Eemods 20.05 10h

0:02

S… Speaker 1 (Eemods 20.05 10h)

So,

0:03

S… Speaker 1 (Eemods 20.05 10h)

already this morning,

0:04

S… Speaker 1 (Eemods 20.05 10h)

used in the IEC Research International,

0:07

S… Speaker 1 (Eemods 20.05 10h)

Lili Barberwitz works for Energy

0:11

S… Speaker 1 (Eemods 20.05 10h)

Consulting and she

0:15

S… Speaker 1 (Eemods 20.05 10h)

has actually a Bachelor of

0:20

S… Speaker 1 (Eemods 20.05 10h)

Science in Google Physics and Mathematics from RPI

0:24

S… Speaker 1 (Eemods 20.05 10h)

in the US and she's a...

0:30

S… Speaker 1 (Eemods 20.05 10h)

a student programmer and analyst.

0:32

S… Speaker 1 (Eemods 20.05 10h)

She was in the team of programming and analysis of testing

0:36

S… Speaker 1 (Eemods 20.05 10h)

data.

0:36

S… Speaker 1 (Eemods 20.05 10h)

She has a strong background in computational mathematics and physics.

0:40

S… Speaker 1 (Eemods 20.05 10h)

Her applications range from radio projects for

0:44

S… Speaker 1 (Eemods 20.05 10h)

Air Force Office in Scientific Research to work on the ENERGISTAR

0:49

S… Speaker 1 (Eemods 20.05 10h)

program for the U .S.

0:50

S… Speaker 1 (Eemods 20.05 10h)

EPA.

0:51

S… Speaker 1 (Eemods 20.05 10h)

And her key additional clients include the

0:55

S… Speaker 1 (Eemods 20.05 10h)

National Resources Canada,

0:56

S… Speaker 1 (Eemods 20.05 10h)

Air Resources and also E -Six Utility.

1:02

S… Speaker 1 (Eemods 20.05 10h)

Key efficiency analysis areas for her are pumps,

1:06

S… Speaker 1 (Eemods 20.05 10h)

lighting and connected cameras.

1:08

S… Speaker 1 (Eemods 20.05 10h)

One

1:21

S… Speaker 1 (Eemods 20.05 10h)

fun

1:50

S… Speaker 1 (Eemods 20.05 10h)

fact is that

1:54

S… Speaker 1 (Eemods 20.05 10h)

the results you're going to see are based on test

1:58

S… Speaker 1 (Eemods 20.05 10h)

results or measurements which Sandy conducted during

2:03

S… Speaker 1 (Eemods 20.05 10h)

our Robin Robinson virtual measurement initiative during all of

2:07

S… Speaker 1 (Eemods 20.05 10h)

that.

2:28

S… Speaker 1 (Eemods 20.05 10h)

This is Paolo,

2:32

S… Speaker 1 (Eemods 20.05 10h)

I can share the screen and then I'll leave you to

2:36

S… Speaker 1 (Eemods 20.05 10h)

Andrea to moderate it.

2:44

S… Speaker 1 (Eemods 20.05 10h)

Thank you for getting help.

3:07

S… Speaker 1 (Eemods 20.05 10h)

Where is she in the phone room?

3:09

S… Speaker 1 (Eemods 20.05 10h)

No,

3:10

S… Speaker 1 (Eemods 20.05 10h)

New York.

3:10

S… Speaker 1 (Eemods 20.05 10h)

Only six hours.

3:22

S… Speaker 1 (Eemods 20.05 10h)

No, it's not.

3:23

S… Speaker 1 (Eemods 20.05 10h)

No, it's good.

3:23

S… Speaker 1 (Eemods 20.05 10h)

Perfect.

3:24

S… Speaker 1 (Eemods 20.05 10h)

Not yet.

3:25

S… Speaker 1 (Eemods 20.05 10h)

I think I'll go from this screen.

3:29

S… Speaker 1 (Eemods 20.05 10h)

You go to the bottom right.

3:30

S… Speaker 1 (Eemods 20.05 10h)

The bottom right is this small icon.

3:32

S… Speaker 1 (Eemods 20.05 10h)

Ready?

3:34

S… Speaker 1 (Eemods 20.05 10h)

Get it on the left.

3:35

S… Speaker 1 (Eemods 20.05 10h)

There.

3:37

S… Speaker 1 (Eemods 20.05 10h)

Look in there.

3:38

S… Speaker 1 (Eemods 20.05 10h)

She's sharing the wrong screen.

3:42

S… Speaker 1 (Eemods 20.05 10h)

Now you have to change the screen.

3:44

S… Speaker 1 (Eemods 20.05 10h)

Sorry.

3:45

S… Speaker 1 (Eemods 20.05 10h)

And then say close presenter view.

3:56

S… Speaker 1 (Eemods 20.05 10h)

Yeah, I think I'd better go with this video post

4:00

S… Speaker 1 (Eemods 20.05 10h)

than mine.

4:01

S… Speaker 1 (Eemods 20.05 10h)

Sorry about that.

4:01

S… Speaker 1 (Eemods 20.05 10h)

No, I can tell you.

4:03

S… Speaker 1 (Eemods 20.05 10h)

Go in presentation mode again.

4:04

S… Speaker 1 (Eemods 20.05 10h)

Ah,

4:07

S… Speaker 1 (Eemods 20.05 10h)

here

4:12

S… Speaker 1 (Eemods 20.05 10h)

I am.

4:12

S… Speaker 1 (Eemods 20.05 10h)

Yeah,

4:14

S… Speaker 1 (Eemods 20.05 10h)

that's perfect.

4:15

S… Speaker 1 (Eemods 20.05 10h)

Okay.

4:17

S… Speaker 1 (Eemods 20.05 10h)

Nice.

4:19

S… Speaker 1 (Eemods 20.05 10h)

Did it.

4:21

S… Speaker 1 (Eemods 20.05 10h)

I gave already the introduction.

