No announcement yet.

Pid algorithm

  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by Mike Nash View Post
    But to be fair, you won't have any analog I/O until you get to $314, which gets you 1 In and 1 Out. And that price is AC in and Relay out. And you have to buy terminal blocks or ZIPLinks separately. But it includes Ethernet and SD Card slot. Just saying...
    I thought analog was extra on the Click too, which would make the comparison apples-apples.

    As for Siemens? Thanks, but no.
    Dude, Siemens rocks! I felt the same way about the S7-300/400s, but the 1200/1500 are awesome. User-defined types, user defined blocks, libraries, tons of memory, professional programming environment, available in a safety version, and blindingly fast. Good multi-language support (LAD and ST, which is what I use anyway, are excellent, SFC and FBD are a little rough, and no person in their right mind would program a PLC in STL/assembler anyway). Same environment (and mostly transferable user library items) from the $200 model up to the one that will smoke a Control Logix or a S7-318. It's like having all the advantages of a Control Logix, but better (and access to all the blocks you've built) in a PLC that costs the same as a Micrologix 1000. And you can't even buy a ML 1000 any more anyway.


    • #17
      Originally posted by ControlsGuy View Post
      Same environment
      It's probably not as bad as I suspect, but it was the Step 5 "environment" that turned me away. I've seen others using the S7 software and it didn't look too bad, but then, they already knew what they were doing.


      • #18
        Oh yeah, different universes. Even the original S7 families (300/400 from the early-mid 90's) were a vast improvement over the S5, and those were the ones I was trashing in comparison to today's stuff (S7-1200/1500), which has been released within the last five years or so, so TWO generations of huge improvement over the beige stuff.

        The software is now TIA Portal, which in addition to the 1200/1500s, will also program the 300/400's, but not the 200s. Better and easier to navigate than the original Step7, though not because of the tight integration Siemens and everyone else is selling as a benefit, just good solid PLC programming software. I personally could care less about the tight integration and in fact, think it's a mistake in some respects. Early versions of Portal, up to about 11 or 12, were buggy and very slow. As of about 13, it's quite usable, and they're on 14 now. Still a bit of a resource hog, but decent.

        But, even more than the software, the improvements in the 1200/1500 over the 300/400 is in instruction and memory architecture. Greatly upgraded. Much more symbolic addressing for your own DBs (meaning it's much harder to break data access, but it's still as efficient as table access), and performance. Here's a performance head-to-head between a 1518 and a 319; That 319 the 1518's blowing away goes for about $5K

        Ten years ago I'd have used a Control Logix or a Compact Logix over an S7 every time. Today, I say Siemens is eating AB's lunch. Ten years from now, who knows?


        • #19
          Why we want PID in a click. $150 with 2 analog in I can control two heaters or chillers. I would have so many projects for this cost level with a PID and analog.


          • #20
            1) As noted before, many many temp control applications don't need and will not benefit from PID due to the rate at which the temperature changes at full control effort combined with the required temp control accuracy, and will in fact be harmed by its use (by making the system more complicated and more difficult to troubleshoot). Temp control does not inherently imply PID, especially in cases where you have no staging capability and must time proportion an actuator that's inherently on/off. The PID will end up settling at the same duty cycle that on/off control will, and if you have your timebase adjusted on the PID and your differential/timers on the on/off adjusted to protect the actuators without being overprotective, the timebase and max error will end up the same as well.

            2) IMO, it's a bad idea to control multiple systems whose operation are independent with a single controller, especially one that costs the same as a nice dinner. When the controller breaks, it breaks two heaters or chillers instead of just one. No redundancy, just to save $150. Not good.

            3) If this is an application for sale, as opposed to in-house use, and bearing (1) and (2) in mind, only extremely cheap and unreasonable customers will insist on PID in an app that can only pay for a $150 PLC. Do you want cheap and unreasonable customers?

            4) Even if none of the above apply, as noted before you can easily write your own PID, so what's the problem?