[FrisscoCBR]

Okay, so this is my first post here.

Hello, i have a really big problem with what is thought to be interference from vfd, here i will explain how it is manifested.

First thing first, i created my cnc from basically scratch(see attached picture), most parts are 3d printed, only the base and the rails are metal, everything else is either MDF or PLA 3d printed plastic. I thought i would make a small cnc with a makita tool for milling, but as things progressed i created a cnc that is 2m long and 1m wide and a 1.5kw spindle on it.

The brain of the machine is an Arduino uno with a Protoneer CNC shield 3.0 on top of it and 3 A4988 drivers in it for X and 2 Y axis motors, i am using pretty small motors but planing to upgrade to 2.8A 57BYGH 2.1Nm steppers with TB6600 as a driver with power of 36v for the driver, this i have on the Z axis becuase the small motor could not lift the spindle becuase it weights 5kg.

I am using openbuilds control as a control software, and GRBL version 1.1h(latest one) .

Now that the description is done i can explaing the problem, the problem starts when the spindles VFD is turned on, i get noticable noise on my LCD screen and it is actually coming from the graphics cards of the control pc, the main problem is that whenever i turn on the vfd with the software the arduino gets stuck and would not respond untill i disconnect the usb and reconnect it and that happens every time, notihing else would work to get it out of that state.Without the VFD everthing runs smooth as it can be. I also get the unknown error code 31 from the software when that happens. The things that i already tried to fix this are as follows:

1.Installed EMI filter in front of the VFD
2.Seperated vfd from the rest of the controls(it is mounted on the wall in a metal box that is also grounded)
3.Grounded the base and everything metal from the machine
4.Unscrewed the connector on the spindle and connected the ground pin properly to the chasis, because a lot of spindles come without it connected, and mine was ons of that type
5.I am using shielded cable from the VFD to the motor and the shield is connected only to the ground terminal on the side of the VFD.
6.On/off switch between the VFD and the Arduino is connected via shielded cable and galvanically isolated via optocoupler.
7.All the motor power supplys and pc power supplys are grounded and the pc chassis is grounded properly.
8.Because the VFD is single phase powered and i have 3 phase power in my house, the VFD and the power supplys are using different phases.
9.All of the grounds are connected together to a single point.
10.All of the limit swithes are normally closed and galvanically isolated via optocupler as that is the best option.

The thing that i am trying now is that i am making a new proper ground point because the house is old and probably has junky ground connections, my grandfather made them himself who knows how, but i know that i have ground for sure in the house.

The one thing that leads me to believe that i have the bad ground is that when the VFD is of the voltage is 0 V between Ground and Neutral and 230 V between Live and Ground as it should be. BUT when i turn the VFD on i have around 4V AC between Ground and Neutral everywhere.

Also to mention is that i tried turning on the vfd via its own terminal with the control cable disconnected. Arduino is doing the same thing it freezes and i have to go through the whole process every time!

So my question is am i correct and is there any way to solve this?

[Al_The_Man]

Personally I would chose a 3ph filter/choke between VFD and motor over supply filter.
Are the negative terminals of the DC supplies taken to earth ground?
If you have 3ph power, why is the VFD fed with 1ph?
Al.

[GoGoStop!]

Hi, I am new to CNC's. Still designing mine.
I read something online about "ground loops", that explains your problem.
Google "ground loops". There is a couple of Youtube videos about it as well.
The solution is probable to install a separate ground rod for your VFD.

I am planning on installing 2 separate earth ground rods for my CNC machine. One for the VFD and one for the rest of the electrical/electronic equipment.

[FrisscoCBR]

If you have 3ph power, why is the VFD fed with 1ph?

Well i had the opportunity to choose between 1ph and 3ph VFD and i chose with 1ph becuase of easier cabling and it is only 1.5kw VFD. That what i chose. Also to point out is that i am using different phase for vfd and for the supplys and the pc.

Are the negative terminals of the DC supplies taken to earth ground?

Pc power supply negative is connected to earth ground and also the power supply negative for small x and y motors is connected to earth ground(its an atx power supply 500w) , the 36v supply for z axis motor does not have earth ground connected to the negative terminal. That's it.

Personally I would chose a 3ph filter/choke between VFD and motor over supply filter.

I am actually using 1ph filter like in the image attached below, what does that 3ph filter look like and where could i find one?

Thanks for the help, appreciate it!

[FrisscoCBR]

Hi, I am new to CNC's. Still designing mine.
I read something online about "ground loops", that explains your problem.
Google "ground loops". There is a couple of Youtube videos about it as well.
The solution is probable to install a separate ground rod for your VFD.

I am planning on installing 2 separate earth ground rods for my CNC machine. One for the VFD and one for the rest of the electrical/electronic equipment.

Yup i know about possible ground loops and i tried to see if that is the problem.

At first i tried by actually getting a cable run from my neighbours house(he has a seperate ground i measured it) to mine to try and see if the ground loop is the problem. First i used the neighbours cable on the pc and supplys and afterwards for the vfd. nothing actually changed. Btw i am now going the test the new ground i just installed today. Will report how it goes.

Appreciate your help!

[Al_The_Man]

See VFD accessories, reactors , https://www.automationdirect.com/adc...y_drives_(vfd)
Something called equi-potential bonding should be instituted to eleiminate ground loops, which are largely miss-understood.
Al.

[FrisscoCBR]

FINALLY! After all the things i tried it was simply bad or to long usb cable!!!

The connection between my arduino that controls the CNC and the PC was at first like this:

First i connected the included 30 cm cable that came with the arduino from the arduino to the kind of diy breakout board that makes it again usb type b (Its a good solid cable with the 100% braid shielding and i also added ferrites to it)

The i used a unknown quality cable from that breakout board to the pc! (It looks good and also has ferrite chokes on its ends so i thought that it is good!)(idk if it is even shielded)

That unknown cable was the cause of all of these disconnect issues. Idk if it is because of its length(its 1.8m and usb 2.0 can go up to 5m as it is stated in its specs) or because of bad quality cable.


How i realized that is what comes next:

With all the cables connected the error code would show up 99% of the time (the 1% of it not showing up was probably just luck!)

Then i tried to connect only the longer cable and the error was still popping up like 85-90% of the time!

Then today i realized that i never tried just the short one from the PC to the Arudino. Just the short one! (Because it was to short to reach the pc directly)

I took the chair and put the pc on it and then connected it directly with that 30 cm cable!

AND FINALLY!!

The full 4 hour job was complete without a single error or stop! And i even tried different things to see if i can force the error to appear. I tried to vary a VFD frequency while the machine was running (previously when i would do that the error would pop up immediately) now it was just fine like normal continued to do its work as intended! Tried turning the VFD on and off multiple times! The error newer popped up again!

With the initial setup (30 cm + 1.8m cable) with the 4 hour job i was getting like 15-20 errors all the time! Sometimes it would not even let me continue!

Soo, CONCLUSION!

Check if you have a good quality(and not to long too) usb cable first, because as it was in my case, you will waste a lot of time, effort and nerves (and money) for nothing!!!