4:23

S… Speaker 1 (Eemods 20.05 10h)

And so we can go.

4:26

S… Speaker 1 (Eemods 20.05 10h)

That sounds

4:30

S… Speaker 1 (Eemods 20.05 10h)

perfect.

4:31

S… Speaker 1 (Eemods 20.05 10h)

Alrighty.

4:31

S… Speaker 1 (Eemods 20.05 10h)

Well, thank you for wearing with me on that.

4:34

S… Speaker 1 (Eemods 20.05 10h)

All righty,

4:35

S… Speaker 1 (Eemods 20.05 10h)

so I'll dive into it then.

4:36

S… Speaker 1 (Eemods 20.05 10h)

So this is a paper that we put together

4:41

S… Speaker 1 (Eemods 20.05 10h)

with Andrea's support

4:45

S… Speaker 1 (Eemods 20.05 10h)

and Jenny Nielsen's support sending us the data set.

4:48

S… Speaker 1 (Eemods 20.05 10h)

So this analysis is based on the

4:53

S… Speaker 1 (Eemods 20.05 10h)

round -robin of converters data set,

4:56

S… Speaker 1 (Eemods 20.05 10h)

one line per drive.

5:00

S… Speaker 1 (Eemods 20.05 10h)

to have sort of a filtered version of the original

5:04

S… Speaker 1 (Eemods 20.05 10h)

RRC data set.

5:05

S… Speaker 1 (Eemods 20.05 10h)

So there aren't difficulties in this.

5:08

S… Speaker 1 (Eemods 20.05 10h)

And thanks for sending us them for preparing that for us.

5:12

S… Speaker 2 (Eemods 20.05 10h)

So let me just think about,

5:15

S… Speaker 1 (Eemods 20.05 10h)

I might have to compress this a little bit just from the time.

5:19

S… Speaker 1 (Eemods 20.05 10h)

So really the biggest points here,

5:22

S… Speaker 1 (Eemods 20.05 10h)

if folks are used to IDC 61800,

5:28

S… Speaker 1 (Eemods 20.05 10h)

A lot of the IC methods have relied on testing

5:32

S… Speaker 1 (Eemods 20.05 10h)

up to the 9 ,100 point.

5:34

S… Speaker 1 (Eemods 20.05 10h)

And that has its advantages and

5:38

S… Speaker 2 (Eemods 20.05 10h)

downsides.

5:38

S… Speaker 1 (Eemods 20.05 10h)

But really,

5:40

S… Speaker 1 (Eemods 20.05 10h)

the goal of this paper was to see just how much

5:45

S… Speaker 1 (Eemods 20.05 10h)

we could rely on the 9 ,100 point in the context

5:49

S… Speaker 1 (Eemods 20.05 10h)

of 50 hertz equipment.

5:54

S… Speaker 1 (Eemods 20.05 10h)

And we do use a couple methods or techniques here to try

5:58

S… Speaker 1 (Eemods 20.05 10h)

to understand what the CVMs are doing a

6:02

S… Speaker 2 (Eemods 20.05 10h)

little bit more closely.

6:03

S… Speaker 1 (Eemods 20.05 10h)

So method one is a way

6:08

S… Speaker 1 (Eemods 20.05 10h)

to try to disaggregate what the PWM method

6:12

S… Speaker 1 (Eemods 20.05 10h)

is at the 50 -100 point.

6:15

S… Speaker 1 (Eemods 20.05 10h)

And then we also use the technique that we've been messing

6:19

S… Speaker 1 (Eemods 20.05 10h)

with in the US side.

6:22

S… Speaker 1 (Eemods 20.05 10h)

at 90 hundreds to see if you can separate out third

6:26

S… Speaker 1 (Eemods 20.05 10h)

harmonic ejection or not.

6:27

S… Speaker 3 (Eemods 20.05 10h)

So just

6:31

S… Speaker 1 (Eemods 20.05 10h)

a brief note here.

6:33

S… Speaker 1 (Eemods 20.05 10h)

So we talked about this a little bit.

6:34

S… Speaker 1 (Eemods 20.05 10h)

There is sort of a companion or a complementary analysis

6:40

S… Speaker 1 (Eemods 20.05 10h)

for U .S.

6:41

S… Speaker 2 (Eemods 20.05 10h)

equipment.

6:42

S… Speaker 1 (Eemods 20.05 10h)

And I will actually do that later on.

6:45

S… Speaker 2 (Eemods 20.05 10h)

And actually we'll find some interesting differences.

6:49

S… Speaker 3 (Eemods 20.05 10h)

So first thing,

6:52

S… Speaker 1 (Eemods 20.05 10h)

we do have to filter the data set.

6:57

S… Speaker 1 (Eemods 20.05 10h)

We decided let's group everything

7:01

S… Speaker 2 (Eemods 20.05 10h)

to be EU -centric.

7:03

S… Speaker 1 (Eemods 20.05 10h)

So we filtered it down to 400 volts in.

7:06

S… Speaker 1 (Eemods 20.05 10h)

We wanted to make sure that we were looking at the 5100

7:11

S… Speaker 1 (Eemods 20.05 10h)

point and 50 hertz and 25 hertz.

7:14

S… Speaker 1 (Eemods 20.05 10h)

There were some drives that were missing,

7:19

S… Speaker 2 (Eemods 20.05 10h)

so fundamental,

7:19

S… Speaker 2 (Eemods 20.05 10h)

voltage fundamental.

7:22

S… Speaker 2 (Eemods 20.05 10h)

We do rely on that,

7:24

S… Speaker 2 (Eemods 20.05 10h)

so we can look at those.

7:27

S… Speaker 1 (Eemods 20.05 10h)

And then there was a couple items that were,

7:30

S… Speaker 2 (Eemods 20.05 10h)

you know, been in the weeds,

7:31

S… Speaker 1 (Eemods 20.05 10h)

but we had one product for the third year.

