Howto: PLC Reef Controller II

[Stbringer]

Wow! Plc's at last!!
Will be HAPPY to offer any help on logic design for reefs, data collection, suggestions, etc.
One caution, most plc contacts are rated like 10a for a relay output. There is a catch though, and that is most contacts share a common.
This is not always the case, but often the sum of the contact amperage is greater than the amperage per common. In simple terms, the make/break characteristics of the relay contacts are collectively higher than the ampacity of the board traces that provide power to the line side of the contact. Again, not always the case, but in MANY cases.

I have seen a great many good ideas and implementations here, and am a bit behind. I was thinking of trying a tc/mv card or an RTD card for PH but never got around to checking the electrical characteristics of the probe.

It seems the S word is taboo here, but I am very experienced with accumulated data, data logging, and conversion to excel, etc. to plot reef params.

Again, just an offer for assistance, and I can definitely learn a thing or two here. PM me if I can help!

[Stbringer]

OH, don't know if this has been handled or not, BUT for those of you using backup power, always keep a digital input available for your normal power ok, and one for "backup power on". This will allow you to determine alternative logic for low power, etc. It may even be of use for the plc to switch back to standard power after a preset time saving the battery but sailing through repeated brownouts, etc.

For items like DIN rail, oiltight/watertight buttons and switches, check your local yellow pages for electrical distributors that carry Siemens/A-B/ etc. plc's. They will have it and more.

[inerratum]

Bump, dont want this project to die.

I have the programming cable on the way, thanks easttn, so I can finally start to get mine setup. Still need the thermocouple module and the analog in module but at least I can start getting the basic timers setup soon.

Does anyone have a link to thermocouples that are saltwater safe and will work well for this application?

[Stbringer]

I recommend a Type J thermocouple for the temperature range in this application. They are made of a juncture of Iron/Constantan and in the US color coded Red/Wht. In most applications that need isolation of the probe, a threaded thermowell made of Inconel is used. Others have small already encapsulated probes. Try www.omega.com .
Oh, for accurate readings, you will need thermocouple wire also. Copper wire introduces error and temperature resistivity differences.
I have only purchased for industrial applications before, guess I shoud get looking!

[jgwinner]

Wow, great thread! I bought some uC's a long time ago to build something myself but never finished the project.

These PLC's are programmed in 'ladder logic' or C code?

'soft start' would be nice too, is there a way to syncronize the digital out with a zero crossing?

== John ==

[javajaws]

They are coded with RLL (ladder logic). I'm not familiar with a "zero crossing"...what's that refer to (I'm assuming some phase condition on the AC line)?

[Stbringer]

Naw, most of the outputs are Triac anyway, and only switch on the 0 cross. The relays response, etc. and the time from output set high in data table to output table update to backplane pretty much rule out digitally programmed soft start... Unless you want to bang it a few times on the way to speed.

Most do ladder logic, statement logic, and Siemens has Pascal available. Some small off brands and other controllers may do the C.

[jgwinner]

Zero crossing means a timer signal from an optocoupler based on an AC in (usually the power supply). In other words, it would be a trigger whenever the incoming a/c hits 0. What you would do is latch a signal saying 'turn on' to the pumps, and then when the zero crossing triggers, actually turn on the outputs.

Some solid state relays do this automatically, which is handy. those are probably 5-10 bucks a piece, can't remember current prices. It's been a while since I looked at this stuff.

One thing I've always wondered is if you could run an aquarium pump on 'half speed' by only passing through say every other waveform (DC if you will) or doing 2 on 2 off kind of a thing. The zero crossing detector would allow messing around with that too.

Edit: Triacs don't wait for 0 crossing, they just don't turn OFF until you have a zero crossing. Triac's are two back to back SCR's which is a diode with an 'input' - once you pass in a signal to the input the diode turns on, permanently until you remove power. Handy little things, no heat disipation at all.

Dimmers use triac's that trigger 'on' somewhere halfway up the sine wave. They then turn off at the zero crossing. I've made dimmers for DJ's based on this.

== John ==

[Stbringer]

Oh, javajaws, yea, you hit it.. the ac as it crosses 0 on the up or on the down is the 0 Cross. Most solid state DC outputs are not limited by this, but the actual Hz you can get out of a plc is sort of flaky in most cases, as it relys on scan time, etc. Most of the industrial motor starters/vfd's that would be used with a PLC contain the soft start so they generally don't build that into em.

[Stbringer]

jgwinner, you know your stuff!
(Guess we are typin at the same time, and you are faster!)
Best thing might be to use the hardware for the soft-start if you can rig it.

[Stbringer]

jg, what about firing a triac output to a 1/2 wave bridge set for the upper half, then later kicking on another 1/2 wave bridge set for the lower half??? Been a long time since I did the board level (20yrs) but know the speed limitations of most plc response pretty well, even if using PII interrupts. If you know of some kind of rig that would take a PWM output and convert each pulse peak and valley as a 1/2 cycle trigger, some of the smaller units do have a pwm out...

