Honda CR-Z Hybrid Car Forums banner

61 - 80 of 263 Posts

·
Registered
Joined
·
339 Posts
Discussion Starter #67 (Edited)
Lithium Install first test and drive.

The install went pretty smoothly, it's all in the preparation.
Take your time and get everything ready before you begin, check and double check stuff.

I've done lots of IMA Lithium conversions but this was the first go with a CRZ, and there are several things I would now do differently regarding mounting etc.
I make no claims to be a fabricator and my intention here is prove the concept works not produce a piece of fine art.
Hopefully others with relevant skills can now follow my lead and make some significant packaging and mounting improvements.

62585


The major PITA is the removal of the rear interior trim, rear seats, conduits etc...
It took me a good hour and I did break a couple of clips. (Can be fixed.) I hate trim so it was not my forte.
I now have a trim mountain and won't be putting any back until it's thoroughly tested and I have finalised a mounting solution.

62586


The old pack is fairly heavy and won't come out no matter how much you yank at it, unless you remember to disconnect the phase wires! Doh.

I transferred the mounting brackets from this pack to my modified one and it went in very easily.
The empty pack is much easier to lift in than the old one filled with Nimh batteries was to lift out.

62587


All wired up and set for standard voltage range (48 LTO Cells) with no voltage or current hacking at this stage.
I added my BMS monitor device but need to add a termination resistor to the bus that I forgot.

I'm also still waiting for that wireless amp/voltmeter to arrive so I can monitor unmolested instantaneous info from the drivers seat.

After installing and switching on I cranked up my HDS to perform a couple of procedures.
This first switch on was crucial, would there be any errors? Phew not so far.

In the HDS I grabbed an IMA parameter data shot and completed the following housekeeping tasks.

1) Rotor calibration as I had used the MCM from the new pack not my old one. (If you swap in your old MCM you don't need to do this)

2) Battery replacement procedure which erases all the old battery info and resets the useable capacity to 75%. (My old pack had 34% useable.)
This useable capacity is not the same as the instant segment SOC.

3) I also set the instant SOC to 75% (8 segments) which is what I had charged the LTO blocks to before installing them.
You must leave some capacity headroom in any pack you install as the car will charge them immediately on starting to work out the SOC.
Also if you have to do the rotor calibration (takes about 5 minutes) it involves high rpm and warming the car up which might easily overcharge an already nearly full pack.

The car started instantly with the IMA and all the cell voltages are normal.
I went for a short drive to the supermarket about 3 miles without incident.

Performance as your would expect was very sprightly and upto full fresh new from the showroom unrestricted OEM levels.
Full SOC useable and instant capacity, optimal faked battery temperature, and lower internal resistance etc.

It did exactly what it says on the tin.

Longer road testing is now required to see how the car reacts to the seemingly bottomless capacity and perfect balance of the pack.
I was up very early as the weather forecast was good for working outside, it's been a long morning and I'm knackered, so may do a road test later or tomorrow.

I won't start the voltage/and or current hacking until I have a good handle on how it's working as is, so that will be a few days.

Video 10 (Lightweight IPU fitted)


Video 11 (Batteries lashed in and connected)


Video 12 (Running on the drive)


Thanks for the support and encouragement. Fingers crossed it keeps going ok and the hacking pays off later.

Peter
 

·
Registered
Joined
·
281 Posts
Very interesting thread!! I have enjoyed watching your modifications and watching and reading all the details. Great work!!
 

·
Registered
Joined
·
339 Posts
Discussion Starter #69 (Edited)
Lithium Install first test and drive part 2.

The attached pdf capture from the HDS with the Lithium pack can be compared with the Nimh capture I uploaded a few days ago.
You will see some significant differences and improvements, and we haven't started hacking yet!

As I had reset the MCM and told it the car had a new battery the assist and regen limits have been set to the maximum for the OEM system.
That's 13kw for assist and 10.1kw for regen. Previously my old Nimh pack was restricted to 9.2kw Assist and 5.7kw for regen.
I expect most people's Nimh packs unless very new will be operating in a derated fashion as they age.

13kw - 9.2kw = 3.8kw = 5hp of Assist and probably ~10nm of torque has been regained without any hacking, just by restoring full OEM capability.

The useable SOC is now 75% compared to my previous 34% useable SOC. That's over double the useable capacity.
Of course the LTO cells store much more than this, so we shall see how the system reacts in due course.

The 9.1k fake temperature resistors are giving a fixed battery temp of 27C which is right in the Goldilocks zone for max performance at all times.
The BCM Fooler gives us 16.2/3V for each voltage tap, so the car thinks everything is nicely balanced, again good for performance.

