PHEV experimental

It gets pretty complicated, your best resource is probably:
PriusPlus - EAA-PHEV

They also have a mailing list where most of the people working on this hang out and can answer questions:
Maillist - EAA-PHEV

The basic idea is this. You can't just monitor voltage, because voltage is a poor indicator of SOC on a NimH battery. Some have tried just simply paralleling under acceleration, but under light loads you still run the risk of over charging the oem pack. What you really need is a controller that monitors the CAN bus traffic between the battery controller and HV controller and intelligently switches the secondary pack in and out based on SOC and driving conditions. The second part of this is to encourage the Prius to use as much extra electricity as possible, which requires different SOCs in different conditions. The HV controller is pretty picky about what conditions it will do this in, so control has to be fairly precise. The current way to do all this is a closed source controller called the CAN-View. This reads the CAN traffic, displays lots of handy info on the built in screen (for '04-'05 models) and can be configured to switch a series of relays based on the states of various CAN parameters. These relays are then used to switch in and out the secondary pack via a large contactor. This contactor is much larger than the internal ones, as it must be able to connect and disconnect under full load. The secondary pack voltage is high enough (>240V) that it causes the battery controller to go into a recalibration routine (SOC drift) which is the only real way to convince it to use the extra power available from the second battery. Normally it just measures current in and out of the oem pack, and puts back what it takes out. Since it knows how big the battery should be, it will never take out more than the OEM battery can hold (a really only a fraction of that). It also monitors voltage, temperatures etc, but its SOC determination is primarily by coulomb counting, not voltage. Thats why it will simply ignore the current coming in from the second battery unless you get the voltage high enough to recalibrate. Ideally you would probably run the secondary pack current through the current sensor and have it counted just like regen current, however the car does not allow this. It has mutliple current sensors around the car so it knows what every component is using/producing. Like a GFI if it can't make all the currents add up it immediately disables the whole hybrid system as a safety precaution. An open source controller is in the works that will not have the fancy graphic interface, but will add an important feature, SOC spoofing. Rather than physically connecting the second battery to drag the oem battery SOC around, it simply maintains the battery at ~60% as the oem controller does. At the same time, it intercepts the battery controller's reported SOC, and tells the HV controller that the SOC is something else based on driving conditions. This lets you instantly react to different driving conditions, without having to wait 10-30 seconds to drag the actual battery SOC around. Maintaining the oem battery at 60% should also be safer for its health. Lastly, the professional converters often just do away with the oem battery and controller, and make their own from scratch. This has advantages in that you don't have to play as many games to trick the Toyota controllers into being cooperative, but has so far be too complicated for the open source community. One nice thing about the current paralleling scheme is that if anything goes wrong the car still just reverts to acting like a standard Prius. This is a nice feature for the do it yourselfer who is tinkering with his daily driver!

The dc:dc converter plan has its advantages and disadvantages. The people doing it currently do so with the argument that you can then use fewer larger more efficient/ideal batteries and upconvert to 240V. For a given weight this can give you more capacity and potentially range, and potentially a less expensive pack that lasts longer. Of course you have the added expense of the converter to account for. The other problem is that a bi-directional converter would be much more complicated. Without that, you can only pull current out of the secondary pack, you can't put any back in from regen. In the straight paralleling arrangement when the secondary pack is depleted it remains in parallel with the oem pack. This improves overall efficiency by lowering the effective pack resistance, and allowing you to capture more regenerative energy. Its not a huge effect, but it does seem to more than cancel out the negative impact of the weight of the conversion. In the existing dc:dc converter based solution this is not possible.
PiPrius - EAA-PHEV

Rob

 

Deleted. miscrms's post above makes mine superfluous.

 

Sorry, got a bit carried away

rob

 

Not at all! Your post covered more information, and covered it better, than mine, which cross-posted with yours, so to avoid confusion I decided to delete mine.

 

maybe put in a diode to prevent energie flowing back into the PHEV battery pack from the hv prius pack? when the voltage from the extra battery pack is lower then the prius pack?

 

Agreed, I like the dc:dc idea, but to pricey for me.

Like most open-source projects, the cal-cars documents are not always up to date. They are in a transition phase between the old controller and new one, so there is an interim control board that may not be documented there yet. If you are seriously thinking about this I would read up on whats there, and then get on the mailing list and ask a few questions about the status of things. I believe there is a good working recipe in place, but its not fully documented as everyone is sort of waiting for the new controller design to be finished up. From my experience so far, the folks on the list are very helpful.

Rob

 

In general you want just the opposite. Keeping the secondary pack in parallel with the oem pack once its been run down actually improves the efficiency of charging and discharging the oem pack. This results in a small mpg gain even after the secondary pack is run down. Its not a lot, but it seems to more than cancel out the loss due to carrying around an extra ~300lbs of batteries. This is one of the disadvantages to the current dc:dc converter solution, once the secondary pack is run down you're just hauling around dead weight. Your mpg will actually be a bit lower during this period than it was previously. In theory the dc:dc converter would allow you to get more charge out of the secondary pack, as you could maintain an ideal output voltage longer. In practice, the existing paralleling scheme runs the 240V pack down under 200V, which is already as low, or maybe even a little lower than you want it from a battery cycle life perspective.

