Wiring HuangYang VFD without grounding! How dangerous is this?

[driftboat]

If I understand the power sources the spindle is powered from a separate 115v

[Goemon]

At the risk of getting my head bitten off.... it's not super dangerous to run a spindle without ground being connected. That doesn't mean it is advisable though. It has the potential to be dangerous if there is a problem with the VFD.

I briefly tested my spindle without a ground connection before I installed the newer style outlets (which all have ground now). It worked fine. It wasn't that long ago that two pin outlets (with no ground) were the norm and within code in America. One of the vendors for Chinese VFDs and spindles has a YouTube video where they state that ground is optional.

A ground connection is a safety feature to prevent you getting shocked in the event there is a problem with your VFD. So.., if all else is well, it should never need to be used. I don't have that level of trust in Chinese VFDs though.... plus you might have an issue with your wiring...

Given the voltages involved and the potential consequences of a problem, however unlikely, it would be stupid not to connect to ground where a ground connection is available. It cost me less than $20 to install a new grounded 50amp 240v outlet. Surely it is better to fix the wiring problem instead of skipping a ground wire....

I found that the screw terminals on my Huanyang VFD to be far too close togeather for the gauge of wire needed for my spindle. I was worried about a short from a stray bit of copper wire etc so, I individually wrapped the ends of each cable with electrical tape and copper EMI shielding tape (so there is no unshielded part at the ends of the cable and no possibility of unwanted wire contact. It works very well.

[mactec54]

At the risk of getting my head bitten off.... it's not super dangerous to run a spindle without ground being connected. That doesn't mean it is advisable though. It has the potential to be dangerous if there is a problem with the VFD.

I briefly tested my spindle without a ground connection before I installed the newer style outlets (which all have ground now). It worked fine. It wasn't that long ago that two pin outlets (with no ground) were the norm and within code in America. One of the vendors for Chinese VFDs and spindles has a YouTube video where they state that ground is optional.

A ground connection is a safety feature to prevent you getting shocked in the event there is a problem with your VFD. So.., if all else is well, it should never need to be used. I don't have that level of trust in Chinese VFDs though.... plus you might have an issue with your wiring...

Given the voltages involved and the potential consequences of a problem, however unlikely, it would be stupid not to connect to ground where a ground connection is available. It cost me less than $20 to install a new grounded 50amp 240v outlet. Surely it is better to fix the wiring problem instead of skipping a ground wire....

I found that the screw terminals on my Huanyang VFD to be far too close togeather for the gauge of wire needed for my spindle. I was worried about a short from a stray bit of copper wire etc so, I individually wrapped the ends of each cable with electrical tape and copper EMI shielding tape (so there is no unshielded part at the ends of the cable and no possibility of unwanted wire contact. It works very well.

You almost got it right, the Ground wire is not only need as a safety wire, far from it, without having a Ground wire, you can do some serious damage to your spindle / motor and sometimes the electronics near by

The Ground wire is used to drain the Rotor Voltage which is always there when running any 3 Ph motor with a VFD Drive, there is nowhere for this voltage to go, but through your machine or the operator, as many that are on the Zone have found out the hard way

Watching Chinese videos showing that the Ground is optional, is only for China as it is a way they use it to regulate there population they have no regulation regarding Grounding, that's also the reason a lot of electronics coming from China are unsafe to use, the Common Router Spindles being one of them that are sold without a Ground connection, the better quality router spindles usually have a Ground connection

[skrubol]

Rotor voltage? On many spindles the rotor isn't even grounded to the case, so I'm not sure how a ground wire would drain it. Plus, I believe any potential induced into the rotor (other than the couple volts from one side of the rotor to the other that generates the field for the motor) doesn't have any current behind it, it's just a static charge that could do damage to electronics (and I believe in some cases can damage bearings over time,) but is similar to walking across a rug in winter. It's not going to injure you, and certainly not going to damage a spindle.

