How to safely/legally install and wire a VFD?

[Moded1952 (Lucas)]

I’m close to the point where I need to wire up my VFD but I’m not super experienced with wiring >24V electricity and the safety concerns of larger spindles, so wanted to consult the hivemind.

I have:

• VFD: Omron MX2 3G3MX2-AB007-E
• Braking resistor: Omron AX-REM00K1200-IE
• Line filter: Omron AX-FIM1014-RE
• Spindle: Mechatron HFP-6508-42

Looking at the manual, installation seems simple enough:

There’s an even clearer image too:

It looks like it’s mostly a matter of grounding everything and using large enough wires, which I feel comfortable doing myself.

However:

• There’re a ton of warnings in the inverter and spindle manuals saying installation should be done by a professional.
• The manual recommends grouding the EMC filter with a 10mm²/7AWG conductor which seems huge.
• The manual mentions some extra things I don’t have, like “reactors” and “chokes”.
• There’s a lot of talk of safety standards and extra safety devices like these things.

Is all of this just CYA on the manufacturer’s part or is there serious danger to look out for?

Do any of the Europeans around here know if there’s some legal requirement to have a professional do it?

Should I consider some of these extra safety devices? Should I think about adding an RCD, some circuit breakers or some fuses?

Overall, how do I set up this thing and stay safe? Should I just give up and buy a pre-assembled electronics box?

I’m posting this in the Shapeoko forum since you guys need to deal with VFDs for spindles but I’m actually installing this on a Nomad.

[LiamN (Liam Newcombe)]

Firstly,

If you’re not familiar with wiring higher AC voltages you might want to get a professional to do it.

The warnings on the VFD are there for a good reason, the incoming 220V is rectified and stored internally at closer to 400V, this persists even after input power has been removed until the capacitors are discharged.

Adding an RCD is usually a good plan, but it cannot protect you downstream of the VFD because a ) it can’t see through the VFD and b ) the VFD has more than enough stored energy to end you anyway

Getting the grounding right is key for both interference and safety.

Grounding your Shapeoko

Why would you want to ground your Shapeoko? This isn’t something everybody should do, but if you have either of the following issues, you should consider it; Dust & chip accumulation (on rails and V wheels causing movement problems) Random disconnects (when cutting plastics or when extraction is on) There’s nothing ‘wrong’ with the way the Shapeoko is delivered and there’s no safety issue WRT grounding on the machines (if you add a VFD spindle that must be properly grounded but that’s not for…

Grounding the shell of the spindle is critical for safety, as yours is a decent German unit I would expect that to be reasonably easy. This grounding is what will trip out the breaker or VFD if there is some sort of fault.

Grounding the screen of the spindle power cable is important for minimising interference.

Grounding the shell of your EMI filter is important for both safety and interference.

The reactors and chokes thing, the noise filter for the input is important, you can add some ferrite chokes on the output wiring, but I’d only bother if you have an interference problem. I personally wouldn’t bother with the AC input reactor unless I had a noise problem.

A nice metal box with cable gland cutouts will make it easy to mount all this stuff. If your VFD display detaches you can put that in the box too, with appropriate ventilation. The box will also likely have earthing studs etc. and let you put some DIN rail inside to mount a breaker and some terminals for your wiring if you want to. These are cheap because they’re mass produced and most electrical factors carry them.

[LiamN (Liam Newcombe)]

Probably the biggest risk is that you’ll end up turning a simple electrical installation into a complete project which you spend days on and end up doing something like this instead of just wiring it up and machining things, it’s really quite dangerous

Reminds me, must install the Angus Young sticker too…

[Moded1952 (Lucas)]

@LiamN

If you’re not familiar with wiring higher AC voltages you might want to get a professional to do it.

How much familiarity would you recommend? I’ve done some small stuff myself like installing and wiring AC/DC converters and installing switches, so I’d say I know enough not to do anything monumentally stupid with <240V.

Would you say that’s enough or do I really need the professional?

On the one hand, I have a healthy amount of caution for AC but on the other, any kind of labor in Switzerland is spectacularly expensive, even moreso if it’s skilled labor…

@LiamN

The warnings on the VFD are there for a good reason, the incoming 220V is rectified and stored internally at closer to 400V

Is that true even of my VFD, which outputs the same voltage?

@LiamN

you can add some ferrite chokes on the output wiring, but I’d only bother if you have an interference problem. I personally wouldn’t bother with the AC input reactor unless I had a noise problem.

