RMC70 class

[Peter Nachtwey]

Actually we have a test bench set up with a RMC 100 and a panelview so the techs can test our temposonic transduacers. (we have had some come back "repaired" that done work. they are internal to the cylinder and are a huge pain to replace)

That is smart. I have been working with Tempo rod for over 20 years.

I have used the RMC as a AB PLC to message to the camera.

This is what I don't understand. The RMCs can't initiate messages, just respond.

I concidered using the DVT's Ethernet I/O (it does have it) to do as you have suggested but it works better to grab all the yz points for an entire row of material. The Message is done with a simple Java type script. DVT has some ABRead/Write commands in the script tool.

That makes it easy to initiate transfers from or to the RMC.

Just as a note I might do future camera projects with a Cognex camera. since Cognex bought DVT I am afraid that the DVT end might shut down.

Hopefully the Cognex cameras will be Ethernet/IP certified and be as easy to work with.

I mentioned this thread to other people at Delta and I told them what I saw at the Rockwell Automation show with the robotic pick and place machines that picked up little pucks. We could do that too.

 

[CharlesM]

(we have had some come back "repaired" that done work. they are internal to the cylinder and are a huge pain to replace)

I had this problem to. I found that when Temposonics repairs they check them on a test stand. The test stand just moves back and forth and gets count readings. I think as long as they get counts they call it good. The bad thing is they can only take so many samples and depending on the length of the sensor the samples may be 0.005-0.025 apart. I wish they had a controller and could use the sensor for closed loop control. Then we would know they are good. However I must say that I have only had 2 go bad in the past 3 years and I have 40-50 in the field running.

 

[SMOKE]

How do you think the RMC would work in this application.
It has to do with vibe bowl and a scale.
material is ran through the vibe bowl to fill a part sitting ona scale.
The fill weight is pretty tolerance is tight to mg.
I am using a PID to control the bowl but I cant get it to run like I want. I have to pulse the bowl when I get close to the end.
I think PID work great on things like temp control but not filling parts.

I think the RMC might work great for this.

 

[Peter Nachtwey]

How do you think the RMC would work in this application.
It has to do with vibe bowl and a scale.
material is ran through the vibe bowl to fill a part sitting ona scale.
The fill weight is pretty tolerance is tight to mg.
I am using a PID to control the bowl but I cant get it to run like I want. I have to pulse the bowl when I get close to the end.
I think PID work great on things like temp control but not filling parts.

Agreed. Fill applications seem to be more of vibration rate as a function of weight or volume left to fill. This formula or table is usually not linear. A PID will work but the tuning must be critically damped or over damped or else too much material will be added. If over damped, the filling will take too long. Tuning the PID to be exactly critically damped is not easy and almost impossible if the system is non-linear.

I think the RMC might work great for this.

It would. The RMC70 is gross overkill but I don't know of any PLCs that would respond faster. You could also consider using a Rabbit semiconductor board. It would be cheaper but perhaps not easier to implement. I would make table or formula that would vary the intensity of the vibration as a function of the weight left to add. As the weight approaches the final fill rate the vibration would be reduced. You can set the scan time to 0.5 milliseconds so the response will be very fast at the end. The faster A to D sample time and higher resolution, 16 bit instead of 14 bit, would help too. From time to time we get applications like this where people just have a need for speed.