Programmable Relay/ mini PLC as control for boiler + storage + fallback heat?

[pybyr]

Or, maybe I should have titled this thread "Smart Relays for the Digitally Impaired?"

For intro/ context: I need to finish constructing my storage and hooking it to the wood boiler (which is all back underway)-- most parts for that are on hand or on the way. I also need to develop a final plan for a control system.

From what I have seen so far from others' approaches and posts here, control systems seem to run to one extreme (sequenced electromechanical relays) to another (the Nofossil or other similar software-oriented systems- some using some fairly sophisticated industrial PLCs and programming, LCD screens, etc)-- with relatively little in the mid-tech-zone.

I'm fully comfortable with the technology and logic layout of the electromechanical multi-relay-based approach (with a solar-type differential controller thrown in to send water from boiler to storage tank only when the boiler is hotter than the tank)- and I don't pretend to have the software savvy to replicate anything resembling the home-written software approach. I can wrap my mind around analog interfaces and dynamics much more quickly and easily than the innermost workings of digital/code-based systems.

My main aversion to relays is that it seems as if one will inevitably have a real birds-nest of wiring to lay out in the controller box (or to eventually work with in the event of system/ topology changes or repairs)-and, if it is like other electrical/ electronic type projects that I've done before, before you're done, the aggregate total cost of the various multiple parts begins to add up to more than you'd ever expected.

I recently came across a category of device referred to as a "programmable relay" or pico-PLC (programmable logic controller) and they seem to be something that one could deploy to control the various modes of pumps &/ or valves to control wood boiler, storage charge & discharge, + other fall-back heat source.

Some of them have the apparent ability to accept signal inputs that are both on/off (such as thermostat or other inputs)and analog (such as temperature monitoring inputs) which makes me wonder if they could also handle the monitoring/control for the differential function for the storage tank- eliminating the need for a separate solar differential controller.

Here's one example:
http://www.bb-elec.com/product_multi_family.asp?MultiFamilyId=39&Trail=32&TrailType=Top
and an example of a 'demonstration application' to show its capabilities:
http://www.youtube.com/watch?v=-m2gPkv-0tQ&feature=channel_page

...and there are various other makes and models such as the Rockwell/ Allen-Bradley "Pico," the Omron "Zen" or the Siemens "Logo!"

After looking up some of the manuals for these various makes and models, they all claim to be very easy to program and use, although the documentation also seems to be written by people who already know how to use them for people who already know how to use them.

Can one of you (such as NoFossil or some of the other digitally-savvy- deerefanatic, and others whose names escape me at the moment) who really knows programming and automation control tell me whether I am onto something (or please _do_ tell me if this is a blind alley that others have wisely avoided)?

A smart relay that could be programmed to run a wood boiler, storage, and fall-back heat system all in one module would seem like it'd be an excellent middle ground for many of us between the "box of relays and spaghetti wiring" and the "truly intelligent/ elegant" set-ups exemplified by NoFossil's. Ideally it would be terrific if the smart relay, with the right programming, could even handle the "differential control" that people usually handle by way of a solar - type controller. Even if not, a smart relay on a DIN rail with an Azel DIN-rail mount solar differential controller mounted next to it is seeming like a pretty interesting possibility right now.

looking forward to feedback from those with more knowledge on the specifics- and headed back into the cellar for tank plumbing

Thanks


Trevor

 

[WES999]

I don't have any experience with PLC controlled boilers, but I have done some simple PLC programming in the past.

Here is one you might look at: http://www.trolsystems.com/
It is a basic programmable controller that uses conversational programming. You essentially tell it what to do in plain English.

Another one to check out is (broken link removed)

They have some low cost small PLC's.

 

[Gooserider]

I don't know - looking at the units that link brought me to (the SG2 series) it looked to me like something that would be at least as much of a challenge to program as No-Fossil's setup, without offering the flexibility and expandibility that his unit does... I will admit that I'm biased in favor of the NF controller on multiple grounds, but looking at it from where you seem to be coming from, I don't see that there is a lot of difference between those boxes and a NF setup, except where there is an advantage to the NF box.

I may be missing something, and if I am, I hope NF will step in and correct me, but here is my take on the comparison between the two...

