Regulated Solar Charging

Mitsubishi Outlander PHEV Forum

Help Support Mitsubishi Outlander PHEV Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

MadTechNutter

Well-known member
Joined
Jul 5, 2019
Messages
132
Location
Aussie Land
This is my first post.
I am an electronic enthusiast and this post is mainly addressed to like minded, who are aware NOT to hold a soldering iron by it's business end :)

I will be proud owner of a reasonably priced, used PHEV (2014) next week, which will be out of warranty soon.
I am cash poor and had to save long for this.
I live in a country where there is lots of sun, Australia.
I have solar with a DIY off-grid LiFePO4 battery system but want to beef it up.
For the price of a charge adjustable EVSE cable down to 6A, I can purchase 6 x 250W solar panels here, which I will, bringing my off-grid system to 3kW(realistically 2.5kW). There is a grid tied 1.5kW as well where I might syphon off unused energy to the off-grid battery, but that is not the topic.

The objective is to open the Mitsu charge cable box, locate the 1khz pilot and the +/- 12V, cut the pilot wire and trigger with it a 555 timer chip in monostable configuration so it outputs a pulse that can be adjustable between 50-250µsec, feed it to a buffer amp that swings between the +/-12V and connect it back to the pilot wire that goes into the Mitsu.

Furthermore a current sensor in the PV dc line will regulate this PWM signal depending on Solar output.

I actually want to do this with a PIC micro 12F683, just mentioned the 555 to explain the function better.

My Questions:
1 - Are the required wires accessible in the Mitsu supplied controller box and not fully immersed in resin?
2 - Has anybody done this before?
3 - Could the car charge at lower than 6A if the Pilot pulse is less than 10% of the full cycle?
4 - Does the vehicle respond to changes in the PWM duty cycle while charging?

Practically I can do this for $2 material cost and have designed far more complicated things, currently designing/building a 4WD electric mini tractor/ride-on, all drive by wire, later maybe autonomous.

ATM I have no clue what else I will be doing with this PHEV but I already know that the only way to stop me from 'disfiguring' it,
will be to pry the soldering iron from my cold dead hands.

I might think about changing the way the throttle responds like a 'pure' EV mode which I read here: http://www.myoutlanderphev.com/forum/viewtopic.php?f=10&t=2191
but Vtech ended up selling this https://evtun.com/phev-box.html
To me that just looks like a pre configured throttle potentiometer response curve that I could do with a $1 PIC micro.
I will see how I get on when I can finally drive it, if such a function is actually useful. I am NOT the lead footed type of person but if I find it difficult to drive up the minor slopes which I have on my daily drive without the ICE kicking in, I might look into it.

Unfortunately I am only a hardware geek and know nothing about any engine mapping codes.
It would seem beneficial to have short term higher EV kW output when I need it without the ICE.

In the end it is all about MOSFETs or IBGTs that are controlled with a DC/PWM signal.
My motto: Hardware overrides Software :)
 
Much of this is gobbledegook to me ;) but I can tell you that the wires are not embedded in resin.
However, the charge box and the car software do communicate, so I cannot tell you whether the system will accept changes in the hardware configuration.
 
I found these sites while researching chargers a few years ago. Good info on how Level 2 charging is controlled. Is sounds like you are skilled, so might be of some use to you. My understanding is the charger indicates it's maximum available capacity and the car requests variable amount of current up to the charger's max capacity using the +/-12V square wave pilot signal. The Hydra design was good for adjusting the output based on the available source power.

Open EVSE - https://www.openevse.com/kits.html

J1772 Hydra - https://www.tindie.com/products/nsayer/j1772-hydra/

I had a desire to build an EVSE, then they became available on Amazon for $250, so that went away quick.
 
With that much tinkering, I would tap into the Chademo directly. No DC to AC then AC to DC lost. Perhaps you can charge with (100) LIFEPO4 in series!

