DIY "Mild" Plug-in Prius Project: Grid-powering the 12v side w/ deep cycle batteries

[bwilson4web]

Originally Posted by orange4boy

This has been done before so hopefully I will be able to get some specific info before I do anything:
http://www.veva.bc.ca/articles/PHCdemoArticle080211.pdf

I will try to get some information from the owner, who is a member of VEVA and has rebuilt several Prii:
http://peopleshybrid.com/index.htm

Excellent! I always enjoy using the fruits of proven technology ... <GRINS>

Originally Posted by orange4boy

. . .
I like that idea better. 24V is better because it seems to be easier to find and implement a down converter than a boost converter. Also, I will need at least 2 batteries if the overhead is indeed 300+ watts. I'm not concerned about the extra weight because it's well known that cutting ICE electrical overhead outweighs the added weight of batteries. Ha Ha.

My old man's brain may not always be the swiftest ... but occasionally even a blind pig finds an acorn. <GRINS>

Originally Posted by orange4boy

. . .
Questions to be answered assuming a 12V nominal system:

Will the lower voltage (under 13.5V) set codes?

Once the car starts, the inverter provides enough voltage to run without a battery. Good Prius friend Hobbit performed that experiment.

Originally Posted by orange4boy

. . . My guess is that it might if it gets too low. Prii with dying 12V batteries exhibit all kinds of strange behaviours. Although a dying battery will probably be sub 10V under load. The graham scanner sees 11.3 on mine with the hybrid system off so there seems to be some flexibility.

That is the funny thing. We find intermittent problems that suddenly disappear when a good aux battery is installed. My OEM battery lasted until January of this year ... 4 years? I had to change it before flying to Detroit in January.

Originally Posted by orange4boy

. . .
I noticed, from the EWD that there seem to be quite a few different blocks or busses. I was thinking, If there was a simple way to keep the DC-DC happy by leaving it to run the ECU's, or at least the engine bus then it would be less likely to set any codes. I have noticed on my van that some sensors (eg. coolant temp) read low when the voltage gets low.

All the right questions.

Originally Posted by orange4boy

. . . What is the electrical efficiency of the THS from ICE (35%eff) to MG(?%Eff) to battery(?%Eff) to DC-DC(?%Eff) to 12V? How does it compare to a regular ICE(20%Eff) belt(?%Eff) Alternator(45%Eff) 12V?

To the best of my knowledge, no one has done a through job of researching this. I have fuel consumption figures for my inverter setup but I've not tried to work the problem back to the source(s).

Originally Posted by orange4boy

. . . Is the DC-DC a simple load based unit? Will it just supply what is asked of it and no more?

Once I start drawing 1 kW, the voltage drops off keeping the 1 kW power but now with more amps and less volts.

Originally Posted by orange4boy

. . . What happens to a stock Prius if the DC-DC goes offline while driving? If the car still operates then we know it would be possible to simply substitute the 12V supply. (better if 13.5V)

Will it be a simple/desireable thing to supply some blocks with 12V and some with the DC-DC?

Should I have a relay set up to switch between systems for long trips when the batteries get low?

Fun stuff.

Fun stuff indeed! I've not tried that experiment but you are asking the right questions. <GRINS>

GOOD LUCK!
Bob Wilson

[orange4boy]

Thanks for the above, Bob.

I've been studying the EWDs and there are some mysteries. One thing that seems obvious is that the inverter powers two main branches, One with the 12V on it with some basic functions and another, isolated from the 12V battery with a much higher total load. It may turn out they are all on a communal bus, but that is how it looks from here. It somehow does not seem logical, on reflection because it would mean that the DC-DC has two isolated output branches. Possible for redundancy but not likely.

I need to find out what the function is of the three leads which are mentioned below...

The "Toyota Prius Electrical Wiring Diagram, 2003 Model" is publication EWD893U would be an excellent guide. Looking on pp. 50 and pp. 59, it looks like the main inverter 12 VDC supply and sense lines are:

C5 connector, pin 1 - main supply, WHITE
C4 connector, pin 3 - sense or secondary 5A line from aux B+ terminal, BLACK with RED STRIPE
C4 connector, pin 1 - IGCT unknown, WHITE with GREEN stripe to all ECUs
C4 connector, pin 2 - IDH unknown, YELLOW to A/C amplifier

One is apparently a sense lead but for what? Why does the DC-DC need that when it has the main line right there? Just thinking out loud.

Back to the diagrams...

[orange4boy]

The area of concern:



Mystery one solved.

The green/black wire from the 12V battery block is the main ignition relay wire.

Mystery two:

There is only one main power wire from the DC-DC: White.

What else would the DC-DC wires be?

