B position resurfaced

Anything in particular you want to know that isn't in what Prof. Kelly covered?

The short answer to "What happens mechanically to either the two electric motors and/or the ICE" is: nothing. The two motors and the engine (why does everybody geek out on saying ICE like it's some PriusChat shibboleth?) are connected mechanically by a fixed system of gears in constant mesh, in a relationship that never changes, ever. (The Prime does have an extra part to prevent the engine from being spun backward, allowing the transmission to operate with motor speeds that otherwise would do that, and which non-Prime Prii therefore have to avoid. But even that part is always there, and not something that is mechanically changed from any drive mode to another.)

The two motors are also connected electrically, in much the way Prof. Kelly demonstrated with his three jumper wires, except in real life each of those six terminals is wired to a pair of transistor switches that can rapidly switch it to either of the two bus rails, and the switches are flipped on and off umpty times a second with the timing precisely worked out by computers, and that's where all the magic happens. But that all happens electrically, not mechanically. And what direction the power is flowing turns out to also be simply a matter of timing, as in which switch flips lead or lag which other ones.

For a sense of how precisely that switch timing is calculated, you can observe that each motor includes a "resolver" that reports its shaft position back to the switch-driving computer (and from that real-time report of shaft position, of course, you can compute RPM). Some PriusChatter once looked up the specs of the resolver chip Toyota was using in that circuit, and found it divides each shaft revolution in 4096 parts, so that position is finer than a tenth of a degree. Let that shaft spin up to 5000 or 10000 rpm, and those resolved shaft positions are distinct every one or two microseconds. (Also, nothing in principle says the computer can't interpolate between those!)

Prof. Kelly's explanations and demos tend to be very good. If there's any weakness it's that he's trying to keep the level of the explanation very light on math, so he ends up doing a bit of hand-waving and appeal to intuition about the directions of current and power flow. Which is kind of helpful and kind of not: a lot of people might guess that if you start with his sort of tortured word-salad explanation from 18:30 to 19:40 or so, and then sprinkle in magic words like "complex number", "Euler's identity", or "phasor", you'd only be making it worse, but what really happens is there's a bright flash of light and the whole word salad sucks down onto the page leaving only a few symbols that would remind you of easy high school algebra.

 

Once the traction battery is full, the car will use the engine to slow itself, as the battery has no more capacity to hold incoming charge.

 

I’d suggest starting with the New Car Features books for the Prius Prime and for previous generations (2001, 2004, and 2010), and the Hybrid Systems: Course 071 Technician Handbook training course, all available by subscription to techinfo.toyota.com. None of these documents gives a comprehensive explanation of the control strategy, however, which presumably Toyota considers to be a trade secret.

There is also an extensive patent literature; see, for example, U.S. patents 5,915,801, 6,811,229, 9,669,808, and 9,702,304. These are all assigned to Toyota, though there is of course no guarantee that the inventions disclosed are used in the Prius Prime or any other model.

 

I'm only extrapolating since I rarely use HV mode and have even less opportunity to experience a full pack with substantial regen scenarios.
I'll hazard to say (my sources are recent B mode posts here at PC and my own experience with B mode while in EV only) B mode in the Prime behaves almost exactly like, if not exactly like, pressing the brake pedal lightly enough is only engage regenerative breaking.
So whatever happens when the battery is full and decending a long hill while riding the brake pedal lightly, should (B)e very similar, in not exactly the same as descending the long hill in B mode with a fully charged pack(?).

@Marine Ray
If you notice something that doesn't match any of the experiences other prime or prius drivers experience
it is always your choice to post what you have experienced and / or how it differs from others descriptions of what they have experienced.

For those of us that want to understand B mode in more detail, we are always looking for more specific data on what happens in B mode for every scenario possible, even if we only use one of two of those scenarios on a regular basis.

Also, here at PC you will see posts by members that use the stick and like it, mostly due to it's manual transmission downshifting type driver experience. While other PC members describe B mode from a perspective of it being either too confusing, complicated or flawed by design and/or implementation.
as always ymmv!

