Steve Bailey

There isn't much to repair.The active components are op-amps to turn the single-ended RS232 signals into differential RS485 signals. You could check for broken connections and possibly replace the op-amps. The unit isn't particularly expensive, so if you place any value on your time it will probably be less expensive to replace the device rather than spend the time trying to repair it. Appendix E of GFK-0356 shows a little detail about a different device but should give you an idea of what the miniconverter does. Here's a link to a diagram for a cable that may work without the need for the miniconverter. http://gefanuc.bnsi.net/eforum/showthread.php?t=7026&highlight=cheater+cable

Merry Christmas to all and best wishes for a happy and prosperous 2011. From my own vantage point, I'm hoping for a 2011 that's somewhat less hectic than 2010. Three sons married and a mother-in-law laid to rest.

1. Can you use PME 6.0 to communicate to a CPU 364? If not, what software should be used? PME 6.0 is compatible with a CPU364 2. To communicate between two CPU 364's are EGD's used? (like between Rx3i's) If not, how do two CPU 364's share data? I just created a new PME folder with a CPU364 in the hardware configuration. I added a produced EGD exchange. The target validated successfully. 3. If a 90-30 chassis has an "RTU Master" module, polling Modbus slave devices over RS-485 network, what changes are required to add another device, ie. another RS-485 drop? Is it just some programming changes to add more logic to poll a new Modbus device address and Modbus tag addresses, or is it a change to something like an EGD (but a serial version?) ? Basically is it logic changes or configuration changes? Logic changes to add polling of the new Modbus device address.

Your CPU363 program has three more rungs (598) than the CPU331 version (595). The first discrepancy starts at rung 186, which is right after an ENDMCRN statement. The CPU363 version has a blank rung at that location. There are four ENDMCRN instructions and three of them have a blank rung immediately following. I suggest you remove the blank rungs from the CPU363 version. If that doesn't help, try changing the MCR/ENDMCR instructions in the CPU331 version to MCRN/ENDMCRN instructions in the CPU331 version before converting to CPU363.

For one thing, your scan time will be faster on the CPU363 than on the CPU331, but its hard to picture scan time as contributing to what you describe. Another possibility could be that there might have been some data values in registers that were loaded by the original programmer instead of being initialized by ladder logic. If that was the case, unless the values were saved and downloaded to the CPU363, they are now zero.

I think migrating to a new system would be in your best interest. The IC2000 product has not been offered for over a decade now. I seem to remember that there was a push to convert IC2000 systems to Cimplicity around the time that GE acquired AFE. At the time I was employed by a distributor of GE Fanuc products and I made a sales call with the Cimplicity rep to a facility that had some IC2000 licenses. I believe GE Fanuc/Cimplicity was selling conversion services along with Cimplicity licenses. At any rate, the company was OMYA, and they were a large worlwide user of IC2000. If you can track down an employee at an OMYA facility in Europe, perhaps you can learn something about whether or not the user manuals are available as pdf files. Also, Craig Thorsland, one of the moderators of the Cimplicity subforum at the GEIP forum was the Cimplicity rep I mentioned earlier. You could try sending him a private message to see if he can point you toward a solution.

The IC693CMM311 can only be a Modbus slave. When configured as a slave you don't need ay ladder logic to support Modbus communications. If you need the PLC to be a Modbus master, you will need the IC693CPU363 module which has a serial port. You could also do it with the HE693RTM705 (Modbus master) module from Horner Electric or the Programmable coprocessor module.

For the past year or so, the bulk of my business has been coming from my cutomer's maintenance budgets rather than from their capital budgets. For the two largest capital projects I've been working on the end customers don't seem in any hurry to take delivery, probably because they have been able to fill whatever orders they have with their existing macinery. I have only one piece of evidence that the situation may be about to change. Last week I had a meeting with the controls manager of an OEM who is lining up resources to deal with an expected influx of new orders. He said that all of a sudden there is a higher level of quotation activity and they look serious. He cut his staff to the bone during the downturn and won't be able to handle the orders he feels are most probable with the people he has. If some of the less likely proposals he has out there turn into orders he'll be swamped. I'm getting the feeling there is demand building up in the pipeline but it hasn't yet reached critical mass. I'm thinking "wait until after the election" is less a political statement than an excuse to postpone making a decision. Whe I was doing PLC training for a distributor, we rented a meeting room in the local Holiday Inn Express for each session. If you could make an arrangement like that work for you, it could reduce your overhead.

