Honda Insight Forum banner

161 - 180 of 202 Posts

·
Linsight Designer
Joined
·
1,877 Posts
P1576(10): VPIN is less than 0.156 volts for two seconds.
P1576(11): VPIN is more than 4.756 volts for two seconds.
P1576(12): VPIN differs from "VES" by more than 10 volts for two seconds.

So yeah, we need to verify the voltage is the same as measured by VPIN (PDU), BCM, and MCM. But you probably knew that already.
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #162
Well I did say I was tired and we make mistakes! :sleep:

Looked at it again this morning, :rolleyes: realised I had added my +48V at the IGBT pre driver, but forgot to change my MCM for the straight through VPIN version, and it was also adding +48V via the summing opamp. :mad:

So the voltage differences between VPIN and MCM HV/BCM etc were +48V. Oh dear..

Anyway swapped MCM and it started and ran fine.

It works as expected without IMA errors but the low current 'feature' is still present.


My plan now is to implement a fixed 120V hack for WOT or max assist situations. (I have done this before to good effect.) We know the car already allows maximum current at around 120V under load so this will help negate the low current zone.

The current and voltage feedback control must be in the MCM as the IGBT pre-driver does not have any current sensing inputs asfaik, so it probably isn't aware of current levels.

By implementing a fixed 120V Voltage hack under load and the +40/50% current hack will probably see us achieve the maximum power the ima system will allow with the OEM control modules.

The currents need for the voltage hacks into the IGBT predriver and BCM fooler / MCM HV etc are very small (less than 1ma) so a simple 120V zener/resistor switched into each HV line should do the job.
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #163
I had some 120V zeners kicking about so have built a little test voltage changeover board.

So the plan is the car will normally operate in battery voltage following mode with +48V being added to the actual 48 cell LTO battery voltage and fed in via the two seperate sections we have now.

1) (Actual Bat V + 48V) into BCM fooler and onto taps and MCM HV input. (That's the first section)

2) (Actual BatV + 48V) into IGBT Predriver HV Input and onto VPIN via the 5V adc (That's the second section)

That all works fine.

The max power changeover basically just swaps the (battery voltage +48V) for fixed 120V ones for the duration maximum power is requested.

The 120V is generated from the (Actual battery voltage + 48V) using zener diodes/resistors for both the above stages.

They have to be separate as the car compares voltages at different points in the system when various contactors are open/closed etc
 

·
Linsight Designer
Joined
·
1,877 Posts
Based on your results, I agree the "feature" is implemented in the MCM control loop.
It's odd that the "feature" also engages at higher measured voltages, too.

One thing I would do differently: go back to hacking VPIN, since that doesn't require removing the PDU.
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #165 (Edited)
One thing I would do differently: go back to hacking VPIN, since that doesn't require removing the PDU.
I agree and intend too in due course.
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #166
Todays 48 cell LTO mods ready for testing.


MCM Mods include.

1) +48V Mod.
2) IMAC&C P&P
3) BCM INTERCEPTOR with Flag Control
4) +40% Current Hack
5) Switched fixed 120V Power Hack
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #167 (Edited)
Test drive results. Interesting. Big power gains and slowly we inch forward.

Mods to discuss today.
1) (+48V) mod used for normal assist and regen.
2) Fixed 120V WOT Max assist mod.

With the 120V fixed voltage under max assist load mod, the current sag 'Feature' disappears.. :cool:
You can accelerate at WOT from <1000 LOW rpm in any gear and current rises instantly and stays high till the red line.

Good job I have 175A and 200A main fuses fitted in my two cars. :eek:
Power gains with the 120V fixed voltage mod were excellent.

Current peaked at 155A when assisting with actual battery voltage of 105-110V. = 16.2+kw
(Up by over 35A (3kw) from prev testing) Very lively to drive. It's a click on/off turbo when you mash the throttle.