 

Personal experience in my Prime.

 

Ditto. And more-so in the PiP.

 

Because 'motor' and 'engine' are too vague and interchangeable.

============
From Merriam-Webster dictionary:

 

I'll add that besides MG1 and MG2 in all prii being motors and or generators and in Prime MG1 and MG2 can be both motors and/or generators (from what I've read so far) and the 1.8 liter ICE being an engine, which by now after 5 years of Prius PlugIn driving is getting to be second nature for me, is not so much for other Prius and Prime drivers.
With Primes newer functionality for MG1 and MG2 beyond that of those motor/generator functions in the non plugin models makes it even more of a chore to get the language easily understood when trying to describe the individual drivers experiences for the lay person(s).

 

A bit more salad mixing:
I've noticed that when driving below somewhere around 12 mph, the odometer register differently by a smidge than I've seen on the odometer registers at speeds above around the 12 mph mark.
It's not really to noticeable unless one drives the same route at the same speed often enough to get a handle on the odom at several locations along the route at a typical speed and than also notice the odom readings at those locations while driving at the slower speed where the odom readings change a smidge.

Make me give pause to think about exactly what the calculations are in the chip processing the 4x1024 shaft positions as well as if and when the processing skips some of those 4096 data points.

 

I don't think this is where the odometer senses. I don't have the technical data handy, but I would suspect it senses on wheel rotation.

 

The front wheel pair (at differential input) is mechanically linked via known fixed gear ratios, to the MG2 shaft, so it could be sensed at any convenient point along that path.

Meh. A narrow 12 bit buss, Gray coded so that only a single bit flips at a time, with those flips coming along at no more than 0.0007 GHz. This doesn't hold a candle to the 64-bit wide CPU and graphics processor data busses clocking along at 1 GHz rates in common computers.

And these motor controllers are not the general purpose math processors found in common CPUs, but instead are either dedicated devices custom built (if volume is high enough), or possibly programmable controllers (FPGAs or PLCs or other families I haven't heard about), streamlined around the single function of motor control, all extraneous functionality stripped away. The real-time architecture would be interesting, yes, but the data processing rates would be yawners.

 

Correct. The Repair Manual explains that the speedometer and odometer both use the wheel speed sensors: “The combination meter assembly [...] receives vehicle speed signals from the skid control ECU [...] via CAN communication. The speed sensor detects the wheel speed and sends the appropriate signals to the skid control ECU [...]. The skid control ECU [...] detects vehicle speed signals based on pulses of the voltage.”

Also true, but the wheel speed sensors are needed anyway for ABS, TRAC, and VSC.

Edited to add:

See N. Maeda, “Construction of Hybrid Vehicle Motor Generator ECU,” SAE Technical Paper 2006-01-1341, which gives specifications for the microcomputers in the ECU planned for Toyota’s 2006 models:

Operating frequency: 60 MHz
CPU performance: 83 MIPS
Internal RAM: 12 kB
Internal ROM: 240 kB (flash memory)
A/D Converter: 12bit 11ch x 2
Power supply voltage: Internal: 1.5V, I/O: 5V
Process: 0.15μm CMOS process​

These are general-purpose, stored-program CPUs, though they have some interesting peripheral devices; the paper has a block diagram. The instruction set architecture was not disclosed.

 

I don't have a subscription, so can't see the actual paper. But I'm suspecting that this is not the portion I was referring too, but instead is the ECU that supervises it, from one level higher.

 

Elektroingenieur has already listed the Prius manuals and technician training materials on TIS, which are what I chiefly rely on for building my own understanding of how the Prius works.

As to the 12-bit-ness of the motor shaft resolvers, that was someone posting on PriusChat the number on the IC connected to the resolver inputs, which was enough to look up a data sheet for that IC, and the specs of the result were from the data sheet. I haven't tried to search up the old thread just now, but it ought to be findable. If it were recent enough, I know just whose posts I'd be searching first to find it, but I think the thread was back before Elektroingenieur was around on here.