I have gotten zapped by an ungrounded spindle before, but I'm pretty sure it was mostly due to an imbalance in the phases capacitively coupling a charge into the case.

[Al_The_Man]

For a AC induction motor, the rotor consists of a series of single turn shorted turn conductors the current induced in these turns is what causes the rotor to revolve, i.e. it tries to follow the rotating field.
In very old motors where the bars were made of copper and plates on the end of the rotor were soldered to the bars, if the motor were to overheat, it would fling the solder off and the result would be no rotation and almost zero current in the stator.
There is virtually no voltage present, just high current in the rotor.
The only damage I have seen caused to bearings by rotor leakage in a motor is for a ground occurring on a DC brushed motor.
Al.

.

[Eldon_Joh]

Rotor voltage? On many spindles the rotor isn't even grounded to the case, so I'm not sure how a ground wire would drain it. Plus, I believe any potential induced into the rotor (other than the couple volts from one side of the rotor to the other that generates the field for the motor) doesn't have any current behind it, it's just a static charge that could do damage to electronics (and I believe in some cases can damage bearings over time,) but is similar to walking across a rug in winter. It's not going to injure you, and certainly not going to damage a spindle.

I have gotten zapped by an ungrounded spindle before, but I'm pretty sure it was mostly due to an imbalance in the phases capacitively coupling a charge into the case.

The VFD output voltage is a pwm sine wave, but collectively all of the phases are delivering a modified sine wave at the pwm frequency. the waveform looks similar to this: https://i.imgur.com/lVSO17M.png

anyhow, there is capacitive coupling between the windings and the rotor and the current flows through the bearings or through the tool in the spindle to the rest of the machine.

if you have a wire shorted to the case of the spindle its also bad news. if the stator isn't grounded then yes you will shock yourself touching it, i zapped myself a number of times on a completely isolated ungrounded motor run from a 480v vfd.

[Goemon]

The VFD output voltage is a pwm sine wave, but collectively all of the phases are delivering a modified sine wave at the pwm frequency. the waveform looks similar to this: https://i.imgur.com/lVSO17M.png

anyhow, there is capacitive coupling between the windings and the rotor and the current flows through the bearings or through the tool in the spindle to the rest of the machine.

if you have a wire shorted to the case of the spindle its also bad news. if the stator isn't grounded then yes you will shock yourself touching it, i zapped myself a number of times on a completely isolated ungrounded motor run from a 480v vfd.

To me, the potential wiring problem is the bigger issue here. I wouldn't be happy to assume I could manage without a key safety feature when there is a probable wiring issue (which increases the chances of needing that safety feature).

Also, there is no need or benefit to leaving out the ground connection. I only tested my spindle briefly without ground because my only 240v outlet was the old ungrounded kind. if you have a ground pin, there is no reason not to use it.

I guess "danger" is a matter of perspective. The is the danger of breaking electronics through incorrect wiring which is different to the danger of getting shocked at high voltage. Still, I prefer to avoid both....

[reg.miller]

Hi, first of all, thanks for the heads up about the fourth terminal not being connected to anything. ( All part of the evil chinese plot to destroy western civilisation ).
It seems the default assumption is grounding through the mechanics.

I would have assumed if it was there it would be a neutral or gnd connection and I'm not sure how long it would have taken me to overcome that assumption.

There will be a ground path through the mechanics on most CNC machines ( apart from the cheapo MDF kits ) This means that connecting the fourth pin in the "correct" way will cause a ground loop. Be it "bonded" or not, there is a loop where currents can be induced and non zero "ground" potentials develop. Ground loops are bad practice.

Q1. Is there a case for electrically isolating the spindle where it is physically mounted to the chassis ?
Q2. Typical chinese VFDs are in a plastic enclosure, yet the fast switching IGBTs are a major source of EMI. Is a secondary, ventilated, metallic enclosure desirable?

