[PlanetCNC]

Controller must be powered with external power supply. 12V, at least 600mA is recommended. Actual consumption depends on what is connected.
Without external power, controller will turn on but because of voltage drop USB connection will not work reliably.

[basilevscom]

I use 12V 1.5A power supply for controller, LPH-18-12 Mean Well.
But if the controller can work both power ways, how do I determine if the external power supply is working normally?
The power led will light up in both cases...

If the external power supply of the controller is connected (+12V), the current via the USB cable (+5V) must be zero?
Can I disconnect pin 1 in cable connector (USB Power +5V)?

[basilevscom]

Controller must be powered with external power supply. 12V, at least 600mA is recommended. Actual consumption depends on what is connected.
Without external power, controller will turn on but because of voltage drop USB connection will not work reliably.

Maybe I need to start with replacing the power supply and not with replacing the cable or computer?

[PlanetCNC]

USB cable without 5V can damage controller is external power is not connected.
You can measure voltage on one of 5V pins if you have 5V then power supply is good. At least in electrical way, grounding can still cause issues.

You mentioned you use shielded cables. How and where are shields connected?

[basilevscom]

Once again I will ask a question - when working with an external power source, the controller consumes current via the USB+5V circuit?

[blowlamp]

Once again I will ask a question - when working with an external power source, the controller consumes current via the USB+5V circuit?

I'd say that question has been answered here: USB cable without 5V can damage controller if external power is not connected.

So if the controller is powered by an external power source FIRST, then it would be OK to connect an unpowered USB connection, although slightly risky. This tells me that power is not required via USB, only data.

[PlanetCNC]

No. In ideal situation current should be zero.

[basilevscom]

No. In ideal situation current should be zero.

Thank you, I understand everything, I will check the voltage and current

[basilevscom]

The solution is found, the work is stable (preparing a full report with photos)!

But first a few words about power - reading the manual:
The controller should be powered with an external power supply.
Mk3/4 - 4 Axis controller hardware requires 8 - 24V DC supply.
Power supply should be at least 200mA.
(I use 24V 1500mA power supply)

STEP ONE.
To check the correct controller's power you need to measure the voltage between the pins "+5V" and "GND" of any "AXIS" connector or "OUTPUT" connector. If external power is bad (or not connected) and you power is via USB only - the voltage is 4.5-4.8V. If external power is good - the voltage is 5.0V.

I think the controller's power supply circuit is like this:

(Why is this important to me?
I used my original I/O board that I specifically designed for controller MK3/4.
I use the controller's power +5V to power my board. 50mA - is not much, but I think that it could affect for controller's power supply.
I checked - the DC/DC converter has a large current reserve and my board does not affect the power supply and stable operation of the controller.)

[basilevscom]

The problem of unstable operation and Controller Disconnected error was in the strong noise from the frequency ?onverter and spindle.
I managed to reduce the noise and achieve stable operation without errors. I installed EMF suppression Filters on the cable to the spindle.
I will tell you more about this...

[theRat]

The problem of unstable operation and Controller Disconnected error was in the strong noise from the frequency ?onverter and spindle.
I managed to reduce the noise and achieve stable operation without errors. I installed EMF suppression Filters on the cable to the spindle.
I will tell you more about this...

So are you now saying that it was a noise problem caused by the VFD? Noise from VFD's is a big problem and should never be discounted.

[basilevscom]

So are you now saying that it was a noise problem caused by the VFD? Noise from VFD's is a big problem and should never be discounted.

Yes, I say! But I was offered to solve it by grounding, shielded wires and others. Someone told me to put a filter on the cable? You wrote earlier:
There are two ways to improve noise immunity. 1) Stop it being radiated at its source; 2) stop it being absorbed elsewhere.... Also keep the USB cable as far away from the spindle cable (and not running parallel).
Where is the advice to put the filter on the cable?)))

[theRat]

Where is the advice to put the filter on the cable?)))

Sorry if I didn't say that, but even all the cheap as **** Chinese machines that I have seen photos of have a line filter already built in. All manuals/articles for VFD's I have read also say they are mandatory. So I suspect it was a safe assumption that you already had one installed.

[basilevscom]

Dear TheRat! Please, show me a photo of a Chinese machine with a filter)))
(I don't have a Chinese machine. I make it myself. If you are a great specialist, then help me with specific practical advice if possible)

[theRat]

This is the internals of what is left of my original Chinese controller (including couterfeit PlanetCNC board). The filter was directly installed in the wiring between the mains and the VFD as per the VFD manufacturers guidance. The VFD used to sit in the space at the bottom left. The filter being the little chrome box with wire going into each end.
I am not an expert by any means, but I read a lot (and am sceptical of people like me claiming to be experts).

[basilevscom]

Many experts recommend grounding - it does not help.
Additional filter for frequency converters for input mains 220V do not helps for me too.
Power is good, the cables are shielded, all the details are correct, but there is still an error...

Solution:
STEP TWO - Clamp Filters (Ferrite Core with Case)
I use inverter HUANYANG HY02D223B (2.2 kW 220V)
I installed one filter at the beginning of the cable next to the inverter and one filter at the end of the cable next to the spindle (I used a shrink tube).
Additionally installed one filter on the USB cable (for reliability).
You can use filters TDK ( for example ZCAT1518-0730-BK) or analog.
One filter on the spindle cable is enough to fix the problem.
(I can use any usb cable 1.2 meters long, even the one that always didn't work.)

As a result, there were problems with the power supply of the controller and with noise from the inverter.
These two decisions helped me, and I hope it will help you!

[basilevscom]

This is the internals of what is left of my original Chinese controller (including couterfeit PlanetCNC board). The filter was directly installed in the wiring between the mains and the VFD as per the VFD manufacturers guidance. The VFD used to sit in the space at the bottom left. The filter being the little chrome box with wire going into each end.
I am not an expert by any means, but I read a lot (and am sceptical of people like me claiming to be experts).

This filter is important but will not fix the problem!
This filter use usually for low frequency and to protect the input mains from machine noise.
.

[theRat]

It looks like you have a shielded VFD cable, but are you certain that it is grounded to the spindle? Asking because it looks like a Chinese made spindle and most of them don't have the ground connected (ie. pin 4 on the connector has no connection to the spindle case even if it is connected in the cable.

[basilevscom]

I use China spindle by "G-Penny machines"
https://aliexpress.com/item/32803123786.html



Pin 4 connected to "ground" inverter, I checked it out
The shield of wire is connected to "ground" on the inverter side only.

[PlanetCNC]

Shield should not be connected on both sides to prevent ground loops.