7:33

S… Speaker 1 (Eemods 20.05 10h)

Displacement power factor that just didn't make sense for the motor,

7:37

S… Speaker 1 (Eemods 20.05 10h)

so that got filtered for data quality.

7:39

S… Speaker 1 (Eemods 20.05 10h)

And also a couple of products with some unusually

7:44

S… Speaker 1 (Eemods 20.05 10h)

high data quality issue type

7:49

S… Speaker 2 (Eemods 20.05 10h)

modulating indexes.

7:51

S… Speaker 1 (Eemods 20.05 10h)

So those all got filtered.

7:54

S… Speaker 2 (Eemods 20.05 10h)

We dropped down to 25.

7:59

S… Speaker 3 (Eemods 20.05 10h)

And then we do

8:03

S… Speaker 1 (Eemods 20.05 10h)

have to calculate some intermediate quantities.

8:06

S… Speaker 1 (Eemods 20.05 10h)

So we start with displacement power factor,

8:09

S… Speaker 1 (Eemods 20.05 10h)

real power,

8:10

S… Speaker 1 (Eemods 20.05 10h)

and your output voltage,

8:12

S… Speaker 1 (Eemods 20.05 10h)

RMS,

8:13

S… Speaker 1 (Eemods 20.05 10h)

fundamental voltage,

8:17

S… Speaker 1 (Eemods 20.05 10h)

current RMS.

8:19

S… Speaker 1 (Eemods 20.05 10h)

And we use a few identities to get down to the

8:23

S… Speaker 1 (Eemods 20.05 10h)

estimated harmonic current.

8:26

S… Speaker 1 (Eemods 20.05 10h)

So you can calculate the current power factor,

8:28

S… Speaker 2 (Eemods 20.05 10h)

or I'm sorry,

8:29

S… Speaker 2 (Eemods 20.05 10h)

you can calculate the current power,

8:30

S… Speaker 1 (Eemods 20.05 10h)

true power factor,

8:32

S… Speaker 2 (Eemods 20.05 10h)

distortion factor,

8:34

S… Speaker 1 (Eemods 20.05 10h)

TFCI,

8:36

S… Speaker 1 (Eemods 20.05 10h)

and then finally the harmonic curve.

8:39

S… Speaker 1 (Eemods 20.05 10h)

So it's just a few identities that you ended

8:43

S… Speaker 1 (Eemods 20.05 10h)

up walking through.

8:44

S… Speaker 3 (Eemods 20.05 10h)

So

8:49

S… Speaker 1 (Eemods 20.05 10h)

this is the other piece.

8:50

S… Speaker 1 (Eemods 20.05 10h)

All of the tools that we're using rely on

8:54

S… Speaker 2 (Eemods 20.05 10h)

the modulation index.

8:57

S… Speaker 1 (Eemods 20.05 10h)

So the one that I'm using here is based on the

9:01

S… Speaker 1 (Eemods 20.05 10h)

voltage transfer ratio.

9:05

S… Speaker 1 (Eemods 20.05 10h)

Sometimes you'll see as voltage transfer efficiency,

9:08

S… Speaker 2 (Eemods 20.05 10h)

VTE.

9:08

S… Speaker 1 (Eemods 20.05 10h)

But for the linear region,

9:11

S… Speaker 1 (Eemods 20.05 10h)

sine PWM has 0 .612.

9:16

S… Speaker 1 (Eemods 20.05 10h)

And this is just kind of default efficiency moving voltage

9:21

S… Speaker 1 (Eemods 20.05 10h)

from the VC plus to output.

9:23

S… Speaker 1 (Eemods 20.05 10h)

Fundamental.

9:27

S… Speaker 1 (Eemods 20.05 10h)

This is something you can rely on until you get to the over -modulated region,

9:30

S… Speaker 1 (Eemods 20.05 10h)

so it's pretty consistent.

9:32

S… Speaker 1 (Eemods 20.05 10h)

On the right -hand side here is just a block

9:36

S… Speaker 1 (Eemods 20.05 10h)

showing this section here is

9:40

S… Speaker 1 (Eemods 20.05 10h)

sine pwn modulating indexes for the products

9:45

S… Speaker 1 (Eemods 20.05 10h)

that we filter down to.

9:47

S… Speaker 3 (Eemods 20.05 10h)

I'm sorry,

9:49

S… Speaker 1 (Eemods 20.05 10h)

space factor in this lower section.

9:51

S… Speaker 1 (Eemods 20.05 10h)

Upper section is sine pwn,

9:54

S… Speaker 1 (Eemods 20.05 10h)

which has higher modulating indexes.

9:58

S… Speaker 1 (Eemods 20.05 10h)

on your tax provision.

10:01

S… Speaker 1 (Eemods 20.05 10h)

So we'll see that later on,

10:03

S… Speaker 1 (Eemods 20.05 10h)

and we do use this modulating index quite a bit.

10:06

S… Speaker 1 (Eemods 20.05 10h)

The other thing is,

10:09

S… Speaker 1 (Eemods 20.05 10h)

until you classify the product,

10:11

S… Speaker 1 (Eemods 20.05 10h)

you don't really know which modulating index is right for it,

10:14

S… Speaker 1 (Eemods 20.05 10h)

so you just have to calculate the code.

10:16

S… Speaker 1 (Eemods 20.05 10h)

So

10:22

S… Speaker 1 (Eemods 20.05 10h)

this is really the bar of method one.

10:25

S… Speaker 1 (Eemods 20.05 10h)

So about 20 years ago,

10:28

S… Speaker 1 (Eemods 20.05 10h)

it's a bit of a deep dive.

10:30

S… Speaker 1 (Eemods 20.05 10h)

There's this formula for the

10:35

S… Speaker 1 (Eemods 20.05 10h)

harmonic distortion.

