ground loop + BNC 2120 behaving like antenna(problem)

[tmdar]

Hi,

I am using DAQ 6024E, and using two sensors and one error message "data is overwritten before it could be read
by the system try divide the signal before measurement" --it took me whole day as to what data is going inside that my computer with 900Mhz processor is unable to handle. I started tearing my hardware step by step. I looked at my range sensor which has one LED which blinks as it throws waves out - but the moment my motors gains speed all of a sudden my labview generates error that "data overwritten before it could be read by the system", now at the same time I looked at my range sensor and instead of blinking slowly like persec it was blinking continuously without break so I took the wire out which was going into the BNC 2120 and the sensor becomes stable again.

I thought that may be the wire is picking up RF signal from motors so I changed all the wires with shielded insulated wire but again as I switched my motors on labview generates the same error. If motors stays off there is no error. I felt my BNC is acting like antenna and is catching motor RF signals. I took motor out of the circuit and supplied separate voltage from another power supplier and it confirmed me that moment my motor speeds up the range sensor speed of blinking increases but If I take out the signal wire going into BNC 2120 then there is no effect on range sensor and it works fine steady. moment I connect BNC with sensor and motors ON then sensor behave erratically with led blinking extremely rapidly.

So it is confirmed that nothing is wrong with hard ware of computer but some strange phenomenon.

Now next I powered my controller with a battery and motors with another power supply to break the loop but problem not solved, I think there is some ground loop thing going on and that is making BNC 2120 as antenna for RF signal created by motor.

That's why initially at less voltage/speed VI works fine but as motors speed up VI stops working and generating the error " data is overwritten before it could be read
by the system try divide the signal before measurement"

I hope now you are in better situation to understand what is happening and how can I stop BNC wire to act like this, remember I am using good insulated shielded wire.

Thanks

[johnsold]

What kinds of sensors are you using and what are you sensing? How are they isolated from power lines and from ground? Do they use a commmon ground? What is the size and type of motor? Is the ground for the power to the computer connected to the ground of the motor and controller?

Generally speaking motors and data acquisition devices designed for use in computers are not very friendly to each other, as you are discovering. Try to be as explicit as possible about the hardware aspects of your system and someone on the forum will probably be able to help.

Lynn

[tmdar]

Hi,
I am using two sensors,
one is ping ultra sonic range sensor(#28015) from parallax, other is simple photo reflector P5587 digital encoder.
I am using two DC motors 9V, these gave me high torque like 100 in-lb (www.hoobyengineering .com) motor number item 1205.
I am using two power supplies one for basic stamp micro controller other fro motors. both power supplies and computer and computer are plugged in a socket in wall. So inlarger perspective the garound is same for all three equipment although powered separatley.
Rest story is same if I remove the signal wire which is going into BNC 2120 thenmy range sensor works fine with or without motors but as soon I connect BNC 2120 then range sensor becomes unstable as my motors speed up. I can see this by looking at the LED of range sensor which blinks smoothly initially and as motor speeds up they become continous ( which they should not -- they should continue to blink steadily)
Then I tried to use the batteries 9V instead of the power supplies for motors and basic stamp, but problem stayed as I powered the motors and if signal is connected to BNC 2120 the range sensor picks up the RF and VI stops functioning "data overwritten before it could be read by the system"
if I disconnect BNC 2120 and only basic stamp editor is connected then every thing goes fine.

I hope I have been able to explain my problem

[tmdar]

Hi,
one more thing I am giving the signals to ctr-gate1 and ctr-gate0 (i.e PFI4 and PFI9) if I check the voltage across ctr-gate and gnd I see 5V on them-- is it correct?

[tbob]

I am not familiar with the ultra sonic sensor, but could it be that it is picking up sound vibrations from the motor? It may not be a ground loop problem. Try shielding the motor sounds using acoustic insulation or styrofoam to see if it is a sonic problem. Otherwise, you could use an isolation transformer to separate the power supply for the motor from the rest of the system. This would eliminate any ground loops or feedback from the motor through the power lines. You could put a scope on the DC power lines to see if there is some ripple when the motor comes on.

[AnalogKid2DigitalMan]

The fact that all supplies are plugged into the same strip (AC ground) is not that important. What is important is how all the DC grounds are connected. Your motors are pretty high torque, I suspect they are the culprit (the link you provided was bad, so could not check what type motor they were i.e. bruched, brushless). As they speed up they generate more RFI/EMI, this is probably spiking ground too.

Without a schematic diagram of you wiring, I cannot provide a concrete solution. If you have an oscope available (highly suggested you scope things out), that would be useful to look for source of noise and how to quench it. I would start by tying all DC grounds together, are your shields tied to ground only at one end, how DAQ card is set up (NRSE, differential, etc). If differential then do you have bias resistors from AIN+ to AGND and from AIN- to AGND as suggested in manual. I am assuming the BNC is used for an analog input on the DAQ.

Good Luck

[AnalogKid2DigitalMan]

Darn tbob, you beat me again by a minute, this time regarding oscope.
I don't mean to jack the thread, but...

Just curious tbob, what do you do for a living, you seem quite active here and have a lot of hands on and real world experience

-Pete

[tbob]

I've got fast fingers, I play lead guitar in a rock n roll band (nah, just wish I did, but I do play lead guitar - just a bit rusty at it right now). I've spent the last 20+ years being a test engineer at various places. I learned from some of the best at AT&T and Bell Labs (when they were a big dog in the 80's). Before that I was a mobile radio repair technician for 8 years. Learned RF and electronic repair on the job. Troubleshooting background has helped me tremendously in the test engineering world. Also spent three years doing SCADA engineering so I am familiar with PLCs (Motorola, Allen Bradley, and GE Fanuc) and HMI software (Wonderware and GE Cimplicity). SCADA was fun, but I prefer the RF world.
Getting back to the problem, you give good advice about grounding only one end of the shielded cables. Grounding both ends does cause problems with non RF signals. A scope will go a long way to determine where the problem is coming from. But since he mentioned ultra sonic sensors, I am wondering if the problem is actual sonic noise from the motor. I would not be able to hear it because I lost all hearing above 12KHz from all that loud guitar playing when I WAS in a rock n roll band.
BTW, your name reminds me of the Rush song, Analog Kid, and also of the progression of analog synthesizers to digital ones, and now there is a movement back to the analog dinosaurs again.

[tmdar]

hello to both of you
I am using DAQ 6024 E in differential mode and I am using bnc 2120 as it is. I only tried to see the voltage across PFI9 and DGND and I read 5V across it without being connected to sensor. Is it correct. Should there be 5v in ctrgate1 and dgnd.
Since I am supplying 5v signal to ctrgate1 therefore there should be no output from the ctrgate1.
Should I say that BNC 2120 is wrongly been configured some how.

What do you say?
Thanks for help

[tbob]

If PF19 is a counter gate, you will probably see 5v on it without any connection. Usually those inputs have pull up resistors that bring up the input voltage to 5v in case there is no input connected. If it were not for the pull up resistor, the input would be floating with no connection and the op-amps inside the board would either begin to oscillate or its output would go steady high or low (there is no predicting which way it would go). The pull up resistors ensure that the output will follow the input. Some sensors provide a ground for a low, and an open circuit for a high. With the pull up resistor in place, a ground will cause 0v input, and an open will cause 5v input. So if you see 5v there, that is normal.