Honda Insight Forum banner

1 - 20 of 55 Posts

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #1
QBatt.

This OEM 2khz 5V PWM 10-90% signal is sent from the MCM to the ECM to tell it the SOC.

LTO/Lithium users who want to have some manual control over the ECM charging behaviour might find it useful.

This video explains more.


The PIC 12F683 firmware is below and the PWM output is confined to the ECM valid 10-90% duty region.

Connection etc is explained in the video. (Sorry it's vertical)

I haven't done a schematic yet but it's easy enough to follow.

I won't be testing this in my car (I've done my bit!) so over to one of you.. ;)

IMG_20200909_160938.jpg

87490
 

Attachments

·
Premium Member
Joined
·
894 Posts
Could you turn the SOC all the way down and see if this forces a constant background charge and how long it actually keeps a background charge going.

At the end of the day you probably really need a IMA C&C P&P if you want to control charging.
 

·
Registered
Joined
·
6,961 Posts
Interesting. That signal is similar to 'DVCT' between DCDC and ECM, which controls DCDC output voltage. You can buy a PWM adjustable signal generator for like $6, I've been using one on DVCT for a while now. Or you can use an arduino mini or similar. Here's a thread where I describe the DVCT thing, if anyone would rather buy something off the shelf that should do the job: Mod to Adjust DCDC Converter Output Voltage

hmm... Peter, do you know what else QBATT might affect? For example, you say it would probably impact background charging, but do you know if it impacts anything else, like auto-stop, or maybe idle speed, etc?

I actually have 2 channels on my 'DVCT' PWM device - so I could probably connect one to QBATT at the ECM, just to experiment. Don't have much need for it, though...

edit/note:
QBATT pink D13 at ECM, 16 pin grey connector at 'top'
QBATT pink also at 24 pin lt blue Connector C103, pin 21
 

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #6 (Edited)
QBatt will directly affect all ECM commanded charging behaviour. It's the only SOC info it receives.

If the ECM decides to command idle charging it adjusts the idle speed itself as normal.
If you set QBatt below 15% I understand autostop is disabled.
Remember I haven't tested any of this yet.

You must not got below 10% or above 90% PWM duty as that is an error and you will get a code. The firmware I uploaded does not allow PWM duty outside the above permitted range.

I have just done a second version of the firmware for the standalone QBatt gadget.

This one has an input for the incoming OEM PWM from the MCM as well as the output to the ECM.
This means It will now pass thru OEM signals when the POT is turned right down.

The pot is now a three stage switch/adjuster.

1) Pot position 0-9% (turned right down) = pass thru OEM signal to ECM. (Red Led On)
2) Pot position 10-90% = send adjustable 10-90% duty to ECM. (Green Led On)
3) Pot position 91-100% (turned right up) = fixed duty at 75% to ECM (Red Led On)

(We can change the 'fixed duty' value)

If you forget to attach the incoming MCM signal then the ECM pass thru signal defaults to 50% duty. (We can change this)

This firmware needs a pull up resistor adding to the little pcb on the incoming OEM data line. PIC Pin 6.

The attached zip file (rename from pdf) contains schematics, firmware hex and source files for both versions.

The picbasic_pro compiler I use has a free edition that support the 12F683 PIC if you want to tinker around.


www.piccircuit.com do a neat tiny little developer board for the PIC 12F683.


I suggest get a pic programmer while you are at it if you want to play. ;)

 

Attachments

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #7 (Edited)
I did get a few minutes today to test this in the car. It works exactly as expected. (y)

Lower SOC settings start a background charge when driving or even sat ticking over.
Assist lessens as SOC reduces.

Higher SOC setting stops any regen and increases assist use.

There may well be various nuanced levels between 10/90% but someone else can be a bit more methodical and test it.

I did not find a point where coast regen stops but brake regen is permitted.
So I think that is just not doable without more sophisticated control. IMAC&C P&P etc.

I see this as a cheap and useful gadget for controlling or modifying some OEM behaviours. ;)

I added a bi colour led to my device so I could see when it was at the either end of the pot scale.
When the POT was >90% or <10% the LED would turn Red, at any other time it is Green.

Note 1
If you set the BCMI ALF flag in QBATT pass thru mode regen will start as expected.
But this can now be stopped by turning the SOC up higher even if the ALF is set.
So although assist is locked out you don't have to have regen now.

Note 2
A finer control or multi turn POT might be useful for testing purposes.

Note 3
I'm adding a 16x2 serial lcd to my device so I can monitor the PWM settings. We can then correlate exact PWM % with car and ECM responses.
 

·
Registered
Joined
·
6,961 Posts
^ Sounds potentially useful. I might have to experiment with my PWM thing...
 

·
Registered
Joined
·
177 Posts
QBatt is a signal generated by the MCM based on what it thinks the SOC is, or is there more to it than that? Is this circuit for fooling the ECM because the MCM has bad information or just to increase performance/control?
 

·
Premium Member
Joined
·
894 Posts
Some questions about this:

1) Can this be used simultaneously with the BCM Interceptor ? IF it is possible, would it be able to pass the BCM Interceptor signal by turning this mod down to 0-9%?

