Regenerative Braking Details?

Whether a motor–generator is in the motor mode or the generator mode is determined by the rotation and torque directions according to the new-car features manual.

The translation:

• If a motor–generator is exerting a torque in the direction it is rotating, it is functioning as a motor.
• If a torque is being exerted on a motor–generator in the direction it is rotating, it is functioning as a generator.

They summarize the operation of the motor–generators as follows.

 

The ten-minute Mathworks video with the link given in #57 will actually show you how that happens, and how a moving field vector that always stays 90° ahead of the rotor can be made with just three stator windings that don't move.

Use the same math to position that vector 90° behind the rotor position, and you've got yourself a generator and a regen brake.

The companion video linked in the same post nicely animates how you make that happen using just six on/off IGBTs connected to the three coils.

In the first video, you'll notice the presenter talks about the importance of knowing the exact rotor position at all times, so the ECU can do the math to generate the stator field in the right relationship to it. That position sensing is what the 'resolver' does, described in the last part (h) of what Gokhan posted above.

The "MG1 and MG2 construction" section above has a point (f) that briefly mentions "reluctance torque" without explaining much. There's another thread here somewhere in the last month or so, where somebody posted a video about Tesla's motor construction, which uses the same technique, and includes a better explanation of that part.

Edit: found it ... it was Bob's thread.

 

... which is precisely what I asked.

And that C rate is how much? Or the Prime's maximum battery power is how much?

I vaguely remember some prior post pegging it to around 40 kW, but am not able to find it at the moment and don't have much faith in that vague memory.

 

The combined power of MG1 and MG2 is 76 kW (23 kW + 56 kW). Since the HV battery can drive both of them at the same time, the required HV battery power to drive them close to their power rating is a lot higher than 40 kW, especially given the fact that Toyota is using the 23 kW MG1 to supplement the 56 kW MG2 for high-power EV driving by employing a one-way clutch to lock the engine to the transmission case to allow MG1 to provide positive torque in that scenario. However, 76 kW and the 8.8 kWh HV battery capacity correspond to a C rate of 8.6C, which is a 60 minutes / 8.6 = 6.9-minute charge/discharge time. This could be detrimental to the HV battery cycle life if prolonged, and I don't know what the maximum C rates for charge and discharge Toyota allow for its HV battery under different driving conditions are.

 

They have to be rated for the maximum power they will ever be handling, which covers conditions other than pure EV.

Consider a full-power acceleration with the engine running. You've got 73 kW coming out of the engine, with much of it taking the mechanical path through the tranny, but up to 23 kW of that could be taking the electrical path (any more than that would be too much for MG1), and that would leave MG2 with only 33 kW of capacity remaining to accept extra power from the battery. (That hypothetical comes to 106 kW, which seems a little higher than the published "net system power" for the car, so it seems they hold back some from that particular operating point.)

It's not necessary that the battery alone be able to power both motors to their rated limits.

 

Note that MG1 isn't used as a motor in the HV mode. Only MG2 is used. MG1 and MG2 are both used as motors in the EV mode under high acceleration or high power. In the HV mode, MG1 is used as a generator, which subtracts from the ICE kW. So, a rough estimate of the hybrid-system net power is ICE - MG1 + MG2 = 71 - 23 + 53 kW = 101 kW. Toyota lists it as 90 kW. Perhaps the missing amount is due to losses in the generator and motor as well as some reserve to charge the battery when necessary.

 

Does anyone remember 2 or 4 years ago, wouldn't even try to find it now, a member here complaining about how much of a dog Prime was getting onto a Cali freeway? Again, no proof so I can only imagine how Prime might limit full throttle after a few seconds of max request and/or combined with history data if making those requests from the car on a daily basis.

 

I believe that. See my post above. When the SOC drops too much, MG2 would no longer be used, and the hybrid-system net power would be ICE - MG1 (MG1 acting as a generator to recharge the HV battery), which is 71 - 23 kW = 48 kW = 65 hp.

 

There are two details missing in that.

• In HV mode, most of the time, MG1 isn't used as a motor. It is used as a generator though. The power split through the transmission carries part of the engine's output mechanically, and part electrically. The electrical portion comes out of MG1 (as a generator) and rejoins at MG2.
• In overdrive conditions (light load cruising, low engine rpm, high road speed, low torque), also called heretical mode, the roles switch, and electrical power is actually flowing the way you don't expect: MG2 generates, and it rejoins at MG1, used as a motor. It's a little mind-bending (not called "heretical mode" for nothing), but the algebra works out, and the car definitely does it.

The power limit on MG1 applies whether it is motoring or generating. If 23 kW of engine power are following the electrical path, that is using MG1 to its full capacity as a generator, and using 23 kW of the capacity of MG2.