10:38

S… Speaker 1 (Eemods 20.05 10h)

factor and it's it equates the modulating index

10:42

S… Speaker 1 (Eemods 20.05 10h)

to the harmonic current produced

10:46

S… Speaker 1 (Eemods 20.05 10h)

in the output side of the pdf so you have your cdm output cables

10:51

S… Speaker 1 (Eemods 20.05 10h)

and motor side and this is this is basically

10:55

S… Speaker 1 (Eemods 20.05 10h)

the i hope you want to call it the non

10:59

S… Speaker 1 (Eemods 20.05 10h)

-work producing current as a feature

11:04

S… Speaker 1 (Eemods 20.05 10h)

of the modulation scheme or pw method

11:08

S… Speaker 1 (Eemods 20.05 10h)

And really,

11:09

S… Speaker 1 (Eemods 20.05 10h)

the derivations are pretty extensive,

11:12

S… Speaker 1 (Eemods 20.05 10h)

but what it really boils down to,

11:14

S… Speaker 1 (Eemods 20.05 10h)

the sine PWM just is not as effective

11:19

S… Speaker 1 (Eemods 20.05 10h)

at a moving voltage,

11:21

S… Speaker 1 (Eemods 20.05 10h)

and it causes more harmonics.

11:24

S… Speaker 1 (Eemods 20.05 10h)

And so at this 5 ,100

11:28

S… Speaker 1 (Eemods 20.05 10h)

point, you can expect just from the PWM

11:33

S… Speaker 1 (Eemods 20.05 10h)

definitions that you're going to have,

11:35

S… Speaker 1 (Eemods 20.05 10h)

if it's sine PWM,

11:37

S… Speaker 1 (Eemods 20.05 10h)

15 % more.

11:44

S… Speaker 1 (Eemods 20.05 10h)

And you can calculate this for anywhere in the linear region,

11:47

S… Speaker 1 (Eemods 20.05 10h)

but since the IEC test procedure,

11:50

S… Speaker 1 (Eemods 20.05 10h)

you can go down to 25,

11:53

S… Speaker 1 (Eemods 20.05 10h)

but at that point,

11:57

S… Speaker 1 (Eemods 20.05 10h)

it's pretty close together.

11:58

S… Speaker 1 (Eemods 20.05 10h)

And if you go too far out,

12:01

S… Speaker 1 (Eemods 20.05 10h)

you can overmodulate it.

12:03

S… Speaker 1 (Eemods 20.05 10h)

So 5050 and 5100 are the points where this is

12:07

S… Speaker 1 (Eemods 20.05 10h)

the most powerful.

12:08

S… Speaker 1 (Eemods 20.05 10h)

And then one thing you do notice is that you do have to use different modulating

12:13

S… Speaker 1 (Eemods 20.05 10h)

indexes for sine pw versus a space vector.

12:16

S… Speaker 1 (Eemods 20.05 10h)

So it actually sort of emphasizes this difference a bit more.

12:20

S… Speaker 1 (Eemods 20.05 10h)

So

12:24

S… Speaker 1 (Eemods 20.05 10h)

given that factor,

12:26

S… Speaker 1 (Eemods 20.05 10h)

you can take all the products.

12:29

S… Speaker 1 (Eemods 20.05 10h)

We estimate the originals of an unadjusted

12:34

S… Speaker 1 (Eemods 20.05 10h)

harmonic permant time arm.

12:35

S… Speaker 1 (Eemods 20.05 10h)

And then we can actually apply this adjustment

12:39

S… Speaker 1 (Eemods 20.05 10h)

factor in both directions.

12:41

S… Speaker 1 (Eemods 20.05 10h)

So we haven't categorized the products yet.

12:43

S… Speaker 1 (Eemods 20.05 10h)

We just know it's a product.

12:46

S… Speaker 1 (Eemods 20.05 10h)

And you can increase and decrease the estimated

12:50

S… Speaker 1 (Eemods 20.05 10h)

harmonic current and then re -divide it by the fundamental

12:54

S… Speaker 1 (Eemods 20.05 10h)

current to get some new HDIs.

12:58

S… Speaker 1 (Eemods 20.05 10h)

And really the goal there is just to try to push things out

13:02

S… Speaker 1 (Eemods 20.05 10h)

of the population to spot if

13:06

S… Speaker 1 (Eemods 20.05 10h)

you were starting with a lot of harmonics and then you can increase.

13:10

S… Speaker 1 (Eemods 20.05 10h)

it'll push it out of the usual range.

13:12

S… Speaker 1 (Eemods 20.05 10h)

And so you can classify a product as sine

13:17

S… Speaker 1 (Eemods 20.05 10h)

pw, and in that case,

13:18

S… Speaker 1 (Eemods 20.05 10h)

if you're running high in harmonics,

13:21

S… Speaker 1 (Eemods 20.05 10h)

and if you're running low,

13:23

S… Speaker 1 (Eemods 20.05 10h)

then you can classify it as a big factor.

13:25

S… Speaker 1 (Eemods 20.05 10h)

So it's a bit of a blunt instrument type

13:29

S… Speaker 1 (Eemods 20.05 10h)

method.

13:30

S… Speaker 1 (Eemods 20.05 10h)

But as I move to the next slide,

13:33

S… Speaker 1 (Eemods 20.05 10h)

you can see it actually is reasonably effective.

13:35

S… Speaker 1 (Eemods 20.05 10h)

So we classified

13:40

S… Speaker 1 (Eemods 20.05 10h)

everything.

13:41

S… Speaker 1 (Eemods 20.05 10h)

In general,

13:42

S… Speaker 1 (Eemods 20.05 10h)

you can tell the products fairly quickly

13:47

S… Speaker 1 (Eemods 20.05 10h)

that are signed PWM.

13:49

S… Speaker 1 (Eemods 20.05 10h)

They have just extremely high THBIs

13:54

S… Speaker 1 (Eemods 20.05 10h)

compared to the bulk of the population.

