Steve Bailey

Daniel has sent me two sample files for review. I realize that others can't see the specifics of his system, but my observations may be pertinent to general applications of the DSM module, so I'll post them here. In the email in which he sent the files, Daniel writes: There are three %I points for each axis that must be true before the module can execute any motion commands (other than jog). These are "Axis Enabled", "Position Valid", and "Drive Enabled". In your system, these are %I105, %I106, and %I107 for axis 1 and %I131, %I132, and %I133 for axis 2. The Axis Enabled bit will be set if there are no errors present. You need to pulse the "Clear Error" output bit to get Axis Enabled. The Position Valid bit will be set when the axis has successfully completed a homing procedure. In your case, since you've configured "Move +" as your homing procedure, when you pulse %Q28 for axis 1 or %Q44 for axis 2, the motor should rotate in the positive direction and stop when it sees a marker (Z) pulse from the encoder. At that point, the Position Valid bit should be set. The Drive Enable bit should be set when you turn on %Q27 for axis 1 and %Q43 for axis 2. Daniel, what error codes are you getting? Another observation: Your software overtravel limits are +1000 user units for positive motion and -1000 user units for negative direction. At the same time, your jog velocity is 12000 user units per second. this doesn't give you much time to move before you generate an overtravel fault.

I've never used the Horner Profibus module, but I have used the Horner DeviceNet module. I expect that the rules that apply to DeviceNet addressing will also apply to Profibus. To the CPU, the network interface module looks like a big combo input/output/analog input/analog output module. In PLC configuration, you define a starting address and a length for each of those four data types. The %I addresses for the data from all of the input modules on the network must be contiguous, as must the addresses for %Q, %AI, and %AQ. If you want to leave space for expansion, leave it between the address for the network module and the other I/O modules in the syatem. The module itself will use some %I and %Q address space (possibly some %AI and %AQ too). The %I space will be status bits informing the PLC about the health of the network module and the network itself. The %Q bits will be commands from the PLC to the network module. My experience is that Horner manuals are not models of clarity, but Horner technical support is very good. If at all possible, try to schedule a few hours to set up a network module in a rack with a couple of drops on the network, just to get a feel for how it all goes together. Do this before you commit to an addressing scheme so you don't have to change a lot of drawings later when you realize you left something out. It would be even better if you can get the PLC Specialist from your distributor to come in and help you get it set up.

I'll repeat the offer I made on another forum. Post the pertinent ladder logic and the details of the Horner module configuration and I'll take a look at it to see if I can spot anything that doesn't look right. Another set of eyes may pick up something you've missed. Quick things to check when comparing your logic to the manual's examples, especially if you're copying the example logic verbatim: Make sure that the %R registers used by the COMREQ aren't being used anywhere else in the program. Make sure you've put the correct value for the rack and slot number in the COMREQ instruction. Make sure you trigger the COMREQ with a one-shot. Try putting coil on the FT output of the COMREQ to see if it's reporting a failure to execute. If it's executing without problems, there should be data returned in the status register that might point you in the right direction. Are you sending anything at all out the module's port? Try monitoring the output with Hyperterm or Procomm or something along those lines to see what you're sending.

File Name: RTU Process Simulation File Submitter: Steve Bailey File Submitted: 17 Oct 2004 File Category: Other PLC Demo Software This is a simulation of a tank fill/drain system. The simulation runs on your PC while connected to a PLC. When the PLC program turns on the pump the simulation shows the tank level increasing. The simulation also sets and clears discrete inputs represe Click here to download this file

File Name: GE Fanuc Process Simulations File Submitter: Steve Bailey File Submitted: 17 Oct 2004 File Category: Demo Software Updated 6 December 2004. Per requests I have added ladder logic to go along with the simulation. It's in a separate pdf file so if you want to try writing your own code just ignore it. This is a simulation of a tank fill/drain system. The s Click here to download this file

What you're proposing is called synchronized start, and it may work just fine. The decision to use that instead of follower mode is largely a matter of personal preference. If the motion profile is simple, I try to stay away from using programs 1 - 10 with the DSM302 and the APM modules, simply because it takes a second software module to maintain it. Program 0 is not too bad, but if you're using Logicmaster you still need to switch between programming and configuration.