How would I determine that I have a noise or interference problem? Would this affect other equipment in the room/apartment like computers, phones etc.?

@LiamN

A nice metal box with cable gland cutouts will make it easy to mount all this stuff.

Yeah, been thinking about that, maybe something like M400300210GE. It has a full-height gland so plenty of room for cable egress and maybe enough space for a couple of fans too.

@LiamN

ou’ll end up turning a simple electrical installation into a complete project which you spend days on

That wouldn’t actually be so bad, my wiring right now is kinda terrible… I’ve been planning to redo a lot of it when I install the new Z-axis. I even bought the proper Molex connectors and pre-crimped leads for the stepper!

[LiamN (Liam Newcombe)]

@Moded1952

How much familiarity would you recommend? I’ve done some small stuff myself like installing and wiring AC/DC converters and installing switches, so I’d say I know enough not to do anything monumentally stupid with <240V.

Would you say that’s enough or do I really need the professional?

On the one hand, I have a healthy amount of caution for AC but on the other, any kind of labor in Switzerland is spectacularly expensive, even moreso if it’s skilled labor…

It’s really a question of what you feel comfortable with doing, this is a personal judgement on whether you’re happy that you can earth things properly and not zap yourself or anyone else who touches it. I know that I am but I’ve got a long history of wiring things up and some literal scars to show which mistakes I learned from.

As for legality, Switzerland may be quite strict, if you’re still going to plug the whole thing into an outlet rather than hard-wire it into a breaker in your consumer unit that can mean you don’t need a licensed electrician in some territories but check your local laws.

@Moded1952

@LiamN

The warnings on the VFD are there for a good reason, the incoming 220V is rectified and stored internally at closer to 400V

Is that true even of my VFD, which outputs the same voltage?

Yes, this is the DC voltage that the sine wave AC is rectified to, multiply the AC RMS voltage (220) by sqrt(2) to get the DC and the AC peak.

@Moded1952

@LiamN

you can add some ferrite chokes on the output wiring, but I’d only bother if you have an interference problem. I personally wouldn’t bother with the AC input reactor unless I had a noise problem.

How would I determine that I have a noise or interference problem? Would this affect other equipment in the room/apartment like computers, phones etc.?

The most common symptom is that your machine becomes unreliable and disconnects from the USB during jobs. The input noise filter to the VFD is there to stop it messing with all your other appliances and should work fine.

@Moded1952

@LiamN

A nice metal box with cable gland cutouts will make it easy to mount all this stuff.

Yeah, been thinking about that, maybe something like M400300210GE. It has a full-height gland so plenty of room for cable egress and maybe enough space for a couple of fans too.

Yep, looks like a decent box.

There seems to be a remote operator cable and kit to put a remote keypad on the VFD so you could keep it hidden in the box where it’s safe and chip / dust free.

@Moded1952

@LiamN

ou’ll end up turning a simple electrical installation into a complete project which you spend days on

That wouldn’t actually be so bad, my wiring right now is kinda terrible… I’ve been planning to redo a lot of it when I install the new Z-axis. I even bought the proper Molex connectors and pre-crimped leads for the stepper!

Well, a DIN rail and a few DIN blocks for connecting up the wiring inside the box along with a nice earthing block might make things easier.

[Moded1952 (Lucas)]

I’ve been looking into it some more and I think I’m more comfortable with it now.

One thing that’s been confusing me though is that the manual recommends a 15A class J fuse on the input side of the VFD.

This is an 800W spindle at 240V, so the average current draw should be 3.3A. Even 10A would exceed the current rating for a standard electrical socket.

Anyone know why the manual calls for a fuse ~5x bigger than what it looks like the input current should be?

And why does the fuse need a 200kA rating?

My hope was to stick an RCD like this one in there.

Maybe I should really just talk to an electrician…

[LiamN (Liam Newcombe)]

Great, it’s not that hard, so long as you’re careful.

As for breakers and fuses, that is a rabbit hole all on it’s own.

A residual current (ground fault in American) breaker is of restricted use and may just nuisance trip on a VFD.
RCDs work by using a small transformer whose secondary winding produces a difference of the input live and neutral cores (in single phase). Where it sees a sufficiently large difference (30mA in that specific unit) it will open the power circuit as it has detected a ‘fault’.

