D-Method on really stubborn cars or the A-Method :)

[elm70]

There you go the Diode Dropper for the ADD-method :mrgreen:

Took me less than 5 minutes. Never mind the fuse holder, I just took that wire from another project.
You should to get all those bits for less than 10 bucks at the local Jaycar store, or whatever your electronic parts store is called.
Red goes to the car terminal, black to the AUX battery negative.
I didn't have a 10A Schottkey but you see how they are wired just add or replace one of the normal ones.
You might only need three diodes, just try combinations.

You might not even have to bother about disconnecting the battery when the charger starts.
The way this works is that the charger voltage will be higher than the battery, well strictly speaking lower as the ground terminal will be too low for the diodes to conduct. You could say electrically disconnecting themselves.

Nice :ugeek:

I like the system for connect in series multiple diode

I'm an "electronic" guy ... so I have tons of diode at home .. but they are rated around 1 or 2A .. so quite useless in this application ...

So .. 4 Diode rated 10A will do the trick ?

Maybe we can have a more precise procedure for the D method without needing to over-discharge the 12v battery ..

 

[michael8554]

Regarding my Power Supply plus DC Converter method.

I agree the need for a protection diode on the output, but volts shouldn't vary with current any more or less than the 3 diodes in series method?

Yes a Digital Volt Meter is required.

But hey, you guys have made my method redundant!

The diode method is so simple!

How do you feel about the appearance of the "EV System Error" message as an indication that the correct volts have been reached?

Obviously with no headlights or heater load on.

Michael

 

[MadTechNutter]

There you go the Diode Dropper for the ADD-method :mrgreen:

Took me less than 5 minutes. Never mind the fuse holder, I just took that wire from another project.
You should to get all those bits for less than 10 bucks at the local Jaycar store, or whatever your electronic parts store is called.
Red goes to the car terminal, black to the AUX battery negative.
I didn't have a 10A Schottkey but you see how they are wired just add or replace one of the normal ones.
You might only need three diodes, just try combinations.

You might not even have to bother about disconnecting the battery when the charger starts.
The way this works is that the charger voltage will be higher than the battery, well strictly speaking lower as the ground terminal will be too low for the diodes to conduct. You could say electrically disconnecting themselves.

Nice :ugeek:

I like the system for connect in series multiple diode



I'm an "electronic" guy ... so I have tons of diode at home .. but they are rated around 1 or 2A .. so quite useless in this application ...

So .. 4 Diode rated 10A will do the trick ?

Maybe we can have a more precise procedure for the D method without needing to over-discharge the 12v battery ..

Somebody need to test it.
Right now for me pointless

I am quite sure that it is not only about the battery voltage but how far the voltage changes under load.
A weak battery will drop, in my case from 10.5V open circuit down to 8.5V under a 1.44Ω load (100W lamp at 12V) which is 6A.
This lamp resistance will be a bit less when the battery was 8.5V but not that much, so 6-7A also seems likely the load when the charge process starts up when I measured a drop to about 8.5V
The diode I used in that example has exactly 1V drop at 10A and 0.9V at 6A.
https://www.diodes.com/assets/Datasheets/ds28010.pdf
So a battery charged with typically 12.4V open circuit would need 4 of these diodes and one Schottkey diode with 0.4V drop to get to 8.4V under a 6A load.
If the load is 10A then 4 standard diodes would do the trick.

 

[MadTechNutter]

How do you feel about the appearance of the "EV System Error" message as an indication that the correct volts have been reached?

I would not press the start button as this could exceed 10A and fry the diodes.

I also suggest to switch off the tail gate ceiling light(and no other doors open) to keep things in check.

 

[MadTechNutter]

Super :ugeek:

Let us know how behave your PHEV .. it was 29Ah before reset ... so check that it does not get over discharged when used "hard" in EV mode.

I'm also checking the SOC changes while at rest, for get an idea of real SOH .. I'm losing over 2Ah if I make multiple trips in EV mode without charge and with some rest time over 2h

Today I managed to drive it down to --.--

No troubles, no limp mode, got 50.8km out of it.
Before I was lucky to get 38km, mostly only 36km.