Tai
 
Thanks for the replies :)
Yes I was looking into making one from scratch but the kits aren't that cheap either, although I found cheaper boards but lost the links.

Would love to 'mess around' with Chademo but we Aussie's are not allowed to have that port, just a blank space and no hidden wires behind it either.

I managed to get the supplied charger apart 'carefully' with a chisel and there is a lot more inside it than I thought and not a matter of simply cutting a wire.

Was hoping there was an external oscillator where I could simply change the duty cycle but the signal is internally fixed programmed in an MCU.
A timer to modify the duty cycle between PWM_out from the MCU and the buffer amp could be a solution but the feedback of the modified Pilot
back into PWM_in of the MCU might make it shut down.


Edit found the link, this is the little board I meant for $50:
https://www.ev-power.eu/Electric-Cars/EVSE-kit-for-EV-charging-station-cable.html?listtype=search&searchparam=EVSE
Github has all the info how to use it: https://github.com/kortas87/simple-evse/wiki/SimpleEvseJ1772guideExternal

I particularly like this version :)
However I would never use this without an RCD !!

687474703a2f2f6576726163696e672e637a2f77696b692f73696d706c652d657673652f696d616765732f6f75746c616e6465722f3032365f4361725f4368617267696e672e6a7067
 
MadTechNutter said:
This is my first post.
I am an electronic enthusiast and this post is mainly addressed to like minded, who are aware NOT to hold a soldering iron by it's business end :)

I will be proud owner of a reasonably priced, used PHEV (2014) next week, which will be out of warranty soon.
I am cash poor and had to save long for this.
I live in a country where there is lots of sun, Australia.
I have solar with a DIY off-grid LiFePO4 battery system but want to beef it up.
For the price of a charge adjustable EVSE cable down to 6A, I can purchase 6 x 250W solar panels here, which I will, bringing my off-grid system to 3kW(realistically 2.5kW). There is a grid tied 1.5kW as well where I might syphon off unused energy to the off-grid battery, but that is not the topic.

The objective is to open the Mitsu charge cable box, locate the 1khz pilot and the +/- 12V, cut the pilot wire and trigger with it a 555 timer chip in monostable configuration so it outputs a pulse that can be adjustable between 50-250µsec, feed it to a buffer amp that swings between the +/-12V and connect it back to the pilot wire that goes into the Mitsu.

Furthermore a current sensor in the PV dc line will regulate this PWM signal depending on Solar output.

I actually want to do this with a PIC micro 12F683, just mentioned the 555 to explain the function better.

My Questions:
1 - Are the required wires accessible in the Mitsu supplied controller box and not fully immersed in resin?
2 - Has anybody done this before?
3 - Could the car charge at lower than 6A if the Pilot pulse is less than 10% of the full cycle?
4 - Does the vehicle respond to changes in the PWM duty cycle while charging?

Practically I can do this for $2 material cost and have designed far more complicated things, currently designing/building a 4WD electric mini tractor/ride-on, all drive by wire, later maybe autonomous.

ATM I have no clue what else I will be doing with this PHEV but I already know that the only way to stop me from 'disfiguring' it,
will be to pry the soldering iron from my cold dead hands.

I might think about changing the way the throttle responds like a 'pure' EV mode which I read here: http://www.myoutlanderphev.com/forum/viewtopic.php?f=10&t=2191
but Vtech ended up selling this https://evtun.com/phev-box.html
To me that just looks like a pre configured throttle potentiometer response curve that I could do with a $1 PIC micro.
I will see how I get on when I can finally drive it, if such a function is actually useful. I am NOT the lead footed type of person but if I find it difficult to drive up the minor slopes which I have on my daily drive without the ICE kicking in, I might look into it.

Unfortunately I am only a hardware geek and know nothing about any engine mapping codes.
It would seem beneficial to have short term higher EV kW output when I need it without the ICE.