The pack is at 114V which is in the comfort zone for the car and the LTO.

I had to deliver the groceries to a friend in isolation which was an unexpected opportunity for a 20 mile test.
There are no issues or errors to report and the car has been very lively in all the driving modes.
It is eager to assist as you would expect and when it does you can feel it.
Upping the voltage significantly later should be very exciting.. :devilish:

I check the LTO blocks and connections for any signs of heating after the drive with my infrared thermometer.
Nothing increased at all and there was no IPU fan activity or heating in the fan pipes. :cool:

With the back open I have found I still have some high frequency hearing left!
I can hear the inverter IGBT pwm squealing when assist or regen ramps up and the volume is in proportion to the load.. . I quite like it..

I arrived back with seven SOC segments so probably around 65-70% instant SOC.
I killed a few dashboard trees with hard acceleration though, so need to work on my eco driving to resurrect them.

Not much more to say at the moment apart from it's in and works.

I'll have to do some proper non CRZ work tomorrow. :cry:
But do have to deliver some essentials to another isolator so will keep you updated after that run.

EDIT It looks like the MCM in the spare pack I bought is the same firmware version as the one that was in the car.
 

Attachments

·
Registered
Joined
·
339 Posts
Discussion Starter #72 (Edited)
Todays 25 mile grocery and prescription collection run was uneventful.
The cells remain very well balanced and SOC was going up and down normally. Performance is very good. :cool:

I plugged in my hacking gadget and the car accepted that and ran normally, so that bodes well for adding the extra pack!

In order to effectively use the large 20ah capacity of my packs ideally some method of faking or resetting the SOC needs to be found.
If you only use similar sized Ah Lithium cells to the OEM 5.75Ah Nimh then you won't have to worry about this as much.

In the G1 Insight we could control SOC directly as we had access to the separate BCM and MCM modules in that vehicle. (In the CRZ they are combined in one.)
In the HCH1 Civic we could repeatedly reset the SOC to high using HDS commands and one of my OBDIIC&C devices.

In the CRZ we may need to identify the CAN commands on the OBDII port that set the SOC, and just keep resetting to high until the Lithium pack is exhausted.

Of course you don't need to bother with any of this if you don't want too..

With a big 20ah pack like I'm using the cells will just cycle up a down in a very limited SOC range giving very gentle shallow cycling and good longevity.

We do need a way of correctly disabling assist or regen in case the Lithium pack gets too high or too low.
You cannot just chop the main IMA power with a primitive BMS. You have to interact with the car properly and use one of it's systems.

Several ideas spring to mind.

1) Does pressing the clutch when driving in gear disable regen/assist or both? I need to check this, it does on the G1 Insight.
It's easy to electrically parallel a second switch with the OEM clutch switch.
If the BMS gives a voltage warning simply throw the switch and the car may stop regening/assisting or both.

2) Interface with one of the battery temp sensors and if battery voltage gets too high or low spoof a hot battery scenario and the IMA will shutdown.
The disadvantage of this is the IPU fan will run like a leaf blower when active.
Experimentation with values and temps will be need to find the sweet spot of cutting assist/regen without setting a code.

3) I'll think about it..

Going to fix my trim clips tomorrow as weather looks bad. :cry:


A little wrinkle to increase regen in all cars when coasting down etc is to parallel a switch across the brake pedal switch that feeds the ECM.
When you press the switch the car will regen at the higher level without you having to press the brake at all.
A nice flappy paddle switch would be good. But I used a simple non locking toggle switch on the Insight and Civic.
 

·
Registered
Joined
·
13 Posts
Mine was low milage with a poor service history but when I serviced it myself all the filters had been done recently all like new so I kept them for next time
 

·
Registered
Joined
·
201 Posts
Depressing the clutch does disable the assistance and charging on my 2013 car so probably also on yours. Therefore easy to modify this to give the effect of pressing the pedal when the batteries overheat. (there are 3 switches on the pedal - up, not up, fully down. I will try and determine which pedal position disables the IMA tomorrow.
I'm still working to find which of the Can Ids I've logged so far do what. I've found some ABS and DASH ids as well as engine, but none yet for the IMA.
Surely the MCM will determine the SOC from battery volts and then send this to the gauge, so just altering the can data might just alter the dash display? The voltage will not drop as quickly with the larger batteries so there should not be any need to alter anything.
 

