Page 15 of 16

Re: GS450H Discussion

Posted: Mon Apr 25, 2022 9:44 pm
by ggeter
If you saw my welding job for my battery racks, you'd have a different opinion :)

Re: GS450H Discussion

Posted: Mon Apr 25, 2022 11:50 pm
by ggeter
TheSilverBuick wrote: Mon Apr 25, 2022 3:15 pm Be mindful, with out powering MG1 it "may" be turning backwards as MG2 is spun up, along with turning the mechanical oil pump backwards. Running the electric pump would keep the system supplied in oil.
I never experienced the system slipping white driving once underway, so I believe I had plenty of pressure. I can see what you are saying, though, and probably need to do more analysis.

Re: GS450H Discussion

Posted: Tue Apr 26, 2022 2:54 pm
by TheSilverBuick
ggeter wrote: Mon Apr 25, 2022 11:50 pm
TheSilverBuick wrote: Mon Apr 25, 2022 3:15 pm Be mindful, with out powering MG1 it "may" be turning backwards as MG2 is spun up, along with turning the mechanical oil pump backwards. Running the electric pump would keep the system supplied in oil.
I never experienced the system slipping white driving once underway, so I believe I had plenty of pressure. I can see what you are saying, though, and probably need to do more analysis.
My quote was for Greg as you are powering and managing MG1 speed, he is intending to leave it unplugged an no input shaft lock. I'm curious to see his results. I like your methodology on using MG1 to manage the mechanical pump and cycle the electric on only when at a stand still or very slow. It's what I would like to do, or as you mention forego the electric pump altogether and use MG1 exclusively since we have the option the drive it versus Toyota's original design of only being driven by the ICE. My GS450H transmission and matching inverter on on a pallet now heading my way.

Re: GS450H Discussion

Posted: Thu Apr 28, 2022 5:07 pm
by ggeter
Wrote code last night to spin up MG1 to 1000 RPM and hold it there when MG2 <1000 RPM. So at a stop sign, MG1 spins up and keeps oil pressure up. Once underway, MG1 spools down and freewheels again. Consumes about 3amps holding 1000 RPM.

No oil pump needed. Even works in reverse, but only up to a point. If I back up too quickly, MG2 reverse spin overcomes MG1 and oil flow stops.

During testing, MG1 spun up pretty high by itself a few times and the car never lurched forward. I was expecting it to from internal frictions at least, but no forward movement at all.

Next: put negative torque on MG1 if RPM exceeds 6K or 8K to keep it from getting above 10K.

Re: GS450H Discussion

Posted: Thu Apr 28, 2022 5:09 pm
by Gregski
ggeter wrote: Thu Apr 28, 2022 5:07 pm Wrote code last night to spin up MG1 to 1000 RPM and hold it there when MG2 <1000 RPM. So at a stop sign, MG1 spins up and keeps oil pressure up. Once underway, MG1 spools down and freewheels again. Consumes about 3amps holding 1000 RPM.

No oil pump needed. Even works in reverse, but only up to a point. If I back up too quickly, MG2 reverse spin overcomes MG1 and oil flow stops.

During testing, MG1 spun up pretty high by itself a few times and the car never lurched forward. I was expecting it to from internal frictions at least, but no forward movement at all.

Next: put negative torque on MG1 if RPM exceeds 6K or 8K to keep it from getting above 10K.
awesome, but MG1 is spinning opposite direction than MG2 correct? (it's the magic of the planetary gear design that converts that motion to "forward" motion?)

Re: GS450H Discussion

Posted: Thu Apr 28, 2022 7:44 pm
by ggeter
Gregski wrote: Thu Apr 28, 2022 5:09 pm
ggeter wrote: Thu Apr 28, 2022 5:07 pm Wrote code last night to spin up MG1 to 1000 RPM and hold it there when MG2 <1000 RPM. So at a stop sign, MG1 spins up and keeps oil pressure up. Once underway, MG1 spools down and freewheels again. Consumes about 3amps holding 1000 RPM.

No oil pump needed. Even works in reverse, but only up to a point. If I back up too quickly, MG2 reverse spin overcomes MG1 and oil flow stops.

During testing, MG1 spun up pretty high by itself a few times and the car never lurched forward. I was expecting it to from internal frictions at least, but no forward movement at all.

