What's up with this DRO?

[PaulL]

Ok, I'm fuddled. I haven't used the DRO on the lathe for anything precise since I transfered it over. Today, wanting to hit a measurement I put in my current diameter, and it set the display to a near, but not the same number. And I don't means the bounds of precision, I mean a few thou out. Video attached demonstrating. Starts at .500, enter .500, and see it jump to .4988. I can wind the axis back to .500, through a few stops so it's not an encoder precision issue.

Any clues?

 

[DavidR8]

Does it do the same thing everywhere along the length of the scale?

 

[RobinHood]

INC vs ABS?

 

[YotaBota]

A hidden offset in the programming somewhere?

 

[jorogi]

What type of scales ?

 

[PaulL]

Magnetic scales.
Same along the whole scale.
Absolute.

So I dropped into metric and the problem goes away.

And now it works in inch as well.

I feel like I stumbled I to some painful conversion edge case? It no longer reproduces the mess from the video.

Mystery. I hate mysteries. Much prefer SF.

 

[PaulL]

Magnetic scales.
Same along the whole scale.
Absolute.

So I dropped into metric and the problem goes away.

And now it works in inch as well.

I feel like I stumbled I to some painful conversion edge case? It no longer reproduces the mess from the video.

Mystery. I hate mysteries. Much prefer SF.

And found another weirdness. It translate .100 inch to 2.45mm instead of 2.54. If that's not a software error hiding somewhere in the unit conversion code paths I'll eat my hat.
It went away after a factory reset and recalibrate.
Fixed for now, I guess. TouchDRO incoming for the lathe as well now.

 

[DavidR8]

TouchDRO FTW!

 

[Susquatch]

Any clues?

I'd put money on a reader head gap issue. The tape prolly isn't level. This will come and go with temperature and how pissed off SWMBO'd is. She prolly has clout with all the God's but especially the one responsible for the length of things.

Other thoughts are cross-slide backlash and swarf in the head / tape gap.

One other thing. Do you have input / output filtering on your VFD's? That would explain the need for a reset.

 

[Ironman]

As I am in a metal building and built a Touch DRO I had similar problems and also problems would appear when under movement. I have a Oscilloscope and it showed the whole shop was full of various frequencies and not 60 cycle either. I spent a tremendous amount of time fussing with shielding and turning off electrical stuff and filtering power. I finally removed the dro from the lathe and went old school. That worked. Eventually I installed the Touchdro on a friends lathe in the US and it worked well for him.

 

[PaulL]

I'd put money on a reader head gap issue.

This is what I checked and corrected before this current issue showed up. I found I had forced an insufficient gap when I recently lifted my cross slide to inspect the cross slide nut (no, it hasn't required or replaced itself since last time).
This appeared to lead to significant non-linearity at the 20-thou range.
Added some shim stock to set the gap better and now it matches the hand wheels.
The strange input behavior came after that :-(

Do you have input / output filtering on your VFD's?

No VFD on the lathe. The mill VFD may have filtering - there's more little gray boxes in that cabinet than I've had the patience to fully trace. Though, honestly, I'd be surprised is there was. This VFD is a higher end unit though.

 

[Susquatch]

I have a Oscilloscope and it showed the whole shop was full of various frequencies and not 60 cycle either.

My shop is metal walls and roof too. I have not checked the grid in there but I did filter the mains before the VFDs cuz they were noisy. I plan to put VFD cable to the motors when I can find some cuz they are quite noisy too.

 

[Susquatch]

This is what I checked and corrected before this current issue showed up. I found I had forced an insufficient gap when I recently lifted my cross slide to inspect the cross slide nut (no, it hasn't required or replaced itself since last time).
This appeared to lead to significant non-linearity at the 20-thou range.
Added some shim stock to set the gap better and now it matches the hand wheels.
The strange input behavior came after that :-(


No VFD on the lathe. The mill VFD may have filtering - there's more little gray boxes in that cabinet than I've had the patience to fully trace. Though, honestly, I'd be surprised is there was. This VFD is a higher end unit though.

Have you checked the gap carefully at various temps?

The VFD itself doesnt need filtering. It's the cause. If your lathe DRO Plugs into the same grid as the Mill VFD and the VFD and DRO were ever both on at the same time (just powered on, not necessarily operating), the feedback through your shop wiring could affect your lathe DRO.

 

[PaulL]

Have you checked the gap carefully at various temps?

I'm lucky, my shop temperature seems pretty stable.
It's hitting dimensions this morning, so I think it's root-caused.
And the TouchDRO is already ordered. I both trust it better than the legacy unit, and find it's functionality better.

 

[Susquatch]

And the TouchDRO is already ordered. I both trust it better than the legacy unit, and find it's functionality better.

Hard to beat isn't it! Plus a pretty darn simple upgrade process.

 

[gerritv]

Strangely the TouchDRO cost landed plus tablet is not too far off the cost of an Aikron LCD. I got fed up with TouchDRO due to emi and the 3+ years of suffereing with bugs and no updates so moved on with 0 regrets.

 

[gerritv]

Plus I would put my money on having too many digits after the decimal point, even with 1micron scales it is pointless having 5 digits.

 

[Ironman]

My shop is metal walls and roof too. I have not checked the grid in there but I did filter the mains before the VFDs cuz they were noisy. I plan to put VFD cable to the motors when I can find some cuz they are quite noisy too.

Well, the electrical noise does not affect me, or the 3 axis scales on my mill. And it does not affect the Igaging scales I used on the old mill. I suspect the problem is when the Bluetooth receiver in Touchdro is sucking up a bunch of vfd blather that the errors show up. Sort of like when your microwave clips the music in you BT headphones.
On another front for those who hate magnetism (where would we be without it?) I noticed when I brought home my plasma table and my welders, now are all inverter type, that's when the hair on my tools appeared, and my shop was a wood shack then.

 

[Susquatch]

Plus I would put my money on having too many digits after the decimal point, even with 1micron scales it is pointless having 5 digits.

For what it's worth, you can choose how many digits to display on all the recent DROs I have seen. I keep my Ditron 1 micron system set to 4.

 

[neer724]

Hmmm - I wonder if the DRO and scales are setup correctly? Are the scales 1 micron or 5 microns? Do you have the DRO setup to match the scale's precision?