·
Registered
Joined
·
339 Posts
Discussion Starter #75
Depressing the clutch does disable the assistance and charging on my 2013 car so probably also on yours.
Therefore easy to modify this to give the effect of pressing the pedal when the batteries overheat. (there are 3 switches on the pedal - up, not up, fully down.
I will try and determine which pedal position disables the IMA tomorrow.
If that holds true for mine, and keeping the clutch depressed when driving does not result in a fault code then that's sorted and can be easily faked if the battery gets too hot/full or empty.

I'm still working to find which of the Can Ids I've logged so far do what. I've found some ABS and DASH ids as well as engine, but none yet for the IMA.
Keep at it..

Surely the MCM will determine the SOC from battery volts and then send this to the gauge, so just altering the can data might just alter the dash display?
The voltage will not drop as quickly with the larger batteries so there should not be any need to alter anything.
The MCM counts current in and out of the pack as well as looking at the total battery voltage, the voltage tap balance and the temperature.
It does not base it's SOC determination purely on pack voltage. It knows exactly what X ah has been counted in or out.

If you keep assisting with a big pack like mine the OEM SOC will deplete but then hang for a while at a low level around 30%, while
the MCM waits for one of the OEM voltage taps to go out of balance when a NIMH cell reaches empty and it's voltage plummets in comparison to the rest.
It then cuts assist and force charges for some time.

Of course in my 20ah balanced pack with the BCM Fooler a tap will never go out of balance.
So the MCM continues to count current out until eventually it says it must be empty now whatever the battery pack voltage and taps say, and it cuts assist as above.

In the Insight this happens after about 3 to 4ah have been counted out beyond it's predicted empty point.

The MCM does send the SOC to the ECM and the Gauge segments cluster.
 

·
Registered
Joined
·
201 Posts
So the MCM continues to count current out until eventually it says it must be empty now whatever the battery pack voltage and taps say, and it cuts assist as above.

I thought you were shunting the current sense resistors so the MCM to alter its prediction of 'capacity too low.'
 

·
Registered
Joined
·
201 Posts
So as I understand the MCM operation, in simple terms, it monitors banks of cell voltage and using a current sensing shunt, the total amps out of or into them. Depending on engine load (throttle position) it then modulates the current out via the pwm duty cycle of the IGBJT gates.
You mentioned in an earlier post, a voltage centered around 2.5v which varies either way depending on charge or discharge. This can easily be recalibrated to fool the MCM - something I think you did on the Insights, so is it worth a try on the CRZ?
 

·
Registered
Joined
·
339 Posts
Discussion Starter #78 (Edited)
The current sensing systems will be modified later to reduce the sensed current. They are unchanged at present.
Modifying the current sensing is mainly being done to encourage the MCM to push more current for more power.

A side effect of say a 10% reduction in sensed current will be the SOC goes down 10% slower, that's good for me.
5ah actually expended would be counted as 4.5ah, however that's a tiny amount compared to the capacity of my pack.

So the issue in the previous post still exists albeit slightly reduced..
The SOC will count down, then hang, then reset to empty when the MCM says enough already..

If we are very lucky and fit a bigger IMA fuse we might achieve a 30% current sense reduction in due course...
A 30% reduction in sensed current means a 40% increase in current/power from the system.

Remember the MCM control logic uses current and voltage to achieve and maintain a requested Kw power level.

Let's say the system (ECM) is requesting 10kw of assist.

At 100V that would be 100A.

We reduce the MCM sensed current by 30% so it only see's 70A of the 100A that is actually flowing.
It determines 100 V x 70A = 7kw 'I'm under target', so it ramps up the current.

In order to get back to 100A from 70A it has to increase the current by ~40%
However we are actually already at 100A so we get 140A

If the batteries can sustain the extra load without sagging then we end up with about 14kw a power gain of ~4kw or about 5hp. :eek:
 

·
Registered
Joined
·
201 Posts
Makes sense. I wonder if the MCM would complain if the ECM asked for 15Kw (by a MITM changing the ECM's assistance request)?
 

·
Registered
Joined
·
339 Posts
Discussion Starter #80
Yes it probably would complain and set a fault code.
The assist /regen requests in the Insight and Civic had to be within the allowed range or they faulted.

That doesn't matter though as we will be increasing power using voltage and current hacking.
If we can develop a MITM device then we can scale requests from the ECM up or down (within the limits) as appropriate.
The Lithium and Nimh CRZ will likely have the same numerical range in the CAN data for assist/regen but it will be scaled in the ECM or MCM to account for the 2kw power difference.
 
61 - 80 of 263 Posts
Top