Why didn't anyone warn me about mag scales on the DRO?

[pontiac428]

Can any of you that are having these problems check if your cable ground is connected to signal ground? Sounds a little bit like some kind of grounding issue. The cable shield shouldn't be grounded at both ends. That can cause a lot of issues. Also check your display end

I'm fairly certain the metallic conduit and connectors for these read heads are all grounded into loops. It's an all metallic wire harness, from plug to head. I was sure the whole thing would be grounded through the machine, which is bonded to panel and RPC earth. I drove and bonded a half dozen ground rods right there at the stem wall where the mill is sitting. I installed the ground wire from the head to the machine's casting. So everything shares common ground, nothing is insulated, everything is metal to metal. The shields don't appear correct for draining interference, but the twisted pair bundle may have a shield and drain wire in it, I haven't taken it apart to look.

I'm not too keen on fixing this problem with anything other than optical scales, to be honest. I'm just going to throw my wallet at the problem and it'll probably go away. I'm not going to work into the wee hours insulating and taping and shielding and testing until it does or does not go away, I am over it- I bought cheap chicom mag scales, and got what I paid for over Mitutoyo, etc. And with the magnetic technology, I'm thinking this is the fool me once, next time it's really on me.

 

[WobblyHand]

I'm fairly certain the metallic conduit and connectors for these read heads are all grounded into loops. It's an all metallic wire harness, from plug to head. I was sure the whole thing would be grounded through the machine, which is bonded to panel and RPC earth. I drove and bonded a half dozen ground rods right there at the stem wall where the mill is sitting. I installed the ground wire from the head to the machine's casting. So everything shares common ground, nothing is insulated, everything is metal to metal. The shields don't appear correct for draining interference, but the twisted pair bundle may have a shield and drain wire in it, I haven't taken it apart to look.

I'm not too keen on fixing this problem with anything other than optical scales, to be honest. I'm just going to throw my wallet at the problem and it'll probably go away. I'm not going to work into the wee hours insulating and taping and shielding and testing until it does or does not go away, I am over it- I bought cheap chicom mag scales, and got what I paid for over Mitutoyo, etc. And with the magnetic technology, I'm thinking this is the fool me once, next time it's really on me.

Can't really disagree with you, but if you can make a cheap fix, you can spend the cash you saved on something else.

Ground loops are a bear to find, but try some easy things first. You could make a cable extender (M & F DB9) and break the cable ground connection at the display end. Might work, might not, but it's cheap to try. If it works, you know what to do. If not, buy some glass scales, but get them as a matching set with a display. Who knows, the different manufacturers might ground them on opposite ends.

On my rig, the read head is isolated from the mill. (3d printed holder and spacer.) I have an isolated (standard double insulated) power supply for my ELS and DRO readout. Stable to the micron. Sino K500 1um X scale, and a Ditron, forgot the model number, 1um Z scale. Both are glass.

 

[RaisedByWolves]

That's pretty helpful. I wouldn't mind if it were easy to predict or just a matter of not placing mag bases near the heads. I think one should be able to wave a silly mag base over and around the table like an exorcist with a vial of holy water without affecting the reading.

Ditron is not picking up their chat line on Alibaba, I have been trying since Saturday. Must be Chinese Labor Day or nighttime over there or something. I think those folks sleep with their laptops to answer questions at all hours, normally. Actually, I'm being facetious, people deserve weekends and off time. Usually they're on the spot with communication.

I'll look up Electronica, if there is actually something to be gained. With China, you never know, it could be just variations of the same knockoff.

Showed my lead man and he was shocked. Then we started brainstorming, and taken with your situation came up with all kinds on non desirable situations you could get into.

None of this even touched on shop pranksters loosely wire tying a magnet near the read head.

We might be changing them out soon.

 

[homebrewed]

Just to clarify the shielding issue, do nearby magnets cause the problem if they are not waved around? The various experiments aren't clear in this regard.

The reason I'm asking is that I expect the sensor to be a hall effect or GMR type magnetic sensor (because inductive pickups' sensitivity would depend on the feed rate). These types of sensors detect the magnitude of the local magnetic field, even if it's static. Electrostatic shielding is transparent to DC magnetic fields so it would be a waste of time to mess around with that as a possible solution.

 

[RaisedByWolves]

Just to clarify the shielding issue, do nearby magnets cause the problem if they are not waved around? The various experiments aren't clear in this regard.

The reason I'm asking is that I expect the sensor to be a hall effect or GMR type magnetic sensor (because inductive pickups' sensitivity would depend on the feed rate). These types of sensors detect the magnitude of the local magnetic field, even if it's static. Electrostatic shielding is transparent to DC magnetic fields so it would be a waste of time to mess around with that as a possible solution.