I interpret your first question as 'Can I use program 0 to achieve the same thing as follower mode'? You could try setting up axis 1 to follow axis 2 and writing a program 0 to control axis 1. Remember, program 0 is limited to 20 instructions. With two axes, you'll use up those instructions quickly. There is motion programmer add-on for Logicmaster software that lets you create program 1 through program 10 and download them to the module. Program 1 - 10 are not subject to the 20 step limitation. You can download the motion programmer from the same website as the manuals. Question 2: For each axis, one of the %Q bits assigned to the module determines whether to follow the internal encoder or an external encoder. For axis 1, it's bit offset 29, for axis 2 it's bit offset 45. If you have addressed the module at %Q0001, setting %Q0029 puts axis 1 into follower and setting %Q0030 tells it to follow the internal master. Remember, you also have to set a master frequency by loading commands and data into the correct %AQ registers. Question 3: Without a home switch, configure your "home Mode" as either "Move +" or Move -". Then, to home the axis, you will need to jog it to within one motor revolution of the home position and set the "Find home" %Q bit (offset 19 and 35 for axis 1 and 2). The axis will rotate in the specified direction at the configured "Find Home Velocity" until the next marker pulse. Then it will switch to the configured "Final Home Velocity", rotate through the configured "Home Offset", and stop.

When you configure the DSM302, you define a %AI address. This address is the first of at least 40 and as many as 64 words of data returned from the module to the CPU. If you configure the starting address as %AI0001, then the actual position of axis 1 will be a 32-bit integer at %AI0009, %AI0010. This is all covered in the manual GFK-1464. If you don't have the manual, you can download it from the website. http://www.geindustrial.com/cwc/gefanuc/su...Solutions-m.htm

I don't see it as negative at all. I think that Jay was simply pointing out that the same problem was posted at Phil's site and that some folks had already posted responses. It gives anyone who might be considering posting an answer here a chance to see what suggestions others have already made. There are lots of people who frequent both sites.

I'm not sure what the dip switches are for, but on my card, they're all down. Check your email. I sent you the file about a half hour ago.

Jay sent this same message to me via PM. I sent a message to camel this morning. I'll be away from my desk today, but post the details of your problem and I'll try to help.

Please post the configuration data for the HSC module. Is it set up as Type A, Type B, or Type C? Quadrature encoder or Pulse/Direction? Are you using ladder logic to reset or enable counting? Are you using any of the hard wired signals to control the counter? Let me make sure that I understand your test setup. You have an encoder connected simultaneously to the GE Fanuc module and the Mitsubishi. Both units start with the same accumulated count. You rotate the encoder by some amount and then reverse it back to the starting point. When you get back to the starting point you find that the Mitsubishi's accumulated count is back to its original value, but the GE Fanuc's is one count higher than it was originally. If you repeat the test, the GE Fanuc's accumulated count will gain one count on the Mitsubishi for each cycle. If you change the amount of rotation do you get a larger difference? Does the discrepancy show up after the first direction reversal?, after each direction reversal?, only on reversal from CW to CCW (or CCW to CW)? Are you looking at accumulated count values or is there any scaling being done in ladder logic? Why are you convinced that the Mitsubishi's value is correct and the GE Fanuc's value is the incorrect one? It looks like our schedules are such that we are only going to get one exchange per day. Given that, please provide as much information about your system as possible. Otherwise, this could drag on for days. If you find yourself in a position where you have some time to experiment with the system, contact me through this forum and we can work out how to communicate in real time.

I've done plenty of work with the 90-30 HSC. I even have one available to practice on. Olease define 'erratic counting'. Post your wiring, the configuration details of the HSC module and the ladder logic you're questioning and I'll try to help you get it straightened out.

