openinverter forum

I am not sure if this tpic already exists but I need some support for my project.

So far I managed to spin the motor with the 5.00.R-Sine SW in manual mode.

When I tryed to configure the 5.00.R-FOC, I observd that the parameter "angle" shows a value of 8.21 degrees and when I tryed to plot the vallue and spin the rotor manually, the vallue remaind the same.

I've disconected the resolver and the vallue remaind unchange.

I have also chaned the Logic board from the inverter and the results are the same for both Logic boards (V1c).

The inverter displayed the following errors :

[13]: WARN - HIRESOFS
[20335]: WARN - LORESAMP

Attached is the parameters file.

Thanks in advance for the support.

HIRESOFFSET - resolver mid value is too far away from 1.65V. Please check voltage on the S1S4 pin, should be around 1.6V
LORESAMP - amplitude returned by resolver is too low, less than 0.8V peak-to-peak . Pull-ups removed? Checked excitation? On a multimeter on AC volts mode it should be about 3V RMS and about 0.5-1V RMS on one of the resolver return coils.

johu wrote: ↑Sun May 02, 2021 3:51 pm HIRESOFFSET - resolver mid value is too far away from 1.65V. Please check voltage on the S1S4 pin, should be around 1.6V
LORESAMP - amplitude returned by resolver is too low, less than 0.8V peak-to-peak . Pull-ups removed? Checked excitation? On a multimeter on AC volts mode it should be about 3V RMS and about 0.5-1V RMS on one of the resolver return coils.

Thank you for your fast reply!

I have measured the excitation voltage with the resolver connected and the value is 2.5V RMS.The measured values for the SIN and COS returns are almost 0V (25mV max) .
I am not sure what do you mean by measuring the S1S4 pin? Also what pull-ups should I remove?

I have finally managed to set up the inverter and the Leaf gen 1 motor connections for the resolver.

I've connected the inverter to a 18S Nissan Leaf battery pack with a Chinese BMS which can provide 300A.

You can see in the video bellow how smooth the motor is running.



Now my question is, why is the motor drowning so much power when no load is attached? On the phone, can observe the maximul DC current (45A) and the max power (3170W) when the motor is at full throttle.

Are this values normal, considering that the DC buss voltage have such a low value (aprox 70V)?

After testing the setup for a while without a cooling circuit, I decided to connect the Nissan Leaf OEM radiator and two of the recirculation pumps in order to cool off a bit the inverter and the motor.

Then I decided to test them for about an hour to see how the temperature gradient will gonna change.

I've started the test with the battery pack fully charged at about 74V and a capacity of about 53Ah. I wanted to measure the real capacity of the battery and in the same time I wanted to do some measurements and observe the temperature of different components of the setup.

In the video bellow we can observe the temperatures of the different components while the motor was running at 172A RMS, phase current.



After only half an hour I wanted to measure again the temperature of the main capacitor and I observed that this had a crack which was not there before starting the test.
Now the temp of the cap was well above 90 degrees Celsius.

After this I immediately stopped the Inverter and waited for the capacitor to cool down.

Meanwhile I managed to make some measurements of the DC current, motor phase current and motor line voltage with my oscilloscope.


The waveform of the DC current absorbed by the inverter

The waveform of the motor phase current which resembles a Sine wave (somehow)
The waveform of the motor line voltage Now I have some personal observations regarding the Prius Gen.3 inverter:

- The buss bars for the MG1 and MG2 are way to thin to support the maximum rated current (especially for the MG2).
In my video at 0:58 it is very clear that the buss bar temp is almost 100 degrees C at a 172A RMS phase current. I really don't want to thing about pulling 600A RMS from the MG2...
- I am not sure why the main Cap is behaving like that when I run the inverter continuously at a not so great 72V and at a very negligible 45A on the DC buss bar directly connected to the main DC CAP.

Maybe someone with more knowledge can provide me some answers.

