Seems like this should have already been discussed at length but I'm not seeing it. I'm pondering a DRO for the lathe and the first thing to decide on is glass or magnetic scales. It looks to me like either will get the job done. Anyone have a good reason to prefer one over the other? My shop in unheated so the temps will vary a lot. From my reading the thermal expansion of the magnetic scales might be an issue. Is this a real concern? Ignoring that the magnetic look easier to install and maintain. Do I need to worry about magnetic bases messing with the magnetic scales? Are the glass more accurate? I'll be turning a lot of steel, any probs with steel swarf sticking to the scales? thanks.
-
Welcome back Guest! Did you know you can mentor other members here at H-M? If not, please check out our Relaunch of Hobby Machinist Mentoring Program!
Glass or magnetic scales
- Thread starter jtr67
- Start date
- Joined
- Apr 4, 2013
- Messages
- 884
In short, you get more bang for the buck with glass scales. The main appeal of magnetic scales is the ability to cut them to length and smaller profile. I.e. they fit tight spaces easier. The drawback is the higher price and somewhat lover practical accuracy.
Long version:
Magnetic scales (good ones) can deal with swarf and ambient magnetic fields pretty well, plus you would want to have a wiper of some sorts to reduce the amount of stuff between the scale and the reading head. They are VERY sensitive to misalignment or distance fluctuations. Better ones (starting with the level of Electronica/EMS and up) come with pretty beefy extrusions that keep the magnetic tape straight. Cheaper Chinese scales that come with a roll of tape are more "complicated". First of all, the tape is THE place where the manufacturers cut corners. Good tape that is temperature-stable and accurate is expensive, but the reading head will "work-ish" with fridge tape, as long as the pitch is around 5mm.
Second, the circuit that interpolates the sine/cosine wave can be implemented several different ways, ranging from very cheap and flaky to very expensive and bullet proof. Mid-range stuff is "slow but accurate". As you go up in price, you get better error rejection and interpolation speed. Cheap stuff is generally "OK" under good conditions but doesn't handle errors very well. Basically you can get the interpolation modules (sensors, interpolation chip, amplifier, quadrature driver) for prices ranging from $5 to $100+ (wholesale), and in this case you get what you pay for.
In practice, you will usually get OK accuracy with some amount of non-linear error under good conditions and once the tape gets old and starts delaminating (the 3M tape most Chinese scales come with is not oil resistant), the accuracy goes down quickly.
Glass scales are [at this point] hard to mess up. The two important parts are the encoder strip and the reading head. Neither can be done in a "home shop", so even the bottom-of-the barrel scales come with decent strips and heads. The place to cut corners is the amplifier circuitry, extrusion, end caps, seals and cable. Those don't matter for accuracy, but might affect long-term reliability. You are also much more likely to get a complete dud than a flaky scale. I.e. if the scale works, it will be accurate; if the scale is bad, it will usually be completely dead. I've seen only one example when a scale flaked out from vibration because the reading head was missing the guide ball bearings.
Finally, in my experience the whole thing about glass scales being sensitive to dirt ingestion and magnetic scales being "bullet proof" is, to put it politely, a marketing exaggeration. If you submerge a glass scale into your flood coolant tank, it will likely fail, but if mounted correctly, you should not be getting coolant into the scale.
All that said, my "standard" setup for a lathe is a 5um glass scale mounted upside down for the Z axis (apron travel) and a 1um scale on the cross slide. On my old Jet lathe I had Ditron magnetic scale, and it held up OK. On the "new" Rockwell I have a 1um Electronica scale. I don't trust the scales to be actually accurate/repeatable to 1micron, but they are more than good enough for the work I do.
Hope this helps
Yuriy
- Joined
- Feb 8, 2014
- Messages
- 11,176
I prefer magnetic scales. Bulletproof, compact, and easy to install.
www.hobby-machinist.com
Lathe DRO
Last weekend my son informed me that it is time to put a DRO on the lathe. I guess he hasn't figured out what all those little numbers are for on the dials. :rolleyes: He has a short production run for a product he is developing so figured it would be easier to do it with a DRO. I suggested...
www.hobby-machinist.com
- Joined
- Apr 10, 2021
- Messages
- 621
Depending on your machine, one type or the other might be easier to mount. A glass scale worked well for the Z axis and mounted on the back of the bed easily. The cross slide worked much better with magnetic. Glass scales cost 1.5" of travel from my tailstock where magnetic cost about 1/2". Dave
- Joined
- Apr 4, 2013
- Messages
- 884
Well, iGaging scales are not designed to be used in a DRO setup. Back in 2012 (when they first came out) the cost of glass scales was still pretty high (on the order of $250-$300 per scale for Chinese scales, much more than that for something with a decent brand name. iGaging scales sold for around $35, and that was a bargain. They offer about 10 micron resolution, but have two HUGE shortcomings:
1. They are VERY sensitive to noise and glitch a lot. Native displays hide this by averaging a bunch of values and ignoring outliners or sudden jumps.
