Sign in to follow this  
Followers 0
Bob A.

Interposing Relays or Direct Drive Solenoid Coils?

17 posts in this topic

Hi Guys, I'm having a wresting match with myself again. My practice has always been to use interposing relays rather than have the PLC Outputs drive loads like hydraulic and pneumatic solenoids and I always use supression on every coil, both loads and relays. I'm wondering how many of you will judge this to be be unnecessary and overly conservative. In my current job, there will be close to 100 such loads and so the cost for the relays will add up to an expensive number. The IO is planned to be a combination of AB Flex and Point IO as it is all Remote over EtherNet/IP. The Flex IO will be used where there is a concentration of Outputs in a single location and the Point IO will fill in the speciality stuff in varioius Remote IO situations. The question is whether to use Interposing Relays between the Output elements and the solenoid coils. The process is a manufacturing sequence of about 20 steps that cycle in a pattern about 3 times a minute. All the solenoid coils are 24 VDC and the ratings are in the 20 to 35 Watt range Thanks in advance for any opinions. Best Regards, Bob A.

Share this post


Link to post
Share on other sites
Use interposing relays - for small solenoids I do not even bother with suppression - the relay can normally handle small ones without any problems. I only ever drive LED indicating lights directly from transistor outputs and only if the lights have flywheel diodes built in. I normally use Omron G2RV relays these days - they are small, inexpensive and the commons can be bridged very simply with bridging bars. A lot easier to replace an external relay in the middle of the night than a transistor module - or a PCB mounted relay for that matter. Just make sure the relays have flywheel diodes in them.

Share this post


Link to post
Share on other sites
If they were low wattage solenoids, maybe, I would connect directly, but at the 20-30 watt range, relays always.

Share this post


Link to post
Share on other sites
Hi Bob, Thanks for your input. I appreciate the benefit of your experience. My practice has always been Ice Cube Relays with the 8 or 11 pin "tube" sockets. Since 1N4007s are cheap, they just go right across the coil terminals on the socket. I just learned that about half of the coil loads have the benefit of a diode mounted right in the pigtail plug (not sure whether I like that or not). So far, I have tracked down costs for 4 different interposing relay schemes and at this point, the G2RV is about the same cost as the Ice Cube / socket combo (about US $40) and I have several others that get down as low as about $20. I'll work on better numbers when the business week returns on Monday. Best Regards, Bob A.

Share this post


Link to post
Share on other sites
Hi Ken, I spent some time looking at the solid state outputs and it looks like they are a better choice for long term use with direct to solenoid coil wiring in this wattage range, but the related cost is as much as a relay. So it basicially boils down to panel space and ease of troubleshooting. Best Regards, Bob A.

Share this post


Link to post
Share on other sites
Omron have the G2RV relays on special in Australia at the moment for $7.50 AU each. I normally only pay $7.50 AU anyway from Omron - although I do use a lot of them. Much less space than an ice block type as well.

Share this post


Link to post
Share on other sites
every component costs money and more components means more opportunity for something to go bad. all inductive DC loads, including relays and solenoids, get the suppressor but i use interposing relays when situation calls for it, i do not just place relay on every output. i check the ratings of all components and use them accordingly to their design limits. if something is not necessary, why do it? in about 20 years and several hundreds of machines i have seen way more relay related failures than failures of plc output cards, and i'm not even placing relays everywhere. but this is just my score, someone else's mileage may wary. my reasoning is cost, reliability and time to complete the project. if the cost of relays, panel space, documentation, procurement, labour, testing etc. easily exceeds the cost of the very card we are trying to protect (and maybe by 3-4 times), is this then really sound engineering decision? the cheapest relays are down to 20-25 a piece (relay + base) and i question idea of using cheapest product to protect something but that is not the point. 16pt card costs me 200-300. 16x$20=$320 which is already higher cost (still not including any labor etc.). decent relay cost some $40. 16x$40=$640, looks like i could get two or even three spare 16pt cards for the price of just one set of interposing relays (still not including any labor etc.). 32 point cards cost less per point so this contrast becomes even greater. btw, i leave spare terminals fuses, relays (wired to spare outputs) etc in every panel. helps me out when comissioning if something unexpected crops up and it helps customer to have few replacements for consumables. some never replace what they use and those are the ones more likely to call me for spare light bulb or relay or whatever, instad of taling to their supplier (even when they make me use parts from their "prefered parts list"). i see no problem with replacing anything on a machine (bulb, relay, button, plc card, servo drive or whatever) - if one has the spare part. replacing something that is worn out or bad is simply part of doing business... throwing money out of window is not good business. in my opinion the best protection is careful design and using components within their design limits. my suspicion is that many people do not bother to read the datasheets or calculate. could they be the ones easily convinced that the solution is very simple - just blindly throw in relays everywhere. i've had machine for major automotive customer who insisted that every input and output will go through interposing relay. yes inputs too. in adition they wanted hardwired function to bypass PLC and home machine in proper sequence from any positon, using just one keyswitch and relay logic - in a rare event that PLC may be down (other than couple of servos, everything was on pneumatic). btw they have many hundreds of PLCs, large maintenance team and tons of spare parts. replacing any part or entire PLC and downloading program is a snap. why they insisted on this is beyond me but they got just what they asked. they have replaced bunch of relays, one cable, controller for light curtain etc. but PLC never failed... oh well...
1 person likes this

