Design Considerations For "Larger" System

[RBPLC]

I'm soliciting feedback and opinions regarding the use of BRX on a "larger" system. The system I'm looking at would potentially need a BRX and one EBC100 for the "local" control rack with an additional four or eight (this hasn't been decided yet) EBC100's for distributed I/O. Wiring is already present at the "local" location for all machine control and I do not seek to change this. The local rack will need 64 24VDC digital inputs, 144 relay outputs, a couple of TC inputs, a couple 0-10VDC inputs and 12 4-20mA outputs. The remote racks will be added to provide machine diagnostics and will consist of TC inputs, 0-10 VDC inputs and 120VAC inputs. These remote racks will be input only and will provide additional diagnostic information but will not be used for direct machine control. One of the initial issues I came across in initially thinking about this system was the limited power available on the BRX to power all 144 relay outputs. These relay outputs would have to be split across the main BRX and the local EBC100. Concerning a system of this type, are there special considerations with the BRX that I need to take into account and what has been the general experience of people who have installed a BRX system of this type? This is a fairly broad question but I'm just seeking input from others who have built similar systems and some of the issues they've ran into that they might not have anticipated going in to the project. (This is a refrigeration system with standard solenoid/contactor/SSR based controls and will have no high speed I/O or control.)

[brucek]

I think you will be fine. We have systems with the H2 that have 8-10 remote racks and on the company network for remote capabilities. We have started using the brx but it wasn't out when we started converting the DL 260's to Do-more. We have used shielded and and unshielded cat 5 and haven't noticed a difference though now we use more shielded just to be safe. I've noticed the BRX remote I/O is more rugged than the Blue Field I/O where we have had blown inputs on the field I/O ( none on the BRX cards) but we didn't have the snubbers in the aux with the coil feeding power to the aux.

We do have a few BRX units with 5 remote racks running for about a year now with no problems at all that i can remember.

It would really rock if they could somehow make the BRX link to the next rack of BRX, if you need 12 cards you don't need a remote rack, kinda like the DL 450 racks had a cable in the back plane that connected them or add a power supply like the field I/O.

[davidbgtx]

Like brucek, I mostly have H2's in my larger systems as they were preexisting systems using WINPLC's as the CPU. I do have several BRX, but only single rack systems.  However one of my larger H2 Do-More systems has 4 expansion racks, 2 EDRV100's, and communicates to Cmore panel, 2 serial comms, and modbus to Kepware for SQL database logging. Still around 6ms scan time. When I run into power budget constraints for relay output modules I go with the 16TDxx modules "ziplinked" to the zip link relay modules, higher relay contact current rating, with low power budget usage. No worries

[jwbaker3]

We built a system with 3 32 point (96 points) dc input cards 3 32 point (96 points) dc output cards with interposing relays 8 channel analog in and 8 channel analog out on the main rack, 1st remote Brx, 2 32 point inputs (64 points) and   3 32 point (96 points) outputs  with interposing relays, 2nd remote Brx 3 32 (96 points) point inputs 2 32 (64 points) we talk to 17 Yaskawa vfd's with ethernet I/P. We also have 2 BX DM 36 plc's to run 2 robots with 3 encoders (this is why we used the 36 for the high speed inputs) and 2 Yaskawa vfd's and 3 servos on each robot, which we message between all of them. We use interposing relays  (slim interface relays) $6.25 each for the outputs and if the power for the inputs is not DC we use them on the inputs also so we can have the higher input point count per card. This system is working without a problem. We have built many systems with 3 and 4 remote racks with the H2 DoMore  CPU's one was installed in 2013. I just visited the site yesterday and upgraded the firmware on the CPU (so we can use ethernet I/P) and installed a new Yaskawa drive. I know 144 interposing relays is $900 up front but it is cheaper and easer to replace a 7 dollar relay than a $100 - 150 output card later. (IMO) We use them on all of our builds on the outputs even if we don't need the higher point count cards.

JW

[RBPLC]

Thanks for the feedback. jwbaker, what type of slim style interposing relays have you had good luck with?

