[WIP] Mercedes 310E (310D campervan) - GS450H/GS300H

[Jacobsmess]

So I guess it's finally time I started a project thread, especially as I finally bit the bullet and bought a Lexus GS450H transmission, inverter and zombieverter....

Rough plan is to find an IPace pack as they seem to be getting cheaper and cheaper probably given they're so big and breakers don't know what to do with them after a while.... Charging no idea yet but fast charging is a must given the likelihood that I'll be achieving somewhere between 1.5-2miles/kWh so I'll need to work out some decent battery cooling.

Originally the engine put out 97hp and 192nm through a gearbox with a high gear of 4.4 and diff of 4.8 so the either the GS300H or GS450H would be adequate....

Ideally I'd go with the GS300H because it's lighter and simpler (no oil pump nor extra gearing) but given their gear ratio of 3.33 would be limited on top speed however, my plan whilst the GS450H is on the workbench is to experiment with a 3rd motor coupled to the ICE input. In the original Lexus setup the MG1 modulates the output shaft speed and gear ratio that is applied to the ICE. I would like to see if this is something I can implement myself, coupling a thin motor (MG1 from a Prius transaxle perhaps) to the front to act as an overdrive for higher top speeds. If I can get this working on the GS450H then I can move it to the GS300H and have a more versatile setup, albeit with a little extra complexity, but I'm willing to experiment.

Any tips/advice/abuse is welcome...

[21j]

I'll be watching closely. Good luck.

[barracuda816]

Sounds like a great project.

I have read up on the L210 (300h box) and also feel the gearing if off for most diff setups. How does adding an electric motor in place of the original ice act like an overdrive? Also how does welding the mg2 planetary gears make a difference compared to locking the input? Sorry if this should be its own topic but I have read up on this motor and im still not 100% and it looks like you have a good handle on it.

Kind regards, Luke

[Aragorn]

Not sure what exactly your trying to achieve with the third motor...

On both transmissions, MG2 is connected directly to the output shaft, and doesnt touch the PSD. MG2 to Output ratio is 1.9 for 450h and 3.3 for 300h

The GS450h with locked engine shaft runs MG1 faster than MG2 (about 1.2x), thus MG1 sets the overall speed limit. trying to use the PSD to slow down MG1 means your no longer using MG1 for traction, and instead are using it as a brake/generator. your adding significant complexity when you could just not use MG1 at all, or weld the PSD and run it at output shaft speed.

The GS300h with locked engine shaft runs MG1 slower than MG2, so doesnt have the same MG1 issue that the GS450h has. Your RPM limit is instead set by MG2. messing about with the PSD here doesnt get you any more RPM's.

Keep it simple. Run the GS450h as is. If you need more RPM's then either dont use MG1 for traction (leave it unlocked), or investigate welding the PSD instead.

[Scrappyjoe]

This might be a useful validator of your energy requirement calculations

https://ev-database.org/#sort:path~type ... g:number=9

Basically, find the most inefficient car you can find. Looks to be an eVito, at ~300wh/km. That's about 2 miles / kmh so... spot on!

[Jacobsmess]

This might be a useful validator of your energy requirement calculations

https://ev-database.org/#sort:path~type ... g:number=9

Basically, find the most inefficient car you can find. Looks to be an eVito, at ~300wh/km. That's about 2 miles / kmh so... spot on!

Yes, originally I was advised to use 1.5miles/kWh. I've hope it will be a little better than this based on the Ford Etransit and Maxus Edeliver 9 both achieving around 2miles/kWh.

[Jacobsmess]

Small wins as of today, I'm slowly collecting parts, the olimex board that was in the zombieverter didn't seem to be doing anything so I removed it and have just (with a lot of help from BigPie) got the OI web interface acccessible....

I followed the various information here (can someone with better editing powers here perhaps rename this page to something more generic as it's not Olimex only instructions?)
https://openinverter.org/wiki/Olimex_MOD-WIFI-ESP8266

Further updates, I've wired up most of the GS450H transmission/inverter connections to the Zombieverter, I need to work out how I'll connect this to the propshaft input flange, which is a smaller diameter of 90mm with an 8 bolt pattern. I'm hoping I can make up an adater that connects to the GS450H guibo (rubber propshaft connector thing).

[Jacobsmess]

Not sure what exactly your trying to achieve with the third motor...