When you configure the DSM302 for analog operation (as opposed to digital), the module accepts a feedback signal from an incremental quadrature encoder and ouputs a +/- 10 VDC velocity command. You need to set up several tuning and scaling parameters in the configuration of the module. Once that's done, you need to perform a homing operation to get each axis to a known position. There are three %AQ addresses assigned to each axis. To command an axis to move to a specific position, you put the 32-bit value for the position into the upper two %AQ registers while loading the code for the type of motion into the first %AQ register. If your motion profile is repetitive, it might make sense to create a motion program and store it in the DSM302. If you do that, to run the stored program, you need only set one of the %Q bits assigned to the module. That's a quick overview. As always, the devil is in the details.

Every year when we had our meeting to set goals and strategy, the GE Fanuc factory rep would come with all sorts of stats, including market share data. I don't know where they got the information, probably from some outfit that does market research. My guess is that if you wanted the same information , you'd have to pay for it.

I agree that most of the wounds suffered in the process of switcing between different PLC brands are self-inflicted. They arise out of trying to force the new brand to behave like the brand you're accustomed to.

I would guess that the dearth of GE questions is more a measure of their market share than their reliability and ease of use. They are reliable, but no more so than any of the major brands. Ease of use is a pretty subjective term. What's easy and intuitive to me may seem downright obscure to you. If you've been trained on one brand, and all of your experience has been with that brand, making the switch to another brand will be frustrating. I spent several years as an applications engineer for for a distributor of GE PLCs. According to the numbers we got from GE Fanuc, for the North American PLC market, AB has about a 40 percent share, with Schneider, GE Fanuc, Siemens Automation Direct and Omron in the high single-digits or low teens. I've been out of the distribution business for four years, so the stats may have changed, but I'm sure its still AB in first place by a substantial margin.

Neither VersaPro nor Cimplicity (recently renamed Proficy) Machine Edition are going to help you with your Series Six PLCs. DOS based LM6 is as good as it will ever get for those PLCs. You should seriously consider a replacement strategy if you're planning to keep the machinery controlled by the Series Six PLCs in production. GE Fanuc is no longer manufacturing Series Six components. Parts are only available from third party suppliers. LM6 will not run on newer computers with an OS newer than NT. I keep an old computer with a 386 processor on hand to support my customers who still have Series Six. When that dies, I'm not sure what I'll do. If I'm lucky, I'll reach retirement age before the computer.

Any Series Six communications through a CCM module is serial. The parallel interface is through the top 37-pin port on the I/O control module. That's the module just to the left of the power supply in the CPU rack. There was a parallel interface card for PCs, but it requires two full length ISA slots in your computer. There was also a Micro Channel version for the IBM PS2 computer. There are two versions of Logicmaster Six, one version for parallel communications, the other for serial communications. Yes, you can run run LM6 under WIN98. It won't run on NT. I've never tried it on ME or XP, but I don't expect it would run. Even under WIN98 or WIN95, I've had problems with Pentium processors. The most annoying is that if you attempt to use LM6 to check the directory for a list of LAD files, it will crash. I'm pretty sure you're right about not being able to do online edits with serial communications, but I believe that parallel is faster than serial. I never had any problems with the speed of updates when using parallel.

The status bit %S0011 (OVR_PRE) is on whenever an override is active. In addition there is Service Request #18, but that only checks the %I and %Q tables for overrides. Unfortunately, there is no blanket "Clear all overrides" command.

What GE calls the "Isolated 24 VDC supply" is the one that's available from the terminals of the 90-30's rack power supply. That's what I was referring to. If you let the smoke out of that supply, you also lose the 5 VDC backplane supply. I prefer to use an independent DC supply for the I/O, not the one provided by the PLC.

You're loading that rack with some power hogs. The CPU364 uses 7.55 watts, and each DSM314 uses 4 watts. You should be using the high capacity power supply (IC693PWR330). It's still rated for 30 watts total from all three supplies, but the 5 VDC supply can handle the full 30 watts. In the case of the MDL645s, the software is assuming that you're going to power the inputs from the PLC's power supply. The 24 VDC isolated supply refers to the terminals on the power supply module. If you don't wire the inputs to it, the module won't draw from it, and you'll be OK. The same applies to the analog modules. Personally, I don't like using the isolated 24 VDC supply to power the inputs. If the field wiring gets compromised, and you draw excessive current from the supply, the 5VDC goes down also, and the PLC shuts down.