EV Conversion using prius gen 2 inverter/transaxle

[voti]

I'm home this weekend. And I'm here for the tests.

OK. I did two test of test for angle. One with the resolver connected and the other with it disconnected. The outputs were almost similar, just the graphs. What I noticed it when I spun the output shafts by hand, the graphs didn't change. But they stayed the same as when not spinning. I've attached pictures belows.
And when I say SHAFT I meant the cv axles and the engine output/input shaft to the transaxle. When doing this tests I was spinning both because none of them was responding so I didn't know what was going on

I have a question. The planetary gear set inside the transaxle is not yet welded to link Mg1 and Mg2 together thus making them rotate together like Damien did. Well not yet. So my question is why is the engine output/input shaft jerking more that the cv axles output. Because in my understanding the engine shaft is only connected to mg1 and since we are working with Mg2 the cv axles outputs should be the ones that are my main focus. But they are not responding. I hope I'm making sense.

At first I thought Maybe the parking pawl was activated hence why the Mg2 is not spinning, but I'm able to spin both the cv axles and also the engine shaft spins by hand.
But when I test spin the motor via the web interface, the motor jerks and the engine shaft seems to be the one more aggressive not the cv axles output.

[RetroZero]

Good to have this info. In order to rotate mg2, I would say you have to rotate both driveshafts in the same direction at the same speed. Input shaft is connected to mg1, correct. Planet gear turns on itself, so since it's not locked, you will not get identical outputs as per other setups.
Path of torque is via the easiest route, least amount of resistance, so planet gear and mg1, thats why the jerky input shaft.
For testing mg2, I suggest locking one driveshaft. The other one is what you will turn by hand (or long spanner) & recheck plot.
You are looking for 0 to 360 waves.

[voti]

Say I lock the engine Input part.
The cv axles are too loose, They spin to easily like they are not connected to anything internally. While the engine input shaft is a bit harder to spin, you can feel that a whole lot of stuff is moving on the inside.

The cv axles on the other hand are way to easy to spin, like a fidget spinner. Is it supposed to be like that?

[RetroZero]

Cv axles are connected to differential, and planet gears. If not turned together in same direction, they turn around themselves.
See previous post. I would say you have to rotate both driveshafts in the same direction at the same speed.

[voti]

I took videos of my test spins and I've uploaded one below. I didn't upload the first one because the graph was speeding so I wasn't able to see clear if I was hitting 0 and 360.

Before you watch the video, there is something I'd like you to please note. For the video linked below. I had my green and brown wires for the resolver mixed together and connected to pin 2 of the control board. When I spun the shafts the graph showed a drop reading like the graph went down. And when I changed the resolver connections and mixed red and brown together and connected them to pin 2. When spinning the motor in the same direction as I was spinning now the graph went up. Also I still have the Delta lights bulbs connected on the inverters three phase output. And for the green and brown combination only two bulbs lit up and those a are the ones connected to phase 2 and 3. When I changed and mixed red and brown the light bulbs also changed and now phase 1 and 2 was lit. For each combination I always get a different set of lit light bulbs, it doesn't happen that all three are lit altogether.

Please watch the video and tel me what you think.

[RetroZero]

Seems to be a problem with the upload of your video, so cannot see what you had recorded.
I'm sure someone will correct me, but with AC, only 2 phases will 'light up as the 3rd is the return to ground in the closed circuit, so normal. I still don't see nice 0-360° graph, so would say the resolver is not yet correct. Plug it into MG1 & see what you get. For clarity, remove the acc pedal for now.

[voti]

I've removed the acc pedal and connected mg1 to the three phase inverter output and also connected the resolver cable to it. What I first noticed with this is that the engine input shaft suddenly become stiff and very hard to turn. Prior when Mg2 was connected I was able to spin it by hand. But with Mg1 connected its hard spinning it even with a pliers, it's very stiff.
What I also noticed I there was a green led blinking on the control board. I also head a different whine sound coming from the transaxle and for a second there I had all three delta connected light bulbs all lit. I'm pointing out all this little details so you guys can be able to know what's going on.

Below I've uploaded a video of how the graph looked when I spun mg1 and I think I might have hit 0 to 360. Please check it out.

[RetroZero]

Ok. Getting somewhere. Do you have mg1 & mg2 phase wires connected to the Inverter? Mg1 difficult to turn means phases are supplied with voltage causing magnets to block the motor/rotor. Remove mg1 phase wires & isolate them. Block one cv output shaft & rotate the other with and without resolver wires connected to mg2. What does the graph look like?

