Q64AD analog-digital converter

[napster 0]

hi guys, what factors can affect the fluctuation in analog-digital conversion? analog voltage in looks stable. (according to my tester) but the digital values that the q64ad converts has a fluctuation. Edited 30 Apr 2014 by napster

[Inntele 60]

A lot of factors, beginning to the grounding. The multimeter is not suitable to look the wave of analog signal, for this purpose is need to use an oscillograph.

[kaare_t 155]

As Inntele says. And I'm also wondering how much is it fluctuating, and what is your current resolution (normal or high resolution) of the input??? Are you using any kind of filter/meanvalue or are you just reading the raw value? How are you checking the signal (in the prog software which updates every 1 second??)?? We need a lot more info to help you...

[napster 0]

i have this video so you guys can have an idea in how much fluctuation it has. unfortunately the limit is 3.5mb. about the resolution the setting is high resolution mode. yes, im reading the value thru the program, and also the hmi has real time monitoring. and for the fluctuation, its playing around 0.2 to 0.3 Note: on the other machines, the value converted is stable. same configuration, same parameters, same wiring Edited 1 May 2014 by napster

[kaare_t 155]

Well, the reason I'm asking is because "how much is it fluctuating"? If you have an indication value between 0.0 - 1000.0 then a fluctuation between 0.2 - 0.3 is nothing. However if the value is to be between 0.1 - 0.5 then a fluctuating value of 0.2 - 0.3 is a lot. So here's a couple of more specific questions: 1. You've set the card in High Resolution mode (~0-16000), but what is the conversion values (what "output" value will be converted from 0-16000 -> ?? what will 0 be and what will 16000 be?) 2. What are you measuring with the measurement equipment (volt/mA) (what should we expect the digital value to be when measuring)? 3. Are you using any mean measurements or are you reading the realtime value (sampling = 1 / time = 0)?

[napster 0]

i got this arithmetic on the ladder but im not sure what it meant.maybe its related on what you are asking for.(its my first time on analog-digital conversion cards,sorry) the digital value on the HMI is on "mm" -10.000 ~ 10.000mm

[kaare_t 155]

OK. But when you say that the value fluctuates 0.2 - 0.3, do you mean 0.2 - 0.3mm (like 5.2 - 5.5mm fluctuation), or is the actual input value (raw value) fluctuating? Just to explain the program you have: First it reads out the digital values (FROM instructions). Then it multiplies with 10 (so that the raw value will be 0 - 160000), then divide by 16 to make the value 0 - 10000 (I guess that the HMI scales the value to -10.000 ~ 10.000)? Also: What programming software are you using? -If GX Developer; find the TO instructions that initialize the card -If GW2, check the parameter settings to find what is initialized in the card

[napster 0]

correct! 5.2 ~ 5.5mm i got the arithmetic but am not really sure about the values (*10 and /16). guess its the "scaling" am i correct? and yes.the HMI scales the values into -10.000 ~ 10.000mm im using GX Dev Version 8.115V and these are the to instructions. the ch were dealing with is Ch2.

[Inntele 60]

You could put the video on Dropbox or Google Disk and publish the link to it here.

[napster 0]

currently uploading it to 4shared. edit: here's the link of the video for your reference. http://www.4shared.com/video/Ui64NuGcba/MAQ02802.html regards Edited 1 May 2014 by napster

[Inntele 60]

Are the similar machines, working excellent, installed the same place?

[napster 0]

yes, correct. they also have fluctuations but not the same as this one. i also tried the card in an another rack. (another set of plc in my bench). i thought the problem is with the card. Use a regulated DC power supply and monitored the the digital values. and to my surprise the card works just fine.

[Inntele 60]

Then it may due to a problem with a power supply module of the rack or with wiring. Once I had a problem caused by new Q62P PS modules (wide input voltage range). Have been forced to replace it with old Q61P-A2.

[napster 0]

thats my another suspect. i tried replacing the Q61P PS module from a machine thats working fine, and still it is fluctuating. and i also tried replacing the noise filter used for the Q61P. still the fluctuation persist. and i dont really have any ideas left.so i turned in to you guys.hehe Edited 1 May 2014 by napster

[Inntele 60]

The searching of such problems is akin to voodoo ;) Can share some samples of my practice. May be it help you. 1. Three mixer-vats, an ultrasonic level sensor installed on cover of each. Two sensors works well, one works bad. The beam of third has been intersected by paddles of mixer. The hole for sensor is drilled with deviation of several mm to the center.

[Inntele 60]

2. High-precision batchers (the accuracy of batching is +-2 gramms on 10 kg scales) were installed at one site on multistorey metallic construction. Got the uncorrelated fluctuations of weight measuring with period of 1~2 min. When connected oscilloscope I saw two successive peaks, the source of which was unknown. The meticulous search has shown that the source of these fluctuation is two bowls of vertical conveyor, which scraped its cover in two points...

[Inntele 60]

3. The same object as in sample two. When three crashers is turned on got a strong fluctuation that can not be filtered. The cause was that all crashers worked on the same frequency, but with deviation on phase (the natural process for machines). In result it leads to heartbeats of 0,2-0,5Hz.

[Edwin 0]

Try to replace other Q64AD from similar machine? will other Q64AD works fine?

[napster 0]

tried that also. the card that i got from a working machine still fluctuates. and then I install the card from the fluctuating machine to the working machine and poof.it works fine. >_<

[napster 0]

I think I have to get my hands on this oscilloscope. Problem is, I don't have any experience on this kinda tools. >_<

[Edwin 0]

check the buffer address directly from the Q64AD card for buffer address K2 & K9: Average time/average number of times (buffer memory addresses 1 to 8: Un\G1 to Un\G8); Averaging process setting (buffer memory address 9: Un\G9). Try to redo the initial setting for the Q64AD card and test again. Probably format the CPU and install the program again.

[kaare_t 155]

I agree with all the above, except for one thing. We've all discussed that there might be a problem in the Q PLC. However I personally doubt it with all the hardware you have tried to change. Have you tried to change/replace the source of the signal (laser?)? Have you tried to re-cable everything? Are you SURE that the grounding is similar in your test machine and the actual working machine (different grounding could be an issue due to noise). Are you sure that the ground cables and cable screens are connected EXACTLY the same in test machine/working machine? I personally think that the problem here is noise in some way. Maybe a bad earth cable, grounding issues or maybe the source of the signal itself. Remember that high frequency noise must be grounded accordingly, with as much area as possible to "distribute" the noise over as large area as possible (e.g. in the cabinet; wire the shield as short as possible and straight to the backplate of the cabinet).

[Inntele 60]

Of course, it's need to check all you mentioned. While I tend to think the fluctuation may caused by working of plant's equipment.

[napster 0]

thank you guys for your replies. i'll try every single one of your suggestions. thanks regards,