Sign in to follow this  
Followers 0
Conor

Point I/O compared to Flex I/O

12 posts in this topic

Hi guys, I have been reading a good few posts lately regarding Point I/O, and I would like to know if any feedback advice on the following. On my site we have a lot of DeviceNet with mainly Compact Blocks, Flex I/O hanging off it for Inputs/Outputs. If I had a new project and as looking to put in a Flex I/O should I consider Point I/O instead? Conor

Share this post


Link to post
Share on other sites
My company has migrated entirely to POINT instead of FLEX because the density and mix of modules that POINT offers is better for our machines. If your installed base of FLEX is significant, there's nothing wrong with sticking with it. Once you understand the quirks of the 1794-ADN, it's reliable and straightforward to keep using it, and it's not going to change. My preferred POINT adapter is the one that's passive; the 1734-PDN. The POINT modules just appear as DeviceNet nodes and I can manage them that way. In general, POINT is where the product development at Rockwell Automation Singapore is being done. I use the serial modules extensively, and the thermocouple modules are the mainstay of my temperature control applications. Any third-party development is probably also being done in 1756 and 1734 platforms, not in the 1794 FLEX platform. FLEX has had a good run; I started using it as a successor to 1791 Block I/O in the mid 1990s. Compact I/O would never have been born if not for the failure of the FlexLogix to catch on with Solar Turbines. And it will probably be around for longer than the SLC platform... you can still use FLEX with PLC-2 controllers !

Share this post


Link to post
Share on other sites
Thanks for that Ken. I suppose as you say on my older systems I may leave well and good alone and use only Flex. On my newer systems that are more Ethernet focused I may look into point Thanks again, Conor

Share this post


Link to post
Share on other sites
Keep in mind that in a POINT IO system, each module creates a CIP connection.

Share this post


Link to post
Share on other sites
I'm going to be a little picky on this: I agree that it is important to understand the way CIP connections are created and counted, but as a general statement it is not true that every POINT module has its own CIP connection. With an EtherNet/IP or ControlNet based POINT I/O system, the 1734-AENT can have a Rack Optimized connection or no connection. You can place as many of the discrete I/O modules connected to the 1734-AENT into the Rack Optimized connection as you wish. Any modules that are not part of the Rack Optimized connection require their own Module connection. This includes analog modules, serial modules, and other specialty modules (SSI, Thermocouple, stepper control, etc). With DeviceNet it gets more complex, since DeviceNet doesn't strictly use the CIP Connection model but rather scanlists. And you can get the 1734-PDN (passive, where every module gets its own scanlist entry) of the 1734-ADN/ADNX, in which the whole dang POINT subnet is a huge Assembly. CIP Connection counting becomes vital with large systems, or with the smaller 1769 CompactLogix controllers, because they have capacity limitations both for packets-per-second bandwidth and for TCP and CIP connections. That's been greatly simplified with the advent of the L7x family CPUs for ControlLogix and CompactLogix.

Share this post


Link to post
Share on other sites
I'm working on a couple systems using the new 5370 Compactlogix processor. AB is on the right track with changing the CIP connection counting scheme. A point I/O rack can have a mix of digital, analog, and specialty modules and only count as 1 "Node" Connection. You don't have to pay attention to TCP or CIP connections anymore. Anything that is configured in the I/O tree (given an IP address) only counts as 1 node connection. The processor models have connection node limits that determine cost. My suggestion would be to go with Point I/O on all new projects. They are inexpensive enough to keep some on hand as spares without upsetting the bean counters. Flex I/O cost will start to rise in the future as AB tries to direct customers to Point I/O.

Share this post


Link to post
Share on other sites
I'm a fan of FlexIO, if PointIO CIP connections can be managed all into a single connection for one rack regardless of specialty modules that would help convert me. The bigger problem for me is wiring them. With Flex you have plenty of terminals to land your filed wires, and in general a 36"x60" panel will accommodate 2 full FlexIO racks and all the power/field wiring/conduit entries PointIO, it all depends on how compact you are trying to make your IO. Sure on PointIO you can cram 8 digital inputs into a terminal, but you have voltage/common terminations that have to move somewhere else, where as on the 2 and 4 PointIO input modules all the wiring can be landed on the PointIO terminal (to an extent). That same 36" x 60" panel could fit more IO using PointIO, but that means more wires, more conduit entries and you have to change your wire ways accordingly and now that 36" x 60" panel seems awfully congested. Gets more complicated when you get into analog IO. 2 point analog cards are great for landing your field wires, but eat up your CIP connections. I think PointIO is good for machine/skid systems, but I don't think I am ready to use it an a large distributed system yet unless I was trying to use a large quantity of control panels of a smaller footprint say 30" x 36".

Share this post


Link to post
Share on other sites
I used Flex on RIO, but Point on ENET. Point is the way to go now for newer systems. Look at the fact that AB has chosen Point to act as a local I/O for some of the new CompactLogix controllers.

Share this post


Link to post
Share on other sites

We also are working on a project where point I/O is being considered and moving forward could be the standard for many projects with this customer. Normally I am a jump on the new bus, normally as the driver, but with this I have few things that I believe should be considered.

Point I/O , especially if you require OE4's or OB2P's be comes a bit of a mess as you have to provide empty slots with CTM/s on either side of the module or there will be temperature issues. There is a tech note on this. Point I/O is also less density per module, adding to the overall costs, etc. Missing as well is a great power distribution terminal system the 1794-TB3 offer, so you have to add terminals in your panel to accomplish this, more realestate...Finally the big kicker is that the wiring on the Point I/O TBs, is more suited for smaller gauge wore, which I'm ok with, as who needs AWG 14 to drive a 0.5 amp 24 VDC coil, however it does happen so to accommodate , one has to add another set of terminals. We tesed using larger wire and the tension, even relieved to the TB could cause the TB to loose connection to  the module, we found the TB actually flexed...

Just some notes.. I'm voting with staying with Flex...

Share this post


Link to post
Share on other sites

My biggest objection to Flex IO in the Logix 5000 world is analog scaling.  Point modules let you scale to engineering  units in the hardware configuration while Flex modules don't.  You have to write your own scaling code, which is made more annoying by the absence of the SCP instruction in Logix 5k. 

Share this post


Link to post
Share on other sites

Joe, I just took over a machine project where there are symmetrical North and South sides, with identical I/O count and signal types.

The North side is where the CompactLogix is installed, so they used an L36 and local Compact I/O because they needed a PLC that could handle 4 axes of CIP motion.

The South side got POINT I/O to save a few bucks and a few centimeters over a 1769-AENTR and Compact I/O.

So I've got Input and Output handling routines that include little stories about the differences in resolution, scaling, and data packing between the two platforms.   I've edited out all my lines about "For reasons that are lost to antiquity...."

1 person likes this

Share this post


Link to post
Share on other sites

Yeah, whenever I do scaling with the Flex IO, I have a long rung comment explaining the "magic math" that's going on so the next guy (usually me a year later...) can decipher what the weird numbers are.  You know, something like, "0-80mm = 4-20mA = 0-16xxx counts". 

I had to really brute force some basic stuff in an S7-300 that should have been really simple. I left some pretty long rung comments there to explain why I went around the house 3 times to get to the front door.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0