openinverter forum

I would like to profusely thank Damien and Johannes for their recent FOC Tuning video! While I had already read and re-read the instructions often enough to grasp the general concepts and achieved a syncofs setting that worked well, I applied Damien's method of dividing the "gap" in half and used that, potentially outside of the intent, to try and figure out where in the "trough" of the graph I was.

The initial syncofs number I found that worked well was 11100.

In an attempt to determine where on the lowest part of the trough was (there are a range of settings that result in zero rotation), I worked OUT from 11100 up and down in 10 digit increments until a MID=5 would result in *some* actual rotation. I then took those numbers and averaged them, hoping to get to the lowest part of the trough.

Testing revealed that the lower rotation threshold was 9850 and the upper 12100, this lead to a "final" trough value of 10975 ((9850+12100)/2), a difference of 125 digits from the value I had from earlier

I do not even know if that is enough of a difference to matter, given the digits range from 0 - 65535, I doubt it was time well spent, but I don't care

I was wondering what ampacity I would need for the HV Battery and motor wiring, so decided to do some testing. What I did not expect, was that the DC current INTO the inverter would be about 1/3 of the AC OUTPUT to the motor phases

I believe Johu himself explained somewhere in Prius Gen2 Inverter thread that output current is more then dc input current. Why is there a massive capacitor hooked up to Inverter in a first place

About your Squareback, does it have standard 'Beetle' suspension, as in NOT independant trailing arms?
Is your plan to remove engine and gearbox and replace it with the MGR?
Reason being, someone is 'offering' me a Beetle, but I know the gearbox tunnel is quite narrow, so Original gearbox and Leaf set-up normally works. Only "easy" Toyota option would be RX transmission? However, If the MGR gets in there, it's an awesome 'drop in' set-up... Your thoughts?

RetroZero wrote: ↑Thu Oct 01, 2020 2:01 pm About your Squareback, does it have standard 'Beetle' suspension, as in NOT independant trailing arms?
Is your plan to remove engine and gearbox and replace it with the MGR?
Reason being, someone is 'offering' me a Beetle, but I know the gearbox tunnel is quite narrow, so Original gearbox and Leaf set-up normally works. Only "easy" Toyota option would be RX transmission? However, If the MGR gets in there, it's an awesome 'drop in' set-up... Your thoughts?

I'm not fully familiar with the two VW aircooled suspension variants. I know one is called "swing axle" and the early ones uses the transmission as a location/pivot point

With the Squareback, I understand correctly, the 1968 and later are the "IRS" types, so the wheels are located by only the suspension parts. Therefore, my '71 (if the damned thing ever gets released by the DMV) should be a bolt in affair.

The MGR is reportedly nearly the same flange-to-flange width as the 4spd aircooled trans., but I do not have a dimension for the VW parts. I do know there will be inboard flange adapting to do, but have access to a BIG lathe and an excellent welder (person + equipment), so that *should be* academic.

konstantin8818 wrote: ↑Thu Oct 01, 2020 12:48 pm I believe Johu himself explained somewhere in Prius Gen2 Inverter thread that output current is more then dc input current. Why is there a massive capacitor hooked up to Inverter in a first place

I seem to recall something from my courses that the same number of electrons measured RMS will always have a greater magnitude than with a DC
measurement???? But that is REALLY foggy.... and I certainly did not expect a 1:3 ratio!

The results also call into question where the 50kW rating is coming from. I know that AC & DC ampacity are unequal and also that the frequency is an influencer, but am not knowledgeable enough to wade thru the finer details of same.

50kW is: 100A @ 500V, ~225A @ 225A, or even 500A @ 100V..... The phase wires for the MGR are 8ga (~4mm) multi-strand and have high dielectric cross-link insulation. Most of the 'net suggests that at <1m distances, this can carry between 70 & 200A @12VDC.

For HV loads, the presumption for all ratings is 100% duty cycle, in a traceway or "romex" and I KNOW this provides a wide safety margin.