Rob

 

In the case as described previously of using a large 2nd battery in parallel a big problem of trying to make a Prius a PHEV this way I guess is accommodating the need of the car to charge the primary battery. In order to simplify things when your in the EV mode since the engine-generator is then disabled why not also disable the regenerative feature of the braking system, this would allow you to parallel wire the batteries for as long as you are in the EV mode using a couple of relays and diodes (the diodes would be used, one for each battery so one battery wouldn't drain into the other). When you exit the EV mode the regenerative feature and battery wiring would return to it's original configuration. Disabling the regenerative feature of the braking system ostensively could be done by doing whatever it takes to fool the Prius that the brakes aren't being used if that was possible while still retaining the use of the car's conventional braking function.

 

the problem with hooking up a battery in paralell especially a different one then stock, is the tendency for the power to balance out, even if it was 201 volts on both packs, if one pack had higher capacity than the other it will end up burning up the OEM battery, what i would do is completely abandon the stock battery for a PHEV project, the stock battery will lead to a bottleneck in my opinion

 

Notice I added a skematic to my post that shows how both batteries use diodes to block draining of one battery into the other one while in the EV mode so burning up the OEM battery wouldn't be a problem.

 

when you prevent flow going into the prius HV battery then data from the battery ecu wil not be correct because the HV battery wil nog show any current drawn but the car is driving in EV mode and thats not normal
maybe it wil even exit ev mode automatically
i think you will get a red triangle on the dash and in any case a fault code generated.

do you have any prove of that
because when the voltage is the same on bot battery packs there no longer be any current flow
only if the bigger battery with the higher capacity is on a higher voltage then this battery wil discharge into the lower voltage prius battery and when it stay in that state it wil overcharge the prius battery

IMO

 

This is the novelty of my idea is I'm preventing flow going into the Prius HV battery but because I'm obeying it's rules it's letting me do it.
I think you think the diode in the skematic because it could prevent flow going in is but actually it isn't it only prevents flow between the batteries.
Example if you drove your Prius in EV mode without using your brakes you also would be preventing flow going into the Prius HV battery.

 

but when you don't take your PHEV battery current to the donut shaped current sensor then the battery ecu will not take into account the extra current and will present the prius with less info about the available energy and therefor the prius Will not use it and go out off evmode and start the engine


i have read this on a callcars mail list site
you need to run the extra current Thu the current sensor from the prius

 

in car stereos for people who have separate batteries, they use a battery isolator. i would assume its just a diode for high amps. stinger makes some. ive seen them go from like 80 amps to 240 amps. then you get a high ouput alternator and it will charge the extra batteries. this is in a typical 12v system. i would assume that you would have to rig something with a lil dif wiring for the prius.

 

I think you are getting off on the wrong track here. What has been shown to work is only adding current from the secondary battery into the primary battery. The Dutchman is right on, if you try to add current direct from the secondary pack to the inverter it will throw an error and shut down the hybrid system. It monitors current at the battery and at the inverters and motors, and if they aren't all equal, it assumes there is a short somewhere and turns off the whole system as a safety precaution.

It gets much more complicated than that, as the primary battery & controller will only accept charge from the secondary battery if you pull its voltage high enough to trigger the battery controller to recalibrate. This happens at about 240V. If your secondary pack is lower in voltage than this, current will flow into the primary battery but the controller will completely ignore it. As a result, you won't get any mileage improvement. Since you have to run a fairly high voltage secondary pack, you can't just leave it connected to the primary, you have to dynamically switch it in and out as appropriate or you will over charge the primary battery.

The last confusing variable is that unless you maintain the perceived SOC in a certain range, the car won't really use any more electricity than it does normally. This is ok if you are in EV mode below 34mph. If you want to drive faster than that you won't see any improvement. By carefully keeping the SOC within a few % window, you can persuade the HV controller to use a lot more electric drive which is where you get your 100mpg performance from. Unfortunately the ideal SOC to make this happens varies with temperature and driving conditions, so you really need a controller to figure out when to connect and disconnect the secondary battery to keep the primary in the ideal charge range.

Just getting rid of the whole oem battery and controller is an option. So far only the commercial converters have gone this route. Its proven too complicated/expensive/risky for the DIY crowd. The challenge is that if your new controller doesn't give all the right answers to all the questions asked by the HV controller under all conditions it will cause all sorts of problems, and can create dangerous situations. You are basically messing around with the com bus that runs the entire car including abs, air bags, steering, braking etc. Since Toyota basically publishes no info on this (beyond the basic protocol), you are left to try and reverse engineer a very complicated computer network.

Rob