Thx

[skrubol]

There will be a ground path through the mechanics on most CNC machines ( apart from the cheapo MDF kits ) This means that connecting the fourth pin in the "correct" way will cause a ground loop. Be it "bonded" or not, there is a loop where currents can be induced and non zero "ground" potentials develop. Ground loops are bad practice.

Q1. Is there a case for electrically isolating the spindle where it is physically mounted to the chassis ?
Q2. Typical chinese VFDs are in a plastic enclosure, yet the fast switching IGBTs are a major source of EMI. Is a secondary, ventilated, metallic enclosure desirable?

Thx

The ground path through most metal structured routers and mills is pretty poor. The thin layer of oil or grease on the ways or in the bearings doesn't conduct well. I wouldn't worry about a ground loop like this. A poor ground is generally much worse than a potential ground loop.

In the electronics realm, the switching speed of VFD IGBT's is very slow. EMI is a function of frequency as well as wire length. In a 5 GHz wifi module you need to worry about every mm. At 5kHz the equivalent would be 1 km. That's a bit of an exaggeration, as the power levels are very different between wifi and a VFD, but it still means it's generally only lengths of wire that emit much EMI (inductors/chokes, transformers and motor windings have long lengths of wire in them.) It's probably not necessary, but I would tend to try and isolate all of my high power stuff from signal stuff in my wiring enclosures, either by using a separate enclosure or putting some sort of shield inside the enclosure (piece of sheet metal, metal screening, or whatever.)

[Al_The_Man]

.

Q1. Is there a case for electrically isolating the spindle where it is physically mounted to the chassis ?
Thx

You should not rely on the mechanical mounting of the motor for earth grounding, a dedicated GND wire should be connected to the motor frame, It is also a good idea to earth ground any moving parts of a machine, such as a gantry etc, this is called equi-potential bonding of the machine and is the main way to reduce or eliminate any ground loops that may exist or have the potential of existing.
Al.

[reg.miller]

The ground path through most metal structured routers and mills is pretty poor. The thin layer of oil or grease on the ways or in the bearings doesn't conduct well. I wouldn't worry about a ground loop like this. A poor ground is generally much worse than a potential ground loop.

In the electronics realm, the switching speed of VFD IGBT's is very slow. EMI is a function of frequency as well as wire length. In a 5 GHz wifi module you need to worry about every mm. At 5kHz the equivalent would be 1 km. That's a bit of an exaggeration, as the power levels are very different between wifi and a VFD, but it still means it's generally only lengths of wire that emit much EMI (inductors/chokes, transformers and motor windings have long lengths of wire in them.) It's probably not necessary, but I would tend to try and isolate all of my high power stuff from signal stuff in my wiring enclosures, either by using a separate enclosure or putting some sort of shield inside the enclosure (piece of sheet metal, metal screening, or whatever.)

Thanks. What order of frequencies are used? The VFD has a max sine wave freq of 400Hz but this seems to be done by pwm techniques switching way faster. Presumably this involves some very fast transients.

BTW, I have been getting a 503 from cnczone for the past 5 days. This a.m. was the first time I have been able to get a proper response, but I don't see any announcement. Has anyone else been seeing this?

[Al_The_Man]

but I don't see any announcement. Has anyone else been seeing this?

Site was down.
Maintenance?
Al.

[reg.miller]

Site was down.
Maintenance?
Al.

I've started another thread about this, so as not to divert this one. Others have confirmed the issue .
https://www.cnczone.com/forums/forum...ml#post2208288

[Eldon_Joh]

Thanks. What order of frequencies are used? The VFD has a max sine wave freq of 400Hz but this seems to be done by pwm techniques switching way faster. Presumably this involves some very fast transients.

BTW, I have been getting a 503 from cnczone for the past 5 days. This a.m. was the first time I have been able to get a proper response, but I don't see any announcement. Has anyone else been seeing this?

click the imgur link in my prior post and you'll see just how fast the transients are.

skrubol is not correct suggesting that the switching speed of VFD IGBT's is very slow. the 5khz switching frequency is very slow, but, we don't really care about the 5khz switching frequency.

its the harmonic content of the square wave which causes the problems. the rising and falling edge of each square wave is a dv/dt on the order of 340vdc in maybe 1 uS corresponding to megahertz range.