 

Note that when figuring device power ratings, motor or generator mode is generally "same same".

Then there is 'heretical mode', where I believe MG1 is motoring.

==========
Oops, Chapman's post now comes in and covers it.

 

I've only pushed the Primes acceleration up near the limit once in EV with SOC showing around 75%. I wish I'd had time and enough experience to check the voltage sage during that uphill hwy entrance lane than over the hill merging onto the hwy where the graph of DrPrius when Red and the app showed 100Amp for a second or two.
I've made plenty of lower speed observations of voltage sage for a lot of different scenarios of throttle / SOC to gain a better understanding of the traction pack charge/discharge dynamics. The dynamics are close to linear, but off slightly from what most would consider normal when comparing SOC to pack voltage.
The low end of SOC is also interesting as there is around a 5 volt buffer below where the engine will typically fire at 330 volts, but the 5volt sage will rebound back to 330 volts if the driver gets off the go pedal in time.
Low SOC behaves way different than Fully Charged but it's only noticeable if the drivers paying close attention and slowing down what most consider normal driving.
You've got to see it to believe it.

 

That's only sort-of true.

And temperature, and state of charge, and ESR which is partly controlled by degradation. And other things.

Prius and Prius Prime have *entirely* different batteries.

I've personally measured *average* rates of charge (for up to 7-8 minutes) up to 19.2kW on my Prime, and much more than that instantaneously.

 

There is no reference to "heretic" in the new-car features manual, but I know of this.

 

That's 2.2C (19.2 kW / 8.8 kWh), which is a 2.2 / 60 = 3.6% increase in SOC every minute. I typically see about a 2% increase in one minute during charging, which is about 1C. I think I saw regenerative-braking C rates around 4C. However, I don't have a precise measurement on either.

 

No there isn't; "heretical mode" naturally is the humorous name given by the earliest folks outside Toyota who figured out what was happening in that mode.

By great good fortune, you're posting in a forum that has been around since 2003, nearly as long as the Prius itself, which contains a wealth of information from a long string of people who have also studied the NCF manuals, the Oak Ridge teardowns, the Toyota technical trainings, and so on, and have constructed a lot of good explanations for how the system works.

... it had been a while since I watched that video Niels made with Blender, and I didn't remember whether he covered the overdrive operation, but yes, he did.

 

A cursory glance at all of the above information enables one one to understand why a Prius can be best understood as a mobile set of computers.

I am still reminded of the AAMCO TV commercials, which brag (with an exploded view) that your automatic transmission has over 800 parts, and AAMCO can maintain them all. OTOH, the Prius "transmission" has a few dozen parts, and hundreds of thousands of lines of computer code in several control units.

 

Conceptually, it's not that complex. There are a series of gears that can be laid out as planetary or parallel shaft and those gears are connected to two motor/generators and one internal combustion engine. One motor/generator is connected to the wheels, through a fixed gear ratio and a differential. The ICE and the other motor/generator are connected to each other and to the one connected to the wheels, through fixed gear ratios. Electronics control the *torque* on the electric motors with the purpose of controlling the speed and torque on all the elements. Those electronics are difficult in detail to implement, but conceptually, they're just knobs controlling electric machine torque (twisting force or moment). Another bit of electronics controls the injection, valve timing and ignition timing of the ICE. Again, the details are difficult but the concept is not - controlling the speed and output of the ICE.

 

It comes back to that idea of lots of different levels of describing the same thing being "ok" for different audiences and levels of interest. A description like the above satisfies the goal of "I'd like to more or less understand what the pieces of my Prius are for."

It starts to break down some if you start asking slightly deeper questions like "what are the paths the power is taking under such-and-such driving conditions", and then a slightly deeper explanation can get you those answers. But that one will start to break down on questions like "ok, how does the car make the power follow those paths when it wants to under those conditions?", and again there's a slightly deeper explanation that doesn't break down there, somewhere around the level of the ten-minute Mathworks video.

Eventually if your curiosity gets all the way to "no really, how does the ECU calculate the pulse timings driving the IGBTs to make that happen", then you're looking at something like the longer TI video. And at that point, it's fair to say CharlesH isn't wrong, that a lot of the Prius transmission exists in the form of ECU code.

 

Where did we find this Prius Prime new-car features manual?

 

Elementary, my dear Watson, elementary. This is textbook stuff in EE books.

The hard part for Toyota was to figure out the constituents and configuration of the hybrid synergy drive, namely the two separate hybrid permanent magnet/induction motor–generators instead of one, the planetary gear unit power splitter, and the one-way clutch. Once the pieces are in place, it's fairly easy to control them through standard straightforward electronics and algorithms.