EV Conversion using prius gen 2 inverter/transaxle

[voti]

Again, verbally describing things that you might have verbally misunderstood is challenging to diagnose.

Take pictures. Maybe you welded the wrong thing?

Like I said. I welded the planetary gears. But after careful examination after disassembling the transaxle again, I noticed that the sun gear was sitting a bit high, hence maybe the reason why things got stuck after assembling. I removed the thrust washer that sits inside the planetary gear and allows the sun gear to rest on it. I then oiled the sun gear and put it back in and hammered it a bit for it to slide and mesh with the welded planetary gears fully. Reason for removing the inside washer was because I feel its not useful anymore because the sun gear and the planetary gears all now rotate together so there won't be any friction inside the whole assembly. After this I put everything back together and my shafts are rotating again.

Below I attached pictures of welded planetary gears. Not mine but they show what I also welded.

Thank you.

[T1Terry]

Good point, might be worth installing the sun gear and maybe even assembling the planetary set into the ring gear, before welding, at least the first tack on each planetary gear to the carrier, that way teeth alignment will at least remain once the planetary set is welded up.

At this point, confusion sets in for me. MG1 will now be spinning at the same speed as MG2, so very little of that low RPM MG2 hp/kw is going to be assisted MG1 hp/kw, but all the MG1 torque will be available, so that's a plus, at around 50mph, MG1 is going to reach it's peak of 10,000rpm so all 57hp/38kw from MG1 will be added to MG2 ..... but MG2 can't spin any faster that 10,000rpm now because MG1 is locked to it, and faster will over speed MG1 .... So what was really gained? A drive system that can hold 50 mph better because of the added input from MG1 ..... but the max speed is still limited to 50mph .....

viewtopic.php?p=31602&hilit=list+of+Pri ... les#p31602

T1 Terry

[RetroZero]

Very interesting info. Will definitely read up on it, thanks.

[johu]

but the max speed is still limited to 50mph ....

It is a Gen2 transaxle, at 60 mph, if I remember correctly it revved lazy at 4000 rpm or so. With the OI logic and sufficient voltage you can make it go 80 mph or more.
Without the gear welded MG1 will overspeed.

[T1Terry]

Good point, might be worth installing the sun gear and maybe even assembling the planetary set into the ring gear, before welding, at least the first tack on each planetary gear to the carrier, that way teeth alignment will at least remain once the planetary set is welded up.

At this point, confusion sets in for me. MG1 will now be spinning at the same speed as MG2, so very little of that low RPM MG2 hp/kw is going to be assisted MG1 hp/kw, but all the MG1 torque will be available, so that's a plus, at around 50mph, MG1 is going to reach it's peak of 10,000rpm so all 57hp/38kw from MG1 will be added to MG2 ..... but MG2 can't spin any faster that 10,000rpm ....... now because MG1 is locked to it, and faster will over speed MG1 .... So what was really gained? A drive system that can hold 50 mph better because of the added input from MG1 ..... but the max speed is still limited to 50mph .....

viewtopic.php?p=31602&hilit=list+of+Pri ... les#p31602

T1 Terry

From about the point I changed the text to red ..... my brain left the building

Somewhere I got my wires crossed between locking the shaft to the case using a one way sprag clutch, and welding the gearset causing the same thing as locking the shaft to the case ..... which isn't what happens at all

I went and measured the dia of the tyre on my Gen 2 and it's 0.6mtr x 3.14 = 1.884 mtrs per rpm.
80km/h (roughly 50mph) = 1,333.3 mtrs per min. = 707.7rpm tyre speed and therefore crown wheel speed.
707.7 x 4.1:1 ratio of MG2 to the crown wheel = 2,900 MG2 rpm @ 80km/h ..... Rough maths comes out to 275km/h before MG2 and there for MG1 in a welded/locked rotor set, would reach the max 10,000 rpm ......


So, MG1 will theoretically never add the potential 50hp/38kw .... until some where around 260km/h

Where I was going with that, before my brain left building, was:

The input shaft with a sprag clutch (one direction locked, the other direction free wheel) anchored to the case, so when MG1 can no longer drive the planetary carrier against normal engine rotation causing the sprag clutch to lock, the input shaft and therefore the planetary gear carrier would then freewheel and spin with the ring gear at anywhere up to the same speed as MG2.
When MG1 can no longer add any torque or hp/kw to MG2, there is no value in powering it to spin, all the drive would then be for MG2.

With this set up, the off the mark performance would be greatly improved, but by 80km/h (50mph) only MG2 would be driving the wheels .......

Is the 45nm/33lbf of torque up to 6,000rpm worth the 75 amps it would take to achieve it? So, where is the value in welding the gearset and powering MG1?

There is every chance my brain has never returned to the building, so probably best if some checks my ramblings ....

T1 Terry

[RetroZero]

Keep rambling, it's interesting brain fuzz material.