2. They are relatively slow. Practical refresh rate is about 10-20Hz at best.
In term of TouchDRO, I've added a LOT of workarounds to squeeze as much performance out of these as I could. Some years ago it was worth it. Subjectively, they were 50% as good as glass scales at 10% the price. Today you can get very passable glass scales for $50-90, while iGaging EZ-View scales start at around $40 and go up from there. At that price they are a very poor value. iGaging Absolute (stainless steel version, which is not actually absolute) is just a bad idea - more expensive than glass and still 50% as good.
From the data points I get from my customers, I almost never hear about problems with glass scales (over the years I've been selling pre-made boards, I've had one case where a scale was flaking out due to vibration and one magnetic scale that was skipping pulses for some reason). With iGaging scales I used to get at least one email per week. (The new 32-bit version of the adapter seems to be fairing better, but still not 100% bullet proof)
The bottom line is that if you already have them, they are better than not having a DRO, but I would not buy new iGaging scales for a DRO with todays prices.
Note (before I get in trouble) - these issues are not inherent to ALL capacitive scales. Mitutoyo made very capable DRO scales that use capacitive encoders, but those were almost two orders of magnitude more expensive. I.e. it's possible to have good capacitive DRO scales, but the involved modules are very expensive.
- Joined
- Oct 16, 2019
- Messages
- 6,876
Thanks I appreciate your expertise. Now I just need to figure out how to fit glass scales to my South Bend.
- Joined
- Apr 4, 2013
- Messages
- 884
I don't have an SB nearby to look, but on my Rockwell the cast bed has slightly tapered sides. I 3D printed thin wedges to get the scale close enough. With glass scales you have some wiggle room. The reading head (the thing that is inside the extrusion is decoupled from the large rectangular thing that rides on top of it. The encoder rides on the glass strip on small roller bearings. If you scroll half-way down this page: https://www.touchdro.com/resources/scales/glass/overview.html you can see a picture of the head. The bottom line is that you can be a degree or two out of parallel and the scale won't care. Any error this introduces can be calibrated out in TouchDRO (or most other decent DROs).
For the cross slide I would look at a magnetic scale. Two things that I would suggest:
1. No matter what scale you get, mount it on the chuck side of the cross slide. It will feel weird, since you can easily damage the scale, but in practice if you haven run your chuck into the compound, you won't run it into the scale. On the flip side, you will appreciate the extra travel of the tailstock more than the capacity over the scale (if you loose any).
2. Get a 1-micron scale. When you remove material, you really care about the diameter (usually) and in diameter reading more 5-micron scale becomes 10 micron scale (0.0002" vs. 0.0004"). 1 micron might seem like an overkill, but especially if you get a magnetic scale, the error on a 1-micron scale will be much smaller than on 5-micron (2mm pitch tape vs. 5mm pitch).
Regardless of what DRO you get (shameless plug - get a TouchDRO
), you can mix and match modern scales. If you get a name-brand DRO head, you will loose the reference mark functionality if you get a different brand scale. With Chinese DROs the reference mark is not used (even if by some freak accident its wired internally). Regards
Yuriy
- Joined
- Jul 19, 2020
- Messages
- 91
I installed a DROPros system a couple years ago on a mill. It was one of their “internet specials” so quite a bit cheaper than off the shelf systems. I cut the magnetic scales to length on the miter saw. With careful installation it was easy to align… the brackets have slots to allow you to correct minor misalignment. No problems since installation. I find chips attached to the scales occasionally, I wipe them off.
EDITED: I forgot to add that DROpros has a lifetime warranty on magnetic scales. Not on glass.
EDITED: I forgot to add that DROpros has a lifetime warranty on magnetic scales. Not on glass.
Last edited: