[badhabit]

Hi,

I have a 3 axis CNC machine with servos all around. This means that i have 3x 2,2kW VFD's for the 3 axis and one 7,5kW VFD for the spindle.

The VFD manual has this depiction of how to have breaker, contactor, reactor and filters connected to the VFD:
Attachment 485422
....but how should i setup this if i have 4 VFDs like in my setup? I was hoping to share some of the components between the different VFDs, eg. the reactor and filter for the axis VFDs and then maybe a seperate set for the spindle VFD...but then i'd have to have to contactor after the filters, is this ok? And what about sharing the reactor between multiple VFDs? I suspect i cannot change the order of reactor vs. filter, so if reactor sharing is a bad idea, then filter sharing is also out of the question...

Reactors are rather large and would fill a lot of sparse control cabinet space...besides the extra cost... the same goes for the filters...so "less is more"...

But how is this done "normally"? What have you guys done? What is the correct solution here?

Appreciate any input!

/Thomas

[joeavaerage]

Hi,
line reactors are a good idea, they tame the input current to both servo drives and VFD's. Both servo drives and VFDs will work perfectly well without line reactors but can cause conducted noise issues
eleswhere in the machine.

In your case I rather think a reactor for each servo drive is overkill, however having one decent line reactor on the VFD input only may well save a lot of nuisance breaker trips and may also save
some line side voltage disturbance to other sensitive components.

A 7.5kW VFD I assume you are running three phase power? In which case follow the VFD manufacturers advice.

Some VFDs allow the use of a DC choke INSTEAD of a line reactor. You would know this because the terminals for a DC choke have to be bought out to the terminal strip. It may prove
that a single DC choke will prove cheaper than a three phase reactor, having said that three phase reactors are widely available and mostly pretty cheaply.

I would recommend (two stage) line filters, certainly one for the VFD, probably one each for the servo drives, or if you want to save some money maybe just one large filter to do all
three servo drives, and one line filter for the sensitive stuff, like the PC and motion control electronics. The PC and motion electronics will draw very little current, probably no more than a few amps
and so the line filter for that could be even be a three stage type without breaking the bank.

Craig

[badhabit]

line reactors are a good idea, they tame the input current to both servo drives and VFD's. Both servo drives and VFDs will work perfectly well without line reactors but can cause conducted noise issues
eleswhere in the machine.

In your case I rather think a reactor for each servo drive is overkill, however having one decent line reactor on the VFD input only may well save a lot of nuisance breaker trips and may also save
some line side voltage disturbance to other sensitive components.

A 7.5kW VFD I assume you are running three phase power? In which case follow the VFD manufacturers advice.

Yes, i should have mentioned i live in europe so i'm using 3-phase, 400V for the system...and i think i might have mixed the terms "VFD" and "driver" a bit up in my first post... in any case, i have only servo motors with servo drivers, no regular AC motors with regular VFD, sorry about that!

The reason for my question is that the manufacturer does not state how to connect multiple drivers with either one or multiple sets of filtering components. They only give an example with one driver and one set of filtering components.

I have had issues with noise before during my testing of the drivers and motors and the introduction of a "strong" groundplane(backplate in a chassis) and a 2-stage filter solved (most) of my noise issues and i ended up making this for my testing purposes:
Attachment 485432
...its a bit hard to see, but i've also added a big ferrite core to the wires going to the motor just below each driver

...this works "fairly well". I still have some issues with random breaker trips and especially tripping of my RCD when i unplug my testing setup...i'm not sure why....yet...the leak current from the filter is supposed to be 18mA and i think this is just too much with all the other stuff i have in my shop. If i cannot find any other solution i'm going to make a dedicated RCD-circuit just for this machine...

Some VFDs allow the use of a DC choke INSTEAD of a line reactor. You would know this because the terminals for a DC choke have to be bought out to the terminal strip. It may prove
that a single DC choke will prove cheaper than a three phase reactor, having said that three phase reactors are widely available and mostly pretty cheaply.

I would recommend (two stage) line filters, certainly one for the VFD, probably one each for the servo drives, or if you want to save some money maybe just one large filter to do all
three servo drives, and one line filter for the sensitive stuff, like the PC and motion control electronics. The PC and motion electronics will draw very little current, probably no more than a few amps
and so the line filter for that could be even be a three stage type without breaking the bank.

