Try some ferrite chokes on the USB cable.
Well, after replacing my relay board and my digital to analog board and a bunch more testing, the problem is a "noise" related problem. The new relay board did nothing nor did the digital to analog board.
What did help was changing the "debounce" setting for E-stop in the ESS configuration settings. That had an immediate effect and improvement. The noise is, from what I can determine, causing false signals to be interpreted by the I-Pac USB keyboard encoder.
The I-Pac is in a separate enclosure (about 4 feed from the main enclosure) connected to the PC via a USB cable. I also have a cable running from the main enclosure to that enclosure with other I/O to power LEDs, and handle encoders and the MPG.
If I disconnect that cable, everything runs fine with no false signals. When I connect that cable, I start getting random signals through the I-PAC. I suspect that the "noise" is coming through on that cable somehow.
Determining the exact solution is a mystery to me other than just trial and error.
Chris
Try some ferrite chokes on the USB cable.
Gerry
UCCNC 2017 Screenset
[URL]http://www.thecncwoodworker.com/2017.html[/URL]
Mach3 2010 Screenset
[URL]http://www.thecncwoodworker.com/2010.html[/URL]
JointCAM - CNC Dovetails & Box Joints
[URL]http://www.g-forcecnc.com/jointcam.html[/URL]
(Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)
Hi Gerry,
Already on the cable.
Chris
Mactec54
Hmm, this is very strange and beyond my understanding....
The other cabinet is the operator panel with buttons, switches, encoders and MPG. There is one circuit board holding all the components. There are TWO circuits on the circuit board.
#1 circuit is the IPAC keyboard encoder which is powered by the USB (Ground or 0V and +5 V) and all signals pass through to the PC.
#2 circuit is connected to the Smooth Stepper board and is powered by the power supply within the CNC cabinet (0 Volts and 5 Volts).
The grounds are NOT common to each other but that is where things get really strange.....
If I measure voltage between 0 Volts on the CNC Cabinet and 0 Volts on the IPAC, I get a 4 volt differential. If I measure voltage between 0 Volts on the CNC side and to a logic 1 on the IPAC, I measure 9 Volts. Makes sense, 4 volt differential between grounds so that would add up to 9 volts.
As mentioned, each circuit has its own 0 volt and they are not connected.
The CNC Cabinet is plugged into the same AC outlet as the Laptop computer.
I can almost understand what is happening, and can almost grasp why this is problem, but it is certainly exceeding my ability to resolve the problem if it is one.
As for double shielded USB cable, I believe it is, however, I have so many USB cables laying around and they all look the same, I am not sure anymore. I will gladly purchase another one to make sure - do you have a source?
Totally confused Chris -- looking forward to some electronics schooling on this one!
Ferrite choke is molded onto the cable.
I will try to take photos and send them.
This is a photo of the front of the operator panel.
THis photo shows the inside of the CNC electrical cabinet.
This photo shows what is behind the operator panel. There is a LOT of wires in here. You can see the IPAC board on the left, that is connected to the laptop via a USB cable. On the right, you can see a basic DB25 pin BOB which is connected to the CNC cabinet via another BOB just like it.
Chris D
It looks like you don't have much room in your main cabinet, in the operator panel if you can twist all the pairs of wires it will no only tidy it up it will help to cancel any noise if there is any at that end
The main cabinet is a major noise machine the VFD Drive and power supplies should have a EMI filter on the input side, is that a HY VFD and do you have the R and T input terminals connected, the output to the spindle is this a shielded and the shield terminated to Ground at both ends
The VFD low voltage signal wires need to be Twisted or a shielded cable needs to be used
Have you checked the spindle 4th Pin to see if it is Grounded to the spindle body, you have many things that can be the problem, it is hard to know where to start
You could try to add an aluminum shield over the top of the Breakout Boards Etc and down the side next to the VFD this would be fastened to the Grounded Cabinet this must not come in contact with any wires or connections that could cause a short
Are all Ground going to one Star Point Power supplies VFD Stepper Drives Etc
Mactec54
See answers below...
Correct, the main cabinet is only 18" x 18", the maximum size I can bend sheet metal. Silicone wires are hard to twist but I could give it a try or rewire with different wires.
Does each power supply need an EMI Filter? Will that filter take care of the noise created by the VFD?
Yes, it is a HY brand VFD, and no, currently R and S. The output cable to the spindle is shielded and was grounded at both ends (however, I need to verify it still is).
Will try replacing VFD logic wires with shielded.
No, the 4 pin isn't connect to the body and the cover of the motor is plastic. This is why I grounded the spindle external through the motor bracket. This is verified back to the outlet as being grounding.
