Problem with defective board

[BladesIIB]

Morning,
Hope somebody can help me with a problem. One of my scales is jumping all over the place. Not just a few thou, but over an inch.
I tried emailing Yuriy but he can't or doesn't wan to resolve this.
I tried a different tablet, wrapped the cable with aluminum foil, mounted the scale on 3D printed standoffs but it still jumps. I'm at the end of my rope and if nobody can help me I will just thrash the DRO and get something from eBay. Doubt that the service will be any worse.
Kev.

Your original post says it jumps over an inch. Your last post says .001” which is it? Those seem pretty different? Trouble shooting is about clearly defining the problem and narrowing it down to a repeatable problem so you can ultimately resolve the problem.

 

[kevinismo]

yes, it's not jut the thousandths that are jumping. When I stop at 2.000", the inches digit is jumping too.

 

[ycroosh]

OK, this makes a lot more sense. So the position is not "jumping by an inch". What you really meant was "the display is jumping, and when the scale stops at an inch boundary all digits are jumping/flickering". I suspect that you are not seeing this with the other scales because they jump less frequently. I have one scale in my "library" that jumps up and down by 7 ticks at one spot. I wonder if yours is doing something silimar.
If you don't mind, do the following:
Set CPI for that axis to 1000. Then move the axis to a spot where the jumping is really bad and tell me what the min and max values are.

Regards
Yuriy

 

[JimDawson]

No. This is exactly what my scale is doing most of the time. It's not only the last digit. When I'm at 0.500" the display jumps to 0.499 and
when I'm at 1.000" it jumps to 0.999" so it's all four digits jump up and down. This never happened with the igaging display.
Kev.

Depending on the resolution of the system, that is only 1 encoder count. That seems pretty normal.

 

[BladesIIB]

yes, it's not jut the thousandths that are jumping. When I stop at 2.000", the inches digit is jumping too.

Ok, but jumping from 1.999” to 2.000” is still just 0.001” or 1 thou and not inches. That sounds like what the seller is trying to explain above? Just an outside observer here trying to help interpret. But that is what I am reading.

 

[kevinismo]

OK, this makes a lot more sense. So the position is not "jumping by an inch". What you really meant was "the display is jumping, and when the scale stops at an inch boundary all digits are jumping/flickering". I suspect that you are not seeing this with the other scales because they jump less frequently. I have one scale in my "library" that jumps up and down by 7 ticks at one spot. I wonder if yours is doing something silimar.
If you don't mind, do the following:
Set CPI for that axis to 1000. Then move the axis to a spot where the jumping is really bad and tell me what the min and max values are.

Regards
Yuriy

There is no one spot where the reading jumps. I moved all the way from left to the right and there are many spots. I don't want to list all of them but they are all the same. For example in one spot the scale flips between 1.436" and 1.439". Another spot is 2.757" and 2.760". Does this help?
What does 1000 CPI do? I had 2562 before. Should I change back?
Kev.

 

[ycroosh]

Hmm, this is more than one bit.
Changing the CPI to 1000 makes it so you can see the raw scale readout (ignoring the decimal period).
So the reading in binary would be as follows:
1436 - 10110011100
1439 - 10110011111
2757 - 101011000101
2760 - 101011001000

Notice that the last two or three bits are different between those two numbers. The second set of numbers is interesting. I have one scale that tends to flicker the last three bits, but I've only seen it go between ...111 and ...000. I think yours might be jumping the last two bits, and the second set of numbers has some offset in TouchDRO so you ended with 101 and 000.

The scales have a native resolution of a tad less than 10 microns (10 microns would be 2540 CPI, iGaging scales are 2650 CPI). This means that each increment is about 0.00039". I don't want to bore you with how to convert binary to decimal, but 1 in binary equals 1 in decimal 11 in binary equals 3 (2+1) in decimal, and 111 equals 7 (4+2+1). With the last two bits, your scale is jumping by 3 "ticks" or about 0.0013". If it's the last 3 bits, then your jumps are 0.0027.

EZ-View display averages 8 or 16 values for each position refresh, so you are not seeing this when the scale is connected to it. I don't know if this is OK for the type of work you do on your mill, but I wouldn't be comfortable with this scale.

Hope this makes sense
Yuriy

 

[ycroosh]

Kevin, you successfully "nerd-sniped" me to dig through my box of problematic scales and find one that has the same behavior as yours. I made an animated .gif of a series of oscilloscope screen captures that show this in action. In this particular case, the scale was plugged into its original display and was completely stationary. I did this on my kitchen counter, away from any machinery, running motors, etc. Second .gif shows a more common flickering of a single bit that would result in jumps by 0.0004" or 0.01mm. That one is an older DigiMag scale plugged into its original display too. Curiously, in both cases, the iGaging displays were completely still.

 

[markba633csi]

The clock is the yellow right? Why so different between the two cases?
Is that expected?
-Mark

 

[ycroosh]

The clock is the yellow right? Why so different between the two cases?
They must be using some type of clock multiplier since I see falling edge jitter on the second case- I don't know what's going on with the first one
-Mark

Mark,
iGaging is constantly tweaking their scales, so I see a lot of differences between different scale and display revisions. Some early display units had a 50Hz refresh rate with equally spaced data packets. The latest EZ-View runs four bursts of 16 closely spaced packets with a much faster clock speed.
Regards
Yuriy