1. Hook-up, about the same... In both cases you will have digital and analog outputs that you will need to wire up to the boxes terminals in an appropriate way. There might be some differences in the types of terminal blocks used, or maybe the sensors, but no major differences - either way, you'll have "gozintas" from the sensors and "comezoutas" that will turn on pumps, zone-valves, boilers, etc... The only "extra" you would need with the NF box is possibly a network cable.

However with the NF box you can expand it (using standard off the shelf hardware available from many vendors) to include a very large number of inputs of different sorts, and an equally large number of outputs - I forget the effective limits, but they are probably more than you would need to run a medium large office building, let alone a house... Everything in the NF system is available as generic off the shelf hardware, you don't have to get it from him...

With the B&B;boxes you will have a very much smaller set of available inputs and outputs - that page seems to say a maximum combo of 44 inputs and outputs. If you get less than the maximum I/O at the start, or want to change things later, you are probably locked into their hardware platform, and will have to buy expansion from them, at their prices, IF they are still supporting the platform....

2. Software - the B&B;boxes come with some, they don't say where it comes from, so it is presumably proprietary - meaning you don't get the "source code" and are dependent on them to fix any bugs or other problems - if they want to... The NF box also comes with software - Linux based, with some added programs written by NoFossil - you get the "source" for everything (you don't have to give NF a dime for it, though you should at least hopefully offer him a case of adult beverage... :coolsmile: ) This means that you can just load it up and run, but if there are bugs or other problems, you can look at it yourself, or get any programmer you can find to look it and fix or customize it to your needs - you aren't depending on NF to fix problems, although he's a nice guy and might be willing to do so...

3. Programming and configuration - With the B&B;box you program it either via that little keypad and screen (an interface like programming your VCR - I'd hate to try doing something as complex as a heating system with that kind of interface) or on a PC using some extra cost software, and download it to the box w/ some extra cost hardware. With the NF box, you connect to it with your PC over a network (presumably your local home network, but could be done from anywhere via Internet) and configure everything with your web browser using the server interface loaded into the NF box.

I'm not sure just how the programming interface works on the B&B;boxes, looks like they require you to use either a ladder logic diagram, or some sort of flowchart type interface...

With the NF box, he gives some pretty good documentation on his website and the interface looks like a really simple thing to set up... You start by assigning names to all the inputs and outputs - so you don't have to remember that "Input 0x1A" is the solar panel output sensor and "Output 0x8B" is the boiler circ pump, etc... Then you create a set of "rules" that describe what the system does, using "Boolean" logic... i.e. "IF thermal panel sensor more than storage tank sensor AND storage tank sensor less than storage tank max safe temp THEN turn on solar circ pump" - these rules can be very simple, or quite complex depending on needs and how complex the system being controlled is... The only limit is the size of the memory on the NF controller board (which IIRC can be expanded...)

4. Monitoring and Logging - No mention of this on the B&B;boxes, any of us that have followed the threads here for long have seen examples of how much NF's box allows... Of course, just because you CAN put your house on-line doesn't mean you MUST... But I do like the idea of being able to see what the system is doing from the comfort of my office chair.

5. Cost - Looks like the B&B;setup will run somewheres between $100-500 depending on the options you choose. I think the NF box is a bit more, but not that much, and one can save quite a bit by building some or all of it yourself, however I think the NF box gives a lot more value...

Gooserider

 

[Nofossil]

Pybyr is right - there's a gap between simple relays and full-blown PLCs. The gap is typically populated by systems with a combination of multiple simple controllers (such as Tecmar differential controllers), aquastats, relays, and other devices. The units that Pybyr found fit into that gap as well, and appear to offer some nice capabilities.

I started out with very much that sort of system. What pushed me over the edge was that it's really hard to keep track of what's happening with multiple devices that don't really integrate together.

My solution is a good deal more than $500, though. I had hoped to be closer to that price and fill in the gap, but I can't get there without much larger volume.

I'd like to think that I have an attractive solution for complex systems, but the gap is still there.