In the end the ICE did kick in for a short time up a short hill but switched off again when it was flat so that 0.8km doesn't count.
I did press a bit harder than normal on that hill because I wanted to see if I get limp mode.
After that I always stayed within the fully green bars on the energy usage indicator(EUI) for the last 5 km and also used the handbrake to avoid regen as I was driving in very quiet and slow but a bit hilly streets. That way I tried to emulate losses from a more faster drive. The rear brakes where quite hot when I got home.

Interesting is that the "Max In" value has gone up to -50.6 hp, almost double than driving before the reset.

The cell voltages where measured standing on my property, with the hand brake pulled as hard as it gets and the accelerator pressed a bit to get the needle showing a few bars on the EUI. Car felt a bit like a lowrider lifting up :mrgreen:
It is illegal in my country to drive with a phone visible next to me.

Also very happy that the cell voltage difference at this 'depleted' state is only 0.008V

 

[elm70]

So .. the screenshot showing 3.68v and 30.4% SOC is at the end of trip with a bit of "gas" ... it would have been more significant without any "gas" in my view ...

What for me is relevant is to check how much SOC drop, and to which voltage the cell are bouncing back after 2h rest.

This will give you an idea how much further then 70% battery usage is really happening in the car.

Still 3.68 with a bit of gas .. is not looking too bad ... I hope this was above 3.75v without gas ... and hopefully it should bounce back above 3.78v or more after rest.

In these assumptions .. I think you probably just use 5% more extra battery capacity then intended (down to 25% vs 30% .. but this is not really an issue, since when driving slow, it is allowed to go down to 25%) ... while before BMU was over conservative allowing to use only 60% of real capacity ...

So ... apparently the BMU reset is a good win for your case ... and I can't see much of risk on overdischarge

 

[MadTechNutter]

The idea behind the 'bit of gas' was to reveal weaker cells.
This becomes more apparent when the battery is low and there is a load.

I think the problem that some had with turtle mode could be individual cells hitting 2.7V under load or whatever threshold is set in the battery management.
As said I can't read the dog while driving. Cops are everywhere here on the roads grabbing revenue rather than fighting REAL crime.

Another problem is I can't wait for 2h for anything to settle. That test was in the morning and it is drive or solar charge all day for me. Sometimes I need to go to town 3 times a day but I am usually quickly back. I work from home but have to pick up and deliver.
The extra driving range is heaven sent for me and I could now also manage a quiet cloudy day without charging.

For me it is not only about near 100% EV driving but also not having to pay for it :mrgreen:

...although I do like to occasionally floor it at 75km/h when all three motors give traction and me a nice kick.
Somehow I will eventually have to replace that fossil liquid that came with the car :twisted:

 

[michael8554]

said I can't read the dog while driving. Cops are everywhere here on the roads grabbing revenue rather than fighting REAL crime.

EV Andy has asked for 245 documented offences to be taken into account........?

Michael

 

[greendwarf]

said I can't read the dog while driving. Cops are everywhere here on the roads grabbing revenue rather than fighting REAL crime.

EV Andy has asked for 245 documented offences to be taken into account........?

Michael

I blame Mel Gibson - they all think they're Mad Max

 

[elm70]

The idea behind the 'bit of gas' was to reveal weaker cells.

Yes it make sense ...

About driving and looking the Dog ... yes don't do it ... (not because the Police, since for some German/Australian driver is not a mayor issue ... not only for looking on the Dog, but as well, for making video & moving the camera while driving :? ) ...
Still ... remember the dog also show the trip record .. with very good info .. like max / min voltage per cell ... max delta voltage, etc.

Anyhow .. as said before ... all is looking quite fine in your PHEV, still only the time will tell if Mitsubishi has been over conservative (maybe due to a bugged firmware), or battery degradation is impacted by discharging down to 15% / 20% real SOC

 

[mellwaters100]

Good to read your experience @MadTechNutter. I reset my BMU a few days ago and have driven nearly 400 miles since and was wondering if the behaviour I have been seeing is "as expected".

Some things I have observed:

1) Prior to reset (32 Ah capacity), cruising at @ 90km/h, the ICE would kick-in at 10.1 Ah and switch off at 10.6 Ah. Since reset, that has changed. It now kicks in at 11.7 Ah and cuts off at 12.2 Ah.

