DRO is driving me crazy.

I have always had problems with the cheap WARCO two axis DRO on my WM 16 mill but of late it has become much worse - like unusable! Setting a datum point using x0 - y0 and without even moving the x and y axis I can - with the machine plugged in - press "ON" and "OFF" repeatedly while watching the readout. Invariably and within the space of six cycles either X or Y will loose its zero reading and display a random measurement up to 30mm or so out of kilter. Sometimes BOTH axis change at the same time. One thing remains constant though, the problem only manifests itself when pressing the "OFF" button - never the "ON".

I suspected some sort of "noise" problem so I have dispensed with the original USB mini B leads and fitted heavier duty shielded leads keeping them as short as possible but the problem persists. The scales are squeaky clean and the calliper readouts, unlike the main display, remain accurate. I tried connecting an earth lead to the central screw on the control panel marked with an earth symbol and the other end to the mill casting and that makes no difference either. Finally I relocated the readout a couple of feet away from the machine - still the problem is there. Having googled at length I can see that this is quite a common problem with these budget systems but I appreciate that with all the variables a common fix is probably not possible. However, I live in hope! I would add that the electrical installation in my workshop was carried out by a pro electrician this year and is certificated.

That is as far as I can go with my very limited electrical knowledge unless some kind soul can advise me. In the meantime I am back to using the graduated hand wheels.

Rik

Not 100% clear what you mean. I assume it's the machine not the display you are switching on and off?

By 'calipers' do you mean the displays on the read heads keep the correct reading?

The problem could be spikes on the mains power supply, it may be that the supply filter on the mill needs replacing.

I fixed a similar problem with my homebrew setup by fitting 10uf capacitors in the battery compartments of my read heads.

Neil

Thats why I took mine off and threw them away, fitted decent glass scales and have not had any problems since.

Not sure what remote display you are using but I did find the magnetic mounts on mine attracted fine cast iron dust which gets inside the box.

J

Neil - Its the "ON" and "OFF" button on the milling machine I am switching on and of.

"do you mean the displays on the read heads keep the correct reading?"

Exactly so.

To limit spike problems the milling machine takes its juice through a surge protector.

Rik

Rik, Have you temporarily done the same tests with the surge protector removed?

If you fitted the surge protector in order to cure the problem originally (and it never actually helped) then its not likely to help now. Can you give a little more information about your setup to help diagnose the problem. I have used iGaging type displays and found them completely trouble free so cheap/crude does not automatically mean they are no use.

Like Jason I found the display boxes are transparent to swarf ingress (mainly brass in my case) as the buttons are not sealed to the front panel. I glued a sheet of clear PVC over the front face and have not had any problems in the last 3 years or so. Never seen any glitches either.

Ian

I wouldn't place too much faith in a surge protector myself to filter noise generated by the milling machine getting into the mains. The problem may well be some degradation in the speed control in the mill, for example a suppression capacitor failing so more noise escapes. Or even the brushes? A better bet may be to improve the filtering of the power supply to the scales/readout, possibly by simply a larger smoothing capacitor.

Same results with or without surge protector.

Disconnect USB leads from read heads but NOT from main display and motor start/stop produces nil errors.

Reconnect USB leads and errors resume.

Rik

Hello Rik,

I would separate the dro power source from the machine supply. Feed it from a different socket to the machine, a source that might not be affected by the mill.

It may be worth changing the dro power supply to one that is more resistant to mains borne interference.

Regards,

Michael

You could have an earth loop. Could you use a double insulated PSU that has no earth connection for the display?

Neil

Michael - The DRO display and read heads are powered by lithium cells and the cells are new.

Rik

Edited By Rik Shaw on 26/12/2015 21:39:49

How is the DRO powered?

One thing I would try is to temporarily unmount the display box and put it on the swarf tray near the tailstock end of the bed with the leads to the scales kept away from the lathe mains wiring. Its possible that the existing route and layout of the DRO wiring form a 'loop' that picks up the mains spikes.

