UC100 error lost connection

[cncdrive]

As I described earlier if you running a high power large and electrically noisy machine and or if you working in industrial environment then using USB for motion control is not recommended. USB is generally for hobby use. There is no cable which will solve that the USB is ground referenced high speed communication which is always sensitive to electric interferences.
You should use an ethernet controller if the machine works in high cost materials and so then the loss from a lost connection or USB issue is high.
The price difference between our ethernet controllers and USB controller is low, for example the UC100 is 80 Euros and a UC400ETH is 110 Euros.

[Robertspark]

Or maybe the cable suffered mechanical damage or had somekind of unnoticable manufacturing error?!

It looks beautiful and undamaged. I took it with me to the shop where I bought the replacement and he tested it with some sort of box with multiple plugs -- I guess to see that the ground was working and the wires were unbroken. I think it could be some kind of manufacturing error -- it would be a good topic for an electrical engineering student's paper.

I agree with cncdrive the way to go is ethernet .... It's all I recommend.

I suspect as you've noted that they just tested the continuity.
To test a cable you really need a spectrum analyser at one end and a signal generator at the other end with an imposed load of the cable to test a member of frequencies to see the signal degradation in dB
Because the cables are short I suspect that the spectrum analyser would need to be quite sensitive.

An oscilloscope set up with two probes calibrated with a common signal for zero difference between the inputs over a wide spectrum, then apply the the input frequencies and measure the difference between the two probes, again using an imposed resistive load.

Then apply a external imposed noise signal via a calibrated signal generator and measure the impact of external imposed electrical noise

All a bit complex and lab environmentish for most places I suspect.... For a $10 cable.

Usb is fine for a mouse or keyboard but damn it gave me agro when I started out with the usb Smooth Stepper . Never again!

I avoid wireless for similar reasons (but I am running a plasma cutter and two vfd air compressors)

Also consider your mains power supply earthing arrangement, the distance from your earth source as some inductance and noise may be present or induced into the earth over long distances from the earth source

Earthing should be done as a star network topology,

But they are normally wired as a radial (daisy chain).... Not good but acceptable... Most won't have a problem.

Just don't ever wire them as a ring as that is where your ground loops come from
https://www.researchgate.net/profile/Mj ... ystems.png

[Kevin Johnson]

As I described earlier if you running a high power large and electrically noisy machine and or if you working in industrial environment then using USB for motion control is not recommended. USB is generally for hobby use. There is no cable which will solve that the USB is ground referenced high speed communication which is always sensitive to electric interferences.
You should use an ethernet controller if the machine works in high cost materials and so then the loss from a lost connection or USB issue is high.
The price difference between our ethernet controllers and USB controller is low, for example the UC100 is 80 Euros and a UC400ETH is 110 Euros.

You might not be aware that the importer of the K1212 is the one that recommended your product. I would have gladly used whatever system he recommended. I even traveled from Tampa to Chicago to meet with him in person at his company to ensure that the actual machine being delivered to me would run with Mach 3 and the UC100. I was there when the freight company loaded it -- I know about demonstrating one machine and shipping another different one. That was an expensive trip. Then I flew back to meet and accept the machine at the freight terminal in Tampa.

Perhaps if the machine had been placed in the same building as my CNC plasma cutter (with other heavy duty machines, compressors etc.) then the issue with USB interference would have shown up -- but I am thinking not. The system worked fine for five years and still works fine with the original USB cable. I have some server grade computers with raid hard drive arrays that should avoid the original failure point but I have not had time to set them up.

[Kevin Johnson]

All a bit complex and lab environmentish for most places I suspect.... For a $10 cable. ...

A bit off-topic but I did go on a snipe-hunt for an issue with my CNC plasma cutter. The gantry x-axis is driven by a pair of stepper motors. They were randomly getting out of synch with each other.

So...

New computers, new power supplies, new driver boards, new stepper motors, new gear racks and drive gears. The errors persisted. So, thousands of dollars spent and wasted.

Finally I obtained the specification sheet for the Molex connectors on the shielded cables to the motors. I read that they were specced to be disconnected and reconnected a maximum of 30 times. I would rebuild the bushings for the drive system at least twice a year and have been running the machine for 15 years so I greatly exceeded that. New pins costing about $10 solved that problem. When the gantry moved it would slightly tug on the cables and cause data errors at the pins.

I called the manufacturer of the machine and they were not aware that there was a limit on the plug disconnects.

I complain but the gentleman across the street from me had a waterjet. When something went wrong with that a technician would come out and the visit alone would run over a thousand dollars.

[Robertspark]

Wow, smallest things, I would struggle to find those problems (guess it's another of life's million things to catch you out)

[Kevin Johnson]

Wow, smallest things, I would struggle to find those problems (guess it's another of life's million things to catch you out)

When I worked in Detroit with my Grandfather (British, BTW), he rented storage space in the lab to a retired pyrometer technician, Clyde.

Clyde started working on ovens at the River Rouge Ford Plant in the 1920s. He told me that his professor had a final exam for the students where he would deliberately introduce some sort of error in a device and it was up to them to locate it and repair it. Very interesting chap.

[Kevin Johnson]

Today i spoke with a relative who is an engineer and has worked on medical imaging devices and he said that he had problems during prototype development with cables. He said the moulding process on the ends can cause issues so that was the first thing he tried swapping out.

Live and learn!