What's new
What's new

Results/comments after using new DRO for 6-8 months.

tewitt1949

Plastic
Joined
Oct 8, 2013
Location
mich
Its been a while since I posted here but I'll keep this short as possible. A couple years ago I bought a new Chinese DRO because I could save some money. It had glass slides and 3 axis. It seemed to work ok once I figured out the owners manual. (sort of figured it out but is leaves a lot to be deired). Anyways, after using it for a few months it seem like I couldn't trust what it was saying. It wouldn't hold numbers/measurements accurately. The business I bought it from was very good and helpful, he did try but hard to understand as he was foreign speaking. He did try everything to correct my problems. Even replaced the slides and replaced the entire box one time. But it never did correct the problem.

So from other members here they said I should go to DRO PROS Digital Readout beats Heidenhain Newall Acurite Sargon Fagor. I bought their DRO but was a little concerned how trustworthy it was. So to see how accurate the readout was I mounted 2 slide readout (theirs are magnetic slides) on the cross feed of my lathe. See pictures. So when one readout pickup would move, the other readout pickup should move the same amount, and it did. I proved just how accurate they are in the video. I hope I can upload it. Well it won't let me upload the video. But I can run the cross feed full length in both directions, stopping at several points and the numbers are exactually the same. If some can tell me how to upload the video I will do it.

Final word on Dro pro readouts. You will never be sorry if you get one. I couldn't be happier. Magnetic slides are the only way to go in dirty environments. Dirt, metal, water, oil etc won't affect the performance of the pickups.
 

Attachments

  • DSC07183.jpg
    DSC07183.jpg
    95.6 KB · Views: 176
  • DSC07184.jpg
    DSC07184.jpg
    96.7 KB · Views: 167
No criticism but at least a clarification. You've proved that the two scales read the same. They could both be accurate or they could both have exactly the same error. To prove that one is accurate you'd need to test it against a verified standard such as gauge blocks.

Anyway, good deal with your new DRO
 
Or the slide could be letting both scales read incorrectly...

I'm glad the DRO is working well for you...but your post is a little like all the other comparisons where someone tries to show why a knock-off is just as good as the real deal.

I'd post more, but I'm heading outside to finish the Lamborghini bodywork I fitted to my Beetle. Sure, it's not a real Lambo but it's cheaper and it'll go the speed limit just like a Lambo.
 
Hi
I think you're being a little harsh on the OP.

If two different DROs from different suppliers using different technologies are providing the same readings side by side, it would be reasonable to conclude they are both accurate and precise. It is not proof but it is the most likely explanation for the OPs observations.

Any other potential explanation relies on some combination of errors that are the same in each DRO. Possible but highly unlikely.

Dazz
 
Hi
I think you're being a little harsh on the OP.

If two different DROs from different suppliers using different technologies are providing the same readings side by side, it would be reasonable to conclude they are both accurate and precise. It is not proof but it is the most likely explanation for the OPs observations.

Any other potential explanation relies on some combination of errors that are the same in each DRO. Possible but highly unlikely.

Dazz


I think too that the greatest probability is that the scales are accurate. The chance of having identical errors is extremely low.

But it's to the point to think logically about what is "proved" compared to what we can rely on as most likely true. Going back to principles of geometry, if you see two rafter squares in the hardware store and you want to be assured that the one you buy is accurate, how do you test? You can take two squares off the shelf (or three or four or whatever) and compare them to each other. If any two or three or four are identical when matched nested inside and outside, do you have an accurate square? No, you've only shown that all squares checked are identical. They could all be good OR they could all be manufactured with the same error. You need a plane surface to place two square face to face to verify that they nest correctly and also reverse correctly.

Just understand and don't mistake one for the other. That's why I said specifically it wasn't a criticism. "Don't go mistaking paradise for that house along the road." Bob Dylan
 
It's kind of expected. As I stated a few time before, Chinese readouts are not built to last and most scales fail within a few years. Now you replaced one low cost import with another one. Dropros is magnetic scales out of India, not glass. Magnetic encoders are less sensitive to contamination and installation errors, that's why you see no problems with them repeating.

However hey do have some serious drawbacks. Accuracy is not even close to quality glass scales. Magnetic pitch is around 0.1" vs approximately 0.0008" on glass scales. Divide it by 4 and it is your native resolution, everything under is approximation.

Have you tried it with some fine steel chips or cast iron dust? They tend to stick to magnetic scales and accuracy further degrades if some are caught in between the tape and reader.
PM_magneticChips.JPG


Btw, what monitor are you using with those? Original Chinese one or DroPros Electronica? From our experience, Most Chinese monitors with 7 segment LEDs are surprisingly reliable and seem to last until keypad wears out. However we had several customers with failed Electronica monitors, so we even started offering adapter cables to use our USA made DR300 and DR400 series readouts with Electronica scales. Sold over a dozen of those adapters in the past few months alone.

P.S: Looks like OP is not comparing 2 different scales, but 2 reader heads reading off of the sames scale. Absolute accuracy comes from the scale, not the reader. With this method, scale can have an error of 0.1" per 1" and both reader heads will still read the same number.
 
In measurement, there are two different things, accuracy and precision. This demonstration shows that the dro is very precise, but not necessarily accurate.

My dad worked as an engineer in heavy industry in the boonies(southern tier of NY). When he needed to teach these concepts to the shop floor, he brought a pellet rifle to work. Different times.

