I have an AC filter coming for my controller build, but also wondered if a filter for the DC side might also be an advantage.
The type I had in mind was
Can anyone with experience shed some light on this subject please.
Itās unlikely to be needed unless youāre trying to solve a specific problem. Itās also not uncommon for poorly designed EMI filters to cause issues, so Iād be a little wary of adding something like that.
Specifically, that looks like it adds significant inductance to the ground path which isnāt necessarily a good idea. Typically filters like that would be of the common-mode choke style, not a pair of differential mode chokes. This will still let common mode through, which is the most common form of EMI, potentially even making it worse by giving you a differential-mode to common-mode transformation at the inductors.
The main reason that I am building a new controller is because I have issues with some EMI.
What has been happening in some jobs, when the VFD turns off and the spindle slows down to almost stopped, I hear a click coming from the spindle. It sounds like a throw-out switch returning. When that happens I get an error signal and Estlcam tells me to use F11 to move from position. I donāt need to press F11, I just move any axis and the error goes away, but it is annoying because the spindle usually stops while the tool is returning to home.
The controller case that I have at present is made of 6mm MDF, so any interference can pass though easily, my new case is all metal.
Sometimes I have been getting USB errors also, and that is a major problem because the only thing I can do when that happens is to restart Estlcam as it wont recover by itself.
I have a shielded USB cable and also ferrite rings on the cable as well.
The new metal case I have used a central earthing point so all negative leads are all connected to earth, and also used shielded cables for as many connections as possible, all only earthed at one end to prevent earth looping.
It is getting close to being able to test it out on my machine and see if it makes any difference.
I went through three USB cables until I found one that was working, Estlcam is really finicky with it.
The one I have now is one of those ugly ones that look really cheap (those are recommended by Christian, but you canāt access that subpage of Estlcam.de with an English browserā¦), after having bought some that were much more expensive:
Hello,
those cheap, transparent USB cables with visible shielding braid are the best option - mainly because this way you can be really sure that there in fact is proper shielding.
The big issue with USB cables is that some are just intended to charge devices - but not actual USB communication - and have no shielding at all. They still work reasonably well in an office environment - but not in an electrically noisy environment close to a CNC machine.
Other big issue is improper or competely missing grounding on the machine. If you have a static discharge and the machine is not propery grounded the discharge will find a way though the motor cables into the machines electronics and then through the USB port to the computer. This inevitably disconnects the USB port and also isnāt really healthy for your computer in the long run.
Proper grounding means: all thouchable metal parts of the machines need a low impedance connection to ground / earth. Take into considerations that guide rails often isolate axes from each other electrically - even if made from metal due to the bearings oil film.
And finally there are often issues with electricaly noisy spindle motor drivers. They need proper line filters which are often not included to cut costs.
Christian
1 Like
While my final solution (to random USB errors) was earthing (@ PS) all isolated metal on the machine, Iām pretty sure using all shielded wiring (with the shields only connected to PS earth) and optoisolating most signaling (relays, VFD, etc. - a given w/ the wishlist Estlcam board) were also important. The EMI filters (vac, VFD and PS) I added had no noticeable effect. I use the blue tinted USB cables that have visible shielding.
I would like more information about how to filter EMI coming from the VFD and spindle, I think that is possibly my main source of the problem.
I presently have line filters for the mains power feeding the machine, but no filters for the power after leaving the VFD going to the spindle. There seems to be some back EMF coming from the spindle on shutdown of the spindle.
Also my DIY machine has Hiwin rails and ballscrews, so I would like some ideas on how to earth the spindle carriage to the rest of the machine due to the isolation caused by the bearing surfaces.
I was wondering about the possibility of static discharge that could come from the vacuum hose fitted to my machine. The hose is made of plastic and is 100mm in diameter, it follows the spindle carriage that it is attached to and being plastic could develop a charge with just the air and particles travelling through it at high speed. How to earth this static charge that could build up.
I would like more information about how to filter EMI coming from the VFD and spindle
Iām not an expert on VFD EMI filtering:
- I know filters for the VFDs output side exist - but Iāve never seen anyone using or needing one.
- A shielded cable between VFD and spindle is usually everything you need there to avoid issues
- The problem is the VFDs input side - and those filters are really important and need to match the VFDs power rating. And: they are large and relatively expensive which is why they are often skipped: https://www.sorotec.de/shop/Netzfilter-1-Phasig-f-r-FU-1-5-bis-2-2KW-ableitarme-Version.html?language=de
- Itās not those little filters you find when googling āline filterā - they are basicallly useless for this kind of application.
I was wondering about the possibility of static discharge that could come from the vacuum hose fitted to my machine
That is a really big issue because those discharges are powerful.
Iām not sure if in this case cables next to it even need to be double shielded. The outer layer to conduct the discharge to ground, the inner layer to block the field the outer layer creates during this current flow.
But again: not exactly an expert on this.
However: there are vacuum hoses with a metal spiral and slightly conductive plastic material that are specifically adressing this issue. The metal spiral can be grounded and the charges can no longer build up.
Galvanic isolation also helps a lot with this kind of problems but is just suppressing the symptoms without solving the issue itself. The discharges still happen - now limited to a different part of the whole circuit with less EMV sensitive parts - but they may still damage components there over time.
Christian
Maybe a coating of conductive paint would help with static build-up on the hose?
Conductive paint is rather expensive to buy, but can be made cheaply by DIY method.
Some research gave me this YouTube video that could be helpful.
Just google āanti static vacuum hoseā - they are pretty common and not expensive.
I think I have found the answer to why I have been getting an error when the spindle stops.
After doing a search for information I found this|:
Braking resistors in VFDs play a crucial role in the systemās efficiency by converting regenerated electric energy into thermal energy during the braking process, preventing it from feeding back to the power supply. The resistance capacity utilization and braking effect are optimized when the appropriate resistor is selected based on the brake unitās maximum current and the specific application, such as heavy inertia load or rapid braking loads. The use of braking resistors helps protect VFDs by consuming regenerative electric energy, thus preventing potential faults or damage to the VFDs. Additionally, braking resistors increase the braking torque capability, allowing for faster and more controlled braking, which can be particularly beneficial in various applications, including elevators, cranes, and locomotives
So maybe fitting a DC braking resistor could stop my problem.
Braking resistors usually do not help with EMI issues - but they are cheap and may be worth a shot.
The thing they are for is āback EMFā - sounds similar but has little to do with the often significantly increased EMI during a motors power up and down phase.
When stopping the spindle acts as generator feeding power back to the VFD. This increases the voltage in the VFDs capacitors - and may force it to shut down even damage it if the voltage rises too high. This is where the resistor comes into play: it just consumes the power that is fed back.
It is usually not an issue because spindle motors have very little inertia and so the power fed back is rather miniscule. Totally different situation to e.g. an elevator where a lot of energy needs to be taken back to stop the motor from turning.
Christian
I found this video very interesting and may help to solve some problem with interferences.
1 Like