Switching from GEnius bus to SNPX

[cm-nonwov]

Hello,
I have a Ge Fanuc 90-30 system with a Cutler Hammer PM 3000 Power series HMI. Besides the HMI, the customer is happy with the controls side of the system, so we'd like to keep the 90-30 in place.

The HMI is basically worn out, and rather then trying to recondition the thing, I'd like to replace it with a Red Lion model (probably the CR3000).

The current PanelMate is connected through a GEnius Bus Controller.

Can I just leave the Genius Bus Controller in place, disconnect the old HMI (PanelMate), and connect the new HMI (Red Lion) via one of the 3 open serial ports (2x 15-pin RS-485 and 1x RS-232) using SNPX? (after programming, of course)

I believe the firmware in the 90-30 is new enough to support SNPX (vs just SNP), so I don't think that will be an issue, but I'm concerned that the addressing wont work via SNPX if that GEnius Bus Controller is present.

Would I have to make any changes to the PLC if I wanted to go SNPX?

This is possible, but I'd rather not make any changes if they can be avoided.

I was hoping to find a decent HMI that supported the GEnius bus, but I haven't come up with anything.

I've found a GEnius bus gateway device, but it's twice as much as the HMI! Also, it seems rather silly to throw more gear in there if SNPX will do the trick.

If anyone has experience with the GE Fanuc world (or what's left of it) let me know what you think.

Thanks

 

[Steve Bailey]

I used to work for a distributor that carried both GE Fanuc and PanelMate. I don't think I ever sold a PanelMate with a Genius interface.

Can I just leave the Genius Bus Controller in place, disconnect the old HMI (PanelMate), and connect the new HMI (Red Lion) via one of the 3 open serial ports (2x 15-pin RS-485 and 1x RS-232) using SNPX? (after programming, of course)

Yes.

Would I have to make any changes to the PLC if I wanted to go SNPX

No.

Unless you delete the Genius bus controller and its hardware configuration, You'll need to have to start from a clean sheet of paper with the new HMI. Otherwise, any bits that are set by the old HMI will continue to be under the control of the GBC. Data that is controlled by the PLC program and displayed on the old HMI will be available to both the old and new.
Where in Mass. are you located?

 

[cm-nonwov]

Just outside of Springfield. It's a small world

PanelMate appears to support the GEnius bus (as far as the driver in the the transfer software reports).
Just opened it up to check and there is a driver for the "Genius I/O V4.97 PanelMate Power Series"
There also appears to be one, and only one 3 wire connection going from the GBC to the PanelMate. I will be confirming all this just to triple check

Just to be clear, I want to remove the the old HMI, then use the new one to address the 90-30 via the 15-pin rs-485/SNPX and not interact with the GBC at all. Just kinda leave the GBC in the rack.

After perusing the REF VU in LM, there are only two devices that report back as anything to do with GBC. One of them is SBA 31 (which I believe is the GBC controller itself) and the other is SBA 02, which I think is the HMI. The only wire connected to the GBC is going straight to the HMI.

I will grab the schematics just to make sure, and will probably attempt to trace that wire to be %100 sure.

So if nothing else is talking on the GBC bus, and all the other IO is normal rack modules, I figured I could use a new HMI with SNPX (Red Lion has the CR3000) and address them that way.

I can change the config, it's just that I'd prefer to not have to.

Is there anywhere else in LM I should look for GBC configuration info? There do not appear to be any mapping for the IO to the GBC bus. I thought that was kinda the point, to just use the GBC to connect the HMI into the system, then use the normal %I, %q, %AI, %AQ, and so on that were mapped to the different IO blocks.

I have found two different gateway devices, one that does GBC to EtherIP and one that does GBC to ModBus (trying to avoid that one) if keeping the GBC is absolutely necessary.

I should be able to pull the config from the panelmate within the next few weeks when the machine has some down time, that may shed some light on things.

Thanks for the help, neighbor!

