Gen 3 Prius Transaxle P410 Makes Clunky Noise.

[Gregski]

I would really appreciate your help trying to figure out what is making the clunky noise in this Gen 3 Prius P410 transaxle as I try and spin the motors for the first time. Damien informed me that the Zombie spins both MG1 and MG2. My power split device is NOT welded, the guts of the transaxle are 100% undisturbed and stock. Parking Pawl is NOT engaged, this has already been confirmed. (When engaged manually by hand the transaxle is locked and won't spin at all). By hand the output shafts spin in opposite directions in either the Parking Pawl locked or open positions.

I am able to spin both output shafts by hand in the same direction and it feels smoot, no noise, however I am doing this at a glacier speed compared to how fast the motors spin them electronically.

Transaxle and Inverter are from the same vehicle a 2010 Prius

VCU = ZombieVerter ver. 2.05A

Power Source = 300 watt, 30 volt, 10 amp bench power supply

I will provide more updates in additional posts this is just a call for help.

[Gregski]

short video of my first spin attempt and the first time I discover the noise

[Gregski]

direct link to Professor Kelley of WeberAuto demonstrating how the Parking Pawl works





Parking Pawl

Parking Gear

Parking Pawl & Gear Engaged

[tom91]

So you are saying that spinning it by hand you feel no weird resistance or noise?

It is very possible your power supply is limiting or you getting a big drop across your tiny test wires into the inverter causing it to drop in and out.

Is there oil inside you gearbox? As it could be MG1 and MG2 fighting through the PSD.

[Gregski]

So you are saying that spinning it by hand you feel no weird resistance or noise?

correct spinning either of the output shafts I hear no noise, however that is after the gear reduction gear etc, I think the noise is clunking before that, but it's difficult to tell

It is very possible your power supply is limiting or you getting a big drop across your tiny test wires into the inverter causing it to drop in and out.

Thank you Tom, tried it with a single battery module from a BMW 530e, these are roughly 60 volt modules with a big punch of current and thicker wires, still same noise

[Gregski]

Is there oil inside you gearbox? As it could be MG1 and MG2 fighting through the PSD.

no fluid inside, adding some now

fingers crossed.jpg (9.65 KiB) Viewed 902 times

Update from the future: 4 quarts in and still makes noise !!!

[tom91]

Also MG1 has no way of transferring any torque, due to not having an engine in it.

Also this means no oil pump working, so locking the PSD is the only way to get the oil pump working and MG1 capable of providing torque.

[Gregski]

Also MG1 has no way of transferring any torque, due to not having an engine in it.

Also this means no oil pump working, so locking the PSD is the only way to get the oil pump working and MG1 capable of providing torque.

per Damien, the ZombieVerter VCU spins both motors, MG1 and MG2 and he claims MG1 spins the oil pump (no need to lock the input shaft, etc.)

I was under the impression per Professor Kelly of Weber Auto that only the input shaft spins the oil pump, but what do I know.

I will try and post the direct link to that section in the video here:

[tom91]

https://toyota-club.net/files/faq/21-12 ... n.htm#P410

Input shaft does spin oil pump and the PSD. So if you weld the PSD it means if you lock the PSD, input shaft spins with the MG1.

Here is the mechanical system diagram. gets quite complex due to the PSD

Good document with pictures and graphs and drawings.
https://digital.library.unt.edu/ark:/67 ... 432/m1/47/ https://digital.library.unt.edu/ark:/67 ... 007833.pdf

[Gregski]

Update 3000:

i can lay under the car and spin both output shafts in same direction and watch the input shaft spinning, all smoothly no clunking noise

at-the-end-of-my-rope.jpg (51.13 KiB) Viewed 848 times

[tom91]

In the weber video the oil pump drive shaft.

[Gregski]

Also MG1 has no way of transferring any torque, due to not having an engine in it.

Also this means no oil pump working, so locking the PSD is the only way to get the oil pump working and MG1 capable of providing torque.