2) Before resetting, I never took much notice of the cell voltage through these cycles but, post reset something strange has been happening!! On the first hybrid drive (about 70 miles), the 11.7Ah equalled 3.65v; 12.2 Ah = 3.75v. Just done a 180 mile hybrid drive today (3 days after reset) and the values are: 11.7Ah / 3.75v ICE cut-in; 12.2Ah/3.85v ICE cut out.

3) After a prolonged cruise at 90km/h, driving in the city the Ah drops to around 9.7Ah before the ICE kicks-in, but the cell voltage remains at 3.75v (at this 9.7Ah level).

4) If I park up without charging, leaving the car for an hour while shopping, switching on the ICE kicks in immediately. In the 2 years of owning the car, this has never happened. In the Dog, Ah had dropped to 7.9Ah.

5) This Ah drop is consistent now after parking up and not charging. Did a 170 mile drive today, got home and had 10.7Ah / 3.75v. 2.5 hours later, showing 9.5Ah but still 3.75v

However, efficiency does seem to have improved even in hybrid mode. The first trip after reset with full charge:

Ttl Distance: 101km.
EV distance: 78km
L/100km: 2.6
Avg speed: 62.2km/h

90% of this journey was at 80 & 90 km/h.

Just done a 170 mile trip; started without a charge. About 160 miles at 90km/h cruise and finished with a reported 51.7mpg in the Dog! Previously, such a trip I would expect 43mpg - 47mpg. Admittedly, this makes no sense, why a pure hybrid drive would be more efficient now, but, there ya go!

Battery capacity has dropped 0.4ah in the last 400 miles (106,500 miles on the clock).

 

[MadTechNutter]

@mellwaters100, I have practically no experience driving in Hybrid mode(yet) but your info is very interesting.
Hopefully somebody else might be able to confirm or compare the Hybrid behaviour you are experiencing.

 

[michael8554]

At 8.7 volts on load to the car, I got the "System Service" message, but the relays wouldn't click.

Removed diodes one by one until I got to 10.4 volts, where relays clicked, no error message, charging started with the Mitsu Granny charger, no change in capacity of course.

This was with two Diodes, so I added a Schottsky to get a small drop to 9.9 volts, got the error message, but no relay action or charging.

The 10 amp diodes came a day earlier than I expected, so the car was 85% charged - is that my problem?

Should I be doing this with the car in Accessory Mode ( 2 button pushes without braking)?

2014 3H, 20,000 miles, 81% and 31aH.

 

[michael8554]

So today I ran the car down to the point where the ICE kept coming on even when stationary, 11.2aH, 36%.

Pretty much the same results.

Below 10 volts the error message appeared, but the best I got was 1 relay click and no charging.

10 volts and above no error message and multiple clicks and charging started.

10 volts was with 2 diodes.

Closest I could get to just below 10 volts with the diodes I had was 9.7 volts with 3 "diodes" in series - 2 ordinary diodes plus 2 Schottsky in parallel.

The forward biased diodes were connected between the Aux battery Plus terminal and the removed Red clamp on the cable feeding the car.

Measurements were between the clamp and earth.

Why isn't 8.5 volts operating the relays on my car?

Michael

 

[Darkflow]

Why isn't 8.5 volts operating the relays on my car?

I'm not that surprised that it is not enough, especially for the contactors in the battery pack.
I've been surprised that people have been able to initiate charging with such a low voltage.

As to why a battery with low charge works, but a string of diodes connected to a battery with higher charge doesn't - I guess in this case the difference matters.
The low charge battery will tend to recover to a higher terminal voltage and then sink under load. The charged battery with diode string will not, at least to the same degree.

 

[michael8554]

Thanks Darkflow

I've been through all of this thread again, and I'm not sure, did madtechnutter's post (with the image of his diodes and leads) actually work, or was he showing how it COULD be done?

Has he attached the diodes ACROSS the battery? That would be almost a short circuit, so don't think so.

I removed the plastic floor (quite easy, just 5 posidrive screws) to get to the Pos terminal so I could more easily measure the volts.

Diodes placed in the Pos feed to the car, rather than in the earth lead.

 

[obiuquido144]

1) Prior to reset (32 Ah capacity), cruising at @ 90km/h, the ICE would kick-in at 10.1 Ah [...]. Since reset, that has changed. It now kicks in at 11.7 Ah [...].

