tmoulder

Thanks everybody! TM

Greetings all, I've got a system with a V280 and 4 devices communicating over modbus TCP. Each requires a TCP connect, so each has it's own socket init, it's own IP Config block, it's own modbus busy bit. I'm trying to optimize the communication traffic pattern. In CanOpen, the PDOs are buffered, and comms are handled at the chip level. The socket arrangment in Modbus TCP seems to bear some resemblance to this. So here's a question: If I were to send a PHR message on all four sockets within the same ladder scan, what would happen? 1. The socket traffic is independent and buffered in OS/hardware, ala CanOpen. Concurrent messages are not an issue. 2. The first message gets through and the rest get dumped. 3. Strange behavior and error messages. 4. V280 screen displays "Self-Destruct Sequence Initiated". Evacuate immediately. Thanks, TM

Good decision

Background - I don't know if they had this show in Israel in any form, but when I was a kid, there was this program called "The Great Space Coaster" that I watched every morning before school. It was a "Sesame Street" style variety show with puppets, songs, and animated sequences, and it was alot of fun. One of the segments was a newscaster who got really perturbed if there was anything to report. Which is to say that there is an entire segment of american men to whom, should you say "No gnus is good gnus" ("g is not silent") they will immediately reply "With Gary Gnu". I'm still amazed after all these years I randomly run into people that remember this

Hi thomas, Use a negative transition contact -INI- related to the button. I think the M91 has those, it s been a loooong time... TM

Thanks Cara, I added the poll to most of the posts. Although I've had alot of downloads, relatively little feedback about how helpful the material has been. But only one vote for "buggy junk" this year, so I'm assuming "no gnus is good gnus"

interesting point simon, explains something I always noticed, that a black screen seems to glow in a dark room, even though theoretically it should be dark. Thanks, Mr Wizard! TM

Greetings all! Motoman is a Yaskawa-based robot company that offers CanOpen as one of their protocols. The documentation for their card is good for the wiring and installation of the CanOpen adapter card, but documentation about how the robot interacts with the card, and how CanOpen commands are processed is non-existent. Literally, it doesn't exist. I've begun a project requiring interfacing a V570 to a Motoman robot via CanOpen, and the attachments here cover my findings. There's an example program (based on the Wago one posted to my blog), and a user guide that explains how the interfacing works, and what I learned about the Motoman CanOpen inplementation. Hopefully, it will give the next guy some guidance I sorely missed out on. As an added bonus (?), the Motoman card is non-retentive, and a heartbeat has to be set every time the robot is powered up. So if you were longing for a bit about SDO, it's here. Hope it helps! TM Motoman CanOpen Module.vlp Motoman CanOpen.pdf

It's not obligatory, but it is a good idea. The canbus ports of all the PLCs and RC1s would share a single power supply, and that power supply is exclusive to that purpose. TM

Hi Toddster Since the beginning, Unitronics has always recommended that a contact bit not be referenced in the same net as it's related output. Just recently, they added this as a warning condition in Visilogic, just to raise awareness of it. The reason for this has to do with how ladder nets are processed. When a net is processed, all input bits are evaluated, then output bits are set accordingly. If you set an output bit for MB 0 in net 10 (for instance), and use an input for the same bit in net 11, MB0 in net 11 will show as being energized, because it was set in the preceeding net. By contrast, if you set MB0 in net 10, then use a contact for MB0 as well within net 10, then the input bit will not show energized until the NEXT ladder scan, when the input bits of net 10 are checked again. As long as the logic works and causes no trouble, then all's good. But when you reach the point in your programming where you need things to take place in sequence within a single ladder scan, it can cause racing conditions and major headaches, because the processing does not behave in the manner you expect. For this reason, Unitronics recommends you avoid it entirely by breaking the net up and not using a bit in the same net as it's coil. TM

Hi Jared, Thanks, I'm familiar with this method. I was hoping Unitronics could automate it a bit by adding a capability into Visilogic. Thanks again, TM

I've got an idea for pop-up screens that will look really neat, but whenever I try to copy a screen from Visilogic (ctrl-alt-print screen) to use as a background image on another screen, it gets the scaling wrong. Two ideas: 1. How hard would it be to have a menu option in visilogic to export an HMI to an image file? 2. For a gold star, how hard to assign one screen to be the background image of another? Mind you, the background screen is nothing - just an image, no functionality. Thanks, TM

Posted to blog. By the way, I want to take a second to thank Emil for posting that circuit. In mine, the 10K resistor is in series with both the Zener Diode and the Analog input. It got me to thinking about the voltage drop on the resistor and how it would affect the analog reading. But looking at his circuit, I see the input impedence on the input is over 1 Megohm, and so shouldn't pose any issues, a 0.01 attenuation. If you need more accuracy than that, you probably shouldn't be using a potentiometer TM