1) 5A black/red: Main start up power? Battery charging wire? This would only make sense if the battery could be isolated. Most alternator systems just vary the whole system voltage which the battery happens to be attached to.

2) Voltage sense wire.

3) amp sense wire.

4) HV supply. This must be the shielded HV wires.

5) Temperature sense wire.

[orange4boy]

Exciting news!

I tried a few simple experiments with the NHW11.

I disconnected the main 12V supply wire from the inverter. The larger grey connector/white wire below the MG2 main cables at the back of the inverter. Started the car and got P3004 Power Cable Malfunction (Prius), P3125 Converter & Inverter Assembly Malfunction (Prius), triangle etc. Still ran.

Hooked up my 50A switch mode battery charger to the 12V battery and set it to 50A. ~ 14.5V Codes and triangle remained. Cleared codes with Graham scanner... turned off car and started again... Ran fine, no codes, no triangle.

Running load showed a little less than 20 amps. I tried fans and lights and some other loads which added to that but not as much as my van. Apparently they have designed some pretty efficient electrics in the Prius. I will try to do a detailed study of the loads if there is no other reference.

Then I shut off the car, disconnected the charger and started it on the 12V battery alone. It worked and the codes did not come back. All the way down to 11.3V

Then turned off the car and plugged the inverter connector back in. All good and showing 13.8V at the battery.

My question is why did the codes not come back when I ran with just the 12V battery the second time? Does the ECU ignore faults for a while after you clear them?

Conclusion: The car runs on just the 12V battery with the main inverter supply wire disconnected. This means it should be a fairly simple thing to supply the 12V system with plug in juice. I may ultimately need a good 100A DC-DC converter for whichever battery pack I decide to use. Hopefully the voltage is the trigger in which case one simply must keep the supply above x-volts to keep the ECU happy.

We shall see.

[bwilson4web]

Originally Posted by orange4boy

Exciting news!
. . .
Hooked up my 50A switch mode battery charger to the 12V battery and set it to 50A. ~ 14.5V Codes and triangle remained. Cleared codes with Graham scanner... turned off car and started again... Ran fine, no codes, no triangle.

Running load showed a little less than 20 amps. I tried fans and lights and some other loads which added to that but not as much as my van. Apparently they have designed some pretty efficient electrics in the Prius. I will try to do a detailed study of the loads if there is no other reference.
. . .

Excellent news!

If you get a chance, see what the Graham scanner reports for the traction battery current with and without the inverter 20 A. (14.5 * 20 = 290 W.) A detailed load study is an excellent idea. You wouldn't want the control computers to 'lose their mind' if someone flipped on a high current load, say window defrosters and windshield wipers, at the same time. Probably anything with a fuse greater than the 20 A. would be a good candidate and as well as the brakes (electric accumulator.) You are also in a position to see how low the 12 VDC system can go before obvious codes are throw.

Hummmm, 290/745 ~= ~.39 hp, nominal savings.

GOOD WORK!

LATE THOUGHT:

. . . I may ultimately need a good 100A DC-DC converter for whichever battery pack I decide to use. Hopefully the voltage is the trigger in which case one simply must keep the supply above x-volts to keep the ECU happy. . . .

This mode, where the deep discharge battery and charger provides the operating current provides a 'fail safe' with one risk ... over charging the existing AUX battery. I had a generator voltage regulator once fail and the 12 VDC battery got hot enough to emit acidic steam ... an interesting problem.

Bob Wilson

[bwilson4web]

Driving into work, it occurred to me there is a very simple circuit that could easily work:

1. 100 A diode - inverter output to 'white wire' - this will typically drop about 1.2 V making the output voltage closer to 12.6 VDC if any current flows from the inverter
2. 13.9 VDC DC-DC converter output voltage - except for any possible voltage drop, the deep-discharge battery should provide the bulk of the power. Then when it goes 'dead', the vehicle inverter picks up the load.

You'd still need some sort of indicator that the deep-discharge battery is drained. Also, you'll need a 12 V. battery to provide starting power if the deep-discharge battery is 'out of juice.' ... Just a thought.

Bob Wilson

[orange4boy]

This mode, where the deep discharge battery and charger provides the operating current provides a 'fail safe' with one risk ... over charging the existing AUX battery.

I did think of this and have had the issue before. I was experimenting with alternator delete and supplying 14.5V to the system so I installed a battery isolating solenoid to take it out of the system. But that's a manual system. If the voltage was kept =or<13.5V then you would not have to worry about overcharging. Perhaps another big diode on the 12V battery?

Your idea is more elegant, simpler and provides some measure of automation. Would the voltage after the diode still be lower when the Prius converter takes over? I'm sure there is a way to design a circuit that would solve that problem automatically but I'm not the man for that design although I could build it, given a diagram. That could be done later when we get proof of concept. The math is good but the real world always has the final say. As the kids say..."Reality bites".