That's because your battery is degraded and the real capacity is still around 32Ah and the engine still kicks in at the same actual SOC.
Your SOC numbers got shifted due to the reset (and look "prettier") but the underlying battery chemistry is the same and the (probably) voltage-based ICE start control mechanisms don't look at the pretty numbers.

Count with me:
10.1Ah/32Ah = 31.5%
11.7Ah/38Ah = 30.8%

ICE is still kicking in at the same SoC, the reset just divided the range into more segments.
Just a thought

 

[elm70]

1) Prior to reset (32 Ah capacity), cruising at @ 90km/h, the ICE would kick-in at 10.1 Ah [...]. Since reset, that has changed. It now kicks in at 11.7 Ah [...].

That's because your battery is degraded and the real capacity is still around 32Ah and the engine still kicks in at the same actual SOC.
Your SOC numbers got shifted due to the reset (and look "prettier") but the underlying battery chemistry is the same and the (probably) voltage-based ICE start control mechanisms don't look at the pretty numbers.

Count with me:
10.1Ah/32Ah = 31.5%
11.7Ah/38Ah = 30.8%

ICE is still kicking in at the same SoC, the reset just divided the range into more segments.
Just a thought

Let me polish this a bit:

That's because your battery is degraded and the real capacity is still around 32Ah

Absolutely NOT ... the PHEV believe in the new SOH that is bigger now after the BMU reset.
Possibly the battery was having a better SOH compared to 32Ah ... but definitely since battery degradation it is a real thing, 38Ah as SOH after a BMU reset, it is a bigger number then the real SOH.

Count with me:
10.1Ah/32Ah = 31.5%
11.7Ah/38Ah = 30.8%

That's correct ... around 30% SOC is what kick in the ICE to start .. 30% of 38Ah is a different value of 30% of 32Ah

What is more relevant ..

From 32Ah down to 10.1Ah ... only 21.9Ah was usable for EV mode.
From 38Ah down to 11.7Ah .... it is 26.3Ah usable energy .. so .. more then 20% EV range increase

Since the real battery SOH is not changed .. I expect that the 30% SOC after BMU reset equal to a much lower voltage level, since further 4.4Ah are consumed from the battery ... if the real SOH would be close to 26Ah ... this could even become a problem.

 

[Darkflow]

@michael8554:

or was he showing how it COULD be done?

Pretty sure that one ^^^. I would also have expected it to work, as long as there was enough voltage to pull in the contactors.

Your experience indicates that things aren't quite that straightforward.

Has he attached the diodes ACROSS the battery?

Definitely not - they're intended to be forward biassed in series with the positive lead, as you've done.
Across the battery would let out the magic smoke.

 

[obiuquido144]

I expect that the 30% SOC after BMU reset equal to a much lower voltage level

This is where our hypotheses differ.

In my experience the ICE start voltage (not under load) is always just under 300V, both pre and post reset.

Also, I believe these precise 4.4Ah are actually NOT fully consumed, I think the display counters just go down slightly quicker, to cover the nominally wider SOH space. Yes there is some extra real energy consumed (which gives the extra range) but not precisely 4.4Ah and probably a lot less.
I believe the absolute SOC Ah number we see in the Dog is not from the coulomb counter, and that the ICE kick in mechanics are independent (voltage based)

In other words, my belief/observation is that the reset-induced higher SOH number affects the charging behavior, but during driving the absolute Ah SOC value is just a meaningless pretty "display" number that's not used for ICE control, and is calculated on the fly to make it fit/add up.

Anyway, at least the ICE kick in voltage levels pre and post reset can be observed with this experiment:
1) Do a data log in the Dog of a trip, ideally not over 60kmh, until ICE kicks in. Note voltages.
2) Do a DBCAM, this will reinstate the SOH close to the pre-reset value, say 32Ah.
3) Repeat step 1 and compare voltage results.

Obviously this can also be done in the opposite order if a reset hasn't yet been performed, and then there's no need for a DBCAM.
In my, less methodical, observations the voltages I'm seeing are not "extra-4Ah-depleted" different, I'd call them almost identical.

Stay tuned, I might have some data in a week or so.