-
Content count
783 -
Joined
-
Last visited
Posts posted by pturmel
-
-
Or just a bog-standard router between the two subnets. The instance of Kepware will need to be configured to point at the MicroLogix. There's no magic to make the connection happen without that part.
1 person likes this -
Those are supposed to be treated as equivalent, in both new and old versions.
-
I so hate proprietary names for common functionality. Depending on whether VLANs are involved, "NIC Teaming" would be either "bridging" and/or "trunking" and/or "bonding" in ordinary managed switch terms.
FWIW, it is common for hypervisor hardware to have either a single trunk to an upstream switch, or redundant (like Rapid Spanning Tree Protocol) trunks, while all virtual NICs in the guest VMs are bridged in the hypervisor to the appropriate VLANs (if any, in the trunks).
You aren't crazy at all, except for being stuck with "walled garden" hypervisor vendor.
-
Pretty cool!
2 hours ago, mikeexplorer said:utilize these skills to update older equipment
I find such work really satisfying, too.
-
Did you enable any kind of routing on the remote end so packets from your machine know how to navigate to the PLC and back? A VPN itself doesn't provide that.
-
Sounds like you need "register extension" to make a 64-bit register. If you are doing the adding in code, you would use your PLC's arithmetic carry bit to indicate when to add one to the 2nd register. But if the register is external, (like an external digital encoder), you will need to monitor changes in the top two bits of the register to decide when to overflow/underflow into the 2nd register. The pythonish pseudocode would be like this, where E is the external encoder, and V0 and V1 are the two 32bit registers composing the 64-bit result:
# Catch forward rollover if V0.30 and V0.31 and not E.30 and not E.31: V1 += 1 # Catch reverse rollover if E.30 and E.31 and not V0.30 and not V0.31: V1 -= 1 # Copy low 32 bits V0 = E
-
Share the EDS files, please.
-
5 hours ago, BobLfoot said:venerable
!!
Hmm. Good word. Old but experienced. (:
-
I have no idea what you are talking about.
-
It definitely cannot use RIO. Nowhere in the docs does it show support for any PLC-5 devices. Just SLC and MicroLogix.
1 person likes this -
You have to look at the specs for the product. Other common chipset are from "Prolific" and have "PLxxxx" part numbers. These do not behave like real COM ports as well as the FTDI-based devices. I'm sure there are other chipsets. YMMV.
-
If the USB->Serial converter doesn't have an FTDI chipset, expect problems.
-
Yeah, pretty sure you will need to specify your final two angles as 370 and 390 to get the behavior you are after.
1 person likes this -
When you cam an axis to a master axis, the cam function has to be applied for every point the master goes through from one position to another. I don't see how what you describe could ever be possible in a cam function.
-
Sounds like you need a fast heartbeat included in the traffic in each direction, with short preset timers resetting on every heartbeat change. If the timer expires, stop.
After you've fixed that basic oversight in the original design, you can pursue ways to make the signalling itself more robust. (Though you probably just need new high-flex cables. They tend to have short lifespans.)
-
1) Should have been fixable with full documentation.
2) Concur. And they are apparently switching to an subscription model, where your install stops working if you stop paying.
3) Pretty wimpy OPC server, from what I've seen.
4) What? There are plenty of OPC products that will reliably obtain and report data changes down to the single-digit millisecond range.
Aside from all that, the NJ/NX family supports EtherNet/IP for tag access, and is well documented. (I recently included high-performance support in an Ignition driver, fwiw.) There are open source libraries that will help you create the appropriate ethernet traffic, and not just on Windows.
-
Are you using the proper GE cable? The ports on these processors are not RS-232. (I'm traveling and do not have that cable with me, but it has electronics in one end.)
-
Open your factorytalk license admin package. Show everything in your current licenses page to Rockwell and they should be able to help you trace your license back to original.
-
It allows you to re-interpret the bytes that hold integer or boolean data as another format because the variables' memory allocation overlaps. Similar to what can be done with byte array copies. I don't find the unions to be very helpful, mainly because you are limited which data types are allowed.
2 people like this -
19 hours ago, BE said:The down side is everyone invariably uses a different program for VPN access
Not only do they use their own VPN software, but most high-end VPN solutions reroute all network traffic through that company's firewall (in many cases, even LAN traffic to your own file servers). Unless you are diligent about not opening the files of other clients while a particular client's VPN is running, you are likely violating any non-disclosure agreements you've signed with those other clients, and exposing your own corporate secrets, too. I am extremely picky about the client VPN technologies I allow in my infra, and never allow anything that impacts my own LAN traffic. Most clients end up approving the the very locked-down VPN I run through my own cloud server.
-
In general, the U.S. National Electric code requires conductors to be protected from overload at the power source (external to your panel). UL cannot waive this requirement, but might add more requirements.
More importantly, most codes require externally powered conductors to be yellow or have a yellow stripe, to signify that they are not cut by the panel's own disconnect. And, above 60V, will need separate arc-flash treatment.
-
For each bound, and for the current time, convert hour and minute to minute-of-day. (zero to 1439.) Then you can do normal bounds checking on those integers.
1 person likes this -
7 hours ago, skyfox said:IT group's blessing
Here in the U.S., I don't think you would.
-
Yes, that's how I would interpret that.
SNTP omron plc
in CX-One
Posted
Network Time Protocol has broadcast modes (send and receive) and peer synchronization modes, in addition to polling. Simple Network Time Protocol is NTP abbreviated to just polling. I'm not familiar with those specific processors, but it wouldn't surprise me if they implemented the broadcast modes of NTP.