FYI, my current specs for comparison:

16 sensor inputs
Up to 8 variable speed control outputs
32 discrete inputs (read switch or contact closures)
32 discrete outputs (drive relays or LEDs)

Full differential rules with deadband
Gain and offset calibration for each sensor
Level and Edge-triggered timers
Rule set backup and restore
On-board datalogging of all inputs, outputs, and variables once per minute with 1 year history

 

[pybyr]

The Nofossil controller clearly and completely surpasses the capabilities of the smart-relay-based approach- and I still haven't gotten to the point where I can form my own clear sense of whether the smart relay approach will be inexpensive and easy to use to the point of being worth pursuing. Some of the smart relays start in the $80-150 range, but then one also needs to get the programming software and cable (some makes and models come with those, some don't) and perhaps also the separate differential control.

 

[CMAC]

I have programmed/installed PLC's in my past position as an automation technician. I am now in management of an electrical distributor.
I am new to the wood boiler game (lasy year) and my automation roots have been gnawing/tugging at me.
In my opinion, the gap has closed over over the five years between high end industrial PLC's and the micro relays.
Actually we(Liteco) have a summer deal on now where customers can buy a base LOGO from Siemens for $125 Cnd. Unbelievable really.
I believe the Siemens LOGO to be the best micro PLC on the market today, actually the micro PLC's probably all roll off of the same assembly line-its the Software that makes the differerence. The LOGOsoft is very very good. I won't use the word user friendly because that depend upon user.
On the the micro PLC's the limiting items are: communications (networking),advanced math functions,and the I/O count. Although for most home applications the I/O count should be do-able with expansion modules.
Anyway- My plan is to use either a LOGO or S7-200 (or Omron mid range processor) for my application in the next year or so.
thanks

 

[Nofossil]

I also spent many years designing and integrating systems incorporating PLCs and other computers. PLCs in particular have gotten much better. However, when I set out to automate my own installation there were too many things I couldn't get done with available PLCs at a reasonable cost.

My approach (and I'd suggest it to anyone going down this path) is to start out cheap and simple, and understand the problem that you really need to solve. Then look at solutions starting with cheap & simple, working your way up the food chain until you find the right combination of cost, functionality, and simplicity. If you can do it with straight relays, don't get programmable relays. If programmable relays will solve it, don't get a mini PLC, and so on.

Don't even think about an NFCS system unless you really need the I/O count or the datalogging or the remote access or the web based monitoring and programming ;-)

In all seriousness, the simplest solution is almost always the best.

I have programmed/installed PLC's in my past position as an automation technician. I am now in management of an electrical distributor.
I am new to the wood boiler game (lasy year) and my automation roots have been gnawing/tugging at me.
In my opinion, the gap has closed over over the five years between high end industrial PLC's and the micro relays.
Actually we(Liteco) have a summer deal on now where customers can buy a base LOGO from Siemens for $125 Cnd. Unbelievable really.
I believe the Siemens LOGO to be the best micro PLC on the market today, actually the micro PLC's probably all roll off of the same assembly line-its the Software that makes the differerence. The LOGOsoft is very very good. I won't use the word user friendly because that depend upon user.
On the the micro PLC's the limiting items are: communications (networking),advanced math functions,and the I/O count. Although for most home applications the I/O count should be do-able with expansion modules.
Anyway- My plan is to use either a LOGO or S7-200 (or Omron mid range processor) for my application in the next year or so.
thanks

 

[pybyr]

The NFCS is certainly the apex of approaches- but as I have dug into this further, it does seem as if the tiny PLC/ Smart Relay systems could be a good mid-tech approach. The Siemens Logo! units and their software do, from all that I've looked into so far, look to be the most versatile.

 

[Medman]

I'm sitting in my shop now going over the catalogs and specs for mini PLC units. I had initially thought of using a unit like this for my system, but I was able to accomplish what I needed at the time with standard relays and snap-disc temp sensors. Now that I am looking to reconfigure the system I am being increasingly drawn to this type of device. For a limited number of inputs (say, 4) and outputs (say, 4) which Siemens LOGO would be best?

 

[Nofossil]

My initial solution was exactly what you did - relays and snap switches. I'll also agree with your implied issue with PLCs - the literature seems aimed at people who work with them full-time. There's a substantial learning curve just to understand which model and options you need to make a complete system, and then another hurdle to figure out what it costs.