Next: put negative torque on MG1 if RPM exceeds 6K or 8K to keep it from getting above 10K.
awesome, but MG1 is spinning opposite direction than MG2 correct? (it's the magic of the planetary gear design that converts that motion to "forward" motion?)
No -- MG1 spins same direction as MG2 relative to the centerline of the transmission. In Damien's code, he spins it opposite because the code assumes the input is locked. In my code, I apply torque in the "forward" direction to get the mechanical oil pump to work. MG1 in "reverse" (per Damien's code) does not pump oil. When cruising, I'm assuming MG1 continues to spin in the same direction as MG2, but at the very least, the oil pump is still pressurizing, so I'm doing something right :)

Re: GS450H Discussion

Posted: Fri Apr 29, 2022 2:55 am
by AdvancedEnergySys
I know some have had questions with regard to another option for oil pump and think I may have a solution.
I'm currently converting my highlander hybrid to EV and it contains a cute little oil pump that is reasonable and has easy hose connections HERE

I'm also converting a BMW with the full 450h treatment and have had aux pump flanges made at a local engineering machine shop. I should have the design files to post up here tomorrow as well.

Re: GS450H Discussion

Posted: Fri Apr 29, 2022 2:58 am
by ggeter
Will it make 100psi?

Re: GS450H Discussion

Posted: Fri Apr 29, 2022 5:41 pm
by AdvancedEnergySys
I can certainly find that out but I would guess so since it is designed to push atf into bearings and is located after the radiator.
I forget where I found the info but at some point I found posts talking about how bmw has a trans oil pump with the same pump core just different housings as the OEM 450h. I bought one of those and had a flange made for it as well so if the pump in my highlander doesn't make the needed pressure then using it is the plan. I can't find where I bought the bmw pump but once I find out what one it was I'll update on both.

Re: GS450H Discussion

Posted: Fri Apr 29, 2022 6:17 pm
by ggeter
AdvancedEnergySys wrote: Fri Apr 29, 2022 5:41 pm I can certainly find that out but I would guess so since it is designed to push atf into bearings and is located after the radiator.
I forget where I found the info but at some point I found posts talking about how bmw has a trans oil pump with the same pump core just different housings as the OEM 450h. I bought one of those and had a flange made for it as well so if the pump in my highlander doesn't make the needed pressure then using it is the plan. I can't find where I bought the bmw pump but once I find out what one it was I'll update on both.
Keep us posted -- would at the very least be cheaper/easier to replace when/if it goes bad!

Re: GS450H Discussion

Posted: Wed May 04, 2022 2:47 pm
by ggeter
Gregski wrote: Thu Apr 28, 2022 5:09 pm
ggeter wrote: Thu Apr 28, 2022 5:07 pm Wrote code last night to spin up MG1 to 1000 RPM and hold it there when MG2 <1000 RPM. So at a stop sign, MG1 spins up and keeps oil pressure up. Once underway, MG1 spools down and freewheels again. Consumes about 3amps holding 1000 RPM.

No oil pump needed. Even works in reverse, but only up to a point. If I back up too quickly, MG2 reverse spin overcomes MG1 and oil flow stops.

During testing, MG1 spun up pretty high by itself a few times and the car never lurched forward. I was expecting it to from internal frictions at least, but no forward movement at all.

Next: put negative torque on MG1 if RPM exceeds 6K or 8K to keep it from getting above 10K.
awesome, but MG1 is spinning opposite direction than MG2 correct? (it's the magic of the planetary gear design that converts that motion to "forward" motion?)
I watched MG1 rotation speed live yesterday while driving. In my setup, I spin it to 1000 RPM until MG2 hits 1000 RPM then I set torque to 0 on MG1. I expected MG1 to keep turning, even speed up as it was being dragged by MG2 and mechanical oil pump. But, MG1 actually stopped rotating altogether even as I accelerated past 40 mph.

I thought this was odd.

Re: GS450H Discussion

Posted: Sat Jul 23, 2022 12:46 pm
by evMacGyver
Bryson wrote: Mon Apr 25, 2022 2:56 pm I’ve been playing with the software in mine lately, I’ve got only MG2 active until 20% throttle (to 50% of it’s available torque), then MG1 ramps up from 20-50% throttle while keeping MG2 constant, then they both ramp up together after 50% as the original firmware had them.

This keeps the car more efficient in most use cases where you’re in light throttle, using one motor at higher torque/efficiency versus two motors at lower torque/efficiencies. This also eliminated some gear lash noise I was getting around 20mph (~2k RPM).