On both transmissions, MG2 is connected directly to the output shaft, and doesnt touch the PSD. MG2 to Output ratio is 1.9 for 450h and 3.3 for 300h

The GS450h with locked engine shaft runs MG1 faster than MG2 (about 1.2x), thus MG1 sets the overall speed limit. trying to use the PSD to slow down MG1 means your no longer using MG1 for traction, and instead are using it as a brake/generator. your adding significant complexity when you could just not use MG1 at all, or weld the PSD and run it at output shaft speed.

The GS300h with locked engine shaft runs MG1 slower than MG2, so doesnt have the same MG1 issue that the GS450h has. Your RPM limit is instead set by MG2. messing about with the PSD here doesnt get you any more RPM's.

Keep it simple. Run the GS450h as is. If you need more RPM's then either dont use MG1 for traction (leave it unlocked), or investigate welding the PSD instead.

After rereading this for the 3rd time I think I've realised something perhaps a little obvious that would alleviate all my worries with either transmission.... Welding or locking the PSD will mean that MG1 spins the output shaft at 1:1, is this right?
If so, there is no issue with top speed but perhaps the issue is that you then limit mg2 top rpm which means limited power from it. Unless it's possible to spin only mg1 with mg2 unpowered unless there a need for torque.

Also, how is it that mg1 slings slower on the gs300H than on the Gs450H?

As a project update, I've been mostly welding my wheel arch and fabricating parts which is interesting/a pain in the arse... Although it's always nice for new metal to be in place of very old metal/rust/dust.

[Bigpie]

Thread needs photos/videos.

[Jacobsmess]

Thread needs photos/videos.

But then there's evidence of my welding abilities....?!

[Jacobsmess]

Can anyone explain to me, how and why the welded planetary gearset is different to the locked input shaft for NG1 speed ratios.
Also, why mg1 is the limiting factor with a welded gearset for gs450H but not the GS300H, as I understood the input shaft was not directly coupled to the output in either transmission. Is this correct? And as a result mg1 the s not directly coupled to mg2 nor the output (one can spin independently of the other). Can they spin at different speeds? E.g mg2 at 10000rpm and mg1 at 6500rpm?
How is the output torque shared in this case?
These things are very confusing and complex!

[tom91]

Picture of the 450h gear layout.

Blue is input shaft, it is the planet carrier. You lock it so it does not freely rotate the planet gears and prohibits torque transfer.

Now the reduction is the sun (MG1) to planets (which are now stationairy) to ring gear output. Welding/locking the planet gears to the ring gear means that the sun gear, which is MG1 will spin as fast as the output shaft which is connected to the ring gear.

Things are not that complex as you might think, everything is designed with torque and speed requirements. Depending on the motors physical size it will have a lower top speed.

A very good slide deck to review, which gets referenced on the wiki. https://slideplayer.com/slide/14432904/

[Aragorn]

I think i covered that in my last post but i'll break it down.

The planetary has three connections, MG1, input shaft (engine), output shaft(prop). The planetary gearset ratios are arranged in such a way that if you lock the engine input, then MG1 is forced to turn at (iirc) ~2.3 times the output shaft speed. If you welded the planetary instead, its now a solid lump and MG1, the input shaft and output shaft are all turning at the same speed, ie 1:1.

MG1 is the limiting factor on the GS450h, because MG2 is geared at 1.9:1, and the MG1 (with input locked) is geared at around 2.3:1 IE MG2 is turning slower than MG1 would be. On the GS300h MG2 is geared at 3.3:1, so its turning faster than MG1, and thus MG2 RPM limit is reached first.

[andybpowell]

But then there's evidence of my welding abilities....?!

don't worry you'll be in good company, I was taught by a coded welder and mine's still crap

[Jacobsmess]

Picture of the 450h gear layout.

image.png

Blue is input shaft, it is the planet carrier. You lock it so it does not freely rotate the planet gears and prohibits torque transfer.

Now the reduction is the sun (MG1) to planets (which are now stationairy) to ring gear output. Welding/locking the planet gears to the ring gear means that the sun gear, which is MG1 will spin as fast as the output shaft which is connected to the ring gear.

Things are not that complex as you might think, everything is designed with torque and speed requirements. Depending on the motors physical size it will have a lower top speed.

A very good slide deck to review, which gets referenced on the wiki. https://slideplayer.com/slide/14432904/

Hi Tom, do you have the original link for this diagram? I'd like to have a play with it? Thanks

[Jacobsmess]

So with a lot of very generous and patient help from BigPie I've been trying to get my GS450H to spin on the bench.

I'm yet to get precharge setup so just raising the current slowly from 0 on my branch power supply.