Also if anyone have a spare Gen.3 inverter Cap I will be more than happy to buy it for a reasonable price.

Drw91 wrote: ↑Wed May 12, 2021 4:32 pm Now I have some personal observations regarding the Prius Gen.3 inverter:

- The buss bars for the MG1 and MG2 are way to thin to support the maximum rated current (especially for the MG2).
In my video at 0:58 it is very clear that the buss bar temp is almost 100 degrees C at a 172A RMS phase current. I really don't want to thing about pulling 600A RMS from the MG2...
- I am not sure why the main Cap is behaving like that when I run the inverter continuously at a not so great 72V and at a very negligible 45A on the DC buss bar directly connected to the main DC CAP.

I think most EV components are not rated for continuous load. For example the main DC rail in a solar inverter is like 100x20 mm and it will happily conduct 1000A all day. The DC rail of a Tesla inverter is like 20x3 mm and it will only conduct 1000A for limited time.

Now running the motor at 3000 rpm and 70V will need quite some field weakening current which puts additional load on the inverter and a unfavourable power factor, i.e. energy swings back and forth between motor and bus cap.

Drw91 wrote: ↑Wed May 12, 2021 4:32 pm- I am not sure why the main Cap is behaving like that when I run the inverter continuously at a not so great 72V and at a very negligible 45A on the DC buss bar directly connected to the main DC CAP.

On the 3rd video Damien linked above, he concludes that he had his deadtime set to zero, causing "shoot-through" on that particular inverter and a resulting significant voltage spike there, which ruined the capacitor. And he sacrificed a smaller capacitor to demonstrate that, and the fix for it.

This is only true for some types and years of inverters, not true for all of them, but I was a bit lost which ones that was referring to.

This is one of the things I'm weakest on, but, ideally there would be a set of working parameters to copy, so not everyone has to reinvent the wheel. Were you using a known set of working parameters?

- The buss bars for the MG1 and MG2 are way to thin to support the maximum rated current (especially for the MG2).
In my video at 0:58 it is very clear that the buss bar temp is almost 100 degrees C at a 172A RMS phase current. I really don't want to thing about pulling 600A RMS from the MG2...

Might want to beef them up then I suppose. Damien pulled 200 amps off his for ~10 minutes before he murdered the coolant pump, but I'd say the wires were dangerously hot during that anyways.

You could pull 600 amps, but, honestly what driving conditions would allow that, and what battery pack could sustain that?

Drw91 wrote: ↑Wed May 12, 2021 4:32 pm After only half an hour I wanted to measure again the temperature of the main capacitor and I observed that this had a crack which was not there before starting the test.

When looking at your parameters posted in first post, "deadtime" is set to 0. This is what Damien had as well when he blew his up.
To my understanding, setting "deadtime" to 0 only "works" in the prius gen2 inverters.
I'm using a "deadtime" of the 63 in my prius gen3 inverter and it is working fine so far. I have not pushed the inverter hard and have not scoped igbt voltages either so only guessed value. Maybe someone else have a more tested value to share?

Jack Bauer wrote: ↑Wed May 12, 2021 4:44 pm

Thank you Damien for remembering me how I was supposed to connect the Battery pack to my inverter, but unfortunately I completely forgot about your mini series of "Can't kill a Toyota Inverter".

I saw all the videos months ago but now I was so exited to spin the motor that I've omitted some important steps in order to make the inverter work properly.

I am not sure if my Cap can be used as it is. The value is the same as it was before. Now I also need to check the little cap that in your case popped.

johu wrote: ↑Thu May 13, 2021 5:22 am
I think most EV components are not rated for continuous load. For example the main DC rail in a solar inverter is like 100x20 mm and it will happily conduct 1000A all day. The DC rail of a Tesla inverter is like 20x3 mm and it will only conduct 1000A for limited time.

You are right about this and for sure is like that. Almost all of the manufacturers are concerned about cost reductions and budgets.