It would be a GREAT service to this community if someone could cut through all that and say "here's everything that you need to buy to get a complete system with xxx capabilities, here's what it costs, and here's how to order it".

In all fairness, it's not just PLCs that have this problem. I spent a lot of time on relays and relay sockets to figure out exactly which pieces go together.

I'm sitting in my shop now going over the catalogs and specs for mini PLC units. I had initially thought of using a unit like this for my system, but I was able to accomplish what I needed at the time with standard relays and snap-disc temp sensors. Now that I am looking to reconfigure the system I am being increasingly drawn to this type of device. For a limited number of inputs (say, 4) and outputs (say, 4) which Siemens LOGO would be best?

 

[pybyr]

This is looking increasingly promising:

"Klaus Hiller lives happily with his family in Reicheltshofen in the beautiful Oberpfalz region. But he did have one problem or another with the cold time of the year. This was less to do with the weather than with his heating system, consisting of a conventional oil boiler plus a heating controller, a wood boiler with storage tank and a hot water tank. The heating controller responsible for the heating circuits and the control of the wood boiler was not optimally adjusted. There were deficits regarding the operation and convenience. But a complete replacement was out of the questions for Hiller. That’s when the idea of simply adding LOGO! 12/24 RC, an analog module AM2 PT100 and an extension module with four inputs and outputs to the heating system came up"

see full item at-

(broken link removed to http://www.automation.siemens.com/_en/portal/news/products_detail.htm?rssItemURL=/detail_rss.php3?template_id=23&id=7960)

I spent the last couple of evenings, after getting my son to bed, reading and trying to digest the various Siemens LOGO! tech manuals for both hardware and software
< ... that's what passes for a wild night around here these days ... :coolsmirk: ... >

These units- _if_ set up right- seem to have a lot of potential to do a lot of aspects of system control, including perhaps differential control for boiler-to-tank charging, and perhaps even responding to and controlling heat calls from the heated area (s).

But it definitely seems as if each situation would require a specific choice of modules and programming.

The "Basic" units seem to have at least 4 inputs and 4 outputs- and you can choose various models that run off of 120/230 V AC or 12/24 DC, as well as choices between analog and digital inputs. There's even an expansion module ( AM 2 Pt100 ) that accepts input from 2 temperature sensors [and a modest number of the various extension modules can be 'stacked' onto the main control unit for additional inputs and outputs.

I'm tentatively thinking that a basic module, plus a pair of the temp input extension modules (allowing sensing of boiler temp, bottom storage temp, top storage temp, and stack temp, might offer a really decent platform for interfacing my boiler, storage and heating. I just want to get more comfortable that I will indeed be able to get the differential control for storage working, and that overall I'll be comfortable/ vaguely competent at laying out the programming's functional block diagrams to get my desired overall function.

For others looking into or considering this, here are some links to the manuals- I also continue to be interested in others' insights as to how this might best be done, and whether it's worth the effort

manual for the Siemens software (includes use for heating function at page 153:

http://support.automation.siemens.c...ib.csFetch&nodeid=24891957&forcedownload=true

Overview of capabilities:
http://www.google.com/url?sa=t&sour...FA+PS4&usg=AFQjCNHPMu9iGpc30zCAzUaCTJzyAT6IxQ

Logo system manual:

https://www.automation.siemens.com/...sionId=15285755147&Language=en&GuiLanguage=en

 

[overshot]

The first thing I would do is - decide what you want your "system" to do. This is how I start all my control designs. If you dont determine what you want your system to do first - how can you buy a controller?

 

[phantom]

I might suggest another possibility as well for a simple controller, the Arduino open source devices units which WoodNotOil turned me on to at the beginning of the year.

I am using one right now to control my entire system. They have several pros going for them including cost around $30 for the basic unit, easy to program using Basic, and very flexible.

My control system consists of the Arduino cpu, 10 solid state relays for output, 5 standard relays for input, several OneWire temperature sensors, and an LCD screen. My solution definitely is not as elegant as others but it works great, was cheap in total probably under $170 and it's easy to reconfigure things in software.