When referring to ground and star point, are you referring to earth ground for mains or zero volts for DC? Mains is common ground, DC zero volts is NOT connected to earth ground.
For further understanding of the specific source of the noise....
I attached my scope to the input for the Reset button which goes through the I-PAC. Voltage is stable with very little "static or ripple". When starting the spindle you can see as much as 3 volts of spikes/drops, once the speed comes up, the spikes/drops are minimized or gone.
Chris
Normally yes for the EMI filters but if you don't have room concentrate on the VFD Drive input EMI Filter
For the VFD Logic a simple cat6 piece of cable will do the job
How have you Grounded the spindle externally
R and S connections work but R and T are more reliable on these HY VFD Drives
Star Ground is a single point where all Main Power supply 240v Ground /Earth connection connect to, this can be a Bus or a single mounting stud
How did you Ground your Shields
Here are some snips on Grounding
Mactec54
A ground wire, connected to earth, runs with the shielded cable to the spindle clamp. I pull the ground wire out of the cable before the connector at the motor.
Just tried the R-T connection, it made no difference.
Yes, my AC grounds are common to what would be referred to as a buss bar.
Shields wires are soldered to the connector which contacts the enclosure which is grounded. The spindle shield also is grounded on both ends, attaching the shield to the same ground wire I use for grounding the spindle bracket.
While trying the R-T connections, I observed the scope some more to learn more about the noise. The noise signals are about 2 to 2.5 volts, very consistent and I would guess they are from the output side of the VFD as they come and go with that running or not running. The duration of the spilke is about 1 microsecond give or take a few nanos. Based on that, I configured the ESS input delays for 1 micro second and the problems went away. I set the delays back to o and the problems came back. I set the delays to 3 microseconds and let it run for about 20 minutes while watching the scope and looking for problems on the machine. All ran perfectly good.
I also determined that something is hinky with one of my rotary encoder switches. Somehow that is "leaking" voltage onto the PCB that it is attached to. This area of the PCB is open and not connected to anything, But knowing that encoder switch should not do that and fearing that it is leaking to the 0 volts area as well, my next step is to replace that. Hopefully will have time tomorrow or Wednesday.
Chris D.
Wow you have a lot of electronic bits in there.....
R and T will not make any notable difference it just a more reliable connection for these HY VFD Drives
You can not attach a Ground wire to the shield wire, soldering is not used for shield Grounding this is a bad Idea so I believe you have a lot of little things to look at
Mactec54
Soldering of ground shields has been done for many years, military used on air craft and .missiles and that is where I first used it over 50 years ago. The kids today do not know how to solder. Some of the aircraft I worked on was from WW2, some of the B-52H bombers I worked on are still being flown today on missions.
1000x750 Workbee CNC - Mach4 - PMDX USB - Windows 10 Pro
I did a lot of the same work and that is not how it was done, I have worked on those same war birds and have had many flights in them also A4 Skyhawks and FA/18 Hornet was the last one I worked on and where my favorite planes as far as fighter jets go, I even have my name on some flight records
No inspectors would pass shoddy work like that on aircraft even back then
Last edited by ger21; 05-20-2019 at 08:52 PM.
Mactec54
WIth shielding on a cable like the spindle cable, is that there to contain the radiated noise only? Does it help at all with conductive noise?
Chris D
A shield will protect against all induced noises, will be totally worthless if not terminated correctly,Cable Shields terminated at one end provide only electrostatic protection against low-frequency, capacitive coupled interference and emission
Use a cable suited to the application. Cables that experience repeated flexing usually use a spirally wrapped shield rather than a braid. Avoid foil-only shielding on flex cables since continuous flexing can tear the foil. 3. Make sure the equipment that the cable is connected is properly grounded. Use an earth ground wherever possible and check the connection between the ground point and the equipment. Eliminating noise depends on a low resistance path to ground. 4. Most connector designs allow full 360° termination of the shield. Make sure the connector offers shielding effectiveness equal to that of the cable. For example, many common connectors are offered with metal-coated plastic, cast zinc, or aluminum backshells. Avoid both over specifying and paying for more than you need or underspecifying and getting poor shielding performance.A shielded system is only as good as its weakest component. A high-quality cable is defeated by a low-quality connector. Similarly, a great connector can’t do anything to improve a poor cable.
Mactec54
All good information, thank you.
I am still trying to determine what the source of my "noise" is. It could be radiated or it could be conductive. Assuming that the shield on the spindle cable is effective, then I would be looking elsewhere for the source of the noise.
I could spend days and tons of money shielding all my cables differently, only to discover that all the noise problems I have are conductive.
In other words, I am trying to avoid a shotgun approach by changing everything.