[voti]

I'll be going home next weekend to continue with the tests. This weekend I'm on duty

[johu]

Just found this, maybe it helps viewtopic.php?p=37706#p37706

[voti]

Ok. Getting somewhere. Do you have mg1 & mg2 phase wires connected to the Inverter? Mg1 difficult to turn means phases are supplied with voltage causing magnets to block the motor/rotor. Remove mg1 phase wires & isolate them. Block one cv output shaft & rotate the other with and without resolver wires connected to mg2. What does the graph look like?

I removed mg1 phase wires, connected the resolver cable back to mg2 and tested again. The graph didn't hit 0 to 360. I've attached a picture of the graph below. When I disconnected the resolver l, the graphs showed no changes when the engine input shaft was rotated. What I also noticed is, I'm only able to get this readings when spinning the engine input shat. When I spin the CV drives shafts and lock the engine shaft nothings happens.

Just found this, maybe it helps viewtopic.php?p=37706#p37706

I confirmed my resolver configuration based on this post and found that I pretty much matches the way I have already have it connected. My brown and green are combined and connected to pin 2 while yellow is connected to pin 1 and red to pin 3. Then it's the exciter ones pins 4 and 5. What I noticed is that when I set pinswap to 7 like the post suggested l, I wasn't able to hit 0 to 360 with either MG1 and MG, while with pinswap set to 5 I was able to hit it with MG1.

[RetroZero]

I might be wrong, but:
Mg1 is connected directly to input shaft which in turn is connected to planet gear input.
Mg2 is connected to output of planet gear set , and never directly connected to mg1.
To rotate mg2 if the planet gear is not yet locked up, requires turning both cv axles in the same direction at the same speed.
The input shaft can be left untouched.

[RetroZero]

Your graphs plot differently, pinswap 5 doesn't have the same scaling, it's not 0 to 300 like pinswap 7.
If this is automatic, pinswap 7 graph is what you want.
So with pinswap 7 and turning both cv joints, what do you get as a plot?

[voti]

Oooooooohh man oh man. Now I get it.when you say lock one cv output it's because they rotate in different directions right????. Well for now but once connected to the wheel they turn in the same direction due to the force of the wheel.

Cool, I did that and man oh man that shaft was hard to spin. Infact I couldn't get it to turn on bit. Is it suppose to be this hard. Like it seems locked or some sort.

[RetroZero]

Are you sure the Park lock is not engaged? It should be difficult to turn, but it should turn from the cv joints. ( in same direction at the same speed)

[voti]

I was thinking the same thing too.

I was also thinking of removing it from the transaxle for this test purposes. What do you suggest?

[RetroZero]

I unscrewed the electronic mechanism and manually unlocked it with a pliers by turning the shaft.

[voti]

Turns out my parking pawl was engaged all this time.
I was able to unlocked it and the shaft started spinning. Please see the attached below. I think k this is the saw tooth that's supposed to show and a clean 0 to 360.
Please also check the json files should need be for any parameters corrections

If so what's the next step now?

[RetroZero]

I would say, remove light bulb circuit, connect up phase wires and see if you get to spin with SINE in manual mode. Then it's on to resolver and FOC tuning.

[voti]

Since I'm using version 5.35.R-FOC
Do I need to change back to vesion 5.35.R-SINE for this test spin?

[RetroZero]

Yes

[voti]

The motor is finally spinning
Please watch the video below to see what happens when I increased fslipspnt

[johu]

Good to see! Sine firmware needs some tweaking to turn a sync motor at higher speed. I say don't worry about it and back to FOC now

[voti]

Cool. When the next step when I get to FOC, should I continue with the tutorials on the video of Damien and yourself or there's something different now ?

[voti]

So I'm home this weekend again, and I'm back on FOC.
I followed Damiens instructions and set my SYNCOFS value to 10120. I went to test manualid a set it all the way up 100 amps but still no spin. Then I out in 0.1 in manualid and 3 on manualiq and the motor spun in the forward direction. Thing is mine didn't go reverse when I put in -3.

Damien in the video save his parameters to flash but I didn't because I'm still testing stuff.

Now I've connected my throttle pedal and I'm stuck with that for now as I can't spin the motor with it. Each time I put in 0.1 in manualid and 3 in manualiq the motor just spins without me presing the throttle.

Please help me with that