When the WOT Max assist voltage mod is disengaged we get 102A when regening with actual battery voltage of 115V. = 11.7kw

I'm tempted now to try the fixed 120V WOT Max assist voltage mod when any assist level is requested (instead of just WOT) to see if that gets rid of the 'feature'. It's an easy change in the IMAC&C P&P slave software to implement.

I suspect we are slowly homing in on the optimum settings for the OEM BCM & MCM controllers.
There might be a bit more power to come with the +50% current hack and a tiny bit of further voltage tweaking.

Our 48 Cell LTO pack has an operating voltage range of ~96-124V. (0-100% SOC)

By using the +48V mod we keep the system detected voltage above the minimum OEM IMA Battery operating voltage threshold of ~ 145V throughout the entire LTO 96-124V Range.
If we drag the resting voltage down below 96V the OEM system will naturally disable assist and IMA start because it thinks the pack has fallen below 145V.

If we push it high to ~125V then with +48V we are getting into the 172V+ OEM disable regen range.. So again there is a natural level of OEM protection.

Of course we also have the CAN OBDIIC&C watching cells now and setting alf/rlf flags well before anything dies. In fact it set one today when I was regening but watching something else and the pack was already full. A cell momentarily went above 2.6V and beep, Rlf flag set, goodbye regen. (y)

The Max Assist Power fixed 120V VPin voltage mod has to be adjusted carefully to just above 120V.

If you go below 120V under heavy assist, it rapidly ramps down power after a couple of seconds as the system cuts current back to nothing if needed in an attempt to try and keep the battery voltage above 120V.

Stay just above 120V and you can pull 150A+ until the motor melts, the IGBT fries or your battery freaks out.

Anyway I need to build up a complete system for the CVT now based on all this latest work to give that some more oomph. I do have a spare flywheel and CVT gearbox, fingers crossed. Also need to add the +50% Current hack board.
 

·
Premium Member
Chicago & Detroit
Joined
·
847 Posts
great work.
do you have the OBD2C&C code for a 72 cell LTO system that queries the different sensor board addresses and monitors the individual cell voltages?
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #169
great work.
do you have the OBD2C&C code for a 72 cell LTO system that queries the different sensor board addresses and monitors the individual cell voltages?
Yes that's discussed in the other LTO OBDIIC&C CAN thread.
 

·
Premium Member
Chicago & Detroit
Joined
·
847 Posts
your 48V LTO system seems to puts out more current than the 72-cell package. I'm not interested but I expect others might be, can the capability of the 48V system be used with the 72-cell package?
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #171
your 48V LTO system seems to puts out more current than the 72-cell package. I'm not interested but I expect others might be, can the capability of the 48V system be used with the 72-cell package?
All the basic hacking methods developed over the years work on all voltages/packs.
 

·
Registered
Joined
·
6,747 Posts
So, this is great news. But - Why/how does the fixed 120V get rid of the current sag/lag?

Wasn't the faked voltage about 106V + 48V = 154V? I'm thinking, OK, so if commanded power were say 10kW, maybe you'd get a max current of 10kW/154V=65 amps. Or 65A X 1.4 for the current hack, so 91 amps. But both of these are still more than you were seeing in that low RPM range...
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #173
No idea and to be honest I'm not going to pursue that much further.
The MCM is a magic box as far as the firmware is concerned, I am just tinkering around the edges.
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #174 (Edited)
All this power hacking (Current and Voltage) is also doable for those with higher voltage packs and more LTO blocks.
If you repeated the same hacks but with 50% more voltage ( 72 cells instead of 48) you would be over 20kw by now. :eek:

Extrapolating from where we are I think the assist power limit for the 48 Cell LTO will be around 17/18kw with current and voltage hacking maxed out. That's pretty impressive from a fairly cheap and readily available Lithium cells. :)

Higher voltage packs are also easier to Voltage Hack with simple potential dividers instead of opamps and extra voltage being added.