[voti]

Is it possible to load throttle parameters in graph form?
I don't have the figures in front of me, but vaguely remember setting throttle. From what I remember:
There are 2 potentiometers.
Pot is for 0 to 4095 (digits)
Pot2 is the reverse 4095 to 0 (digits)

Setting Pot: note value when NOT pressed (example 500) - set your min value LARGER than this (example 550).
Now press throttle completely & note value (example 4500 - set pot max to LESS than this value (example 4200).
Pot 2 is done the same, except its the reverse of values because it reads in the opposite direction 4095 to 0.
Plot your graphs.
What you are trying to create are values where when no pedal is pressed, Pot is at 0, with a slight margin of play. The same is done for potmax.

I'm home this weekend and I'm back to throttle calibration.

I was following the above given instructions . So before pressing the throttle I got pot 963 and pot2 2013. When the throttle was fully press both pot and pot2 gave 4095.
So I did the math exactly as instructed above. Well maybe those numbers from Retro were just examples but I used them as per. The only difference was when it comes to potmin. I put in 3500 since that's the max parameter.

Then I went to throttle parameters and please see attached screeshots of how I loaded the new numbers. For pot2. I had to interchange the min and max values since retro mentioned that for pot2 its vice versa, so I wanted to make sure that i got it right in order to see which configuration works but none of those worked.

The results I got was that when pressing the throttle when it's halfway, the motor jitters, it does rotate but jittery rotation. When I continue pressing the throttle to its full capacity the motor then stops jittering and comes to a stand still. Please not that this results were same even before I changed the parameters, like when I first connected the pedal I did press it to see what will happen and this jittery rotation was the same.

[MattsAwesomeStuff]

I'm only half paying attention, but have you gone through the procedure Damien showed in his video for how to set up SynchOffs properly?

[voti]

I'm only half paying attention, but have you gone through the procedure Damien showed in his video for how to set up SynchOffs properly?

Yes I did Matt, I have my syncoffs set to 10120 like Damien showed on the video.

[RetroZero]

Sorry, I'm away for a few weeks, so cannot really help.
Just to be sure, you have resolver 0 to 360 graph.
You have set up syncoffs as per the video. Did you go through different values to get to 10120 or did you put in Damien's value?
I noticed that gen 2 & gen 3 are not exactly identicle. Throw in different HV setups and that changes the value as well...
I'd try a couple of different syncoffs....

[MattsAwesomeStuff]

Yes I did Matt, I have my syncoffs set to 10120 like Damien showed on the video.

As Retro pointed out... just making sure you aren't blindly choosing 10120. Damien showed the process everyone should follow because all units are different. It was not him demonstrating how he found the "correct" value for all Gen2 Prius's and that everyone should use that number.

[voti]

I think I'll have to redo my resolver connection test and confirm that I do get 0 to 360 plot on the graph. Then from there I'll have to redo the syncofs settings.

At first I had tested my resolvers 0 to 360 plot, and me getting to 10120 I had also gone through other values testing them also, Up until I got 10120. I did put in different numbers on manualid up to a hundred to see if 10120 value was indeed correct. I followed Damiens test as is and it seemed to be working just fine for me also. But Retro mentioned something important that the HV values are not the same for gen 2 and gen 3, plus Damien had higher voltage when he did his test. But my concern is that like I said his values seems to be working just fine for me. Hence I'll have to redo the test to confirm.

But to be sure. This weekend I'll redo everything and confirm all my connections since I had some of them disconnected when i was welding the planetary gear.

[voti]

I'm home this weekend.

So liek I said that I'll go back to resolver testing before continuing to throttle calibration and setup. What I noticed is that after taking the transaxle apart to weld the planetary gear set, I lost my resolver connection. When I plug in the cable back on the transaxle, either I get varying angles or no angle at all. Just now I set automatic refresh and unplugged my resolver as it was stuck to 0°. When I lugged in it again it started showing way to many angles; 48°, 75° , 180°, 336° you name it. When I restart the inverter and reconnect the resolver cable, it's back at 0°.
At some point I was able to have the angle at around 98° and when I proceeded to the resolver plot test and reotayed the transaxle by hand with the HV disconnected, the graph didn't change but it showed this waveform that fluctuated continously. On another test also, when I started the flot it just jumped from 0 to 360 without me having to rotated the shaft. This 0 to 360 pattern also was continous and like I said it just jumped to these plot without me having to spin anything. I watched Damien video to troubleshoot what was going on, all I found as that Damien did mention that just a behavior means hat I should stop because something is wrong, only problem is he didn't say what that problem is or how to solve it.

Anyone with knowledge on this please come to my rescue

[voti]

Anyone, please

[MattsAwesomeStuff]

0 to 360 fluttering makes sense to me, as, they're the same point. 0 and 360 degrees are both the start of the circle.

As to the rest, I'm regretfully useless. I don't know what's supposed to happen, to know why you're not getting it.

[voti]

I'm home this weekend and appreciate any of you guys coming through for me right now.

[johu]

So you had a working resolver before taking the transmission apart. You have a non-working resolver after taking it apart and putting it back together. Correct?

You need to see a clean 0-359°C angle plot (1000 burst, 1000 points) in FOC firmware when spinning the motor by hand. Otherwise all other tuning is pointless