This is just a WeberAuto video showing how the Gen 3 Prius Transaxle uses the final drive ring gear to lubricate itself by splashing the fluid around

[Gregski]

so we went... Def Con 4

with the transaxle split into an open faced sandwich, spinning MG2 by itself with the final drive gear removed we noticed that it pulsates either always in the same spot or maybe two

we thought the Prius throttle pedal may have been bad, so first we recalibrated it, then we tried it in Single Channel mode, then we replaced it with a 2010 Honda Fit pedal and still same results

however when we spin MG1 by itself it spins smoothly no noise no pulse

I hope to upload the video tomorrow

[Bigpie]

Might as well weld up the PSD now it's there

[tom91]

with the transaxle split into an open faced sandwich, spinning MG2 by itself with the final drive gear removed we noticed that it pulsates either always in the same spot or maybe two

Do you hear any abnormal noises from bearings or gears when split apart like this?

There is going to be firmware soon that allows spinning of each Mg separately for even more testing.

I agree with Jamie on welding the PSD. Make sure to plop it in the ring gear on on the MG1 sun gear when welding so the gears are aligned.

[Gregski]

Do you hear any abnormal noises from bearings or gears when split apart like this?

I do not that's what's bizarre, without the final drive gear what I call the differential it spins by hand smoothly like a Swiss watch. The reason I spin it without the diff is two fold, one it is easier, and two the diff wobbles a bit with only bearing support on one end with half the case off.

Also when spinning MG2 with electricity, out of the corner of my eye I saw my 30 volt bench power supply spiking to like 57 volts as if regen was pink ponging it back and forth.

[Gregski]

Might as well weld up the PSD now it's there

baby steps.jpg (5.79 KiB) Viewed 744 times

[Gregski]

tried a complete second Resolver wire harness today, and still same problem

[Gregski]

Happy to report the problem has been fixed, I will leave the technical explanation to the masters. Hope to update a happy motor spinning video soon.

[tom91]

Long and short of it is, The zombie was thinking the inverter was reporting a really high temp so it was cutting throttle.

Turns out the Zombie was looking at the wrong byte in the serial comms with the inverter. This is now being worked on by Damien with my support to resolve this and verify the other information from the Prius Inverter.

"quick fix" is setting temphsmax to 300.

[MattsAwesomeStuff]

Turns out the Zombie was looking at the wrong byte in the serial comms with the inverter. This is now being worked on by Damien with my support to resolve this and verify the other information from the Prius Inverter.

So I'd presume Greg is the first to use the Zombie for a Prius then, as this should be consistent across the platform, no?

This underlines how hard the work Damien is doing is. One little thing here or there, no bueno.

Kudos to Greg for disembowelling his transaxle all over the garage floor chasing down and eliminating possible causes.

[tom91]

So I'd presume Greg is the first to use the Zombie for a Prius then, as this should be consistent across the platform, no?

This is down to the Toyota Prius Gen3 with OEM control yes. Damien only spun it up on the bench.

Damien with my support is getting through bugs alot quicker now, always good to have a second set of eyes who is clued up on the code base.

[MattsAwesomeStuff]

Damien with my support is getting through bugs alot quicker now, always good to have a second set of eyes who is clued up on the code base.

Go team.

IIRC Damien hates programming anyway.

[royhen99]

Did today's testing unequivocally show that the inverter temperature is signed 8 bit?

If so then the GS450 tmphs/temp_inv_water still needs fixing

temp_inv_water=int8_t(mth_data[42]);

[tom91]

Did today's testing unequivocally show that the inverter temperature is signed 8 bit?

If so then the GS450 tmphs/temp_inv_water still needs fixing

temp_inv_water=int8_t(mth_data[42]);

Working through them all one by one, yes the temps are most likely reported in int8_t as IS300H threw problems below zero due to roll over.

Trying to pull all the coding inline, alot is based on knowledge from bench tests where not everything can be fully explored.