As to complex numbers, Euler's identity, phasor, pretty much any source should be fine. They all have entries on wikipedia (almost certainly also on Khan Academy). When I was taking electricity and electronics, the books we used were Purcell for E&M, Horowitz and Hill for electronics. That was some time ago (he wrote, vaguely), but both books are still going concerns that have kept abreast of new things and had new editions out since then.

If you change the o to an e in phasor, you get something that can be set on stun. Euler's identity is widely regarded as one of the most stunningly beautiful discoveries in math. Coincidence? You decide.

If that's the main question you're wanting to get at, there might be some information available, but it might end up being incomplete. The biggest part of the answer would be "it doesn't need to", as there's no point burning gasoline to make more power under conditions where the car is already harvesting more power than it has capacity to absorb.

But we don't know all of the control strategies programmed into the ECU, and I have a feeling the first one of us to say "absolutely! never, under any circumstances!" might get tripped up by some obscure rule that might be in there about the catalyst temperature getting too low, or something like that. But anything like that would be kind of a small codicil to the general rule.

Trick question? If by "regen" you mean harvested energy being sent to the battery, then there isn't any, because it stops being sent to the battery. On the other hand, if by "regen" you just mean energy being harvested from the momentum of the car, that's still happening; it's just getting sent to the engine to pump air and make noise.

 

@ 14:00 "See if we can let the smoke out" *then gets the resistor to start smoking*

This guy reminds me of my High School chemistry teacher "teaching" reactions by throwing sodium metal in water. Probably dangerous, but it makes the topic much more exciting.

 

I have observed that whether the engine runs in B mode is more complicated than the state of charge of the battery.

There seems to be a limit that varies based on time, temperature, and state of charge for how much power the battery can absorb. I assume there are other limitations on B mode such as the temperatures of the motors and inverter. For example, the battery seems to be able to absorb 40kw while braking for a short time. But using B mode for an extended period at high speed will turn on the engine, even if the battery is only partly charged.

Hybrid Assistant indicates the power into the battery in B mode varies based on speed. At 45mph it might be 18kw, and at 65mph it might be 23kw. I'm going by memory here, so these numbers are not totally accurate.

On a long descent at 65mph, the engine seems to come on within a couple minutes in B mode. Apparently the battery can't absorb 23kw for more than a few minutes (or the control algorithm is programmed not to allow it to prevent damage to the battery). But on a long descent at 30mph the engine will probably never come on (assuming you use B mode but no additional braking). I have charged the battery from empty to 80% descending a mountain at 30mph using mostly B mode.

I've observed using Hybrid Assistant that when the engine comes on, at least for a while the power dissipated by the engine is only a few kw, so there's still 18-20kw going into the battery. The car can measure the power dissipated by the engine because engine power is a function of MG1 power.

If the battery is cold that will also tend to cause the engine to be more likely come on in B mode. A cold battery can't be charged as quickly. And presumably if the battery gets too hot the engine will also come on in B mode, but the times I've watched it the battery temperature is still within a typical range when the engine comes on.

Braking with the pedal is similar to B mode, but with some differences. The power vs time limit for the battery seems to be a little different, allowing a little more charging while braking with the pedal before the engine comes on. But eventually it will also come on using the brake pedal (similar to how most automatic transmission cars automatically downshift if they detect that you are going down a hill).

The brake pedal also has several conditions that will activate the friction brakes, such as pressing the brake pedal suddenly, hitting a bump, or reaching a low speed.

To specifically address whether the engine burns fuel in B mode, you can observe the MPG display to find out. I have seen that if the engine is cold and it starts in B mode, it will burn a tiny amount of fuel (the MPG might go from 999 to around 600 on a 10 mile trip, maybe around 0.02 gallons), then it will stop burning fuel the rest of the time it is braking. At the bottom of the hill it will complete the warmup cycle which I estimate uses about 0.1 gallons of gas.