The current hacking is the same whatever voltage and just uses simple resistor techniques.
 

·
Registered
Joined
·
6,747 Posts
If 17-18kW is possible with only 48 cells (2 Fit modules), where cell voltages are holding up extremely well, and where it seems like we're probably getting close to motor limits, is there a good reason to go with more cells - besides maybe less hacking needed?
 

·
Registered
Joined
·
93 Posts
Congratulations Peter! You are leading by example on how to think outside the box! Not only that, but you are also demonstrating the skills to make it happen! Your persistence is astounding and you are surely opening another door. This 48-cell LTO setup provides yet another option for us to extend the useful lives of our vehicles.
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #177 (Edited)
If 17-18kW is possible with only 48 cells (2 Fit modules), where cell voltages are holding up extremely well, and where it seems like we're probably getting close to motor limits, is there a good reason to go with more cells - besides maybe less hacking needed?
I don't know the answer. But it depends what you are after. A few things to consider.

Higher voltages mean less current is required for the same power.
Less heating in wires.
Smaller main fuse.
Lower % or no current hack.

A 96 cell system (double the voltage) would need half the current to develop the same power as the 48 cell system.

Watts = Volts x Amps :)

Voltage hacking a higher voltage pack down to the required operating levels is also very simple.
1) BCM Fooler = 1 resistor in feed wire.
2) VPIN wire = 1 resistor in the wire.

Obviously the weight and Wh capacity of your pack is directly related to how many blocks you have..

You pays your money, you takes your choice.

Personally for my two road cars I'm very happy with the 48 Cell LTO performance.
It's +60% power above standard (16kw versus 10kw) with about 2.2kwh capacity.
It fit securely and safely in the OEM space/modified case and doesn't weight too much..

For the UK CVT Rally car we will probably go for a much higher voltage just under 220V + all the other mods.. :devilish:

We still don't know what the ultimate power limit for the IMA motor is.... Lets find out!!!
 

·
Administrator
Joined
·
11,586 Posts
Discussion Starter #178 (Edited)
Todays LTO jobs.

Change switchable current hack resistors in my CVT BCM to allow +40 or +50% current. (Done)

Convert my CVT MCM to the safer voltage following system and add the switchable 120V max power hack. (Done)

I rationalised and simplified it all onto one messy stripboard. I need to draw up schematic soon to get it made up on a proper pcb.

Build up an old style +50% Current hack board for testing in CVT/MT. (Done)

+40% works well using 1k MCM and 240R BCM resistors.
+50% to be tested using 820R and 180R resistors.
+60% we know causes IGBT overcurrent fail with 680R and 150R resistors. .

Raining now so waiting for dry spell to install and test.

If +50% current fails I will try + 45%, if that fails I'll call it quits at +40%.

 

·
Linsight Designer
Joined
·
1,877 Posts
If 17-18kW is possible with only 48 cells (2 Fit modules), where cell voltages are holding up extremely well, and where it seems like we're probably getting close to motor limits, is there a good reason to go with more cells - besides maybe less hacking needed?
A higher pack voltage is theoretically going to be able to impart more torque at higher RPMs. However, if we've already hit the OEM BLAC motor's power limit (by consuming more power), then it's likely the real-world benefits of higher voltage are minimal. Another possibility is that at higher voltage we can actually impart more power through the motor... Peter is pushing right up to the current limits that the OEM hardware can deliver... but is nowhere near the voltage limits. It's entirely possible that with the correct tom-foolery the OEM BLAC could briefly handle quite a bit more power.
 

·
Linsight Designer
Joined
·
1,877 Posts
Peter, great news. My only caution is that you're getting awfully close to burning out your fuses. Generally you want at LEAST 50% headroom to the nominal current rating. Of course you'll find out whether this is an issue in due time, and the solution is a simple higher current fuse swap, so very low risk.
 
161 - 180 of 202 Posts
Top