13:57

S… Speaker 1 (Eemods 20.05 10h)

Similarly,

13:59

S… Speaker 1 (Eemods 20.05 10h)

you can typically spot products that are space

14:04

S… Speaker 1 (Eemods 20.05 10h)

vector or third -party injection capable.

14:06

S… Speaker 1 (Eemods 20.05 10h)

There is some middle ground where

14:11

S… Speaker 1 (Eemods 20.05 10h)

it is

14:13

S… Speaker 1 (Eemods 20.05 10h)

I would say it's hard to tell,

14:14

S… Speaker 1 (Eemods 20.05 10h)

although realistically,

14:16

S… Speaker 1 (Eemods 20.05 10h)

these are likely thermoponic injection.

14:19

S… Speaker 1 (Eemods 20.05 10h)

They're just right on the edge of the limit

14:24

S… Speaker 1 (Eemods 20.05 10h)

that I was using,

14:25

S… Speaker 1 (Eemods 20.05 10h)

so I figured it was safer to classify them as indeterminate

14:31

S… Speaker 1 (Eemods 20.05 10h)

to be safe.

14:32

S… Speaker 1 (Eemods 20.05 10h)

With that said,

14:33

S… Speaker 1 (Eemods 20.05 10h)

when we did this,

14:34

S… Speaker 1 (Eemods 20.05 10h)

we've seen about 40%,

14:36

S… Speaker 1 (Eemods 20.05 10h)

at least,

14:37

S… Speaker 1 (Eemods 20.05 10h)

of the products tested.

14:41

S… Speaker 1 (Eemods 20.05 10h)

have third harmonic capability.

14:43

S… Speaker 1 (Eemods 20.05 10h)

So that was a nice finding.

14:45

S… Speaker 1 (Eemods 20.05 10h)

We've heard over the years that it's becoming much

14:49

S… Speaker 1 (Eemods 20.05 10h)

more common than your default drive for

14:54

S… Speaker 1 (Eemods 20.05 10h)

just about any application,

14:55

S… Speaker 1 (Eemods 20.05 10h)

and it feels good to see that confirmed.

14:57

S… Speaker 1 (Eemods 20.05 10h)

The other thing...

15:00

S… Speaker 1 (Eemods 20.05 10h)

just a side note modulation index does run ahead and

15:04

S… Speaker 1 (Eemods 20.05 10h)

sign PWM exactly like it is 5 .6 at

15:08

S… Speaker 1 (Eemods 20.05 10h)

5100.

15:09

S… Speaker 1 (Eemods 20.05 10h)

153 is this vector.

15:12

S… Speaker 1 (Eemods 20.05 10h)

Now you can use that to refine things a little bit at

15:16

S… Speaker 1 (Eemods 20.05 10h)

data, but really the biggest finding here is just that there's a decent

15:20

S… Speaker 1 (Eemods 20.05 10h)

share of third harmonic capable equipment.

15:22

S… Speaker 1 (Eemods 20.05 10h)

So

15:28

S… Speaker 1 (Eemods 20.05 10h)

method two.

15:30

S… Speaker 1 (Eemods 20.05 10h)

This one,

15:31

S… Speaker 1 (Eemods 20.05 10h)

I will say,

15:33

S… Speaker 1 (Eemods 20.05 10h)

is a lot more effective in the U .S.

15:35

S… Speaker 1 (Eemods 20.05 10h)

side of the fence.

15:36

S… Speaker 1 (Eemods 20.05 10h)

One of the things that we ran into

15:40

S… Speaker 1 (Eemods 20.05 10h)

when trying to use the IEC tools to project out to 100 -100 is

15:46

S… Speaker 1 (Eemods 20.05 10h)

that we noticed the expected voltage drops and things like that,

15:49

S… Speaker 1 (Eemods 20.05 10h)

expected behavior between 90 -100 and

15:54

S… Speaker 1 (Eemods 20.05 10h)

100 -100.

15:58

S… Speaker 1 (Eemods 20.05 10h)

30 -40 % of products,

16:01

S… Speaker 1 (Eemods 20.05 10h)

the IDC tools worked fairly well,

16:03

S… Speaker 1 (Eemods 20.05 10h)

and the rest of the products did not fit that well.

16:06

S… Speaker 1 (Eemods 20.05 10h)

On the other hand,

16:08

S… Speaker 1 (Eemods 20.05 10h)

what we see in the IDC 50Hz side of the fence,

16:12

S… Speaker 1 (Eemods 20.05 10h)

which I'll get into shortly,

16:13

S… Speaker 1 (Eemods 20.05 10h)

is it did work a lot better for those products.

16:17

S… Speaker 1 (Eemods 20.05 10h)

So by the way,

16:20

S… Speaker 1 (Eemods 20.05 10h)

we did filter everything down to 400V.

16:22

S… Speaker 1 (Eemods 20.05 10h)

At 400V,

16:27

S… Speaker 1 (Eemods 20.05 10h)

DC bus one way through five times the input

16:32

S… Speaker 1 (Eemods 20.05 10h)

voltage.

16:32

S… Speaker 1 (Eemods 20.05 10h)

You can derive the sort

16:37

S… Speaker 1 (Eemods 20.05 10h)

of maximum fundamental voltage that you can get out of the product.

16:41

S… Speaker 1 (Eemods 20.05 10h)

Signed PWM has a,

16:44

S… Speaker 1 (Eemods 20.05 10h)

it runs into O modulation more

16:49

S… Speaker 1 (Eemods 20.05 10h)

quickly.

16:49

S… Speaker 1 (Eemods 20.05 10h)

So

16:54

S… Speaker 1 (Eemods 20.05 10h)

the theory behind this technique is

16:58

S… Speaker 1 (Eemods 20.05 10h)

You can inspect some drives are operating at points where sine

17:02

S… Speaker 1 (Eemods 20.05 10h)

T WM can't reach.