2) Would you even want to use the BCM Interceptor if you use this device?

3) Why did you link this part in your post, what is this part used for? Do I need it for the build?



What I'm thinking may be better for me is just a BCM interceptor profile with the new 1840 chip that sets the SOC state to 10% and charges the car until it hits 80%. This way it forces a background charge when you really need to charge the car, and then once the car is charged you just switch back the BCM Interceptor to the normal profile.

I don't think I would want the 20-90% SOC state, as I would just want to have full regen and assist when I'm not charging.
 

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #11
QBatt is a signal generated by the MCM based on what it thinks the SOC is, or is there more to it than that? Is this circuit for fooling the ECM because the MCM has bad information or just to increase performance/control?
No there isn't any more to it.

Yes the circuit is to see what extra control it gives us and/or override information from the MCM.
 

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #12
1) Can this be used simultaneously with the BCM Interceptor ? IF it is possible, would it be able to pass the BCM Interceptor signal by turning this mod down to 0-9%?
A) Yes, it can be used with the BCMI
B) No, it doesn't interact with the BCMI

2) Would you even want to use the BCM Interceptor if you use this device?
Yes, the BCMI is on a different line controlling different things like voltage, soc, amps, temperature and flags.

3) Why did you link this part in your post, what is this part used for? Do I need it for the build?

It's just a little 12F683 prototype board for tinkerers.
You can use whatever you have on hand to build the QBatt device.

What I'm thinking may be better for me is just a BCM interceptor profile with the new 1840 chip that sets the SOC state to 10% and charges the car until it hits 80%. This way it forces a background charge when you really need to charge the car, and then once the car is charged you just switch back the BCM Interceptor to the normal profile.
I don't think I would want the 20-90% SOC state, as I would just want to have full regen and assist when I'm not charging.
That's probably easy enough to do, but of course with SOC that low you would not have any assist/autopstop etc
The BCMI coding can be adjusted within reason to cover a lot of eventualities.
 

·
Premium Member
Joined
·
894 Posts
I was under the impression that the BCM Interceptor changes the SOC based on how you program it.

If that is the case then, the BCM Interceptor is sending SOC data, then this QBATT is also sending SOC data?

Would one override the other?
 

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #15 (Edited)
IB Please do some background reading and study the relationships of the various modules.
I did post info about that on one of the other threads very recently. You will have to dig around.

1) The BCM manages the battery. It sends info to the MCM. (Soc, Voltage, Amps, Current, Flags)

1A) The BCM Interceptor device sits in between the BCM & MCM.

2) The MCM manages the IMA motor. It sends info to the ECM and BCM. (Qbatt, Fault codes etc)

2A) The QBatt device sits in between the MCM & ECM.

3) The ECM manages the engine it sends info to the MCM (Assist & Regen requests)

3A) The IMAC&C P&P Slave device sits in between the ECM & MCM.

The SOC sent by the BCM to the MCM is a different signal to the SOC sent by the MCM to the ECM. The % values might be roughly the same but it's a different format and signal type.

The ECM does not receive SOC direct from the BCM.

QBatt can override BCM SOC in certain situations and vice versa.
The BCMI overrides BCM SOC and also that can affect QBatt SOC.

Both the BCMI and QBatt devices have a part to play in SOC and performance management of the system.
 

·
Premium Member
Joined
·
894 Posts
HI retepsnikrep, If you have time, and I can see you must be busy with all of these things you are working on, could you upload an updated schematic that has the added features like this 16x2 display and MCM readings.

Is it worthwhile to use a 16x2 screen for this mod?
 

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #17
This is the schematic for the serial LCD and the bi directional red/green led version. It has got a few extra bits.

Note 1.

Serial LCD's are expensive and somewhat flaky compared to standard parallel interface types.
I've only added an LCD to aid with testing. Once I have established the levels I doubt I will retain it.

Note 2.

I might well use a single button system to simply cycle round and round several qbatt levels. Cheaper and more predictable easier to set than the POT on the move.

Startup = OEM pass thru (Green Led On)

1st press = Fixed 75% (Assist and coast/brake regen allowed) (No background charging) (Red Led On)
2nd press = Fixed 80% (Assist allowed, regen disabled) (More use of assist) (Red Led On)
3rd press = Fixed 50% (Assist allowed, regen enabled) (Charging starts depending on engine load) (Red Led On)
4th press = Fixed 20% (Assist disabled, regen enabled) (Forced charging enabled) (Red Led On)

5th press = Back to Startup

Note 3.

It may well be possible to simply use an IMA Boost pcb to do this QBatt mod.
 

Attachments

·
Administrator
Joined
·
12,156 Posts
Discussion Starter #18
Here is the old IMA Boost user manual and schematic.

You can see it's a nice compact design with button and bi-colour led already.
It fits in a standard Hammond abs case and uses the 12F683 PIC.

We can simply omit the two pots and reconfigure the QBATT software to give a stepped button press system.

I also have quite a few pcb's available. ;)
 

Attachments

·
Registered
Joined
·
836 Posts
1 - 20 of 55 Posts
Top