In the end it is all about MOSFETs or IBGTs that are controlled with a DC/PWM signal.
My motto: Hardware overrides Software :)

Time few weeks, and you will master to press the accelerator without having the ICE to start up

It is hard to use 99% of all power without kick in the ICE ... but to get 90% it is not so complicated, and you will master this in a relative short time
 
Thanks elm70.
I discovered if I go just 5km/h over the speed limit shortly before the hill on the highway I get to the top with help of the extra inertia without slowing traffic behind me.
The police generally don't book you here with 5k's over and they always stand on the down hill stretches anyhow.
That is where I am harvesting nicely with the paddles keeping it spot on 80. Never thought keeping to the speed limit is so much fun :lol:
 
Definitely going to build my own charger though, the supplied charger is a design nightmare. Just holding a 1MΩ test probe on the PWM signal switches on the fault LED.
It was good that I took it apart though because a plug and wire on the power relay had signs of heat damage.
Turned out it was corroded bad, possible fire hazard. All the other connectors are perfect and no sign of moisture ingress.
 

Attachments

  • Charge_Box_fault.jpg
    172.4 KB · Views: 92
wws said:
You might consider just building your own EVSE, e.g, from: https://www.openevse.com/ Then you can set the amperage to whatever you please.

Absolutely is now the plan :)
but I decided to design it myself with a self programmed PIC micro and build it inside the J1772 plug that just has a long cable that goes straight into the 230V of my inverter or house mains.

I will update, after the plug arrives that I ordered for a reasonable price here http://evbitz.uk/EVBitz.uk/Type_1_Plug.html and can get on with it.

Still a lot of planing. I am now thinking of using the house connection that I already made from the battery inverter and using an automatic transfer switch to switch it to the J1772 plug in the day and to the house power points in the night that are on a grid connected solar inverter in the day but is too weak for the car and the house together.
Apart from saving a lot of wire (panels and battery and inverter are in the back of the house, PHEV in the front) I have the RCD breaker from the house connection and wouldn't need to worry about any GFCI circuitry(basically the same thing) in the EVSE which would make things messy if I want to fit it all in the plug. Still a bit in deep thought there.
Open for any ideas or requests :)
 
Here is my self designed charger that fits my requirements but couldn't get the guts to fit inside the plug, the extra box is fine.
It has a 20 meter cable to reach the back of the house with my off grid system.
It tells my car to use only 5.3A, at least that is what my clamp meter shows.
...and of all things I found an outdoor RCD safety adaptor still new and originally packed for $10 just last Sunday at the local flea market :D

I have not made the charge current adjustable although that would only be a few lines of code.
At this stage I don't need to adjust it as 5.3A is awesome because my current grid-tied solar system can handle that without using any grid power.
5.3A x 240V = 1272W
Anybody who actually checks their solar system knows that a 1.5kW system never puts out more than 1.3kW in the real world.

My off-grid inverter also runs more efficient on that current too, so when some clouds pass by the off-grid house battery won't get to heavily loaded so I don't see any need for a feedback regulated system right now.
This might change in summer when I have the 12 dedicated solar panels up and running, which I am about to install on the roof soon.

It has a simple one button timer function instead that can be set and reset while it is charging. So if I see clouds coming I can set it to terminate charge between 1-4 hours or cancel that at any time without disrupting the charge.

One strange thing I noticed in the 3 days I am using my slow charger is that I am getting 3-4km more REAL driving range.
Absolutely same distances down to the 100m, same time driving, same traffic, I always watch exactly my econometer when I drive uphill and regen with the same paddle setting downhill.
This does not show on the dog, same 29Ah SoH, which I am now calling the bullshit meter because that is based on some software calculations as I see no true evidence of any measured capacity here.