With the tests I did the numbers would change and remain stable but incorrect until you moved the magnet again.

Regardless of what happened everything returned to normal when the magnets were removed altogether.

 

[WobblyHand]

With the tests I did the numbers would change and remain stable but incorrect until you moved the magnet again.

Regardless of what happened everything returned to normal when the magnets were removed altogether.

Hmm, seems magnetic then. On the other hand, I have been fooled before by ground loops. The GL's behavior can be quite different, depending on the situation. If it's possible, break the cable ground connection at one end of the cable. I usually ground the shield at the sensor end, but sometimes you have to do the opposite end.

 

[verbotenwhisky]

Okay, I'm not holding anyone accountable for this, I attended all of my physics classes sober so it's all pretty obvious. But I had something happen today that justifies dumpster slamming magnetic scales in favor of opticals.

There I was, milling away joyfully and producing a nice part today, when the hot blue steel curls became unbearable on my skin. I set up my chip shield and finished the part. As I was tearing down, I removed the mag base on my chip shield, an aging Starrett with a real weak pull. While moving with the magnet above the table, I noticed the displays fluttering everything to the right of the decimal place on my DRO. I cussed an incantation and waved the magnet all up and down around my mill, watching the numbers change. It would do it 24" away from the scale or sensor somewhat predictably as I moved it up, down, left, and right through the air over the table.

The mag scale readers are metallic and the cables are metal shielded. They may have ground loops, which would explain the sensitivity to the field. To be honest, I don't care why the DRO responds so much, it's the fact that the numbers change in response to a common mill table tool's presence that I can't abide. I don't run the mill on faith, I run it on confidence. If I have to rely on faith that the local magnetic fields are stable in order to make a part, I'm tapping out.

This has me on Alibaba measuring up for some optical scales right now. It's not sitting easy with me. Have any of you observed such a thing?

Yes and I also had issues with an electric field at one point, chased it for weeks.

 

[pontiac428]

Hmm, seems magnetic then. On the other hand, I have been fooled before by ground loops. The GL's behavior can be quite different, depending on the situation. If it's possible, break the cable ground connection at one end of the cable. I usually ground the shield at the sensor end, but sometimes you have to do the opposite end.

I saw a block diagram from Ditron that shows the cable shielding coupled to the pull up resistor at the reader and to the connector pull down resistor at the head unit. Seems legit.

Just to clarify the shielding issue, do nearby magnets cause the problem if they are not waved around? The various experiments aren't clear in this regard.

The reason I'm asking is that I expect the sensor to be a hall effect or GMR type magnetic sensor (because inductive pickups' sensitivity would depend on the feed rate). These types of sensors detect the magnitude of the local magnetic field, even if it's static. Electrostatic shielding is transparent to DC magnetic fields so it would be a waste of time to mess around with that as a possible solution.

If it's TTL then it's not a Hall sensor, Hall requires a variable reluctor to convert from zero crossing to a low/high wave. It's a logic signal.

 

[Harry Knutz]

Gravity waves would affect your part the same way so there would be no error produced!

If you are following the other thread, I am having problems with a new Ditron DRO. My cross slide readings are off. The Ditron rep is now giving me some *******t like "a .002 inch error over 1 inch is normal for magnetic scales."

I had some issues with Ditron when I put my magnetic scale from them on my enco lathe, Their customer service is pathetic, You will get a bunch of empty promises, We will help you tomorrow and tomorrow never comes etc. I finally had to learn how to take care of it myself.

Edit to add: I think their scale that I received from them is actually pretty well made, I have machined some parts with it and they have turned out spot on, And it's mounted on the cross-slide, So I am pretty happy with it, It's just if there is a problem you are on your own with Ditron.

 

[homebrewed]

I saw a block diagram from Ditron that shows the cable shielding coupled to the pull up resistor at the reader and to the connector pull down resistor at the head unit. Seems legit.


If it's TTL then it's not a Hall sensor, Hall requires a variable reluctor to convert from zero crossing to a low/high wave. It's a logic signal.

Not necessarily. There are analog hall effect sensors as well. The Hall effect always produces a small analog signal that is proportional to the strength and direction of the magnetic field passing through the sensor. Since (apparently) the DRO interpolates 1,000 positions between stripes it has to be doing some processing on the mag sensor outputs, which start out as sine/cosine signals. That would explain why the DRO is sensitive to stray DC magnetic fields. Also see an explanation of magnetic DRO scales and factors that affect their accuracy here.