17:03

S… Speaker 1 (Eemods 20.05 10h)

So

17:07

S… Speaker 1 (Eemods 20.05 10h)

that's the goal.

17:09

S… Speaker 1 (Eemods 20.05 10h)

So

17:13

S… Speaker 1 (Eemods 20.05 10h)

here's when

17:19

S… Speaker 1 (Eemods 20.05 10h)

you apply the technique.

17:20

S… Speaker 1 (Eemods 20.05 10h)

Red lines,

17:21

S… Speaker 1 (Eemods 20.05 10h)

or red dots here,

17:22

S… Speaker 1 (Eemods 20.05 10h)

are the entire data set with the

17:27

S… Speaker 1 (Eemods 20.05 10h)

estimated sine.

17:35

S… Speaker 1 (Eemods 20.05 10h)

So you start with things uncategorized.

17:39

S… Speaker 1 (Eemods 20.05 10h)

And then from there,

17:41

S… Speaker 1 (Eemods 20.05 10h)

you're looking for anything that crosses this boundary really

17:46

S… Speaker 1 (Eemods 20.05 10h)

can alleviate you with their harmonic injection.

17:48

S… Speaker 1 (Eemods 20.05 10h)

And I say this is more exciting in

17:52

S… Speaker 1 (Eemods 20.05 10h)

US products,

17:55

S… Speaker 1 (Eemods 20.05 10h)

because when we look at it in 50 Hertz products,

17:58

S… Speaker 1 (Eemods 20.05 10h)

There's actually only two points that

18:03

S… Speaker 1 (Eemods 20.05 10h)

even hit this maximum fundamental frequency

18:07

S… Speaker 1 (Eemods 20.05 10h)

line from the space vector

18:11

S… Speaker 1 (Eemods 20.05 10h)

perspective.

18:12

S… Speaker 1 (Eemods 20.05 10h)

So really,

18:14

S… Speaker 1 (Eemods 20.05 10h)

as far as the sine p wm is concerned,

18:16

S… Speaker 1 (Eemods 20.05 10h)

everything is achievable with sine p wm still.

18:20

S… Speaker 1 (Eemods 20.05 10h)

I will say that as you get close

18:24

S… Speaker 1 (Eemods 20.05 10h)

to this 1 .15 threshold,

18:27

S… Speaker 1 (Eemods 20.05 10h)

that's where things get really tough for sine pw

18:31

S… Speaker 1 (Eemods 20.05 10h)

modulating.

18:32

S… Speaker 1 (Eemods 20.05 10h)

So I would say products tend to try to slow down.

18:36

S… Speaker 1 (Eemods 20.05 10h)

And even before that,

18:39

S… Speaker 1 (Eemods 20.05 10h)

you're still getting a decent amount of square wave saturation.

18:42

S… Speaker 1 (Eemods 20.05 10h)

So these products at the

18:46

S… Speaker 1 (Eemods 20.05 10h)

far right -hand side,

18:48

S… Speaker 1 (Eemods 20.05 10h)

in all likelihood,

18:49

S… Speaker 1 (Eemods 20.05 10h)

are going to be thermo -harmonic injection capable,

18:54

S… Speaker 1 (Eemods 20.05 10h)

simply because you don't want to push a

18:57

S… Speaker 1 (Eemods 20.05 10h)

sign PWM drive quite this hard.

19:00

S… Speaker 1 (Eemods 20.05 10h)

Not that C can't do it,

19:03

S… Speaker 1 (Eemods 20.05 10h)

but they're not likely to have products.

19:05

S… Speaker 1 (Eemods 20.05 10h)

Here you would expect these to be their problem in connection.

19:08

S… Speaker 1 (Eemods 20.05 10h)

Also what we see in the US data and

19:12

S… Speaker 1 (Eemods 20.05 10h)

in general, products that tend to have very low fundamental output

19:16

S… Speaker 1 (Eemods 20.05 10h)

voltage tend to be sign PWM.

19:20

S… Speaker 1 (Eemods 20.05 10h)

So correlation,

19:23

S… Speaker 1 (Eemods 20.05 10h)

not causation there.

19:24

S… Speaker 1 (Eemods 20.05 10h)

And the reason I thought this

19:28

S… Speaker 1 (Eemods 20.05 10h)

out is

19:31

S… Speaker 1 (Eemods 20.05 10h)

I compared what I was seeing in the 50 hertz,

19:35

S… Speaker 1 (Eemods 20.05 10h)

100 and a half a point,

19:36

S… Speaker 1 (Eemods 20.05 10h)

versus the 460 about 60

19:41

S… Speaker 1 (Eemods 20.05 10h)

hertz, 100 and a half a point.

19:42

S… Speaker 1 (Eemods 20.05 10h)

And so in the US side of the fence,

19:45

S… Speaker 1 (Eemods 20.05 10h)

you actually have quite a few products that are operating well

19:50

S… Speaker 1 (Eemods 20.05 10h)

past the sine BWM threshold.

19:52

S… Speaker 1 (Eemods 20.05 10h)

So that was an interesting finding is that products

19:57

S… Speaker 1 (Eemods 20.05 10h)

serving the EMA for ice.

20:00

S… Speaker 1 (Eemods 20.05 10h)

tend not to be over -modulating or

20:05

S… Speaker 1 (Eemods 20.05 10h)

operating in this sort of mortality region as much.

20:09

S… Speaker 1 (Eemods 20.05 10h)

Now,

20:12

S… Speaker 1 (Eemods 20.05 10h)

this is the 100 -100 point,

20:13

S… Speaker 1 (Eemods 20.05 10h)

to be fair.

20:14

S… Speaker 1 (Eemods 20.05 10h)

So there you can see a lot of two points that are classified with this process.

20:18

S… Speaker 1 (Eemods 20.05 10h)

And a few things that are kind of close to the boundaries that you can

20:23

S… Speaker 1 (Eemods 20.05 10h)

say with certainty,

20:24

S… Speaker 1 (Eemods 20.05 10h)

maybe three or four of these products would be classified

20:28

S… Speaker 1 (Eemods 20.05 10h)

as space vector here.