Frustratingly it seems the TPS I bought may be faulty. I got the same one as below as I'm converting from a cable drive and figured this would be flexible but when I give it 5V and grd to the below pins I get no signal from either wire. I do need to have a little play with it but unless the cables themselves are faulty I'm not sure what it could be. These do have a little PCB inside covered in glue so it's not possible to really see what's going on inside.

Been doing a bit of reverse engineering on a Peugeot throttle position sensor. These are cable operated, so ideal for a bike throttle. They're also cheap (only £16 new). Part number is 9643365680. Google for "Peugeot 306 HDi throttle position sensor"
IMG_1361.jpg
Rather than a true potentiometer throttle, it appears these use a contactless sensor to output a pair of analogue signals.
Using a 5V supply and a 1kΩ pull-up resistor, I got the following readings:
0% throttle white: 0.419V yellow: 0.209V
100% throttle white: 3.780V yellow: 1.871V
ueXW9clwTnuvoV6ppHciWg.jpg
I may use two: one on the throttle cable and one on the clutch cable (for regen control )

[crasbe]

Did you add the 1k Pullup resistor to the signal wires as well? That is mandatory for the device to work as far as I remember.

[Jacobsmess]

Did you add the 1k Pullup resistor to the signal wires as well? That is mandatory for the device to work as far as I remember.

I didn't, it was something I thought might be necessary but kind of hoped I could skirt over it. So I need a 1k pull up resistor in series with each signal wire?

[crasbe]

I didn't, it was something I thought might be necessary but kind of hoped I could skirt over it. So I need a 1k pull up resistor in series with each signal wire?

No, if it would have to be in series, it would be called a "series resistor"
The Pullup has to go between the signal wire and +5V, it it used to pull up the signal to the high level. Hence the name.
Since the TPS has two signal outputs, I guess it needs two Pullups. One for each signal wire.

[Jacobsmess]

Thanks, electronic nomenclature is not my forte! So how would the zombie read the signal? Does the resistor just bridge between the signal wires and 5V and the zombie signal wires connect to the TPS signal wires the same?

See my perfect diagram for what I'm asking, as you can see electronics diagrams are also not my forte....

And finally, as Gregski says.... You can't make this stuff up!

[Jacobsmess]

Right, so I'm hitting a bit of a wall ATM, struggling to get the motor spinning.
All wired up correctly as far as I can see, I get inverter whine when I turn it on but no life. I'm just using a potentiometer for the throttle ATM and I've calibrated it in zombie. Tried with a 120V 3A bench top PSU but Bigpie advised I'd be better off with a lower voltage battery. So I've hooked up 4 small 12V 7ah UPS batteries in series for 48V, they're old and cold so may not be up to it but I charged them at 53V last night and tried again this morning.... Nothing.....

Web interface is not showing the HV voltage under udc, is this normal?

Pulled out another 4 UPS batteries that are known to be a bit better (on my cheap battery tester at least)... I'm charging them at the moment.

One thing I did notice is there seems to be more resistance when spinning the output shaft whilst it is trying to run, (at least in one direction) so it seems as though it wants to spin at the very least!

I'll try again at lunch time with the better batteries and see if they have enough juice to get it moving.... Fingers crossed!

[tom91]

Did you manage to plot the pot nom to see if it is actually trying to request something?

Did you also wire the inverter to be powered on by the Zombie? This is what as the issue was for BigPie.

[andybpowell]

Right, so I'm hitting a bit of a wall ATM, struggling to get the motor spinning.
All wired up correctly as far as I can see, I get inverter whine when I turn it on but no life. I'm just using a potentiometer for the throttle ATM and I've calibrated it in zombie. Tried with a 120V 3A bench top PSU but Bigpie advised I'd be better off with a lower voltage battery. So I've hooked up 4 small 12V 7ah UPS batteries in series for 48V, they're old and cold so may not be up to it but I charged them at 53V last night and tried again this morning.... Nothing.....

Web interface is not showing the HV voltage under udc, is this normal?

Pulled out another 4 UPS batteries that are known to be a bit better (on my cheap battery tester at least)... I'm charging them at the moment.

One thing I did notice is there seems to be more resistance when spinning the output shaft whilst it is trying to run, (at least in one direction) so it seems as though it wants to spin at the very least!

I'll try again at lunch time with the better batteries and see if they have enough juice to get it moving.... Fingers crossed!

do you have a shunt connected ?

[Jacobsmess]

No shunt, which now you mention is, makes me realise that's why there's no voltage/current readings in OI.

[andybpowell]

you can frig it to run without the shunt I did during bench testing but I wouldn't recommend it