I think the only way I will "KNOW" what the ampacity of the Prius HV battery cabling is, is to run it to failure

Sounds like a good video, so quote an YooToobe famous member, what could go wrong??? LOL

ZooKeeper wrote: ↑Thu Oct 01, 2020 10:26 pm I seem to recall something from my courses that the same number of electrons measured RMS will always have a greater magnitude than with a DC
measurement???? But that is REALLY foggy.... and I certainly did not expect a 1:3 ratio!

The results also call into question where the 50kW rating is coming from. I know that AC & DC ampacity are unequal and also that the frequency is an influencer, but am not knowledgeable enough to wade thru the finer details of same.

50kW is: 100A @ 500V, ~225A @ 225A, or even 500A @ 100V..... The phase wires for the MGR are 8ga (~4mm) multi-strand and have high dielectric cross-link insulation.

The inverter uses PWM to multiply current when lowering voltage.
If 320VDC is going in, the maximum AC output voltage is 226VAC (since the peak is at 320V).
So let's say the motor is spinning at 1/3 of the fweak speed; the AC voltage will then be 1/3 of the maximum AC voltage (plus boost) so about 80 to 90 volts. This means that AC current will be 3 times the battery current - since power equals voltage times current, if voltage decreases by a factor of 3 then the current must increase by a factor of 3.

The 50kW is a peak rating at the maximum power point. Much like an internal combustion motor, the power peak shows up at a few thousand RPM.
This is because the inverter is current limited -- it may output 200 amps all the way from 0 volts to 226VAC, and so when you hit 226VAC (at the fweak point) you will be outputting maximum power.
200 amps at 226VAC is 45.2kW. With a 360v nominal battery (arguably the most common) you can get 254VAC - that means 196A for 50kW of power.

Note that most vehicles will not require 50kw at cruising speeds - IIRC my 3100lb (though well streamlined) car pulls about 18kw at 60mph (will need to check this again).
So you probably won't be pulling more than 100 phase amps at highway speeds. 8 gauge does seem a little small; remember that this is not exactly designed for continuous use, so larger phase wires might be a good idea for hotrodding efforts.

-Isaac

Based on the EV calculator (by Uve Rick) the SB will require 4hp @ 30mph and 15hp @ 60mph ( 3kW @ 50kph & 11kw @ 100kph), flat ground.

I also know that fine stranded HV cabling (welding cable or similar) has almost 2x the current capacity of "typical" more coarse stranded cables.

My plan is to take two sections of Prius Gen2 HV battery cable and run them to failure this weekend THEN I will KNOW their limits.

Well so much for THAT plan

Thanks to some seriously irresponsible behavior by certain elected officials, we have a SERIOUS covid outbreak locally and I am not going anywhere other than the grocery for a while, so testing on the cables will have to wait.

I did some work on the LEAF relay board for my DIY VOLTswagen Squareback EV project. I attached the pdf I made and linked to the generally crappy video that goes with it.

Enjoy.

I think I have become a "hoarder" of electronics test gear

More testing! Since I cannot, even after reading and re-reading the charging information more than 10x over the past months, get the VCU controlled charging to work at all, I punted and have what should be called a "Third World Charger" working perfectly

Wired up my Leaf relay board, powered it from a switchable 12v source and BINGO! My 28 salvaged Prius Gen2 modules are nearing 220v with a planned final of 235vdc. Since my current plan is a 240v (peak) Leaf based pack, this is looking good!!!!

Video for forum members to follow soon.

<<EDIT>> Added video.

The Prius Gen2 pack + Leaf board works! Yesterday I used the "3rd world charger" set up to get the pack up to scratch. Today I started off by powering the VCU, then applying a PC along with the Prius Inverter logic power and then closed the DC relay. As hoped and expected, I was able to start the VCU and power the MGR up just as prior testing suggested.