The safety issue is twofold

1. The RCD cannot ‘see’ past the rectifier / inverter stage of the VFD, you can have an arbitrarily large leakage current through anything you like to ground on the other side of the VFD and this still appears to be ‘normal’ power to the RCD
2. Even when the fault does cause the RCD to trip the VFD still has full charge on the capacitors behind the rectifier so the RCD trip time won’t save you

Then there’s a false-tripping issue

3. The VFD and the noise filters both work with a network of inductors and capacitors which is connected to power lines and ground. If the noise from the VFD is asymmetric then this will result in a non-zero differential between live and neutral and may trip the RCD, particularly at the high harmonic frequencies input rectifier noise tends to occur at.

Then there are the tripping settings which is what that ‘tripping curve B’ is about in the breaker spec

4. Most VFDs now have a very smooth ‘soft start’ for their own power up, even the HuanYang do this now so the legacy rules about choosing a C or D curve to allow for power-on surges doesn’t apply any more. The VFD also soft starts the motor with a controlled ramp up
5. The fuse spec may well be a ‘save the silicon’ spec where the fast acting fuse can blow and isolate quickly enough to not vaporise the expensive components inside the VFD, these are generally used in addition to a normal current limiting circuit breaker.
6. The breaking capacity is there to ensure that you can actually interrupt the expected fault current in the circuit, it’s largely about the supply circuit, not the size of the VFD.

I have RCDs in the main house supply so I didn’t add an extra one just for the VFD, if there’s a fault it will probably trip a whole series anyway if they’re the same sensitivity. I have not (yet) had any nuisance trips on 30mA trigger, even with my 2.2kW HY VFD and noise filter.

I put a breaker in with a class C trip curve because I hadn’t realised the HY VFD did such a good soft-start. This breaker was rated to about 130% of the expected current at full power.

I didn’t add an extra ‘save the silicon’ fuse, but I’m not a licensed electrician and it is only plugged into a standard socket using a standard fused plug (UK) so I already have the fuse.

I definitely did not put any switching or breaker between the VFD and the motor as that’s quite a good way to cook your VFD if you open that breaker whilst the motor is energised.

[Moded1952 (Lucas)]

Thanks!

@LiamN

The safety issue is twofold

Is there anything that can be done about that? Like putting a 3-phase RCD on the output side of the inverter?

@LiamN

The VFD and the noise filters both work with a network of inductors and capacitors which is connected to power lines and ground. If the noise from the VFD is asymmetric then this will result in a non-zero differential between live and neutral and may trip the RCD, particularly at the high harmonic frequencies input rectifier noise tends to occur at.

I read about this but my line filter has a sticker saying the “nominal leakage current” is 7mA. That should mean it’s fine on a 30mA RCD, right?

@LiamN

The breaking capacity is there to ensure that you can actually interrupt the expected fault current in the circuit, it’s largely about the supply circuit, not the size of the VFD.

Ah, so the capacity is there to ensure that if there’s a short, the supply doesn’t just melt everything? And I suppose the capacity is high because the VFD is usually installed in an industrial environment where a 240V circuit might have a much larger fuse upstream?

[LiamN (Liam Newcombe)]

@Moded1952

Thanks!

NP

@Moded1952

Is there anything that can be done about that? Like putting a 3-phase RCD on the output side of the inverter?

A breaker in the output line is likely to zap the VFD if it trips.

The only thing I found to do was to make sure that my spindle casing, Z axis and machine frame were all properly grounded back to the big fat ground point in my electrical box and rely on the fuses / VFD to trip out on a fault. Test all your earths and check they work.

@Moded1952

I read about this but my line filter has a sticker saying the “nominal leakage current” is 7mA. That should mean it’s fine on a 30mA RCD, right?

It was for me, I think the issue is more the high frequency nature of the noise, as you have a small motor and a decent quality VFD I would only worry if you actually have the problem.

@Moded1952

Ah, so the capacity is there to ensure that if there’s a short, the supply doesn’t just melt everything? And I suppose the capacity is high because the VFD is usually installed in an industrial environment where a 240V circuit might have a much larger fuse upstream?

Yep, it’s a ‘make sure the fault is disconnected’ rating, otherwise in an extreme condition an arc can develop where the fuse used to be and the fault can continue, this is generally considered to be bad

On large three phase switchboards it’s not unusual to see main breakers rated to stop fault currents up to 100KAIC which is 100,000 Amps Interrupting Capacity, this is the sort of power delivery best observed from a safe distance…