Posts by Sean B

Thank you panic mode, that was indeed the problem and the fix!

I enquired with KUKA earlier this year about a system update because we were considering purchasing RSI and they told us they were only selling it for 8.5 and later. Unfortunately the logistics of shipping the controller to KUKA deterred us from doing it.

Do you know if we can update to something like 8.3.42 ourselves without shipping the entire controller to KUKA?

Hi,

KSS 8.3.14, KRC4

I have been using a program to start and stop a trace recording (see pdf, apologies but could only upload one document). Today I changed the $TRACE.NAME[] variable and now I get the message and the .src and .dat files are no longer visible in the directory from the message. I have changed the trace name multiple times in the past without trouble, however, this is the first time I have recorded multiple traces without exporting old ones. The other programs are working fine, it is only this one, nevertheless I would like to find the problem so that it doesn't persist.

I have exported the traces and deleted them then tried a cold start but it didn't help.

Any help would be greatly appreciated.

NOTE: The bool DAQ_VAR is a global variable in global.dat that is modified externally by JOpenShowVar.

Thank you

Hi,

KR C4, KSS 8.3.14

I have been using the trace to measure motor currents. I have noticed while viewing the data in the oscilloscope that it is much preferred when the filter is set to on. I have been exporting the trace data to Matlab but I can only export the unfiltered data. Is it possible to export the filtered data? And if so, how?

Thanks

Thank you for your input SkyeFire and panic mode. I agree that this is far from the way one would do something in a production setting, however, this is a research problem. We have implemented force control with a dynamometer during milling and now we are exploring ways that could possibly achieve the same thing without an external sensor like the dynamometer.

I was also skeptical about there being any response at such a low force but we have been recording a noticeable response at the robots natural frequencies that agree with external accelerometers.

Anyway, thanks for your help!

Yes. Brakes are released and drives are engaged. Robot is "moving" but POV is set to 0 so no motion is actually happening. The force is in the form of a sin wave from a shaker which is why the robot cannot move. The force is approximately 20N.

We are measuring the input force at the end effector with a load cell but we are trying to determine the relationship between motor current and dynamic force at the end effector. We have gotten signals that show a response in the motor currents but the SNR is not very high. As a result we were wondering if the digitally commanded current signal could be obtained because we are assuming that it would have a much higher SNR.

Hi,

KR C4, KSS 8.3.14, KR 90

I have been using the NextGenDrive module and Iststrom channel to record the motor currents. Unfortunately, for this application the robot is not moving and the force applied to the end effector is quite low so the noise in the current measurement is significant. I was wondering if there is a channel that records the commanded current which hopefully would be digital and noiseless? Or even if there is a way to export filtered data? Although I'm not sure how the last option would be any better than filtering during post processing of the data.

Any suggestions would be great.

Thanks

Hello,

KRC4, KSS 8.3.14, KR90 R3100

I am doing some shaker tests to excite the structural modes of the robot and trying to measure the response through the motor currents. I was wondering if there is a way to manually release the motor brakes without starting a motion so that the robot is statically held by the motor torque. I saw on another thread that setting $OV_PRO to 0 will prevent the robot from re-applying the brakes. That solution has been a sufficient workaround but if there is a way to manually change the brakes to on/off that would be beneficial.

Thanks,

Sean

Thanks Fubini,

So if one were to post process the data from the GearPos and DriveMotor variables it would not be as simple as scaling the data with the gear and encoder ratios to get the actual values?

Hi there,

WV 5.0, KSS 8.3.14, KR C4

I have data for motor torques from it seems two different measuring methods.

1. The system variable %TORQUE_AXIS_ACT

2. Using the trace function with DriveMotorTorq_CmdEko

I believe method "1" returns the torque in Nm and method "2" needs to be multiplied with the axis gear ratio to get the motor torque. Does anyone know what the difference is? I'm assuming method "1" is filtered more and is likely less accurate but I haven't been able to find any documentation on the second method.

I have the same questions about position monitoring with $AXIS_ACT_MEAS vs GearPos_CmdFineIpo.

Thanks

Thank you for all your input everyone! I will have a go at it.

Thank you for your response Panic Mode

Quote from panic mode

it also means to never underestimate someone's ability to wreck a good thing..

I very much agree with this, hence my caution. I have been researching KUKAVARPROXY for a few weeks now. I imagine this is hard to answer without looking at the system in question but I'm hoping you have some experience with this.

I was under the impression that KUKAVARPROXY will be able to operate independently from ethernetIP because I have seen papers where KUKAVARPROXY is used in collaboration with RSI. My concern is that KVP and RSI work together because KVP is TCP/IP and RSI is UDP. Is it likely that the KVP server will overwrite/interfere with ethernetIP because they both use TCP/IP? They do not need to work at the same time but it is important that the ethernetIP connection work when KVP is not running (because we do not want to wreck a good thing!).

Hopefully you have some experience/ guidance to share!

Thanks again

Hello,

I am using a KR90 R3100 HA with KR C4 and KSS 8.3.14.

I am planning to install KUKAVARPROXY and Jopenshowvar to stream joint positions to MATLAB. There is already a connection established between the PC, controller, and PLC via EthernetIP. My question is, will using KUKAVARPROXY interfere with or wreck the EthernetIP connection?

Thanks