My problem is not that the buss bars are so thin and they can't withstand the max current that the IGBTs are rated for, my idea is to find a way of making them beefier in order to be able to support for longer periods of time greater currents without overheating.

johu wrote: ↑Thu May 13, 2021 5:22 am Now running the motor at 3000 rpm and 70V will need quite some field weakening current which puts additional load on the inverter and a unfavourable power factor, i.e. energy swings back and forth between motor and bus cap.

Regarding the capacitor problem, now I know what I did wrong, thanks to Damien.

Drw91 wrote: ↑Thu May 13, 2021 4:07 pmI saw all the videos months ago but now I was so exited to spin the motor that I've omitted some important steps in order to make the inverter work properly.

I'm not technically savy enough to contribute to most projects here, but I try to do some smaller lifting on things being archived and documented. In a big picture kind of way, supposing someone new shows up, who hasn't been following along like a blog, could they easily find the correct information to become successful? And if not, what could change to make sure that happens?

It's great to have the information out there, if you knew it existed, and knew what to search for. But if it's not archived and presented well, we can do better.

So in your case, you had seen the content but forgot about it, and skipped some steps perhaps.

In that light, where could or should we have put the information so that this wouldn't have happened to you, and won't happen to the next person?

Should we more or less write a manual or a procedure to follow for "I've got my inverter and my controller replacement, what now?" (wiki perhaps). Would that be helpful? Would that be premature because the design is still evolving? If that was linked on Damien's store page where you bought the item? Where best?

bexander wrote: ↑Thu May 13, 2021 7:16 amWhen looking at your parameters posted in first post, "deadtime" is set to 0. This is what Damien had as well when he blew his up. To my understanding, setting "deadtime" to 0 only "works" in the prius gen2 inverters.

IIRC, on the Gen2 inverter, the deadtime is "set" to zero, but in use, is not zero. It's just not controlled by the controller brain. The controller part only has access to half the cycle. This is abstracted somewhere down the line, and the original inverter decides what the deadtime should be. Nothing you set it to has any affect, I think.

On the Gen3, you have access to both halves, so it will actually follow what you set it to. If you set it to zero, it will actually be zero. You don't want it to be zero, you have to actually set the parameter. This itself doesn't add any extra value, but, having control over it does enable us to get more function out of the Gen3 than the Gen2 in other ways where you need control over both sides rather than the inverter deciding that for you.

That's my caveman understanding of it.

bexander wrote: ↑Thu May 13, 2021 7:16 am When looking at your parameters posted in first post, "deadtime" is set to 0. This is what Damien had as well when he blew his up.
To my understanding, setting "deadtime" to 0 only "works" in the prius gen2 inverters.
I'm using a "deadtime" of the 63 in my prius gen3 inverter and it is working fine so far. I have not pushed the inverter hard and have not scoped igbt voltages either so only guessed value. Maybe someone else have a more tested value to share?

The fun fact is that the first time when I was setting the parameters I was considering using some "deadtime" for the IGBTs. After setting 130 for the "deadtime" I observed that the inverter was consuming about 150W from the battery without the torque command. After trying other values without luck, I've decided to set the "deadtime" to 0 and for that, the inverter was not consuming the 150W anymore.

Drw91 wrote: ↑Thu May 13, 2021 5:15 pm

bexander wrote: ↑Thu May 13, 2021 7:16 am When looking at your parameters posted in first post, "deadtime" is set to 0. This is what Damien had as well when he blew his up.
To my understanding, setting "deadtime" to 0 only "works" in the prius gen2 inverters.
I'm using a "deadtime" of the 63 in my prius gen3 inverter and it is working fine so far. I have not pushed the inverter hard and have not scoped igbt voltages either so only guessed value. Maybe someone else have a more tested value to share?

The fun fact is that the first time when I was setting the parameters I was considering using some "deadtime" for the IGBTs. After setting 130 for the "deadtime" I observed that the inverter was consuming about 150W from the battery without the torque command. After trying other values without luck, I've decided to set the "deadtime" to 0 and for that, the inverter was not consuming the 150W anymore.