[skrubol]

click the imgur link in my prior post and you'll see just how fast the transients are.

skrubol is not correct suggesting that the switching speed of VFD IGBT's is very slow. the 5khz switching frequency is very slow, but, we don't really care about the 5khz switching frequency.

its the harmonic content of the square wave which causes the problems. the rising and falling edge of each square wave is a dv/dt on the order of 340vdc in maybe 1 uS corresponding to megahertz range.

True I neglected the harmonics (which are significant.) 1 MHz is still a 300m wavelength though.

[reg.miller]

Thanks skrubol, that does help put it in perspective. Does that imply that a stepper motor ( which has comparable drive freq, and pwm switching from current limitting driver chips ) could be connected with flat ribbon cable without giving off problematic amounts of EMI?

I'm thinking of using flat cables of a forklift and grounding every other conductor. ie four phase wires to the motor and four (five ) ground wires.

Max length will be under 2m long.

BTW MW radio is that kind of order 300kHz- 3MHz , but I don't need a 300m antenna on my radio.

[Al_The_Man]

Does that imply that a stepper motor ( which has comparable drive freq, and pwm switching from current limitting driver chips ) could be connected with flat ribbon cable without giving off problematic amounts of EMI?
.

You can also get shielded ribbon cable!
Also many of the spurious E-Stops that are seen here and elsewhere when using Mach etc, and put down to EMI etc, and often the attempted cure is the Mach suggested de-bounce circuit, as I have posted many times here before, the way I cure it is to ensure all DC Supply commons are referenced to the Star point ground plate. As of course, the PC power supply common already is.
Al..

[Eldon_Joh]

Thanks skrubol, that does help put it in perspective. Does that imply that a stepper motor ( which has comparable drive freq, and pwm switching from current limitting driver chips ) could be connected with flat ribbon cable without giving off problematic amounts of EMI?

I'm thinking of using flat cables of a forklift and grounding every other conductor. ie four phase wires to the motor and four (five ) ground wires.

.

but the voltage they run at is typically one tenth that of a 240v vfd, (34vdc) so you're dealing with something that is only able to generate 1% of the radiated energy.

[skrubol]

Thanks skrubol, that does help put it in perspective. Does that imply that a stepper motor ( which has comparable drive freq, and pwm switching from current limitting driver chips ) could be connected with flat ribbon cable without giving off problematic amounts of EMI?

I'm thinking of using flat cables of a forklift and grounding every other conductor. ie four phase wires to the motor and four (five ) ground wires.

Max length will be under 2m long.

BTW MW radio is that kind of order 300kHz- 3MHz , but I don't need a 300m antenna on my radio.

As long as the ribbon can handle the current you're probably fine. I've run small steppers this way, and even tried having USB signals on the same ribbon (that didn't work out well. USB would drop out occasionally when the stepper was running.) You're probably actually better off running without a ground between the the two wires for each winding in the stepper, as they should be differential and the closer they are to each other the better their fields cancel out. A ground or 2 between each pair wouldn't hurt though, and a ground on the outside of the ribbon would act as a bit of a shield.
An ideal simple antenna is 1/4 wavelength for a monopole and 1/2 wavelength for a dipole. Shorter work, but usually some tricks are done to make the antenna behave like it's longer.

[reg.miller]

Thanks skrubol. Good point about not putting a ground between the + and - wires to each motor phase. Your comments suggest it's at least worth wiring up this method to see if it gives good enough suppression.

Since this thread is principally about VFD grounding, can anyone comment on the need to box these chinese VFDs in a metal grounded box. I would imaging the circuitry would be a source of EMI and the plastic box does not even have any sign of conductive paint. It's just plastic. I'm going to box in the stack of stepper driver boards and the ARM controller board but does the VFD need boxing as well?

TIA.