20:31

S… Speaker 1 (Eemods 20.05 10h)

whereas in the U .S.

20:33

S… Speaker 1 (Eemods 20.05 10h)

side of the power pit,

20:34

S… Speaker 1 (Eemods 20.05 10h)

this plant cloud actually is about 40 %

20:38

S… Speaker 1 (Eemods 20.05 10h)

of the products.

20:39

S… Speaker 1 (Eemods 20.05 10h)

So we definitely found that the 5100 technique was

20:43

S… Speaker 1 (Eemods 20.05 10h)

more effective for IDC equipment.

20:46

S… Speaker 1 (Eemods 20.05 10h)

I would say that was the biggest finding right

20:50

S… Speaker 1 (Eemods 20.05 10h)

there, is that when you go to IDC's

20:54

S… Speaker 1 (Eemods 20.05 10h)

calculation tools,

20:56

S… Speaker 1 (Eemods 20.05 10h)

where it expects a certain amount of voltage drop,

21:02

S… Speaker 1 (Eemods 20.05 10h)

We are thinking on the U .S.

21:05

S… Speaker 1 (Eemods 20.05 10h)

side that there's really some layers to that.

21:07

S… Speaker 1 (Eemods 20.05 10h)

That the 10 % fundamental voltage drop used in IC

21:12

S… Speaker 1 (Eemods 20.05 10h)

is fairly accurate for a lot of 400 volt

21:16

S… Speaker 1 (Eemods 20.05 10h)

50 hertz products.

21:17

S… Speaker 1 (Eemods 20.05 10h)

It actually seems to be about what we're seeing there.

21:20

S… Speaker 1 (Eemods 20.05 10h)

And I would suspect it's because a lot of these motors are

21:24

S… Speaker 1 (Eemods 20.05 10h)

rated for 380 to 430 volt.

21:29

S… Speaker 1 (Eemods 20.05 10h)

So in a sense,

21:30

S… Speaker 1 (Eemods 20.05 10h)

having products running at this 380 volt RMS

21:36

S… Speaker 1 (Eemods 20.05 10h)

input voltage is not that bad for it.

21:39

S… Speaker 1 (Eemods 20.05 10h)

Whereas in NEMA,

21:41

S… Speaker 1 (Eemods 20.05 10h)

we actually see a lot of drives that are willing to push

21:46

S… Speaker 1 (Eemods 20.05 10h)

all the way to 440,

21:48

S… Speaker 1 (Eemods 20.05 10h)

445,

21:49

S… Speaker 1 (Eemods 20.05 10h)

even 450 RMS input voltage.

21:53

S… Speaker 1 (Eemods 20.05 10h)

It's a different philosophical difference.

21:55

S… Speaker 1 (Eemods 20.05 10h)

I'll

22:00

S… Speaker 1 (Eemods 20.05 10h)

go back to the summary.

22:02

S… Speaker 1 (Eemods 20.05 10h)

One is just,

22:04

S… Speaker 1 (Eemods 20.05 10h)

is this technique useful?

22:05

S… Speaker 1 (Eemods 20.05 10h)

So the HDF method,

22:08

S… Speaker 1 (Eemods 20.05 10h)

were we able to classify things?

22:10

S… Speaker 1 (Eemods 20.05 10h)

Yes.

22:11

S… Speaker 1 (Eemods 20.05 10h)

I would say about 80 % of

22:15

S… Speaker 1 (Eemods 20.05 10h)

the products could be classified.

22:17

S… Speaker 1 (Eemods 20.05 10h)

And we were able to determine

22:21

S… Speaker 1 (Eemods 20.05 10h)

that at least 40 % of the RRC data set had their carbonic capability.

22:26

S… Speaker 1 (Eemods 20.05 10h)

So we found that to be a useful finding that we find it running

22:30

S… Speaker 1 (Eemods 20.05 10h)

out.

22:32

S… Speaker 1 (Eemods 20.05 10h)

The other side of it is the 90,

22:35

S… Speaker 1 (Eemods 20.05 10h)

100, 100 classification technique useful

22:39

S… Speaker 1 (Eemods 20.05 10h)

in the U .S.

22:41

S… Speaker 1 (Eemods 20.05 10h)

very much.

22:42

S… Speaker 1 (Eemods 20.05 10h)

In the IC product world,

22:45

S… Speaker 1 (Eemods 20.05 10h)

not as much.

22:46

S… Speaker 1 (Eemods 20.05 10h)

Products have a lot more repair.

22:48

S… Speaker 1 (Eemods 20.05 10h)

So that was an interesting result,

22:51

S… Speaker 1 (Eemods 20.05 10h)

which I was really meaning.

22:53

S… Speaker 1 (Eemods 20.05 10h)

For

22:59

S… Speaker 1 (Eemods 20.05 10h)

folks that are interested,

23:00

S… Speaker 1 (Eemods 20.05 10h)

these are GSPF questions.

23:03

S… Speaker 1 (Eemods 20.05 10h)

I did mention that they are a function of the modulating index.

23:06

S… Speaker 1 (Eemods 20.05 10h)

The derivations of these are fairly complex,

23:09

S… Speaker 1 (Eemods 20.05 10h)

but they are very useful.

23:11

S… Speaker 1 (Eemods 20.05 10h)

Much thank you to Ahava

23:16

S… Speaker 1 (Eemods 20.05 10h)

and RJ Kirkman for doing this work about 20

23:20

S… Speaker 1 (Eemods 20.05 10h)

years ago.

23:20

S… Speaker 1 (Eemods 20.05 10h)

By the way,

23:22

S… Speaker 1 (Eemods 20.05 10h)

sine PWM is always on its own.

23:26

S… Speaker 1 (Eemods 20.05 10h)

Space vector and third harmonic differ only

23:30

S… Speaker 1 (Eemods 20.05 10h)

by all.