I think the People's Hybrid system used 4 x 50 amp DC-DC converters in parallel which might make the whole thing cheaper. I've tried to find high amp converters and they are $$$$$ but lower ones are $.

Now if we could find out where the ECU get's it's Voltage signal and leave the Prius converter hooked up to that then we don't need any $$$ converters.

Hummmm, 290/745 ~= ~.39 hp, nominal savings.

This is why I was hoping for efficiency numbers on the MGs and converter. In a standard stone age system, it's more than double.

If you get a chance, see what the Graham scanner reports for the traction battery current with and without the inverter 20 A. (14.5 * 20 = 290 W.)

Oh, right...(smacks forehead) Form the current draw on the traction battery I could figure the efficiency of the DC-DC. That's one down...

Terribly exciting. Not as sexy as a PHEV conversion but massively easier and cheaper.

[bwilson4web]

I was thinking about this at lunch and realized the deep-cycle battery only has to provide the 90% load over the time, vehicle base load, and let the vehicle inverter with diode handle the peak loads (aka., rear window defroster, windshield wipers, night driving lights and brake.) Anything over that would drop the output voltage and the inverter would then provide the make-up or peak power. For example, you measured ~20 A. in your first test. Add whatever you consider to be additional, standard load plus 5-10% and that would be your deep-cycle battery DC-DC converter design goal << 100 A. <grins>

This also, in theory, means we can lighten up the power diode, forward current limit and use a less expensive part. However, there are tons of automotive alternator diodes out there that should easily be within this range and quite affordable.

One other thing, the vehicle ECUs use 'cheap' DRAM for persistent memory. What this means is losing the 12 VDC wipes out not only the clock and radio stations but also the ICE Lambda constants. Many have noticed that it takes a couple of runs before the Prius 'relearns' these values and becomes more efficient again.

Bob Wilson

ps. CBS3502412 (300W) or CBS3504812 (348W) would be a very nice converter and Allied sells them $171.53 ... <hummm> Also, the TI PT7750, 15A is another nice part. Then there is is Ebay part: 160381658887 ... sounds too good to be true.

[orange4boy]

More good ideas.

One other thing, the vehicle ECUs use 'cheap' DRAM for persistent memory. What this means is losing the 12 VDC wipes out not only the clock and radio stations but also the ICE Lambda constants. Many have noticed that it takes a couple of runs before the Prius 'relearns' these values and becomes more efficient again.

This only happens when the 12V is disconnected though right? Not if it just drops below 12v.

For now I'm going to test it with my 12V batteries - 750 watt inverter - 50 amp switch-mode charger - 12V Prius. Some losses from the batteries there but it's all the same to the Prius.

In the meantime I should be on the lookout for some big-ass diodes and hopefully a similarly big-ass DC-DC. The diodes reside in the voltage regulators, yes? Or are they mounted on the alternator?

Is the diode simply wired in line with the Prius DC-DC main wire? Which way should the "ring" face.

I'm going to need block heaters to get decent mileage in the winter here. I just did two trips to the Cove and back and the Km/100 bars traced a hyperbolic curve down from the cold drive to the warm.

[bwilson4web]

Originally Posted by orange4boy

. . .
For now I'm going to test it with my 12V batteries - 750 watt inverter - 50 amp switch-mode charger - 12V Prius. Some losses from the batteries there but it's all the same to the Prius.

Excellent approach. You'll get a rough-order feedback on the potential fuel savings without putting a lot of money in the project. If it looks good after and "A" "B" "A" test, then an optimized design makes sense.

Originally Posted by orange4boy

. . .
In the meantime I should be on the lookout for some big-ass diodes and hopefully a similarly big-ass DC-DC. The diodes reside in the voltage regulators, yes? Or are they mounted on the alternator?

It varies. I was looking at Ebay and saw a Ford and Chyrsler external set set of 100 A diodes but I haven't checked Digikey, Allied and Mouser, yet. What really surprised me were some Schottky diodes. These have an unusually low forward voltage drop.

Originally Posted by orange4boy

. . .
Is the diode simply wired in line with the Prius DC-DC main wire?

The 'ring' should be the direction the "+" charges exit. Just check it with a battery and VOM. The positive charges pass in the direction of the arrow towards the base.

Originally Posted by orange4boy

. . .I'm going to need block heaters to get decent mileage in the winter here. I just did two trips to the Cove and back and the Km/100 bars traced a hyperbolic curve down from the cold drive to the warm.

In the USA we have to order block heaters from Canadian dealers.

Bob Wilson