Troubleshooting Glass Scale Issue

MAlcocer

Registered
Registered
Joined
Sep 26, 2014
Messages
61
I hate to be that guy who was just touting his DRO setup on another thread only to turn on said DRO and have it not work. But that is where we are at today.

I went to do some work on the mill and noticed I had left the DRO powered up for who knows how long now. Didn't think much of it until I was trying find the edge of my work piece and noticed that the DRO display was not moving when I turned the wheels. Not only on a single axis but it was intermittently not working on all 3 axis. I did some basic checks, the scales are getting the required 5 volts, everything is clean inside the scales. I tried someones suggestion of "tightening the spring" to force the read head down on glass scale. This helped a bit but was not 100%. There is a read LED that blinks on the the MSP430 when this happens but that is not always the case either.

Here is my setup. I am using Touch DRO with a MSP430 Launchpad controller and a custom voltage shift circuit modeled off of Yuri's X3 build here. This is all for 3 Ditron glass scales, DC10 models.

Here are videos of two of the read heads and what is going on:

Just some back info is the problem arises when I move the head ever so slightly and run it the length of the scale. I get the most consistent results when the reader is flush on the scale. But it will eventually glitch out. When the reader is pulled as far from the head as it can be and run the length of the scale it almost never works. (While watching inside the scale while I move the outer portion of the read head I cannot see any visible movement of the internal read head).

My theory is that I have a short in the wires going into the scale pictured belowRead heads (2).jpg For one this seems like a classic short where the DRO stops receiving signal. Second I moved the wires from the way they are pictured here to have the S bend on the far side of the head. This caused the scale to stop working altogether. What makes me think that it is not a short is that I would be amazed if all the scales went out at the exact same time. So that means maybe I blew out the controller in my DRO or maybe the heads are not sitting dead on right (we have had one hell of a heat wave lately and while in a cool garage the temp has been up, maybe the spring weakened).

Given my limited amount of tools for diagnosing this issue (multi meter and no replacement DRO or scales) anyone have any ideas? Anyone in my area want to let me plug some scales into their DRO and let me rule out them as the issue?
 
If you are having a problem on all three axes, I would not expect it to be a scale problem. It would be highly coincidental for all three scales to go out at the same time. I would look first at areas that are common to all three scales. Power supply, your voltage shift circuit, controller, etc. You may be operating on the ragged edge for some parameter so it doesn't require much of a shift to go from reliable operation to intermittent operation. A short on your power supply may be drawing down the voltage to the point the scales don't operate. I would hook the meter into the power supply and watch it as you try to replicate the problem. I would also reinstall the scales and try moving two scales at once and see if the problem occurs simultaneously or if it is asynchronous.

Troubleshooting is all about dividing the possible causes up so you can isolate them and pin the culprit.

Good luck!

Bob
 
I have had the meter on the power supply during operation. Voltage is good all the way to the head. I am really leaning towards the micro controller at this point but there are two things that don't make sense. Adjusting the tension on the read head springs makes the problem slightly better or worse. Disturbing the natural bend of the wires inside the head makes the problem worse. And the scales run fine for a few passes and then will suddenly stop sending signal. I would love to know a good way to test signal out put of the scales to rule them out as an issue.
 
A meter might be insufficient to diagnose noise on your power supply. Since they are cheap I would try another power supply first.

I am not familiar with the MSP version of this design but the arduino one you can connect via USB to your pc and see the value being sent to the Bluetooth transmitter this would help determine where to look for problems.

Sent from my SM-P605 using Tapatalk
 
I would love to know a good way to test signal out put of the scales to rule them out as an issue.

The only real way to test them is with a oscilloscope. But from what you describe, I would be looking at wires and solder connections. A cold solder joint might cause the intermittent signal.
 
Even a scope would be useless if this is an intermittent issue. You would need a logic analyser

Sent from my SM-P605 using Tapatalk
 
And the scales run fine for a few passes and then will suddenly stop sending signal.

Adjusting the read heads may be a red herring. I'd go back to basics.
The fact that it looks intermittent tells me it's likely a bad connection or broken wire somewhere, as Jim suggests.
You could use the meter on "continuity" setting and make sure all wires are intact end-to-end as you move things around.
And as Bob says, look for something that is common to all the affected axis.

The first video looks like it "stalls" at two points (-7.9xxx and -6.1xxx) on the first right-to-left movement.
In the second video it "stalls" at -9.58xx and -7.91xx again on the right-to-left movement.
It could be the scale, a connection/bad wiring, or even the microprocessor, but if all axis are affected it's likely the second or third of those options.

A little more info could help:
Has the system been running error free for some time before these problems started?
Have any changes been made recently? (adding the third axis, adding new "noisy" equipment nearby, etc.)

-brino
 
The x and y were running problem free for a couple months. The z axis scale was installed to the mill but I had not finished the bracket to the head so it just sat there not plugged in and not moving. I have not added any equipment and they had been working well before this. This may not mean much but it had been left on for maybe a month straight prior to this happening.
It is coincidence that the heads in the video stalled in the same direction. I can get them to stall either way.
If I check for continuity in the wires from the plug all the way down to the read head there is no chance of blowing a circuit correct?

As for power supplies that is not the issue. I have changed them out and even ran the USB power supply off my computer with various cables. The spring adjustments do seem to be an issue because while they can make the problem better they can also make it a lot worse. But that would be the only thing that would make sense for all three to go out at the same time.
 
I have had the meter on the power supply during operation. Voltage is good all the way to the head. I am really leaning towards the micro controller at this point but there are two things that don't make sense. Adjusting the tension on the read head springs makes the problem slightly better or worse. Disturbing the natural bend of the wires inside the head makes the problem worse. And the scales run fine for a few passes and then will suddenly stop sending signal. I would love to know a good way to test signal out put of the scales to rule them out as an issue.
It may be that your output signal is marginal so that any small change in strength can cause a disruption. I am not familiar with your scales as I used the iGaging scales for my lathe. In looking at Yuriy's web page regarding connecting glass scales, he stresses that the voltage presented to your controller is critical and must be no greater than 5 volts or controller damage can occur. Is it possible that your supply is capable of more? Switching power supplies are usually fairly consistent but transformer based supplies can have no load outputs as much as 50% higher than their output at rated current.
I had a problem with my power supplies in that transients from the lathe motor shutting down were coming through them. I tried three different switching supplies and tried installing line chokes as well. I finally solved the problem by running with a 12 volt battery. My symptom was a display reset on power off and not related to what you are experiencing except to point out that these systems are highly susceptible to disturbances.

Since the problem is common to all three, it would point to two possibilities: that is caused by a element external to the displays, damage controller, voltage level shift incorrect, etc. or it may be due to a design or manufacturing flaw in the scales, a wire stretched too tight, a cold solder joint due too a malfunction during manufacturing, etc. My eleven year old glass scale system has flexible metal conduit coming from the pickup heads and the conduit is clamped a short distance from the pickup head so the internal wire flex very little there. It is highly unlikely that you would experience an open or short mid cable, especially on a new system and on all three at about the same time.

Intermittents can be a real bear to sort out. A scope would definitely help. I have seen some software based scopes running from the mike input on your computer which may be enough to help you without taking you too far astray. Bang-good keeps pushing ads at me for cheap oscilloscope kits as well. One of those options may give you the tool you need to figure out what is happening.

Again, good luck.

Bob
 
Looks like I have some good directions to go for testing stuff. Today's the anniversary and I am already getting the evil eye for being on the computer with broken tools in front of me. I will get back with some more results tomorrow.
 
Back
Top