With 1Adc @ 208v fed to the inverter, I measured 3.3A RMS on a phase of the MGR, again matching prior test results

The inverter neither boosted nor bucked the DC feed voltage, which is as I expected. This means that 50kW @ 240V would need: 50/240 = 208A rms >> 208/3.3 = 62 to 208/2.0 = 104 Adc from the pack for peak output. This current level is EASILY tolerated by the factory Leaf relay board and Prius/Toyota HV cables I have.

YEY

Literally, all I NEED to get this thing on the road, is the actual CAR

<EDIT> You can't get something for nothing, conservation of energy etc... If the motor produces 50kW, then you MUST pull 50kW from the battery.

One more item just validated, if "F" & "R" are both tied to 12v at the same time, you get ZERO output from the inverter. This means that even if you decide not to use VCU charge controls, movement can be easily prevented while charging.

So I have confirmed that somehow I damaged the DC relay circuit on my board. The good news is I have a replacement for the shorted 9T18GH mosfet, but it is not like-for-like; what I have is an F1405Z. The ratings are similar, but not identical. I do not anticipate any issue with this substitution, especially since it drives a relay with relatively low (<2.5A) current draw.

I'll be jumping back into project soon and hope to pick up the pieces and fill in gaps that you so nicley filled.
Any idea on DC circuit fault?

Update, I have the Q2 FET fixed and the DC switch control now works as designed as well as the PC that was fixed in firmware last year.

Over the next few days I plan to do some testing on the charge functionality, but I am still very much lacking in understanding of this feature.

Hi Zookeeper, I though you might like to see these pics of an MGR sitting with its axle centres within a few mm of the original VW transaxle. To get it perfect i would need to lose the bottom cooling fin off the resolver. Fabricating mountings looks pretty straightforward and some firm engine mount rubbers should mean there is little movement to cause problems.

Chris

That looks very nice! Since the '71 does not have the "horns", I will have a little more wiggle room, but a more complex rear mount.

Thanks!

Well that was SMELLY

My plan for today was to do some crude charge testing after a quick motor spin, everything is wired as before except the PC fet has been fixed. I had previously manually precharged using my Leaf relay board dozens of times. This time, when I powered on I heard bacon

After disconnecting the battery, I just followed the odor and after making everything safe, opened the relay board enclosure to find the PC resistor split open. That really makes no sense, but will try and figure out what happened.... after I ventilate adequately.

Yup, the PC resistor is burned out and I *think* I know why....

In my haste to get the wiring connected before supper, I forgot to disconnect the do not try this at home line voltage power supply. The problem came because the rectifier is on the load side of the switching, creating a dead-short for the PC resistor, a load it was not made to handle.

Let's try and not do that again.

OK, after reviewing all my notes, photos and video I *think* the error has been discovered. More proof I am neither an EE, Electrician or "Engineer" and do not play one on the 'net...

Apparently when I connected up my wiring, I inadvertently set the VCU to charge mode by tying Pin 31 high WITHOUT CSDN also being high, which if I understand Johu's video correctly, will short HV+ to HV- and consequently fry the PC resistor.

I will try and tie CSDN to the HV+ relay ground, so it will be high when the HV+ is open (ie: during PC) with the key on.

LOL! I *might* be able to retire early....... I am all for opensource and all, but a copyright is a copyright and I was the first to coin "VOLTSwagen".

https://arstechnica.com/cars/2021/03/vw ... ebranding/

Wow John, you had better get in quick:

What's more, while Volkswagen is a live US trademark, USPTO has nothing in its Trademark Electronic Search System database for "Voltswagen." (Which suggests to me that someone could make a quick buck by registering the trademark first, were they so inclined.) And this coming Thursday is, of course, April 1, when people, particularly on the Internet, love to play pranks.

And I have "standing" to do so!

I also found out that, in part due to coffee & doughnuts, my VIN has been verified by a LEO in CA Once the seller gets it signed off by the DMV, I can collect it!

Even more good news, while I lost out on the Leaf battery I was after, another has turned up today. Time to shake the Money Tree!