Tuning the transition from one motor ramping to another forced me to roughly learn the torque/power split between the two. When MG2 is in second gear, MG1 provides approximately 50% more torque/power to the output shaft than MG2, assuming that torque commands are linear.

As MG2’s transmission is a 1.9:1 ratio (from memory, correct me if I’m wrong), that would put MG2 in First gear at 27% more torque/power than MG1.
Staging motors sound interesting. I do have some rumbling noise starting around 4k of MG1 speed and up, putting MG1 to zero torque I could eliminate a bit if transmission is source or not.

I looked motor speeds during drive and if I calculated correctly, absolutely maximum speeds of 11k (?) for either motor in different gears would be around:
Low gear MG1 -6524rpm and MG2 11000rpm
High gear MG1 -11000rpm and MG2 9137rpm
Am I correct?

Just curious how did you end up to these torque factors?

Re: GS450H Discussion

Posted: Sat Jul 23, 2022 3:23 pm
by Bryson
Those gear ratios seem about right to me.

I arrived at the torque factors by butt dyno. I’d steadily and smoothly roll on throttle for various settings and if I detected a change in acceleration when transitioning from one motor to the next, I went back and iterated until I couldn’t detect any bump or dip in torque linearity. It’d be more accurate with a dyno but it was actually not too difficult to feel. The numbers I have could be off by a few percent for sure.

Re: GS450H Discussion

Posted: Wed Sep 21, 2022 3:54 pm
by Solition1DC
Is it possible to run MG2 motor without making any connections to MG1, would like to physically remove it.

Re: GS450H Discussion

Posted: Thu Sep 22, 2022 2:26 pm
by Gregski
Solition1DC wrote: Wed Sep 21, 2022 3:54 pm Is it possible to run MG2 motor without making any connections to MG1, would like to physically remove it.
I'm currently doing that in my truck build, in Phase I, I only plan on driving around on MG2, but I do not plan on removing MG1 just not connecting it, somewhere around this video on my channel I connect only MG2 then MG1 etc if you are interested



Re: GS450H Discussion

Posted: Thu Sep 22, 2022 5:45 pm
by Solition1DC
Impressive and super informative video!
You've confirmed that wiring to MG1 is not necessary at all.

Re: GS450H Discussion

Posted: Wed Oct 05, 2022 4:29 am
by ianhora
Has anyone yet put an electric clutch on the spline of the transmission to lock the shaft, instead of locking it permanently?
Then one would be able to utilize the internal mechanical oil pump by releasing the electric clutch and driving the vehicle only from MG2.
This would save the use of the electric oil pump, which was not designed to be operated continuously.

Re: GS450H Discussion

Posted: Wed Oct 05, 2022 2:27 pm
by PatrcioEV-ATX
ianhora wrote: Wed Oct 05, 2022 4:29 am Has anyone yet put an electric clutch on the spline of the transmission to lock the shaft, instead of locking it permanently?
Then one would be able to utilize the internal mechanical oil pump by releasing the electric clutch and driving the vehicle only from MG2.
This would save the use of the electric oil pump, which was not designed to be operated continuously.
I think it would be simpler to write in some code that only operates the electric pump below a certain RPM. This has been discussed in here and I've put code in my software to shut the electric pump off over 500 RPM of MG2 (I believe), but I don't have mine on the road yet so can't really speak to the success of this approach.

Re: GS450H Discussion

Posted: Wed Oct 12, 2022 12:55 pm
by ggeter
I don't use the electric pump at all. See my previous post on this topic: viewtopic.php?p=39923#p39923

Works like a charm.

I thought about the electric clutch thing, too, but my car is so small, I decided to not use MG1 at all.

Re: GS450H Discussion

Posted: Tue Oct 25, 2022 4:49 pm
by Romale
I have been thinking about installing this transmission in my Hummer H2 for a long time, but I doubt very much if the ability of one mg2 is enough to control the rear axle of a heavy SUV? my goal is not to make a clean electric car. I want to connect the Lexus transmission to the native v8, which will work only at idle and by adjusting the power of the mg1 regen, slightly mix the torque from the native internal combustion engine to mg2. the main idea is to increase mileage without access to outlets with low fuel consumption. also, in the case of such a modification, no inspector on the road even sees that an electric car is in front of him. for the front axle, I would later add another motor. it is possible to SDU away Tesla, but it will be much later.
immediately, the idea was to link the Hummer engine and the Lexus transmission, forcing mg1 to generate energy, and mg2 to push the hummer.
What are your thoughts on this?
I have a rear axle gear ratio of 4.11:1 and a tire circumference of 2.88 meters! I'll probably have to always drive a lexus in low gear

Re: GS450H Discussion

Posted: Sat Nov 26, 2022 12:13 pm
by montyjohn
Hi. I've got some really basic questions as a newbie.
I would like to fit this gearbox to a P38 Range Rover. As an interim solution I would love to also have use of the original engine and treat it similar to a plug in hybrid if possible.
Any answers to the following would be greatly appreciated.

