Best way to accellerate for ICE efficiency?
#61
Re: Best way to accellerate for ICE efficiency?
Hi Bob,
Thank you for your correction of the original drawing.
Back to your original post...
I have never seen such MG2->MG1 power flow on any material.
http://www.toyota.co.jp/en/tech/envi.../ths2/how.html
Would you please show us the reference pointer for the MG2->MG1 power flow?
Thank you for your correction of the original drawing.
Back to your original post...
I have never seen such MG2->MG1 power flow on any material.
http://www.toyota.co.jp/en/tech/envi.../ths2/how.html
Would you please show us the reference pointer for the MG2->MG1 power flow?
Bob Wilson
Last edited by bwilson4web; 02-09-2007 at 11:25 PM.
#62
Re: Best way to accellerate for ICE efficiency?
There are outliers due to the sampling nature of the monitor. Each data item is captured at separate times but to calculate things like fuel burn, you have to combine injector-time and ICE rpm. For power, the product of MG1 Nm and ICE rpm. There are transient states such as acceleration and braking where even with interpolation of the data, it gives an outliner. The data has to be treated as a sets of samples.
There folks doing original work with direct data capture from injector wiring and current sensing of MG1 and MG2. By using such 'real time' data sources, the problem of outliers can be greatly minimized.
Last edited by bwilson4web; 02-09-2007 at 11:31 PM.
#63
Re: Best way to accellerate for ICE efficiency?
Hi Bob,
I'm not sure what does 'heretical' mode mean.
Would you please explain?
Ken@Japan
I'm not sure what does 'heretical' mode mean.
Would you please explain?
Ken@Japan
#64
Re: Best way to accellerate for ICE efficiency?
Now I may not understand all of what they've discussed but one claim is that at high speeds MG2 may work as a generator to make the current needed to drive MG1. The claim is MG1 is driven to keep the ICE in a high-torque, low-rpm mode. That in fact was what my original drawing depicted BUT my experiments and data have yet to demonstrate "heretical mode."
I like the corrected drawing because it closely matches the SAE papers describing Prius operation. There are some subtleties that have to deal with the angular rates and power generation but we have enough data from the Dept. of Energy studies to answer those questions. It is much simpler system if MG1 is always used as a generator and the source of counter-torque for the ICE.
The miniscanner does a great job of showing current in and out of the battery. But it doesn't show AC current flow in and out of MG1 and MG2. However, my experiments with audio recording has shown the signature of generation versus motor mode in the profile. I just haven't tried to find and document the elusive "heretical mode," yet.
One option, I need to check is the availability of MG2 torque from the miniscanner. MG1 torque tells us when it is running as a generator or motor. MG2 torque should do the same. There is a fairly direct relationship between MG1/MG2 torque and current.
Bob Wilson
ps. I've expanded the mechanical description in the corrected drawing. Eventually, I'll work up a series that starts from the schematic and power flows and morphs into the physical layout.
Last edited by bwilson4web; 02-11-2007 at 09:02 PM.
#65
Re: Best way to accellerate for ICE efficiency?
I don't know if he was the first, but Graham Davies describes heretical mode (and others) here:
http://prius.ecrostech.com/original/PriusFrames.htm
http://prius.ecrostech.com/original/PriusFrames.htm
#66
Re: Best way to accellerate for ICE efficiency?
I don't know if he was the first, but Graham Davies describes heretical mode (and others) here:
http://prius.ecrostech.com/original/PriusFrames.htm
http://prius.ecrostech.com/original/PriusFrames.htm
Originally Posted by Graham_Davies
. . .
MG2 generates power for MG1 to be a motor? If you think this sounds all wrong, you're in good company. My efforts to model the Prius powertrain mathematically were stuck for weeks because I failed to realize that this could happen. When I presented it to the Yahoo! discussion group Prius_Technical_Stuff, I was met with skepticism and received one response with the subject "No Way!". As evidence was collected by people who could measure the ICE r.p.m. with OBD-II scanners and the mechanics were rehashed, this mode of operating the power split device slowly became accepted. Because both the ICE and MG1 contribute power which is combined by the PSD, the name "power joiner mode" was suggested. However, the idea of MG2 generating power for MG1 to be a motor was so counter to the way people believed the system worked that the name that has stuck is "heretical mode".
