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

[pontiac428]

Thanks for the explanation, @B2. You're back a ways in line, I promised my scales to another member who asked as soon as I started making a fuss about the issue.

I got my butt handed to me in electrostatics class, so I never really wrapped my head around the subject like I can in others. I still think clamping a mag base to my mill table or head makes a moving field, because the table moves in relation to the sensor. That's why I decided not to mess with it using foil tape and patented shielding materials. I can't see a static situation on the mill other than a quill plunge. The table must move to cut all other features. Then there's the fields from the main motor, servo drive, and RPC. All I see are magnetic fields when I look at it. I can see fluctuations with the mill powered off from waving my magnet around, but how can I be confident of anything the display reads before, during, or after the movement? As far as waving the magnet goes, imagine me standing in the operator's position with a magnet in my hand waving it like a vial of holy water, chanting the power of Christ compels you like the priest from the movie. That's not setting the magnet on the sensor or locking it down on top of the strip, it's movement a foot or two away. And it's a old, worn, tired, weak, dumpster-rescued Starrett base magnet, nothing fancy. Shouldn't be a problem in another week or so, I hope.

 

[B2]

I still think clamping a mag base to my mill table or head makes a moving field, because the table moves in relation to the sensor. That's why I decided no

I have an experiment for you! Take a steel plate, say a little larger in width and length than the magnetic base and say 0.25" in thickness. Clamp the magnetic base to it. Then wave this around your sensor and see if you get any changes in the readings. This then would simulate the base being clamped to your mill table or other moving surface. The flux from the base is then contained inside the steel plate (mill table) and does not reach outside (much at all). The thickness of the steel you are clamping to is not critical as long as it is thick enough that the fields from the base do not saturate the magnetization in the plate. You can easily get a feel for this. First clamp it to a piece of sheet metal, like thin steel air ducting which would be saturate by the base's magnet. Then touch the steel with a non-magnetized screw driver to see if you can feel it sticking. If so the plate is definitely not thick enough. Then up the steel sheet thickness and repeat the experiment. Along the way for each thickness you should wave the assembly around and observe the DRO. The steel the base clamps to really does have a dramatic effect on the distance of the stray field can extend from the base. Not all bases are equal. Many use Ferrite magnets, but strong ones use Neo magnets and are stronger. If when you turn the lever switch your base to the release position there should be very little stray fields from it, but there are some. If it is designed wrong these can be larger than you might like. If the magnetic base is hanging off of the edge of the mill table then it might produce some external fields from where it is exposed. Putting an extra piece of steel plate on the exposed parts should fix this.

If you are not familiar with the concept of magnetic saturation, it is the properties of all magnetic materials that they only magnetize so much at which point any additional applied field penetrates the material. All magnetic materials are composed of small magnetic regions called magnetic domains. These magnetic materials are just that, magnetic, but are not always magnetized as the direction of the magnetization of these regions can be almost random and the flux from one circles back via the others. We call this closure of the flux path. This closure of a magnetic circuit is what happens when the magnetic base clamps onto the steel mill table. The degree of closure is dependent upon the gap (air) between the base and the steel table. If the surfaces do not match well the base may still grab the surface, but not as strongly as some of the flux is leaking out.

As one applies a field to the materials, the magnetization in each domain tends to slide around or rotate so as to be oriented with the applied field. As the field is increased more and more of them aligned to the applied field. Once they are all lined up in the same direction we say the material is saturated. The different materials have different magnetic moments per atom or unit cell and so have different final saturation values. A soft magnetic material simply has domains that are easy to move around or rotate into alignment. A hard magnetic material tends to resist the movement and so is hard to change. A permanent magnet is permanent because it is difficult to change the domains back to a closed (random like) state. For example, we are all now familiar with the Neo magnets. (This is a very complex Neodymium Iron Boron crystal: Nd2Fe14B with 64 atoms per unit cell) Because they have lots of iron in the material they have a high magnetic moment per unit cell and because of the Nd they have a high directional anisotropy energy resulting it in being very magnetically hard. The iron tends to rust so these are usually plated to protect them. The SmCo magnets are even harder (higher anisotropy energy), but do not have as large a saturation magnetization and are more expensive (Cobalt costs much more than the rare earth Sm), but do not rust! Hence it is difficult to demagnetize either type. For SmCo about the only way to demagnetize them is to heat them up! (demagnetize does not mean non-magnetic... it just means having internal closure of the magnetic flux via the domains.) Ferrites ( Barium or Strontium Iron oxide) are a work horse as the materials are cheap (Mostly Iron and oxygen), but the saturation magnetization is low because of the large amount of non-magnetic oxide. Electric car motors all use the Neo magnets.