1. The GS450H box looks like the end can unbolt where the Lexus transfer box would mate up. Is this true, or is it only the LS600H boxes that can accept a transfer box? I would like to fix the Range Rover transfer box but can't really judge how possible this is yet.

2. Does the open VCU contain all the logic to allow this box (or the 600H) to work with an ICE attached? Damien mention locking the input, however, can the two electric motors be configured to prevent the input shaft from spinning.?

3. Does the VCU contain the logic (or does it need the logic) for the CVT to work correctly when running with an ICE?

Thanks

Re: GS450H Discussion

Posted: Thu Feb 16, 2023 2:26 am
by Compassion
Bryson wrote: Sat Jul 23, 2022 3:23 pm Those gear ratios seem about right to me.

I arrived at the torque factors by butt dyno. I’d steadily and smoothly roll on throttle for various settings and if I detected a change in acceleration when transitioning from one motor to the next, I went back and iterated until I couldn’t detect any bump or dip in torque linearity. It’d be more accurate with a dyno but it was actually not too difficult to feel. The numbers I have could be off by a few percent for sure.

(politely asking)

Hey Bryson!

I'm working on doing a GS450h conversion for a Series II XJ6L, and I was wondering how you did up the rails/mounts for the GS450h, and the battery packs. (I'll keep digging since I think I saw the post for the car on another forum.) Just trying to simplify an engineering problem for that portion and the stock driveshaft.

Re: GS450H Discussion

Posted: Thu Feb 16, 2023 11:45 am
by crasbe
montyjohn wrote: Sat Nov 26, 2022 12:13 pm 1. The GS450H box looks like the end can unbolt where the Lexus transfer box would mate up. Is this true, or is it only the LS600H boxes that can accept a transfer box? I would like to fix the Range Rover transfer box but can't really judge how possible this is yet.
I think in the 600h thread are some pictures, they are quite different.
2. Does the open VCU contain all the logic to allow this box (or the 600H) to work with an ICE attached? Damien mention locking the input, however, can the two electric motors be configured to prevent the input shaft from spinning.?
Short answers: No and no.
3. Does the VCU contain the logic (or does it need the logic) for the CVT to work correctly when running with an ICE?
Also no.


Long answer: Building a hybrid and getting the old combustion engine to play nicely with the hybrid gearbox is anything but an interim solution. It requires A LOT of work (not so much mechanical but software engineering: possibly adapting the original ECU, deep dive into control systems, ...) and even more tuning to get a decently driving result.
Possibly expect a broken engine and/or gearbox in the process if anything went wrong in the software, etc.

Re: GS450H Discussion

Posted: Thu Feb 16, 2023 3:12 pm
by Bryson
Compassion wrote: Thu Feb 16, 2023 2:26 am (politely asking)

Hey Bryson!

I'm working on doing a GS450h conversion for a Series II XJ6L, and I was wondering how you did up the rails/mounts for the GS450h, and the battery packs. (I'll keep digging since I think I saw the post for the car on another forum.) Just trying to simplify an engineering problem for that portion and the stock driveshaft.
I welded up some brackets for the rear transmission mount to adapt it to the XJ stock locations, and did some ‘arms’ that point forward to the original engine mounts and come back to a waterjet closeout plate at the front of the gearbox.

For the battery, I used 1x3” steel box section to stack Tesla modules with some brackets welded to bolt into the frame rails directly with nutserts. The rear batteries were easy , just more 1x3 stacking but bolted to the spare tire well. There should be some pictures in this thread here: https://www.jaguarforums.com/forum/xj6- ... ct-184994/

Hope that helps!

edit: custom driveshaft too! I don’t think you can keep the stock one and still package it well.

Re: GS450H Discussion

Posted: Thu Feb 16, 2023 9:01 pm
by nkiernan
Wondering if anyone is running Catphish's GS450H_v8 code in their setup? Does the wifi module need to be programmed specifically to work with this version? Or would a wifi module from something like the Zombieverter already be set up for it's OI web interface?