remove the stock DC-DC and mount the Meanwell on a plate in the space left by the stock converter. I don't want the Meanwell outside the IPU and I also want to lose the weight of the stock DC-DC. So I will take the MDM out tomorrow to sort this.

Sat on the drive ticking over the car was quite happy to charge at around 215V 8A or so.
So the high pack voltage doesn't seem to be a regen limiter.

The OBDIIC&C Canbus worked fine after I changed the parameters from 48/96 cells and added the extra block ID's.

The IMA light is on (non fatal code) as assist/regen etc all still work.

I need to investigate the bypass contactor and might need to speed up the pre-charging as I suspect the car is seeing a delay in it reaching it's expected MDM voltages. I'll talk a bit more about this in later posts.

On the road the car worked normally, IMA start, autostop, assist, regen, LB etc etc.

Assist with the +40% current hack, IMAC&C P&P WOT max assist mode was about 105A at 210V so an easy 22kw!
This is without the switched low voltage fixed 120V hack :) So that's another potential 30A 6-7kw LOL.

Importantly we didn't have the low rpm dead zone we got with the low voltage 48 cell setup..

Regen was strong on coast down and manual mode. 60A at 220V so that's 13kw or so..

The higher pack voltage means the currents are a lot lower so the 175A fuse is less stressed etc.
Pushing for more power is much easier than with the low voltage setup.

You can feel the weight in the back of course, but it's no worse than the first 40ah lithium pack I did over ten years ago, when they were all in the spare wheel well on a very similar dexion mounting.

You can see from the video we still have a decent crush zone before the batteries get involved.
The wheel well mounting is very secure with big load spreading washers etc.

So more testing and tweaking over next few days/weeks. Thanks for your interest.
 

·
Administrator
Joined
·
11,603 Posts
Discussion Starter #18
It's easy to make mistakes and forget things. LOL

I might have been a bit silly here. :unsure:
My non disabling IMA code is P1444' High voltage short circuit' which means voltage leaking to the car body.

I overlooked/forgot about the fact the LTO cases might be live.
Not sure about the metal mounting tabs? I'll check it out.

Might have to revise the wheel well mounting and insulate it from the car.
 

·
Super Moderator
Joined
·
8,156 Posts
Don't you have the packs sitting in their normal plastic trays? In that case, it isn't likely the live cases causing the P1444.

The tabs are totally isolated from the cells, so they can be screwed down to the chassis. In fact all the aluminum plate versions essentially do that.
 
1 - 20 of 42 Posts
Top