I could help you but I'm not within driving distance. I was an applications specialist for a GE distributor for ten years before setting out on my own. If nobody nearer to you shows up and you have enough work to justify a multi-day service call, contact me. BTW, VersaPro is not GE's current programming package. The highest version of VP was (I think) 2.03. Since then the programming software is called Proficy Machine Edition Logic Developer. Version 6.5 of Machine Edition is due out before the end of October.

On first glance, neither of your systems appear to be so large as to need the RX7 platform. For system A I wouldn't recommend a 90-30 solution. The 90-30 is a 20 year old platform and the RX3i is its successor. In summary, I would go with the RX3i for both systems. For the details of the redundancy strategy, there are options limited more by the size of your budget than technical considerations. You can go with a simple system which just has a second CPU synchronized with the primary for bumpless transfer between the two. You can have a redundant sync link between the two CPUs. You can have redundant I/O cables from the CPUs to the field I/O. You can have redundant I/O modules. You can have redundant I/O devices. The current redundancy strategy involves a high speed "Reflective memory" module alongside each CPU. This gives you a fiber optic link between the two CPUs to share data that has to be identical in both CPUs so that either can control the I/O without any hiccups when you change which CPU is in control. The manual GFK2300 is the hardware manual for the reflective memory modles and GFK-2308 is the redundancy manual. The brochure GFA-603 is an overview. There used to be a product called Max-On which involved an ethernet sync link, but I don't know if they are still selling it. The manual for it is GFK-2409.

The Workmaster you have was an IBM XT portable. The key component for connecting to the Series Six is the parallel interface (WSI) unit. That's the two-card set with the big 37-pin D shell connector. I don't see it in your pictures. If you don't have it there's not much point in trying to put things back together. If you do have the parallel interface, then any of the parts you need are standard XT parts which means you're not limited to GE stuff. Your best bet for locating parts is probably Qualitrol. http://www.qualitrol.com/html/prod_results.asp?Product_Line=Series%20Six There is a description of the boards in the GEK-25367 manual available in the GE downloads section here at MrPLC.com. That's a compilation of individual module data sheets. http://forums.mrplc.com/index.php?app=downloads&showfile=557 The data sheets are GEK-90764 and GEK-90765 Here's what it looks like assembled. http://www.obsoletecomputermuseum.org/ge_work/

I show the seal-in logic and the latch/unlatch to attendees at my PLC training sessions. I explain that both methods use the same amount of memory and that an observer of the machine who doesn't have access to the ladder logic will be unable to tell which method is in use. Given that, your choice of which method to use is strictly a matter of personal preference. To me, more important than the method used is that you understand whether your particular PLC brand will treat latched or sealed logic as retentive and if there is a difference in behavior between the two methods when you cycle power to the PLC.

I tend to put all of the elements that break the seal on the top horizontal line to the right of the branch end, but there's nothing wrong with the way its written. What are the two rungs you plan to replace this with? A latch and unlatch?

