Weird noise on travel and losing position

Denford router driven by Mach 3, every now and then the machine loses position and there is a knocking noise on each of the three axis which s mostnotable on the X axis. Here is the noise https://lozharroplures.co.uk/my_images/Xaxis.m4a

I am pretty sure that it isn't a mechanical problem as all axis are affected. The machine can run for weeks with no problem even running non stop for 20 hours. Rebooting the PC and /or the control box doesn't help. The spindle is controlled with a VFD plugin.

https://lozharroplures.co.uk/my_images/1.jpg

Any ideas anyone?

I'm not exactly the most experienced here but sounds like a stepper skipping a step to me.

What motors are you using, what drivers, what PSU, and what are the current and microstepping settings on your driver's?

Does it happen when you're doing a specific type of move in particular?

There are many things it could be and unfortunately, it's a process of elimination which starts with mechanical first. Then a careful examination of electrical connections and wires.
If nothing shows after this then next I would look to the PC. It's not uncommon for PSU in the PC to start to fail and drop out but for fractions of a second which wouldn't be noticed in normal use but for CNC machine and it's critical timings this is a lifetime and would cause this exact thing to happen.
Likewise if using a parallel port card then it could be that dropping out or being interrupted by other software like Virus checkers etc. But I would expect that to show up sooner and more often so I would tip it's a PSU type issue.

After this, we get into the weird and wonderful world of noise and interferences from outside factors like Next doors radio or the Mrs running hair dryer, tumble dryer, etc.
So check the earth to the machine frame, which really should have been done when checking out the wiring etc earlier.

If you go thru the process of elimination the fault should show it's self eventually. I'd tip PC if it's not a mechanical issue.

Thanks that's given me an idea where to start looking.

Quote:

---

Originally Posted by AndyUK

I'm not exactly the most experienced here but sounds like a stepper skipping a step to me.

What motors are you using, what drivers, what PSU, and what are the current and microstepping settings on your driver's?

Does it happen when you're doing a specific type of move in particular?

---

Andy yes it does sound like missed steps but chances of that happening on all axis are very remote unless it's something affecting the pulses which would then affect all Axis.

If it was a new build I'd also say to look at the motor tuning as it's common for people to over tune and use the same settings on all axis. But he's been using the machine for good while now so I doubt it's that either.

Quote:

---

Originally Posted by ngwagwa

Thanks that's given me an idea where to start looking.

---

Replace the Keyboard first, I've had keyboards do all sorts of crazy things like this and it's easy to do.

Well tried a different PSU, nothing else running on the PC. Will build another PC tomorrow to try and rule that out.

Quote:

---

Originally Posted by ngwagwa

Well tried a different PSU, nothing else running on the PC. Will build another PC tomorrow to try and rule that out.

---

Have you checked the mechanicals.? Don't assume just because there's no play etc check them properly for binding because it doesn't take much on these low powered machines.

Quote:

---

Originally Posted by JAZZCNC

Have you checked the mechanicals.? Don't assume just because there's no play etc check them properly for binding because it doesn't take much on these low powered machines.

---

Will do in the morning, thanks.

Well no solution found today.

I am pretty certain it isn't a mechanical issue. I did discover I had a faulty limit switch on the X axis home and thus replaced it (in configuration I saw the switch flashing on and off when the machine was well away from home).

The workshop where the machine is located has its own electricity supply so I think outside interference can be excluded?

The parallel port I am using is a StarTech 2S1P PCI Serial Parallel Combo Card and this was replaced with a new one to no avail.

One thing I noticed was after changing the parallel port card and powering the PC and router bakck on I jogged the router in X,Y and Z almost full travel several times without the problem showing itself, then I sent the machine home after which the problem noise was back again. Maybe coincidencebut it makes me wonder if something is on it's way out and as it gets hotter it starts failing?

I am now putting an image of the working PC that was taken a few months ago onto a spare PC to test on Thursday - at least now I have a parallel port card to use in it.

Quote:

---

Originally Posted by ngwagwa

Well no solution found today.

Maybe coincidencebut it makes me wonder if something is on it's way out and as it gets hotter it starts failing?

I.

---

Check the cooling fan on the computer chip is clean and working properly

Phill

Cheers Phil, fan is in fine fettle - I was thinking it maybe something in the control box that is getting warm.

So, after reading on the Denford forum that the leadscrew nuts need cleaning every now and then I spent almost all of yesterday cleaning the nuts (without removing them from the lead screws). They were pretty well packed solid with wood/mdf dust and after going through several cans of various degreasers, WD40, PCB cleaner, brake cleaner etc plus a bunch of wooded tooth picks I think I got them prettty clean. A liberal spraying of silcone lubrication and the weird noise had mostly gone so I left the router traversing X,Y and Z for a couple of hours.
Turned on the machine this morning and all was sweet jogging along to the soft limits several times so set up the datums for the work and set ti cutting. First piece finished n problem but after G55 the position had been lost in X by about 3mm.
Swapped out the BOB and, there is no other way to describe it, but jogging in all axis the machine sounded sweeter, sort of less strained. Anyway recut the second part followed by the next operation (about an hour each part) with no problem. Have left the machine on the next operation which should finish about 8am in the morning so hopefully all will be well.
So while the gumed up nuts weren't helping changing the BoB appears to haev been the solution, it may have been a lose wire or a faulty board, don't care as long as it is sorted. Will see in the morning.