[jwbaker3]

I was using the AD (KPR-SCE-24VACDC-1) until the stock ran out with no delivery date for new ones. We had to use some from our Schneider distributor (RSL1PRBU) they look almost the same except the plastic is black not white, same size, same rating but more money a little over $17 per relay, every dollar counts on a job that has already been bid. (but better than the AB product that looks the same and the relays are interchangeable $58 each) We also use the card relays from AD (RS6N-DE) they come as 6 relays mounted in a card, they take up a little more room and cost $43.50 for the 6 relay card. (7.25 per relay but you have to get 6 at a time) We have used a lot of these on CIP (clean in place) systems in food plants as interposing relays on the pneumatic solenoid valves. I put in a large system (over 280 interposing relays)  in 2012 with the card relays and have not had to replace any relays yet. We also use the SPDT relays from AD (781-1C-24D) if we need a higher amp rating these are 15 amp and still fairly small but larger that the slim interface relays. The thing we like about the slim interface relays is they take the same space in the panel as a terminal block so we done have to increase the panel size to have the interposing relays. That helps control the cost of the project, most of our competitors don't use interposing relays but we think is is a better, more reliable system, and cost less to maintain over the life of the system.

JW

[Controls Guy]

Don't use the cheapest product in your bid.  Go up a step or two, then if the cheap item is available, it's gravy.   If you bid the cheap item and can't get it, YOU have to pay the difference.  Also, try Factorymation for direct-sale industrial automation parts.   For relays I think they sell Finder (though they used to sell Magnecraft like ADC, only Magnecraft branded).

[Controls Guy]

Also check your Idec distributor for the RV8H series.  I think they might be cheaper than the Schneider ones.

[RBPLC]

Thanks for the information guys. Concerning the KPR-SCE-24VACDC-1 relays from ADC, do they have suppression or opto-isolation circuitry on the input side for DC coils? Literature says they have filters for AC coils but I can't tell if there is some form of coil surge suppression for the DC. 

[Controls Guy]

You don't need suppression for such small relays, at least not for output protection.  I never use suppression even on normal size (10-15A ice cube) relays like the 78x series.  I've done probably thousands of them and to my knowledge have never worn out a single output because of it on any brand of PLC.  Are you thinking for output protection or just general electrical noise mitigation within the machine?

[ADC Product Engineer]

It's not just the outputs that need protected.  That reverse CEMF spike can travel anywhere the power leads go to.  If you aren't suppressing that spike, something is taking the hit.  Eventually it will likely fail.  Semiconductors really hate reverse voltage.

[RBPLC]

It's not just the outputs that need protected.  That reverse CEMF spike can travel anywhere the power leads go to.  If you aren't suppressing that spike, something is taking the hit.  Eventually it will likely fail.  Semiconductors really hate reverse voltage.

What he said. ADC, can you speak to the suppression built in (or not built in) to the KPR-SCE-24VACDC-1?

[ADC Product Engineer]

ADC, can you speak to the suppression built in (or not built in) to the KPR-SCE-24VACDC-1?

The manufacturers documentation is a bit vague looking at it.  It says no RCZ, but it still shows a diode, so... 

I think you will need to shoot an email off to tech support and have them ask the Product Engineer for that part to find out for sure.  I am not in the office or I would do it.    https://www.automationdirect.com/adc/form/user/generalinquiry?ref=web

[Controls Guy]

That reverse CEMF spike can travel anywhere the power leads go to.

Granted.

If you aren't suppressing that spike, something is taking the hit.  Eventually it will likely fail.  Semiconductors really hate reverse voltage.

I believe you, really, I stand ready to be convinced.  Presumably something in my NEXT 3.5 decades will show me the error of my ways.  Maybe what's getting lost is the amount of energy that can be stored in such a tiny mass really isn't all that much.  No idea, I just know I've had zero issues and I've never snubbed.

I do avoid triac outputs, I've seen those fail switching contactors and other significant inductive loads, and solved that by interposing a relay.  Those outputs weren't mine though.  Years before that, I used some on one job (TI545) and they leaked, so I never used another, and nor will I.  If I have AC loads, I'll use either a relay module or DCDO driving an interposing relay.

[ADC Product Engineer]

I'm not here to convince anyone.  I'm here to share 30 years of knowledge.  I always suppress anymore.  YMMV