Share this post


Link to post
Share on other sites
You guys are paying way too much for your relays - see what I pay. In addition the fact that all the commons can be bridged (20 ways at a time) means there is virtually no labour cost for commons - only switching lines. Manual bypass is very common on generator jobs - they even use dual battery chargers and starter motors on many jobs these days. I have even had to put in 2 generator controllers and switch them each month so they are not constantly 'dry'. Dual redundant processor/power supply PLCs are also being utilised much more these days on generator jobs. I am still of the view it is much easier in the middle of the night to replace a relay rather than a PLC or output card. The way I do it the cost is minimal - and if the job is that tight? Ouch!

Share this post


Link to post
Share on other sites
Bob, It appears to me that the currency conversion is almost 1.00. I'll do some more shopping on Monday. I have found that the AB "Equivilent" is about $20 US (700-HLT1Z24). Best Regards, Bob A.

Share this post


Link to post
Share on other sites
PM, I agree with you that the IO outputs can reasonably handle solenoid and starter coils, especially when the proper suppression is utilized which for me is always. But looking at a number of contact specs, it is not all that uncommon to see contact closure ratings on the order of 100K - 300K at rated load. My process will be running at about 3 per minute, so that will get me down to 33K - 100K minutes or 555 - 1665 Hours or 23 - 69 days. With a 35 Watt coil (1.48 Amps), I'm thinking that a 6 Amp contact will last at least a few years. It used to be easy to find contacts rated at 10M operations, but I have not yet found anything close to that now at any price. I'm sure they are still around, but I expect that the price won't be too interesting. As usual, nothing is ever as easy as it ought to be! Best Regards, Bob A. Edited Note: Upon further checking, I have discovered that the 10 Million operations rating refers to "Mechanical Life" and that the "Electrical Life" at rated load is usually in the 100k - 300k range. Edited by Bob A.

Share this post


Link to post
Share on other sites
Another option is Finder - I just do not like the woosey pins - Schneider sell them re-badged Schneider as well. The Omron ones have better pins and the pins are much tougher - they are much more electrician friendly and do not bend as easily.

Share this post


Link to post
Share on other sites
The major concern to look at is will the project work if interposing relays are added? If it is a high speed operation will the additon of the the response time of the relays be fast enough to control the process (in addition to the program scan time + the input card's response time + output card's response time)?

Share this post


Link to post
Share on other sites
Time in this application is not an issue. The repetitive steps cycle about twice a minute. I'm hoping to get under 20 seconds but the speed of a relay is not going to matter. Bob A.

Share this post


Link to post
Share on other sites
Looks like you guys pay to much for relays.... On the flip side.... We pay way to much for Plc hardware !!, in oz that is.....

Share this post


Link to post
Share on other sites
Hi Bob A, Does that not mean that you would actually get more than the 100k-300k out of a relay as you normally wouldn't run a relay at exactly its rated load, it would normally run under this? Also if you are using a dc relay you should look at the ones with the built in diode or if not built in add a diode across the load Conor Edited by Conor

Share this post


Link to post
Share on other sites
Not trying to speak for Bob, but yes, in my opinion, try to use relays with a common socket base, that have built in diodes they are not usually that much more money. I would also say yes, if you already have good relays without diodes, add them, they're cheaper than relays, and do it right with heat shrink and ring terminals ;) As for relay life, we have some 1746-OW16 relay modules that get a consistent 300,000 cycles running 13watt solenoids with diode protection at the solenoids. That's when the first of the 12 in use on that card fails, so individually, a million cycles in a nice clean cool environment with good spike suppression seems reasonable. Edited by OkiePC

Share this post


Link to post
Share on other sites
Hi Okie, Thanks for the statistics. I have been "tied up" with start up "stuff" for a while now. As for the relays, I have about 4500 machine cycles on a dozen relays in the first machine with 35 Watt solonoids, thus far without obvious damage. The diodes currently are on the relay socket but I'm planning to add them to the marshalling strip at the valve stand. The current setup is a "proof of concept" installation and it will be relocated shortly to the permenent location. When I watch the relays, there is still plenty of light coming out of them so I'm thinking that I'll try using diodes on both ends first. Bob A.

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