First, VersaPro is not the current programming software for GE PLCs. The highest numbered version is 2.03 (I think) and there won't be any upgrades or service packs beyond that. The current software is called Proficy Machine Edition. Machine Edition can import VersaPro files so there should not be any migration issues. How to: Search variables within a program—(i.e. NALML). Search descriptors within a program—(i.e. No Alarm Latched). Search instruction types within a program—(i.e. ONDTR_TENTHS) You can search by clicking the "Edit" menu and then "Find in Blocks". Also by <CTRL> + F10. In the popup that appears when you do that, you can select search by Address or Name as well as by instruction. Perform online edits. That depends on which model Versa Max. The Versa Max modular PLC supports online changes. The Versa Max Micro does not. Determine free/available reference addresses when creating a new variable? Every address used in the program gets entered into the variable table. If an address doesn't show up in the table, it's available. If, while you're entering or modifying a program you enter a memory type (%I, %Q, %R etc) without specifying an address, VersaPro will select the next highest available address. It only maintains a pointer to the highest address usage for each type, so if you've already used %R001 and %R100, it will tell you the next available %R address is %R101. It won't touch %R002 through %R099. When making ONLINE logic changes, is there a way to test the logic before assembling the logic, similar to Allen Bradley? The logic will have to pass validation before you can download it, but there is no direct equivalent to AB's "Test Edit". Can inputs, outputs and internal points be forced? Yes Is there an instruction set reference or function block reference available for VersaMax? The VersaPro help files.

At the risk of hijacking this thread and steering it way off-topic... In addition to the issues with permits and building/wiring inspectors, are you willing to limit the potential buyers of your house to geeks like ourselves? Because a house controlled by a PLC might not be the home of everyone's dreams.

A new CPU module will cost you more than $500 and you'll have to re-enter the program rung by rung. However, unless there's an EPROM on the CPU or you can be sure the battery hasn't failed, there's no guarantee on being able to extract the program from the existing CPU. I think you said there's a printed copy of the existing program, so personally I'd go for a new CPU module and re-enter the program manually. Be sure you take note of the I/O modules and check for compatibility between what's in place and the latest version of CPU firmware. BTW, did you have a chance to check into the possibility of the power supply being dragged down by the I/O modules?

To clarify Russ' point, Logicmaster requires that you have a project folder opened in order to upload from the PLC to the PC or to download from the PC to the PLC. The name of the project folder must match the name of the program in the PLC. There is one exception to that rule. You can upload from a PLC to a project folder named "Temp" regardless of the name of the program in the PLC. When you first establish a serial connection between the PC and the PLC, the name of the program in the PLC is displayed on the "Select or Create a Program Folder" page. If you don't already have a program folder with that name, you can create one.

Let's ay you just started Logicmaster and you're looking at the screen where there's a box in the middle of the screen that says "Shift F1 ... Series 90 Micro Programmable Controller", "Shift F2 ... Series 90-20 Programmable Controller", "Shift F3 ... Series 90-30 Programmable Controller", etc. Make sure that "90-30" is highlighted along the top of the screen. If it isn't, press <SHIFT> + F3 to make it so. Once you've done that, press F1 to launch the PLC programming part of Logicmaster. When you do that, you will probably come to a screen that asks you to select a program folder. Down at the bottom of the screen there will be a field that says eithe "Offline" or "Monitor". If your cable between the PC and the PLC is disconnected, it will definitely say "Offline". If you have a cable connected between the PC and the PLC and Logicmaster says "Monitor" it means it has established a connection with the PLC. If it says "Offline", it means no connection. There are many reason for lack of connection. It could mean your COM port is not available. It could mean you don't have the right cable. It could mean that the port setup for Logicmaster doesn't match the way the PLC is set up. I'm in the office this afternoon, so if you want to give me a call, feel free to do so. I gave you my phone number in my PM to you last week.

If there is an EPROM present, I believe you can remove it from one CPU and install it in a different one. Then you would have to configure the CPU with the EPROM to load its program from it on power-up. That's speculation on my part. I've never actually done it. Did you receive a private message from me regarding training?

If the handheld won't communicate with the CPU331, then I don't think there's any way you'll be able to extract anything from it. The handheld has the same ability to clear faults and put the CPU back into run mode, but it can't do anything if it can't establish communications. This is a long shot given that the OK LED is off, but the 15-pin port on the power supply provides 5 VDC to power the handheld or an RS232/485 converter. With a voltmeter you should see 5 VDC between pins 5 and 7 of the port. If you don't, all the more reason to remove all of the I/O modules to see if something is dragging down the 5 VDC supply. Excessive current draw that drags the voltage below approximately 4.5 volts would cause the PLC to drop out of run. I don't know for sure whether it would also drop out the OK LED.