Let's run through it again and vary the point of view. The ICE drives the planet carrier at low speed. MG1 drives the sun gear backwards. This causes the planets to spin forwards and adds more spin to the ring gear. The ring gear still gets only 72% of the torque of the ICE, but the speed at which the ring spins is increased by MG1 motoring backwards. Turning the ring faster allows the car to travel faster for a low ICE r.p.m. MG2, incredibly, resists the car's motion slightly with its generator drag and produces electricity to be fed to MG1. The car is moved along by the remaining mechanical torque from the ICE.
. . .
MG2 generates power for MG1 to be a motor? If you think this sounds all wrong, you're in good company. My efforts to model the Prius powertrain mathematically were stuck for weeks because I failed to realize that this could happen. When I presented it to the Yahoo! discussion group Prius_Technical_Stuff, I was met with skepticism and received one response with the subject "No Way!". As evidence was collected by people who could measure the ICE r.p.m. with OBD-II scanners and the mechanics were rehashed, this mode of operating the power split device slowly became accepted. Because both the ICE and MG1 contribute power which is combined by the PSD, the name "power joiner mode" was suggested. However, the idea of MG2 generating power for MG1 to be a motor was so counter to the way people believed the system worked that the name that has stuck is "heretical mode".
Let's run through it again and vary the point of view. The ICE drives the planet carrier at low speed. MG1 drives the sun gear backwards. This causes the planets to spin forwards and adds more spin to the ring gear. The ring gear still gets only 72% of the torque of the ICE, but the speed at which the ring spins is increased by MG1 motoring backwards. Turning the ring faster allows the car to travel faster for a low ICE r.p.m. MG2, incredibly, resists the car's motion slightly with its generator drag and produces electricity to be fed to MG1. The car is moved along by the remaining mechanical torque from the ICE.
. . .
Bob Wilson
#67
Re: Best way to accellerate for ICE efficiency?
Oops, sorry. I knew that, I just went too hurriedly with the post.
#68
Re: Best way to accellerate for ICE efficiency?
Hi,
This morning, I reconfigured the Graham scanner and directly observed both modes:
What I need to do is refine my Prius schematic and show the various power modes with data samples from the Graham scanner. This will cover the principles of operation and refine the models. Then I need to add a series of charts that morphs the 'linear' Prius schematic into the compressed version and ends with photos of the parts. This will help folks more quickly understand what is going on. I'll probably post the drafts in "Prius_Technical_Stuff" for peer review and then share the final product here.
Of course one interesting question is can we operate our Prius with neutral MG2 power? I don't think so because there would be no power for MG1 to give the counter torque needed by the ICE. However, mapping the operational ranges and transitions between "normal" and "heretical" may lead to new understandings.
Bob Wilson
This morning, I reconfigured the Graham scanner and directly observed both modes:
- normal mode - power split device uses MG1 as a generator to pass power to MG2
- heretical mode - MG2 runs as a generator to make power for MG1, which the PSD combines
What I need to do is refine my Prius schematic and show the various power modes with data samples from the Graham scanner. This will cover the principles of operation and refine the models. Then I need to add a series of charts that morphs the 'linear' Prius schematic into the compressed version and ends with photos of the parts. This will help folks more quickly understand what is going on. I'll probably post the drafts in "Prius_Technical_Stuff" for peer review and then share the final product here.
Of course one interesting question is can we operate our Prius with neutral MG2 power? I don't think so because there would be no power for MG1 to give the counter torque needed by the ICE. However, mapping the operational ranges and transitions between "normal" and "heretical" may lead to new understandings.
Bob Wilson
#69
Re: Best way to accellerate for ICE efficiency?
I think "heretical mode" needs a new name ;O)
What's interesting about the set of spins you describe is that in the Prius PSD simulators online you can actually arrive at this scenario relatively simply in just normal playing with the sliders - but, because none of those demos worry about Torque, it doesn't tell the whole story.
One thing I can't get a straight answer out of my Engineer friends about is what exactly the addition is when the sun gear spins one way and the ICE the other - do you add the torque of the 2? Or does the ICE only provide torque above what the sun gear is already providing?