 

[Steve Bailey]

PanelMate requires an option card to be installed on the HMI to support the Genius bus. In Logicmaster hardware configuration, if you zoom in on the GBC module and look at Genius address 2, you'll see the addresses assigned to the HMI. There are two blocks of memory assigned, one block that originates in the PLC and gets sent to the device at that address, the other block originating in the device at he address that gets assigned to the defined PLC address. The second block of data is the one I was warning you about. It will contain the bits that are set or cleared by the HMI as well as any numeric data originating from the HMI. If you leave the GBC configuration intact, it will write data to those addresses even when the HMI itself is disconnected. So, if you were to try to use the same addresses in the new HMI any data received through the serial port from it would be overwritten by the data from the GBC every I/O scan.
I still have both Logicmaster and PanelMate Power Pro software, so if have specific questions you can send me the project files for both devices and I can help explain what you're seeing. Send me a PM and I'll pass along my email address.
I'm trying to place what company you work for. I take back what I said about never selling a PanelMate with a Genius option card attached. I think I may have sold one to a company called Fairview Machine. I'm not sure if they're still in operation. Their shop was in the Westover Airbase industrial park.

 

[cm-nonwov]

The setup was probably sold to an SI who, in turn, installed it in our place. This all probably took place when I was in grade school, 400 miles south of here. That's my guess anyway. Let me talk to some other people and see if they can pull up the records on this.. Seems like people's memory does not extend that far back around here.

I'll reach out via PM to discuss local things.

However, I'd like to keep the technical discussion in the forum for now, just in case someone else would benefit from our conversation.

Would removing the GBC from the config and then from the physical rack take care of the serial over-writing issue? I could make a change like that, if it's absolutely necessary.

LM reports Device SBA 2 has an Input1 Ref of %I0497 with a Length of 48
and an Output1 Ref of %Q0121 with a Length of 8.

I'm assuming those are the blocks you speak of, and that if the gbc is installed in the configuration, the PLC or the GBC rewrites what's on all of the rest of the %Is and %Qs with what comes through on those bits?

Forgive my ignorance on the subject.

I do have some time to wrap my head around GEnius, so I'll be reading through the manuals, not just for understanding this project, but now I'm interested in the protocol in general, from a purely academic standpoint.

Thanks

 

[cm-nonwov]

Ohh, I think the answer to my questions may be in chapter 2 of GFK1034!
It almost looks like if you physically yank the GBC out, then the status bits for the 2 devices (SBA 2 and SBA 31) will go to zero and nothing will be input or output to the bits/registers in the PLC.
I'm not necessarily advocating doing that without changing the config, but at a cursory glance, that seems like a (quite dirty) "fix".

Let me pour over these manuals while in piece and quiet at home.

I'm a huge nerd, I love this kind of stuff.

 

[Steve Bailey]

Do you have the spares available to test that? If not, I do.

 

[cm-nonwov]

I read over the GBC User's Manual (GFK1034B) and it appears that just pulling the GBC wont work (at least, in theory).

I'm assuming that if I were to pull the GBC without changing the configuration, the associated status bits for GBC devices would all go to 0. Even if they didn't, they might be in an unknown/undefined state, which is also not good.

The manual states that if the status bits go low, then it will either zero out the memory locations associated to the particular device, or it will hold the last valid data that it receive from the GBC on those bits. This behavior is determined based on the GBC "data default" setting. Neither of those states would work in my particular situation.

So, it looks like, in order to switch over to the still supported SNPX protocol, I would need to remove the GBC device from the configuration, like you said originally. I would then be able to program a new HMI with the same memory locations that the old one used, which means no change to the logic.

Thanks for your input. It prompted me to stop assuming and RTFM!

We do not have any spares. I'll PM you regarding that; I'm definitely interested. That's much better than testing on our production machine!

 

[Steve Bailey]

This evening I set up a test using a 90-30 CPU351 and a Genius Bus Controller talking to a Genius Field I/O network interface unit with 8 inputs and 8 outputs. Inputs were assigned to %I00073 - %I00080. Outputs were assigned to Q00017 - %Q00024. When I downloaded the configuration to the PLC the inputs %I00073 - %I00080 mirrored the states of the signals at the Field I/O drop. When I removed the GBC from the rack but left the hardware configuration intact, inputs %I00073 - %I00080 retained their last state and I could not change them.
When I deleted the GBC from the hardware configuration and downloaded the revised configuration to the PLC, I regained the ability to change the states of the inputs even when the GBC was present in the rack.
This means that if you don't change the hardware configuration you will need to change your logic. Every instance of inputs %I00497 - %I00544 will have to be changed to a different address. Presumably that range of inputs are set and cleared by buttons on the HMI.
I still have the hardware on the bench, so if you need to check out anything else, let me know.

 

[cm-nonwov]

Perfect, that approximates to what I thought would happen after consulting the manuals.
It looks like the best path forward is to delete the GBC in the config, just like you did in your test.
I don't want to have to monkey with logic that's worked for 25+ years if I don't have to.
Thanks so much for your help!