I've done some digging in the documentation for the drivers, but it is only drivers above 37kW that have external terminals for the DC-choke. So i think i'm "stuck" with regular line regulators...

This is what my plan is so far:
Attachment 485430

It is by no means perfect! I changed the order of the reactor and the noise-filter from the recommended order according to the manufacturer(as shown in my first post). I also joined the drivers for the axis on one reactor and then put the spindle driver on its own reactor. I'm not sure this is a good idea, but after researching reactors a bit it might be easier/cheaper to find two 10Hp reactors rather than one big one that fits the physical space available. The load from the axis motor might also be more "similar" than with the spindle servo, so it might make sense(?). Any thoughts on this?

I was hoping to put breakers before each driver, but with this setup it seems best to put the breaker in front of the safety contactor instead. Or any inputs on this?

I really appreciate your input and the feedback from this forum, so thanks for any feedback!

/Thomas

[Al_The_Man]

Are these spindles, standard 3ph motors, i.e., originally intended for across line use?
In all my VFD installations over the years, I have primarily fitted 3 ph chokes or inductors between VFD and motor.
Rarely any on the input side.

[joeavaerage]

Hi,
Lineside reactors are primarily to improve power factor and reduce line side voltage disturbance owing to the otherwise unabated harmonic current distortion.

Secondary side reactors (between motor and VFD) are primarily to limit capacitive output loading of the VFD, most commonly associated with long cable runs.
The inter conductor capacitance of a three phase cable used for the motor to VFD interconnect will easily exceed 200pF/m, and thus at long runs say 10m there
is the potential for the output of the VFD to see considerable capacitance, over 1nF. That will vastly increase the switching power loss in the MOSFETS/IGBTs
and anything that prevents power loss in those devices increases reliability.

I think OPs plan is reasonable.

I use 750W servos rather than 2.2kW, but what I've found is that on average they draw less than 100W each, and only very occasionally draw anything like rated power, and so I have
not bothered to put a lineside reactor in front off them, but rather a two stage EMI filter. It has proven adequate. At low to moderate power levels VFDs and servo drives have VERY low power factor,
seldom better than 0.5. While a reactor would improve that to 0.85 or so at low to moderate power levels the power saving does not warrant the expense of a reactor. The situation with lineside voltage
is less clear. The voltage disturbance is most at peak loads, and infrequent event.

I would probably forego the line reactor in front of the servo drives, although would leave room in the cabinet that a reactor could be fitted IF it is determined with in-service testing
to be necessary. A good two stage EMI filter in front of the servo drives would most likely be adequate on its own.

The VFD will spend many hours running at anywhere from low to moderate power levels, but depending on the job at hand could also run near rated power for hours at a time.
A line reactor will improve power factor and reduce nuisance breaker trips but more importantly reduce the harmonic distortion which will therefore be less of a noise challenge to
the more sensitive motion control electronics. A good two stage EMI filter may be adequate on its own but an EMI filter AND a line reactor would be the gold standard.

When I installed a line reactor with my VFD I found the overall electrical noise issues, which had previously been quite bothersome, diminished by 90% or more. My VFD is single phase
and I wound my own reactor on a recycled transformer core so the cost was minimal excepting my time.

Craig

[Al_The_Man]

Yes, I attended a seminar provided by the local power service company and the main thrust was reducing PF for manufacturing companies, VFD's were one area where using inductors on the input improved things, the typical input of any VFD is a very large capacitor bank, so this makes sense.
I use them between VFD and motor to reduce the HF portion of the output, which could potentially harm motors, especially of the older vintage.
It also has a beneficial effect on protecting the VFD itself.

[joeavaerage]

Hi,
in recent years here in rural New Zealand there has been a huge increase in the use of VFD driven down-hole bore pumps for irrigation. They are often 100kW and some
of the bigger ones 500kW. Farmers were very lax at fitting line reactors or anything else until everyones PC and TVs started blowing up because of the power supply 'pollution'.
As a consequence power companies have have to insist that each and every installation meet the standard ( IEC 61000-3-12 ).