Most steel alloys are magnetic and are pretty soft. You can magnetize your screw drivers a little by stroking them against a strong permanent magnet. However, you will notice that they tend to demagnetize over time.... the more you use them. If you want to demagnetize them you can put them in an AC magnetic field which is slowly decreasing in amplitude. This tends to scramble the orientation of the domains in to localized closure. OR... you can just tap the end of the screw driver a lot on some hard surface and the mechanical stresses induced will tend to scramble the domains back to internal closure.

By the way, on my lathe the DRO sensor for the cross feed is mounted on the right a way from from the spindle side where I would clamp a magnetic base, but is still close enough to be of concern. However, for the z feed the sensor is mounted way down below the saddle and so is far from any field associated with a base being on the saddle somewhere and moving with the cross feed. Since the z sensor moves with the saddle a magnetic base on the saddle would not create any changing fields at the sensor.

I will be interested to hear how you come out with this experiment.

Dave L.

 

[rwm]

I tried this experiment just now and it works exactly as you suggest. With a 3/8" plate on the base I get no errors in almost direct contact with the head.
My concern is that I will slide the mag base right over the scale, where it is only separated from the tape by a little aluminum. I'll be careful.

 

[pontiac428]

Dave @B2, thanks for the post, I really enjoyed reading it.

So, forgive me, but what does performing this closed flux test prove? It looks to me as if it is designed to determine scientifically whether or not I am using a magnet. I believe you were telling me that a closed field or internal flux should not "leak" if a base is clamped to the table, but I never got so far after seeing the response on the screen to holding a magnet in my hand. The test stopped there because the result was unacceptable to me and I did not want to apply band-aids to a wet turd by trying anything else. I was aware that I was using an actual magnet during testing, and I determined that wherever the magnet was involved, my numbers whirl and spin. So I'm scientifically certain that a magnet can/will induce changes on my DRO screen during one or more commonly-encountered conditions. No further determination is needed, I have new optical scales on order and will wire in all the negatives for the twisted pairs at the head unit by completing the missing connections simply because I am aware of it and can easily do so.

Maybe I'll re-evaluate mag scales when they make them wireless, voice-activated, and pocket-sized.

 

[RaisedByWolves]

Hi Guys, @pontiac428

You can send me any magnetic DRO you want to get rid of! They are fine.

All of the sensor technologies have limitations. I worked in the field of sensors for years and even longer in the field of magnetism and magnetic recording. The optical sensors (DRO) do not do well in dirty environments. Neither do the capacitive. Magnetic sensor systems are far more tolerant. I have a magnetic DRO on my lathe. I have never gotten around to putting a DRO on my Mill as it is CNC. On big motors moving large paper webs (news print) where the position must be measured accuratly so that the speed can be carefully controlled (do not break the web) they got rid of the optical sensors as they fail from the paper dust. They were replaced with magnetic sensors (magneto-resistive).

I have never tried this experiment of waving a magnet around my DRO sensor. Of course it could have bad effects if the sensor and stripe are not built correctly or well.

Just as you could scratch the optical scale and it would start to fail, you could demagnetize the magnetic stripe, but is is pretty strong, high coercivity (resistance to change in the magnetic state), and you would essentially have to bring a magnet into contact with the stripe to mess it up (you would have to replace the stripe, which is not hard to to if you can get it). The stripe is usually made from a polymer containing ferrite particles with magnetized patterns written into in. (A digital recording pattern.) It should have a coercivity of several hundred oersted or larger. For reference, the earth's magnetic field is about 1-2 Oe. The magnetic stripe on a credit card can be as much as a couple of thousand Oe, but is usually about 700-800. Most better digital tape (video/audio recording tape) is about 800 Oe. MP, metal particle, tapes have even higher coercivity.

The magnetic sensors themselves are not very subject to EMI, but the cables to the readout might feed a time sensitive noise to the readout electronics. If EMI at the cables is the problem it probably does not matter which type of DRO you have. Ideally the sensor head has enough electronics in it to convert the sensor signal into a digital signal before sending it on to the readout electronics ... so has to minimize EMI/noise effects. There is lots of EMI coming from the motor etc of a lathe, especially if you have a VFD. There is a huge amount from a motor with brushes.

The DRO magnetic sensors, Hall Effect or Magnetoresistive (there are several kinds of MR sensors) measure the DC and slowly varying AC magnetic fields from the pattern that is on the stripe. Since the lathe motion is so slow, this is essentially DC. Waving a magnet around would indicate to me that you are changing the DC field slowly (quasi-static field) at the sensor and this could produce a change in the sensor output that is interpreted as steps along the magnetic stripe. One would not normally do this during operation. Your magnetic base is designed to clamp to the steel somewhere (lathe bed). As you rotate the knob or lever of the base the magnetic is turned so that the poles cause the internal magnetic field to point through the base surface and into the steel. The magnetic field (magnetic flux) comes out of one surface of the base, through the steel lathe bed, and then returns into the other side of the base. This way the flux is essentially coming from the north pole and back to the south pole of the magnet that is inside the base. In this magnetic circuit there is virtually no external fields outside of the base and the lathe bed. When it does this the field/flux is contained in the lathe bed and so does not /should not be much of any where else. That is, there are virtually no magnetic fields outside of the device and the steel it is clamped to. Hence, the sensor and DRO should work perfectly fine as it would see no field from the magnetic base.