Jason, Back when the Series 90 product line was relatively new, GE Fanuc made a strong push to get every 90-30 up to a minimum firmware level. If the system you're working with never got upgraded, that may have something to do with your problem. I don't remember the exact firmware level, but I think it was in the 2.xx range. I second Russ' recommendation to use Logicmaster for extracting the program. If you can locate the installation disks for Logicmaster, be aware that versions earlier than about 5.xx could only be installed on PCs with small hard drives (less than 512 meg I think). Another possibility beyond Logicmaster is to see if you can locate a 90-30 handheld programmer. That ought to work with those old CPUs. If the program is very long it will be a tedious process to transcribe the instruction list, but if that's the only way you can get communications established ... If the "OK" LED doesn't come on, your CPU331 is probably shot and the program in it may be unrecoverable. Normal operation is for the OK LED to blink while the CPU does its inital self-check then remain on steadily. One thing you can try with that one is to remove all of the I/O modules so that you only have the CPU and the power supply in the rack and then cycle power to the system to see if you can get the OK LED to come on. If there are any "smart" modules in the rack its possible that there are firmware version issues between the CPU and the smart modules.

They're both capable of floating-point arithmetic, so if you can write a formula for the output value, you can execute that formula in either processor. Many years ago I worked for Beloit Corp. and if I remember correctly we did it in both the PLC3 and GE Series Six and possibly in the Modicon 984. I know we used all of those platforms on winders at one time or another and I'm pretty sure we used the parabolic profile on those winders. I wasn't directly involved with winder controls at the time, so I can't be certain. The way I remember it. the operator would enter core diameter, starting load, peak load, and the roll diameter at which to end parabolic loading. Rider roll load would follow the parabola from the core diameter, peaking at the max load at halfway from the core to the end of parabolic load diameter. Once the diameter reached the end diameter, rider roll load would decrease at a rate equal to the slope of the parabola at that point until the cylinder pressure was just enough to counterbalance the weight of the rider roll beam.

The ease of moving from being a VB programmer to being a PLC programmer depends on how well you understand the systems the PLC will be controlling. How well do you understand variable speed drives and how they are controlled? What do you need to do to start and stop a motor using a drive. What do you need to do to change the speed of a motor using a drive? How does a PLC handle those tasks? What do you know about pneumatic and hydraulic solenoid valves? Do you understand the different types of solenoid valves available? Do you understand how the PLC programs controlling the different types of valves might differ? A liquid level sensor outputs 4 - 20 mA over a range of zero to 10 meters. The PLC receives the 4 - 20 mA data as an integer between zero and 4095. Can you write a formula to scale that raw data to engineering units? Could you recognize the implementation in ladder logic? What would be the difference in PLC implementation between a PLC that could do floating-point calculations and one that was limited to integer arithmetic? The ease of transition also depends on your understanding of how the PLC operates "under the hood" ("beneath the bonnet"?). Misunderstanding the way a PLC "scan" cycle operates is often the biggest hurdle to overcome for a programmer making the transition from programming in VB to programming in ladder logic.

Congratulations Ken, Whatever you do, don't celebrate by lighting one up! There was a time when I had gone more than a year tobacco-free. I convinced myself I could try one for old time's sake. Before I knew it, I was back to a pack or more a day again. It was another five years before I was finally able to quit again. That, thankfully has lasted. I'll be celebrating 13 years without the weed in another month. Like you, I went cold turkey. My gimmick was to count milestones. If you count in small enough steps, you can celebrate a milestone just about any day, some round number of hours, days, weeks. After one day, it's 25 hours, the next day is 50 hours, the fourth day is 100 hours, the fifth day is five days, then 150 hours, then 1 week, and so on. You say to yourself, "I can hold out until the next milestone", then when you reach it, you set a goal of reaching the next one.

You're welcome Rob, I'm glad it worked.