In most cases to meet the standard you have to use not just line reactors but active PFC correction units....worth many thousands of dollars, approaching that of the VFD itself. Farmers hated it
but it had to be done.

For even moderate size CNC machines its not really considered an issue and most therefore have little or no harmonic current reduction equipment but that has a consequence for other
devices sharing the same power supply.

Craig

[joeavaerage]

Hi,
I had not considered the potential to protect motors that are not strictly 'inverter ready'. I've had a couple of cheap motors (one from India and two from China) that blew up
as a result of being VFD driven. I've never had a bother with some of the older British and US 'brick s...thouse' designs, but secondary reactors make sense.

Craig

[mactec54]

This is always a great subject but most of what has been posted does not help the Op some of what has been posted is correct, but some is out there with electrical knowledge behind it.

He has AC Servo Drives so all he needs is an EMI Filter on the logic input Power to each drive, and (1) Reactor on the mains 400v supply to the system nothing else should be needed for this system, this is the norm you will do for any AC Servo system if you have any EMI problems

Add the EMI Filters first for each Drive logic power input, and then the Reactor it needed, make sure all Grounding, /Earth connections are correctly terminated

[badhabit]

He has AC Servo Drives so all he needs is an EMI Filter on the logic input Power to each drive, and (1) Reactor on the mains 400v supply to the system nothing else should be needed for this system, this is the norm you will do for any AC Servo system if you have any EMI problems

Add the EMI Filters first for each Drive logic power input, and then the Reactor it needed, make sure all Grounding, /Earth connections are correctly terminated

Thanks for your input!

Just to clearify: "drive logic power input" is just the 3-phase power input for each drive, right?

So you'd recommend that each of the 4 drivers get their own EMI-filters? All two-stage filters? Sounds fair, but I think i'm going to have a leakage current issue with this setup. Each filter has eg. 10mA leakage times four filters giving 40mA leakage total...i'm a bit concerned with this since i think that we are legally required to have 30mA RCD's(residencial building)...i'll have to investigate this a bit...

Can you recommend any 3phase/400V filters? I currently have a 30A version of this in my test cabinet: https://www.mouser.dk/datasheet/2/41...A2-2075941.pdf

What would you recommend regarding breakers and safety relays? Just one set in front of the reactor or 4 sets in front of each driver?

/Thomas

[Al_The_Man]

This is always a great subject but most of what has been posted does not help the Op some of what has been posted is correct, but some is out there with electrical knowledge behind it.

What is that supposed to Mean??

[mactec54]

Thanks for your input!

Just to clearify: "drive logic power input" is just the 3-phase power input for each drive, right?

So you'd recommend that each of the 4 drivers get their own EMI-filters? All two-stage filters? Sounds fair, but I think i'm going to have a leakage current issue with this setup. Each filter has eg. 10mA leakage times four filters giving 40mA leakage total...i'm a bit concerned with this since i think that we are legally required to have 30mA RCD's(residencial building)...i'll have to investigate this a bit...

Can you recommend any 3phase/400V filters? I currently have a 30A version of this in my test cabinet: https://www.mouser.dk/datasheet/2/41...A2-2075941.pdf

What would you recommend regarding breakers and safety relays? Just one set in front of the reactor or 4 sets in front of each driver?

/Thomas

Would have to see the drive input power schematic to say how you need to best do it.

[mactec54]

What is that supposed to Mean??

It is just as it is posted but missed the (No) sorry if this makes any difference, (but some is out there with No electrical knowledge behind it. )

[badhabit]

Would have to see the drive input power schematic to say how you need to best do it.

Yeah, thats hard to get on these chinese somewhat no-name brands:-( I ended up having a two-stage filter per servo-driver and then having a single reactor in front of the four filters. I'm going to place the safety contactor in front of the reactor.

I ended up using buying Schaffner filters, FN3288-10-44-C26-R65 for the axis drivers and FN3288-16-44-C26-R65 for the spindle driver. For the reactor i had to choose something that is readily available in Europe and TCI seems hard to source, so i ended up buying a LR3 40-4/20 from Block.

Hopefully i'll be able to rewire my test-setup in the next couple of weeks and try it out...

Again, thanks for the input and advice!