The magnetic sensor "housing" should be made from a soft magnetic material to deflect/re-route any external fields away from the sensor itself, but by necessity the sensor must be able to sense the stripe. Ideally the back of the stripe would be of soft magnetic material to also prevent any external field from reaching the sensor. In both cases the soft magnetic material would then essentially short out most of any external field. Even simple steels tend to be pretty good soft magnetic materials for this purpuse. Someone mentioned mu-metal which is an extremely soft (near zero coercivity) magnetic material. This would be good to shield the sensor head as well as to put on the back of the magnetic stripe spaced a short distance away from the magnetic particle materials. (think a layered structure: magnetic particles/non-magnetic materials/soft magnetic base)

Some stainless steels are not magnetic, but some are. Aluminum, brass, copper and many other metals are not magnetic and so would NOT work as shields from external fields.

I had a project a couple of years ago where I needed to observe the magnetization patterns of a material changing with applied field. These are small enough that I needed to do the measurements via a special microscope (polarized light is changed a little bit when it reflects from a magnetic material surface --- Kerr effect). However, the applied field was time dependent and the steel parts of the microscope would be pulled as the field varied. Hence, parts of the microscope wiggled with the varying magnetic field screwing up the optical signal. I had to take the microscope and all of the associated mechanical stages apart and replace any parts made from ferromagnetic materials with parts made from non-magnetic materials. In many cases I used brass or non-magnetic SS for the bolts, nuts, rods, etc. However, in one of the objective housings there were bearings that ran on steel runners. I replace the balls with ceramic balls and the runners (~0.25 wide by 2.5" diameter washer shapes) which were very thin ~0.015" steel with Titanium. It is always hard to clamp and cut sheet material, but it is really hard to machine Ti this thin!

So either shield you sensor head or stop waving magnets around! If you are going to clamp a magnetic base near your magnetic DRO I suggest that you reset your origin/reference after you clamp the magnetic base down. After all the magnetic base should have no effect on the DRO once it is stationary.

On the other hand, just send me the DRO's you are taking off line!

Good luck.

Dave L.

 

[rwm]

Dave @B2, thanks for the post, I really enjoyed reading it.

So, forgive me, but what does performing this closed flux test prove? It looks to me as if it is designed to determine scientifically whether or not I am using a magnet. I believe you were telling me that a closed field or internal flux should not "leak" if a base is clamped to the table, but I never got so far after seeing the response on the screen to holding a magnet in my hand. The test stopped there because the result was unacceptable to me and I did not want to apply band-aids to a wet turd by trying anything else. I was aware that I was using an actual magnet during testing, and I determined that wherever the magnet was involved, my numbers whirl and spin. So I'm scientifically certain that a magnet can/will induce changes on my DRO screen during one or more commonly-encountered conditions. No further determination is needed, I have new optical scales on order and will wire in all the negatives for the twisted pairs at the head unit by completing the missing connections simply because I am aware of it and can easily do so.

Maybe I'll re-evaluate mag scales when they make them wireless, voice-activated, and pocket-sized.

I see your point but it does seem like your read head is unusually sensitive. After some observation, I have determined that I can use a mag base on the front side of the carriage without any concerns. Also, I have no option based on physical constraints...

 

[B2]

Dave @B2, thanks for the post, I really enjoyed reading it.

So, forgive me, but what does performing this closed flux test prove? It looks to me as if it is designed to determine scientifically whether or not I am using a magnet. I believe you were telling me that a closed field or internal flux should not "leak" if a base is clamped to the table, but I never got

I am glad you enjoyed. Of course my point in the experiment is that when the magnetic base is clamped to the mill table there is no external field to affect the DRO. i.e. from the DRO view point there is no magnet! The test was to show this by using a simulated mill table (iron sheet) so that you could wave it about.

The rest of my note was just my babbling on about magnetism! Forgive me. I intentionally purchased one of the first magnetic DROs that Matt (PM) sold because of my technical history in magnetic (tape and hard disk drives) as well as optical (compact disk) recording as well as with rotational optical encoders failing (missing counts) from dust in the air getting on the glass. Like I said all of these sensors have limitations and failure modes. There is a good reason they put the housing over the DRO track. Jim's pictures,

Here is the X axis on my mill

showing the dirty magnetic scales (operating without a cover) and still working is impressive!

Good luck and enjoy your optical scales!

Dave L.