23:32

S… Speaker 1 (Eemods 20.05 10h)

So everything is represented by this one.

23:36

S… Speaker 1 (Eemods 20.05 10h)

So

23:41

S… Speaker 1 (Eemods 20.05 10h)

that is the bulk of that.

23:43

S… Speaker 1 (Eemods 20.05 10h)

I can open up for questions if those have anything that you're interested

23:47

S… Speaker 1 (Eemods 20.05 10h)

to ask about.

23:48

S… Speaker 1 (Eemods 20.05 10h)

But thank you for having me.

23:50

S… Speaker 1 (Eemods 20.05 10h)

Thank

23:56

S… Speaker 1 (Eemods 20.05 10h)

you, Lily.

23:56

S… Speaker 2 (Eemods 20.05 10h)

I think you're here as well.

24:01

S… Speaker 2 (Eemods 20.05 10h)

Questions from the audience?

24:03

S… Speaker 1 (Eemods 20.05 10h)

One

24:09

S… Speaker 2 (Eemods 20.05 10h)

question I had just right now.

24:14

S… Speaker 2 (Eemods 20.05 10h)

The reason why on the IEC we are

24:18

S… Speaker 2 (Eemods 20.05 10h)

talking about the 90 -100 point is that according

24:24

S… Speaker 2 (Eemods 20.05 10h)

to manufacturers in the standardization group,

24:27

S… Speaker 2 (Eemods 20.05 10h)

they want to avoid the 100 -100 point

24:31

S… Speaker 2 (Eemods 20.05 10h)

because they're starting to have flux weakening.

24:34

S… Speaker 2 (Eemods 20.05 10h)

Can you confirm that?

24:36

S… Speaker 2 (Eemods 20.05 10h)

If you look at the data,

24:38

S… Speaker 2 (Eemods 20.05 10h)

so it's actually slide 11.

24:43

S… Speaker 2 (Eemods 20.05 10h)

There's not,

24:44

S… Speaker 2 (Eemods 20.05 10h)

we cannot say that there's no field weakening.

24:47

S… Speaker 2 (Eemods 20.05 10h)

It's just what you say is sine PWM

24:52

S… Speaker 2 (Eemods 20.05 10h)

probably uses third harmonic injection,

24:55

S… Speaker 2 (Eemods 20.05 10h)

which is not yet.

24:55

S… Speaker 2 (Eemods 20.05 10h)

It's a little bit of old modulation.

24:58

S… Speaker 2 (Eemods 20.05 10h)

It's not field weakening.

24:59

S… Speaker 1 (Eemods 20.05 10h)

Yeah,

25:02

S… Speaker 1 (Eemods 20.05 10h)

yeah, I'm coming on that.

25:03

S… Speaker 1 (Eemods 20.05 10h)

So what this sine PWM definition is,

25:07

S… Speaker 1 (Eemods 20.05 10h)

is pure sine PWM with no

25:11

S… Speaker 1 (Eemods 20.05 10h)

third harmonic injection.

25:14

S… Speaker 1 (Eemods 20.05 10h)

And then the equipment that can do third harmonic injection is

25:18

S… Speaker 1 (Eemods 20.05 10h)

covered in the THI envelope.

25:21

S… Speaker 1 (Eemods 20.05 10h)

So THI is a fixed factor.

25:23

S… Speaker 1 (Eemods 20.05 10h)

It has a lot more capability.

25:26

S… Speaker 1 (Eemods 20.05 10h)

So in this sense,

25:28

S… Speaker 1 (Eemods 20.05 10h)

you could easily go up to 370,

25:31

S… Speaker 1 (Eemods 20.05 10h)

380.

25:33

S… Speaker 1 (Eemods 20.05 10h)

380 volts RMS with that DC plus value and

25:38

S… Speaker 1 (Eemods 20.05 10h)

still be in the linear region.

25:40

S… Speaker 1 (Eemods 20.05 10h)

So you do have a lot of extra capability there.

25:43

S… Speaker 1 (Eemods 20.05 10h)

That said,

25:46

S… Speaker 1 (Eemods 20.05 10h)

the products,

25:46

S… Speaker 1 (Eemods 20.05 10h)

I think about the 40 % or so of products that we assume

25:50

S… Speaker 1 (Eemods 20.05 10h)

are signed PWM,

25:52

S… Speaker 1 (Eemods 20.05 10h)

they're over -modulated pretty heavily at this point

25:56

S… Speaker 1 (Eemods 20.05 10h)

because we start over -modulating.

26:10

S… Speaker 1 (Eemods 20.05 10h)

you start over -modulating right here,

26:12

S… Speaker 1 (Eemods 20.05 10h)

and rapidly,

26:15

S… Speaker 1 (Eemods 20.05 10h)

and does it worsen your signal.

26:17

S… Speaker 1 (Eemods 20.05 10h)

But you have quite a bit of saturation at that point.

26:20

S… Speaker 1 (Eemods 20.05 10h)

Yeah,

26:25

S… Speaker 2 (Eemods 20.05 10h)

that was one thing I was,

26:26

S… Speaker 2 (Eemods 20.05 10h)

that's a rather context diagram here,

26:28

S… Speaker 2 (Eemods 20.05 10h)

I had to compare the exchange over twice a minute to get

26:33

S… Speaker 1 (Eemods 20.05 10h)

it done.

26:33

S… Speaker 2 (Eemods 20.05 10h)

Okay?

26:34

S… Speaker 2 (Eemods 20.05 10h)

Any other questions?

Eemods 20.05 10h

ສັງລວມ

ຖາມ AI ກ່ຽວກັບການແປນີ້

Eemods 20.05 10h

ສັງລວມ

ຖາມ AI ກ່ຽວ​ກັບ​ການ​ແປ​ນີ້

ຖາມ AI ກ່ຽວກັບການແປນີ້