As we all know, the Prius will send power from MG1 to MG2 at low speed to allow the engine to operate without the need for a gearbox. But I was wondering if the reverse is the case under some high speed conditions: Does the Prius under some conditions send power to MG1 to allow the engine to spin slower than it would if MG1 were simply locked ?
Short Answer: YES. Long Answer from Graham's site: Now, the higher speed and the low ICE rotation rate puts the power split device in an interesting situation. MG1 must now spin backwards, as in the table and the animation at right. By spinning backwards, it makes the planet gears rotate forwards. The rotation of the planets is added to the rotation of the planet carrier and makes the ring gear move much faster. To compare this situation with the previously discussed situation for acceleration and climbing hills, you may want to open two browser windows on this page and look at the animations side-by-side. The rotation rate of the planet carrier (ICE) is the same in both, as far as I am able to make it. The difference, once again, is that in the earlier case we were happy with a high ICE spin rate to get plenty of power even though we weren't moving that fast. In this new case, we want the ICE to stay at a low spin rate even though we're up to a fair speed so as to deliver the lower power demand with high efficiency. We know, from studying the power split device, that MG1 must exert a backward torque at the sun gear. This is how the ICE achieves leverage to turn the ring gear forward. Without the resistance of MG1, the ICE would just spin up MG1 instead of moving the car. When MG1 was spinning forward, it was easy to see that this backward torque could be supplied by generator drag. Hence, the inverter electronics needed to draw power from MG1 and the backward torque appeared. But, now that MG1 is spinning backwards, how do we arrange to supply this backward torque. Well, how would we make MG1 supply forward torque and spin forwards? By being a motor! It's the same in reverse - if MG1 is spinning backwards and we want torque in that same direction MG1 must be a motor and push using electrical power supplied by the inverter. This is starting to look peculiar. The ICE is pushing, MG1 is pushing, is MG2, then, pushing too? There is no mechanical reason why it couldn't - for a while. This may look attractive at first. Two motors and the ICE are all contributing to motive force. But, we need to remind ourselves that we got into this situation by reducing the ICE spin rate for efficient low-power operation. This would not be an effective way to get a lot of power to the wheels; to do that we should increase the ICE spin rate and go back to the earlier situation of MG1 spinning forwards. Having got that issue out of the way, we need to think about where we're going to get the power to run MG1 as a motor. The battery? For a time, we could do this, but after a while we'd be forced out of this mode of operation with no battery charge left for acceleration or climbing hills. No; we need to get this power continuously, without drawing down the battery charge, so we're forced to the conclusion that the power must come from MG2, which must therefore be acting as a generator.