They set the rifle up out in the yard and gave two different demonstrations. He installed the gun in a vise, tweaked the sight off a bit and fired it a number of times. The rifle was very precise, with a very small grouping, but way off the side of the target (not accurate). He reset the sight, and let the guys take a few shots each. The shooters were much more on target (accurate) but the groupings were larger (less precise). They also had a good laugh when he took his turn with the rifle. Let's say he wasn't accurate or precise.
 
I recommend you pull in your horns at this point.

Suggesting a jump over to a hobby machinist site is like a fart in church on this forum. And knowing that you also have a thread started on the Home Shop Machinist forum really begins to sound like overpromotion.
 
I think too that the greatest probability is that the scales are accurate. The chance of having identical errors is extremely low.

But it's to the point to think logically about what is "proved" compared to what we can rely on as most likely true. Going back to principles of geometry, if you see two rafter squares in the hardware store and you want to be assured that the one you buy is accurate, how do you test? You can take two squares off the shelf (or three or four or whatever) and compare them to each other. If any two or three or four are identical when matched nested inside and outside, do you have an accurate square? No, you've only shown that all squares checked are identical. They could all be good OR they could all be manufactured with the same error. You need a plane surface to place two square face to face to verify that they nest correctly and also reverse correctly.

Just understand and don't mistake one for the other. That's why I said specifically it wasn't a criticism. "Don't go mistaking paradise for that house along the road." Bob Dylan

Occam's razor applies. If two DRO systems from different manufactures using different technologies give identical results side by side, it would be reasonable to conclude that they are both accurate and precise. This is not proof. It does improve confidence in a statistical way. The probability of both scales having the same errors is low, therefore the probability of both scales being accurate and precise is high. There is high statistical confidence.

Dazz
 
Occam's razor applies. If two DRO systems from different manufactures using different technologies give identical results side by side, it would be reasonable to conclude that they are both accurate and precise.
Lemme pull out Occam's Big Sword here and go for broke - what looks reasonable to me is this guy is a spammer who could care less about accuracy or precision :D
 
If two DRO systems from different manufactures using different technologies give identical results side by side, it would be reasonable to conclude that they are both accurate and precise.

He is not comparing 2 different scales from different manufacturers, but 2 identical reader heads reading off of the same scale at the same time.

Every scale with no exception is repeatable to within 1 count of resolution, unless its defective, contaminated, or improperly installed. Accuracy is a different story. There is an absolute accuracy grade which is typically better than 5um/meter for quality glass scales and then varies widely for low cost glass and magnetic, somewhat around 20-50um/meter is typical

And then there is an accuracy within 1 period. It depends on many things such as approximation factor, sensor design, gratings quality, etc. Grating period is 20 um for glass and around 5000 um for the magnetic. So even if ideal magnetic tape is accurate to 0 um/meter, it will be dead on every 5000 um, but in between, DRO will always be reading in the following increments 0-5-10-15-20-25...4095-5000 while actual position can be 0-2-7-9-26-27..4098-5000 um

That's why even quality German made magnetic encoders, which are tested and certified to traceable standards, are often rated at 25um+20um/meter accuracy, e.g. Guided Unit with Linear Encoder, e. g. for press brakes |

So basically because of the nature of magnetic technology, errors of as much as 25um (0.001") can be introduced anywhere within 1 grating period. And that is without contamination or electromagnetic interference. Anyone really believes that low cost Asian imports are better quality and more advanced than German made ones?
 
interesting. what is the grating period your talking about?

Typical glass scale has microscopically etched lines which are 10um wide. Distance between adjacent lines is also 10um. This gives you an alternating pattern of clear and opaque areas with a period of 20um. Here is a more detailed description of optical encoder technology with illustrations: https://www.zsinstruments.com/downloads/OpticalScanning.pdf

Magnetic encoders use a tape which is magnetized in such a way that North and South poles alternate. Distance from one North pole to the next one depends on the particular model, typical values are between 2mm to 5mm, that is your period or pitch. There are some higher end magnetic scales with period as small as 400um but those are primarily intended for the high end NC applications.
 
this is new to me. so how can the 10/20 mu scale achieve 1 mu resolution? and the 2000/5000 mu scale? well, must be a finer grating in the reading head, right? but the magnetic head would need 1000 divisions to read in 5 mu increments?
 
this is new to me. so how can the 10/20 mu scale achieve 1 mu resolution? and the 2000/5000 mu scale? well, must be a finer grating in the reading head, right? but the magnetic head would need 1000 divisions to read in 5 mu increments?

Interpolation. As reader moves, sensor generates a sine/cosine waveform. Electronics does some scaling/signal conditioning, converts it to digital form, and then estimates position based on the stored values for the ideal sin/cos shape. Real world signal is not ideal and includes many harmonics, hence accuracy degrades with larger interpolation factors.

Quadrature scanning gives you a resolution equal to 1/4 of grating period, so 20um glass allows for 5um without approximation. You can comfortably subdivide a well made encoder 10 to 20 times without significant loss of accuracy. Here are your 1um and 0.5um resolutions. Recent advancements in electronics increase that to about a 100x. And while 1000x and even 16,000x electronics is available, accuracy is lost and at some point extra digits become meaningless.

And no matter how good the electronics is, you have to have a quality scale to begin with for interpolation to be accurate. Variations in line width (real lines are not exactly 10um), edge roughness and sensor design will all affect the usable resolution
 








 
Back
Top