This is a bit strange to me?
When not producing torque i.e. no motor currents flowing, there should be no or the opposite effect on power consumption by changing deadtime.
Do you have the DC/DC connected in the inverter? Do you have anything else connected to the HV DC? How did you measure that 150W consumption?

bexander wrote: ↑Fri May 14, 2021 6:51 am

Drw91 wrote: ↑Thu May 13, 2021 5:15 pm

bexander wrote: ↑Thu May 13, 2021 7:16 am When looking at your parameters posted in first post, "deadtime" is set to 0. This is what Damien had as well when he blew his up.
To my understanding, setting "deadtime" to 0 only "works" in the prius gen2 inverters.
I'm using a "deadtime" of the 63 in my prius gen3 inverter and it is working fine so far. I have not pushed the inverter hard and have not scoped igbt voltages either so only guessed value. Maybe someone else have a more tested value to share?

The fun fact is that the first time when I was setting the parameters I was considering using some "deadtime" for the IGBTs. After setting 130 for the "deadtime" I observed that the inverter was consuming about 150W from the battery without the torque command. After trying other values without luck, I've decided to set the "deadtime" to 0 and for that, the inverter was not consuming the 150W anymore.

This is a bit strange to me?
When not producing torque i.e. no motor currents flowing, there should be no or the opposite effect on power consumption by changing deadtime.
Do you have the DC/DC connected in the inverter? Do you have anything else connected to the HV DC? How did you measure that 150W consumption?

It is a bit strange. Anyways, now I've changed the deadtime to 63 and it seems to work just fine. The 150W value was shown by the BMS's mobile app. The only thing that was connected to the battery was the inverter.

BMS current resolution isn't the worlds best. Say you have a +-400A range, then 0.5% jitter is 2A. 2A@300V = 600W

MattsAwesomeStuff wrote: ↑Thu May 13, 2021 5:12 pm

Drw91 wrote: ↑Thu May 13, 2021 4:07 pmI saw all the videos months ago but now I was so exited to spin the motor that I've omitted some important steps in order to make the inverter work properly.

I'm not technically savy enough to contribute to most projects here, but I try to do some smaller lifting on things being archived and documented. In a big picture kind of way, supposing someone new shows up, who hasn't been following along like a blog, could they easily find the correct information to become successful? And if not, what could change to make sure that happens?

It's great to have the information out there, if you knew it existed, and knew what to search for. But if it's not archived and presented well, we can do better.

So in your case, you had seen the content but forgot about it, and skipped some steps perhaps.

In that light, where could or should we have put the information so that this wouldn't have happened to you, and won't happen to the next person?

Should we more or less write a manual or a procedure to follow for "I've got my inverter and my controller replacement, what now?" (wiki perhaps). Would that be helpful? Would that be premature because the design is still evolving? If that was linked on Damien's store page where you bought the item? Where best?

bexander wrote: ↑Thu May 13, 2021 7:16 amWhen looking at your parameters posted in first post, "deadtime" is set to 0. This is what Damien had as well when he blew his up. To my understanding, setting "deadtime" to 0 only "works" in the prius gen2 inverters.

IIRC, on the Gen2 inverter, the deadtime is "set" to zero, but in use, is not zero. It's just not controlled by the controller brain. The controller part only has access to half the cycle. This is abstracted somewhere down the line, and the original inverter decides what the deadtime should be. Nothing you set it to has any affect, I think.

On the Gen3, you have access to both halves, so it will actually follow what you set it to. If you set it to zero, it will actually be zero. You don't want it to be zero, you have to actually set the parameter. This itself doesn't add any extra value, but, having control over it does enable us to get more function out of the Gen3 than the Gen2 in other ways where you need control over both sides rather than the inverter deciding that for you.

That's my caveman understanding of it.

Matt,

This it one of the most well designed and well documented forums I've ever seen in the conversions field.
Thanks to some great people which are working to develop and test all sorts of ECUs and SWs, others like me are able to build their own electric car with just a bit of effort.

Regarding my previous mistakes, I admit that they were made because I was not gathering all the info necessary for me to start this project. To be honest I've had access to the "Toyota Prius Gen3 Board" Wiki page, to all of Damien's and Johannes's videos regarding the Prius Gen.3 inverter and V3c Logic board and all the info that this forum can provide.

The problem was that I rewatch some of the Damien's older videos where he connected the inverter in the wrong way, but obviously this was at a time when he had no idea that is not the correct way of doing this.
After seeing this I totally ignored the wiki page where it is mentioned the correct way of connecting the battery to the inverter.

Regarding the deadtime value, I presumed that 0 will work for all the Toyota inverters. Currently I am running the inverter with a deadtime value of 63 but I am not sure if this is the right value.

Drw91 wrote: ↑Tue May 18, 2021 3:43 pmThis it one of the most well designed and well documented forums I've ever seen in the conversions field.
Thanks to some great people which are working to develop and test all sorts of ECUs and SWs, others like me are able to build their own electric car with just a bit of effort.

Certainly this community is a treasure trove of information. We are, if nothing else, often drowning in detail. This is how progress is made, you can't snap your fingers at a distance, you have to get into the meat of things.

Myself, I'm not able to add to it much, as most of it is beyond me. So I look to contribute in ways of making it more accessible and comprehensible.

Regarding my previous mistakes, I admit that they were made because I was not gathering all the info necessary for me to start this project. To be honest I've had access to the "Toyota Prius Gen3 Board" Wiki page, to all of Damien's and Johannes's videos regarding the Prius Gen.3 inverter and V3c Logic board and all the info that this forum can provide.

The problem was that I rewatch some of the Damien's older videos where he connected the inverter in the wrong way, but obviously this was at a time when he had no idea that is not the correct way of doing this.
After seeing this I totally ignored the wiki page where it is mentioned the correct way of connecting the battery to the inverter.

On one hand, you take accountability for your error (it is nice that you are not blaming anyone else), and are saying that the information is out there.

On the other hand, if you made this mistake, so might others. Functionally, there's no difference between the right information being out there or it not being there, if it's not the in a place it can be used. As a community, I think we can do better, and I'm brainstorming how we (and me specifically) can make progress there.

The heavy lifting is done by the small group of people making this possible. But past that, the accessibility and usability of that information is what limits the degree to which it's used by the people who couldn't have done that work originally.

By necessity, engineering is going to have far more detail during development than is useful or necessary for someone to use that information. In some places I've seen comments like "All your answers can be found in this 30-page thread somewhere." A beginner might not recognize that they are reading the answer when they are reading it, and no one would know without consuming ALL of the information whether that particular piece is the most up-to-date (as happened in your case), and that presumes that they can understand every piece of that development history along the way.

I would like to remove as many of those complications and barriers as possible.

I'm not sure how best to do that. For one, I think the community in general stressing that any new knowledge or context be added to the wiki is a good start. So that people can reliably say "The wiki is up-to-date with the best that we currently know." And then we should also regularly stress that of all the sources of information, people go there to get their answers (so that they don't get an intermediate answer as you did).

And then if the wiki becomes too much of a data dump, we need to compact that knowledge into a more useful form, maybe a "Guide To Using...".

And, ideally on the various Open Inverter and EVBMW webstores or perhaps Githubs or whatnot, link to the particular wiki along with the support thread? The support thread is a good way of asking new questions, but it's a discussion thread, not a great place to find answers other than to just ask new ones.

... Long story short I see a lot of effort put into volunteer engineering, but not a lot into volunteer documenting, and I think that means we don't see nearly as much adoption or people making use of that effort as we might. It doesn't feel as accessible.

*shrugs*. Specifically in this case was hoping for a silver bullet of what could have saved you from blowing up your capacitor.