Contactor (Contactor Based PHEV)

I believe you are correct, the early DTC issues were before they had discovered the hall effect bypass. The cal-cars solution was a pretty brute force affair. You had a big lead acid pack at a higher voltage than the hev pack and used the controller to open and close the contactor dynamically to drive just enough current into the HEV battery to trigger SOC drift, but hopefully not enough to over charge and explode the HEV battery The constant cycling of the contactor and the resulting big current surges can't have been good for the HEV battery in any case. However it was the best solution at the time for a reasonable cost and decent performance. Using a dc:dc converter to control current and voltage dynamically was always deemed a more elegant solution, but at the kind of power levels needed to fully satisfy current demands (15-20kW?) it was never really a cost viable option as manzanita micro's solution illustrated. Enginer threw an interesting wrench into the discussion by showing just how much you could accomplish with a simplified low output converter. This was a really interesting development and can yield impressive results under the right conditions (mainly sustained moderate speed driving I'd say), but has a lot of limitations. In general its hard to do better than just tying the batteries together without an expensive and less than 100% efficient piece of high power electronics between them.

SOC spoofing as done in the BMS and a number of other controllers really made the classic cal-cars contactor switching method irrelevant. Now there is no need to yank the HEV battery voltage up by driving current into it to trigger SOC drift and recal, since you can just "lie" to the hybrid controller and tell it you have done so. That allows your PHEV battery to be more closely matched to the HEV battery voltage, eliminating the need to carefully connect and disconnect it dynamically to protect the HEV battery. In these modern conversions the contactor no longer plays a real role in the PHEV operation, they are really now just a piece of safety equipment just as the contactors in the HEV battery are. Its very important to note this difference between the job of the contactor now vs. in the cal-cars conversion. Its really a completely different conversion method. For the most part you can now just leave the two packs connected together, letting them work together to supply drive current and reducing resistance and increasing capacity to capture regen energy. This is a much more elegant and effective solution, and IMHO also makes the dc:dc converter option largely irrelevant.

 

miscrms,
Did you get my pmail?

 

I always get this confused particularly after so many years, but here's how I recall the hall effect bypass issue. Someone may be able to jump in and confirm if this is right.

There are essentially two main places in the HEV battery pack you can inject PHEV battery current. On the battery side, and on the inverter/motor side.

Putting it in on the inverter side has the advantage that the current going into the HEV battery is seen by the battery hall effect sensor and registers on the coulomb counter in the bms just like regen current. However when the batteries are delivering current to the inverter motor, the current supplied by the PHEV battery is not seen by the Hall Effect sensor and so there is a mismatch between the current measured at the inverter and the current measured out of the battery. If this mismatch is too large (>~14A) the hybrid controller will throw a DTC and disable the hybrid system assuming their is a fault. This is basically how the Enginer kit works. You don't need a fancy controller because the OEM BMS tracks the extra charge supplied by the PHEV battery, but in order to avoid DTCs the total amount of current injected must be kept fairly low relative to the electric drive demand which can be over 100A. The early cal-cars work wanted to use this function of the built in BMS but routinely triggered DTCs due to excess current so they had to find a new solution.

Injecting current at the battery side solves the discharge DTC issue, as now during electric drive the PHEV current is counted along with the HEV battery current and the currents at both the battery and inverter/motors all match. So now you can supply as much current as you like from the PHEV battery without throwing DTCs. However, when the HEV battery is charging, any current coming in from the PHEV battery is not seen by the hall effect sensor and so is not counted by the OEM BMS. This means we need a way to manage the real charge in the battery and get it to report SOC values that will force the hybrid controller to use all this extra electric drive capacity. In the early cal-cars this was accomplished through closing the contactor to raise the voltage to >244V and trigger an SOC drift recalibration in the OEM BMS every time the reported SOC started to fall below ideal values (~72% as I recall). This was later replaced with SOC spoofing and opening and closing the contactor to maintain the HEV battery at a more nominal SOC value. This then evolved into the kind of thing we're seeing now where the PHEV battery is more closely matched to the HEV voltage, and they can just be kept connected in parallel most of the time while the controller deals with feeding the right SOC values to the hybrid controller to maximize EV use.

You can see the difference in how the two approaches tap into the Prius HV system in the pictures below.

First a picture from an Enginer install showing the PHEV taps on the right hand side of the contactors just before the HV wires leave the battery case. This represents the first case, where the connection is made on the inverter/motor side and current flows through the hall effect sensor when the HEV battery is charging.


The second example is from the cal-cars conversion, where the connection is made to the left of the hall effect sensor on the battery side. This results in PHEV current passing through the hall effect sensor when the batteries are discharging into the inverter/motors.


Hope that makes some sense

 

I did, thanks for the reminder. I'll send you a soon pm.

 

This is a more elegant approach to the before Hall sensor electrical tying

 

Not first hand, but maybe first and a half hand. I was on the eaa listserv during the cal-cars development and was helping with mathematical modeling of batteries to estimate the relative performance of different battery brands/models. Nominally the packs had a Voltage of ~240V, and a capacity of 20Ah. An average discharge strength of 60A resulted in an effective capacity of ~11Ah, or 2.6kWh. Not great for a 300+ lb pack that took up more or less the whole rear underfloor area. Range was 15 miles on a good day, but often 10-12 miles under realistic conditions. Due to poor capacity and high weight, packs were routinely discharged to 100% DOD and left tied to the OEM battery in charge sustain mode so the efficiency gain would more or less cancel out the weight penalty. These factors both were very hard on battery life. The batteries used were about the best available and had an estimated life of 300 cycles at 100% DOD. Specified life is for 0.1C charge, 1C discharge, so real conditions were harsher. In practice range would start heading south after a few months, with some packs only lasting 6-8 months and some making it to about a year. The BB brand batteries used seem to currently run $70-80 each for a pack cost of $1400 to $1600. You see cheaper prices advertised, but they are generally for an off brand "compatible" replacement with much lower specs for PHEV/EV use. Its worth noting that the $/kWh for these lead batteries is about the same as lithium now, even without accounting for the 10X+ difference in cycle life. $1400-1600 for 2.6kWh is $538-615/kWh, vs. <$2480 (qty discount unknown) for 4kWh of GBS cells at <$620/kWh.

 

Please note that the PHEV HV connection has live voltage all the time even when the vehicle is off. At the minimum, the positive lead should be connected to inverter/motor side of the relay so that when the vehicle is off or the vehicle is crashed, the current loop is open.

Bypassing the current sensor could mess up the BCU's SOC calculation and could put the traction battery in dangerous situation to be overcharged.

 

Please note that Enginer kit is scalable. We proved it by paralleling three 8kWh kits in a 2001 Prius during the Green Grand Prix competition in the past weekend. It has combined 15kW throughput and 24kWh capacity. It achieved better fuel economy than two PIS equipped Gen 2 Prius vehicles, the Toyota Plug-in Prius and the Chevy Volt.

 

Please!!!!, that was a MPGe competition test drive.
Under regular circumstances no one will have have 3x 8 KW worth of batteries and some converters with an add on weight of over 700 Lb. at least.
I understand that Jack feels very proud of attaining such an accomplishment for his company but, you really want to compare apples to apples, not a fruit salad of energies.

 

That is correct Jack, which is why I use a BMSplus to adjust SOC based on traction voltage and which also allows me to use the full 19 amps output from your new Version 6 DC Converter instead of having to limit the output current to 15 amps to avoid getting the dreaded DTC due to current mismatch.