Hi Emil! That's a great circuit, and will work perfectly if you want to divide down to 10 V for the entire travel of the pot (0-10 VDC full range of motion). The problem I had was a bit different. I needed 0-10 VDC over a *potion* of the potentiometer, to get better resolution over a specific range of physical movement. This is tricky - you can't divide a potentiometer up, obviously. By applying 24 VDC to the pot directly (no R1), and adjusting the physical mechanism, I could get 0-10 VDC over the lower range of the pot while not allowing greater voltages to reach the AI and blow it. The problem is, it's always possible to get a situation where you do end up with >10 VDC at the pot. I had this happen while testing this - loose wire in the connector. Or when the pot is replaced or being adjusted. To prevent this, I used a zener diode and resistor to clip the pot output to 12.7 VDC maximum - overscale, but under smoke point. It provides a safety to the AI in this configuration. Thanks! TM

Caution: This approach can be risky. Analog voltages greater and 15 VDC can fry an analog input on an E1B module. Use with caution, and at your own risk. I had a project where I needed to use a potentiometer with an E1B snap IO to measure an angular position. The problem I ran into was resolution. With 10 VDC applied to the pot, and the E1B resolution of 1024 units for full sweep, the portion of the pot I needed was only seeing 22 units. I had an idea - I put full 24 VDC to the pot, then positioned the pot physicially so I was at the lower end of it's physical scale. This gave me 10VDC at slightly less than half of the pot's travel, which took my measured sweep from 22 units to 92 units. Huge improvement. The drawback was - there's 24 VDC on the pot, and E1B inputs fry at 15 VDC. If someone replaces it incorrectly, gets to playing around with it or even a wire comes loose, poof. So to protect the input, I hunted high and low and came up with this circuit. It uses a 10k resistor in series with a 12V zener diode, all available at your local Radio Shack. It shunts the current in such a manner that the input can never see more than 12 VDC even with a full 24 VDC coming from the pot - overscale, but below smoke point. The zener begins regulating at 11.4 VDC, and below that, behaves like a regular reverse-biased diode. Hope it's helpful! TM

Success, I'll through a drawing up at my blog tomorrow. TM

Digging a bit deeper, it seems the zener might be the way to go. I'm going to pick some up and test them. TM

No, I'm not doing yard work I'm using a potentiometer to pipe 0-10 VDC to an analog input on an E1B snap IO board. The problem is resolution - the active portion of the pot is very small, and since the resolution on the analog input is also tiny, the active area encompasses all of 24 measurement units out of 1024 for full swing. To overcome this, I applied full 24 VDC to the pot, and set up the mechanism so the active portion ranges from 9-10 VDC. This brought my resolution to 92 units, sufficient to my purposes. The problem - at 15VDC or more, my input will go *poof*. Not a big deal right now, but in the future, if somebody changes the pot or spins it somehow without consulting me first, they may well let the smoke out. So I've been exploring ways to limit the potentiometer output by restricting it to 12 VDC or less. A zener does not seem suitable because I need the full voltage range below 10 VDC. I don't want to divide the voltage and lose resolution again. A diode clamping circuit I built was unsuccessful in tests. And lastly, all the documentation I can find on the web is all about clipping and clamping AC signals, not DC outputs. This seems like it would be a common problem in these analog circuits, and I was hoping somebody might have an idea about a solution. Thanks! TM

Beta? I never saw the beta at the FTP site? Am I missing something? Yay pop-ups! How long till we can do our own custom pops?

When other people make typos, their spelling looks bad. When PLC people make typos, things don't run

As a supplement to the example program I posted earlier, I've put together this PDF guide. It explains how the program operates, and goes into some light detail about CanOpen itself, as implemented in Unitronics. Hopefully, it will help give the new CanOpen-er a boost in the right direction. Like everything in this blog, this is open. If you see a correction or think something needs covered that isn't, let me know and I'll add on to it. Hope it helps! TM CanOpen Primer.pdf

http://www.canopensolutions.com/english/about_canopen/guarding_heartbeat.shtml Near the bottom of the page. Glad to help!

Example, albeit for an IO node and not a servo drive. May be helpful though. http://forum.unitronics.com/index.php?/blog/16/entry-51-wago-750-307-canopen-node-example/

I've noticed this too. The top frame of the visilogic IDE window (the "Windows" window) usually displays the name of the project currently open. I don't know how you could get the path in there, I've never seen that. But to display the name without the tilde (~), open visilogic first, then open the project from within it using the Open button or menu item. You get the ~ when you try to open a project by double-clicking it in Windows Explorer. TM

Hi guys, In the bit operand world, we have 3 transitional bits to choose from: 1. P input bits turn on for one ladder scan when the input turns on. 2. N input bits turn on for one ladder scan when the input turns off. 3. X output bits change state when the ladder net leading to them is true. I'd like to see a couple of new additions. I think somebody may have mentioned these before: 1. !P and !N input bits turn off for one ladder scan when the input turns on or off, respectively. 2. X input bit that turns on for one ladder scan when the input transitions, both on and off. 3. !X input bit that turns off for one ladder scan when the input transitions, both on and off. These would mainly be useful in certain timing and comm applications, like looking for the toggle bit in a CanOpen NMT Node Guarding scheme. Thanks! TM