If the power to the DRO scale readers are derived from the display box (which in turn gets its supply from a 'wallwart' than as suggested use a different mains outlet or if possible power it temporarily from a battery.

If the DRO setup is a standard Warco option then surely they should have some responsibility to provide reliable operation.

Ian P

Hi Rik,

I had the same symptoms and assumed it was an electrical problem. It was actually mechanical.

My display was mounted on top of the control unit. Switching off causes the contactor inside the box to drop out with a healthy thump. The vibration shook the usb cables sufficient to cause an intermittent connection. I found my plugs & sockets to be slightly loose at the display end, but not obviously so.

Your set-up looks to be less vulnerable to vibration than mine but it might be worth adding some form of strain relief to the cables to ensure that the plugs can't wobble.

Hope it's as easy as that!

Regards,

Dave

Rik, I've just re-read your original question and noted that you have replaced the original cable with something better.

Whilst I have no knowledge of this Warco unit I would be gobsmacked if the signals between the DRO elements is really a serial protocol conforming to 'USB'. It might use USB type connectors but the pinout and wiring details could be completely different. What you think of as being the 'screen' might actually be a signal line!

Can you revert to the supplied cables to do some diagnostic tests.

Ian P

If I read the original post correctly the displays lose their marbles when the motor is cycled on and off? Presumably this is repeatable? That being the case it is unlikely to be mains borne interference, which by its very nature is random. I also understand that the DRO and displays are independently powered by batteries? I'd suspect either airborne EM interference from the motor drive or the DRO and/or its shielding being connected to the machine metal work and upsetting the differential USB signals via injected noise.

Andrew

PS: Just seen the post by Ian Phillips - makes good sense. The faster USB standards are fairly low impedance, 90ohms differential, so should be fairly robust against interference.

Edited By Andrew Johnston on 26/12/2015 22:06:23

Ian - I have retained the original supplied cables, As I have pointed out I am getting the same duff results no matter which set I use.

Andrew - the cycling is also random. Sometimes the error occurs after just one cycle and rarely it can occur after anything up to twelve cycles.

Rik

Edited By Rik Shaw on 26/12/2015 22:17:51

Hi Rik i had a similar type of readout some years ago that died because of condensation getting inside it with all this damp weather it could cause a problem with it. As a aside i left a mp3 player in the workshop for a couple of days and that suffered the same problem, bought inside and after a couple of days it worked again.

Graham - Thanks but I have a warm, dry insulated workshop and condensation is thankfully something I do not have to suffer.

Rik

Hi Rik,
I think the source of the problem is electrical noise generated when the NVR switch opens. (The NVR switch is the unit with the start and top buttons.) I would first try fitting a spark quencher across the contacts of the NVR switch. Such as this item My next step would be to fit a mains filter (such as this but with a suitable current rating for your machine.) if possible inside the control box of the mill or if there is not room then in a box very close to where the mains cable goes into the control box. There will probably be an earth loop formed by the fact the metalwork of the scales is connected together via the metalwork of the machine and the other end connected in the display unit. You could try just connecting one scale at a time to the display and see if the fault still occurs. If the fault does not occur with just one scale connected then make insulated mountings for one scale so it's metalwork is not connected to the machine. Although the connection to the scales uses USB cables the signals do not use USB protocol. They probably use a protocol called BIN6 which is a serial protocol. the data is on one wire in the usb cable and the clock signal on the other. I do not think the problem is caused by corruption of the data on the link. I think it is the dislplay unit electronics that is being effected.

Les.

Les - I was hoping you would respond - thank you. I intend following your advice through as my gut feeling also tells me that the problem is with electrical noise. Also thanks to all who have offered help on here. I'll get by with a little help from my friends

Rik

Not claiming to be an expert in this matter at all. Only to say one of my friends had a similar problem. Ian took the lot off his mill and set it up on his wooden work bench no trouble at all, put it back on the mill and the problem was back again. So he electrically insulated each component of the digital read out from the metal work on the mill using paper, every thing now works fine. Could this be the answer ? John