Despite those extra kms I have never managed to complete the charge because the charge is ...errr.. slow and doesn't complete by the time for my second trip.
The two trips I do on weekdays are a bit less than 6 hours apart.
When charge is completed with the original charger the guesso-meter shows 45km but I come home with 1-2km left.
Incomplete slow charge showing 42km and I come home after the second trip(which is in the morning) with 5km left.
The car stands overnight half empty.

Tomorrow I will be able to do the complete slow charge but I won't be driving my two exact weekday trips so I will have to monitor this a bit longer to find out what is going on here.
At least I am not charging daily up to 100%, that was my intention.

I am happy to supply more details of my charger if required for the technical minded :ugeek:
 

Attachments

  • My Charger copy.jpg
    My Charger copy.jpg
    557.1 KB · Views: 47
  • My Charger_open copy.jpg
    My Charger_open copy.jpg
    451 KB · Views: 47
MadTechNutter said:
One strange thing I noticed in the 3 days I am using my slow charger is that I am getting 3-4km more REAL driving range.
Absolutely same distances down to the 100m, same time driving, same traffic, I always watch exactly my econometer when I drive uphill and regen with the same paddle setting downhill.
This does not show on the dog, same 29Ah SoH, which I am now calling the bullshit meter because that is based on some software calculations as I see no true evidence of any measured capacity here.

Despite those extra kms I have never managed to complete the charge because the charge is ...errr.. slow and doesn't complete by the time for my second trip.
The two trips I do on weekdays are a bit less than 6 hours apart.
When charge is completed with the original charger the guesso-meter shows 45km but I come home with 1-2km left.
Incomplete slow charge showing 42km and I come home after the second trip(which is in the morning) with 5km left.
The car stands overnight half empty.

Tomorrow I will be able to do the complete slow charge but I won't be driving my two exact weekday trips so I will have to monitor this a bit longer to find out what is going on here.
At least I am not charging daily up to 100%, that was my intention.

Here the result of an attempted 'full' charge using my slow charger.
I got now 9.3kWh available on a battery hat has only 8.7kWh capacity and was charged to only 97.9%.
So I wasn't imagining things.
Battery is definitely not over charged, the 4.095V for a 97.9% SoC looks correct.
Please tell me what going on here?
 

Attachments

  • Screenshot_20190811-143537.jpg
    Screenshot_20190811-143537.jpg
    222.3 KB · Views: 35
  • Screenshot_20190811-142318.jpg
    Screenshot_20190811-142318.jpg
    231 KB · Views: 35
I stand by my opinion that the SoH indicator on the Dog is calculated bullshit(from the BMU) but the displayed energy going into the battery while charging is true.
So this energy must be going somewhere.

I have a theory(just an idea) that by taking almost twice as long to charge that individual cells are getting balanced better. The weakest cells are given more time to absorb the charge without the earlier termination from a standard charge when they hit 4.1V.
This slow charging may have physical/chemical consequences like a possible reduction of their internal resistance although I have not heard of any in regards to Li-Ion

This is however common with other batteries, like NiMH where I have recovered many batteries almost from the dead with a very slow charge.


Edit: ...or the Dog is showing rubbish, I just swapped the kWh displays to Ah and the SoC is now only 28.5Ah and SoH 29Ah
Guess it is a bug.
 
Following this thread, as I am very interested in reducing the charging rate.
Firstly I'd like to charge during the day while my PV's are generating - but I only have 1.5KW (1100 real watts!) grid-tie (we're not allowed any more than this due to the energy provider's influence on government) so if I could reduce the charging rate it would hopefully be covered by the panels.
Secondly if doing so will prolong the life of the batteries it would be well worth doing on the overnight charge when there's no rush.
 
If you have a Type 1 plug I believe you can limit the current drawn by altering a resistor in the plug:

1500 Ω resistor – 13A cable
680 Ω resistor – 20A cable
220 Ω resistor – 32A cable
100 Ω resistor – 63A cable

This seems to be a non-linear relationship, so you will have to draw a graph to determine what resistance you need - or carry out some clever maths !
 
Back
Top