It sounds like what's being accomplished is that the sun gear is in a low-rpm, high torque state and the ICE is in a moderate-rpm, moderate torque state, but when cruising the car doesn't need all that torque so it bleeds off a lot of torque with MG2 with minimal impact to RPM, and a constant flow of electricity that can variably go to MG1 or the battery to provide a range of cruise gear ratios.
This PSD is one simple-looking, complex device.
What's interesting about the set of spins you describe is that in the Prius PSD simulators online you can actually arrive at this scenario relatively simply in just normal playing with the sliders - but, because none of those demos worry about Torque, it doesn't tell the whole story.
One thing I can't get a straight answer out of my Engineer friends about is what exactly the addition is when the sun gear spins one way and the ICE the other - do you add the torque of the 2? Or does the ICE only provide torque above what the sun gear is already providing?
It sounds like what's being accomplished is that the sun gear is in a low-rpm, high torque state and the ICE is in a moderate-rpm, moderate torque state, but when cruising the car doesn't need all that torque so it bleeds off a lot of torque with MG2 with minimal impact to RPM, and a constant flow of electricity that can variably go to MG1 or the battery to provide a range of cruise gear ratios.
This PSD is one simple-looking, complex device.
Last edited by SoopahMan; 02-12-2007 at 03:45 PM.
#70
Re: Best way to accellerate for ICE efficiency?
Hi,
On proposal was "electric overdrive" but in truth, I'm not sure there is a good term to use.
AMEN!!! What is needed is some kind of 'coloration' of the teeth showing torque that move the car forward as opposed to torque that is a consequence of generating shaft power.
This is the 'headache' exercise. The key is the planetary gears are driven by the ICE shaft. As they rotate, the ring-gear side is providing positive torque to move the car forward. The sun-gear side is being balanced by MG1 whose torque is also in the direction of moving the car forward. What I mean is if the ICE shaft broke and the planetary gears froze on their shafts, the forward torque of MG1 would move the car forward.
I'm not sure I follow this. Let me suggest that once I get my charts published and reviewed, I'll bring them here and it will hopefully make more sense.
I suspect the part that will be really confusing for some is that in some modes, the MG1 torque that allows the ICE power to move the car forward comes from generating power. In other modes, the MG1 torque comes from power generated by MG2 and sent to MG1. The reason for this has to do with the reactive force from MG1 being very speed dependent. BUT let me get the charts and data together and hopefully it will become clear.
AMEN!!! Someone once posted something about needing a PHD to understand the HSD/TSD. For sure, it has taken me a while to reach my understanding and Ken's posting helped clarify and lead to better experimental data. It is non-trivial, which means I have no envy for those adjudicating the patent dispute although I'd love to see all of evidence.
Bob Wilson
On proposal was "electric overdrive" but in truth, I'm not sure there is a good term to use.
What's interesting about the set of spins you describe is that in the Prius PSD simulators online you can actually arrive at this scenario relatively simply in just normal playing with the sliders - but, because none of those demos worry about Torque, it doesn't tell the whole story.
One thing I can't get a straight answer out of my Engineer friends about is what exactly the addition is when the sun gear spins one way and the ICE the other - do you add the torque of the 2? Or does the ICE only provide torque above what the sun gear is already providing?
It sounds like what's being accomplished is that the sun gear is in a low-rpm, high torque state and the ICE is in a moderate-rpm, moderate torque state, but when cruising the car doesn't need all that torque so it bleeds off a lot of torque with MG2 with minimal impact to RPM, and a constant flow of electricity that can variably go to MG1 or the battery to provide a range of cruise gear ratios.
I suspect the part that will be really confusing for some is that in some modes, the MG1 torque that allows the ICE power to move the car forward comes from generating power. In other modes, the MG1 torque comes from power generated by MG2 and sent to MG1. The reason for this has to do with the reactive force from MG1 being very speed dependent. BUT let me get the charts and data together and hopefully it will become clear.
AMEN!!! Someone once posted something about needing a PHD to understand the HSD/TSD. For sure, it has taken me a while to reach my understanding and Ken's posting helped clarify and lead to better experimental data. It is non-trivial, which means I have no envy for those adjudicating the patent dispute although I'd love to see all of evidence.
Bob Wilson
Last edited by bwilson4web; 02-12-2007 at 08:39 PM.