[javajaws]

How important is soft start and zero crossing? What benefits would that give us?

[DJ88©]

Less wear and tear on anything you need to start up.

A good analogy is dropping the clutch in a car when the engine is at 3,000 RPM all the time. Something will wear out in time.

Listen to a maxijet powerhead when you plug it in. YOu cna lessen the "chatter" with a soft start.

[jgwinner]

I'm surprised these PLC's have such speed limitations. I had been working with a 68hc11 single board computer a while ago and it had the 0 crossing built in. Geez, maybe I should dig that project out of retirement.

The PLC project seems nice and easy though, most everything is built.

The soft start is apparently disputed somwhat - my understanding is that no matter how you start the pumps they 'clatter' but it definitly would increase eletrical noise in any event.

Digital relays usually have a 0 crossing built in, including the 'hocky puck' relays.

Thanks for the kind words Stbringer, been a while since I did any professional EE work. Mostly doing C++, Oracle, IT stuff now.

[Stbringer]

HAHA! PLC's are fast for most industrial needs, and you deserve the praise. You remember what most of us forget, and clearly at that!

Ya, will get some noise, but maybe save the impeller some wear.

Problem with these little magnet motors is they need enough kick to light off initially or they just vibrate...
You are the man to ask, so say we fire the AC output to an SSR, it will then inherently light off the pump(s) at the 0 cross? The plc code can do the latching on/turning off on internal timers no problem.

Most of the 16pt and even some 8pt ac outs are very low resistive current rated outputs. The ssr would allow easy switching of multiples, larger pumps, etc. with virtually no penalty to the ampacity of the output card itself.

I use SSR's from time to time for resistive heating applications...PID algo. in the PLC. Using an SSR on an output for the heater(s) would allow you to kill them for sure in the event of a runaway temperature event (stuck thermostat in heater).

[jgwinner]

Here's a link to what I mean by hocky puck relay:

http://www.surplussales.com/Relays/RESSPlug-1.html

Although they didn't have much on stock today apparently. It'll give you an idea. The description for these don't say but I believe a lot of those have the 0 crossing built in.

[sailfinreef]

I just got back from pittsburg and I think I may be slightly confused about what you are talking about, i was under the impression when you ended your program it would stop the scan and restart it, if you dont end your program it will scan every chip on th plc then restart the scan. As far as a soft start up, I dont think it would be cost effective to get a frequency drive for a few pumps. the drive would not be worth the cost of replacing a pump every 5 five years. I have been using the old school allen bradelly plc 3 and using all the outputs to controll ice cube relays so I dont burn the board up.I thinki I may have not understood something I read though. if Im not making any sense could someone pm me and reexplain it to me.

[jgwinner]

>>As far as a soft start up, I dont think it would be cost effective to get a frequency drive for a few pumps. <<

Not sure what you mean by this - with a uC I would just activate the digital out for 'on' only at the exact instant the zero crossing bit activates. That's no more or less cost effective than anything else, it's just software if you have the zero crossing detector. What do you mean by a frequency drive?

== John ==

[Stbringer]

With a frequency drive, or a flux vector drive, they use like IGBT's to programmatically totally re-work the output to typically 3 phase motors for optimal slow starting, speed control, etc. In practice, even with a VFD rated 3 phase motor, and a fanstatic drive, you seldom want to run it below 20Hz to avoid heating the windings. Many of us PLC guys are dealing with motors ranging from 1-5000 hp, and 480-13,000V where even the rotor mass at rest can be a significant factor let alone if a load is attatched.

With a PowerFlex drive for example, you can program in a startup ramp rate in 1/10 seconds, etc. They are electrically noisy as heck though you could program an output sine of 120V 5 hz if you wanted to.

In the case of the tiny single phase motors we are dealing with, I am not sure how much impact starting on the 0 cross would have, as we will still be going from 0 to 60hz inside of one cycle....

That is why I was babbling about using two outputs and 2 1/2 wave bridges....which would effectively cut the rms output by 50% for starting (IF the little motor would START on 1/2 wave!) then kicking in the other half of the wave after a delay....Probably half baked, but looking for lower cost options...
Maybe one of the analog input ssr's used for heating applications..but still pricey!??

[Stbringer]

Anything cheap out there that would take an analog signal and use it to pass say 0-120v or 60-120v single phase? You could probably work something there for a much softer start and even give day/night profiles...

[jgwinner]

Right, but if there is an interrupt on the zero crossing then it's all just software.

All this takes is 3 things:

1) a 'zero crossing' signal on a digital input
2) a way to set an interrupt
3) a way to have an 'interrupte routine' so that you can turn on whatever you want only when you get the zero crossing signal.

For output, all you need is a radio shack Triac, maybe a buck. I would use one of the optoisolated ones.

I'll hook all this up as a test IF I can figure out number one and two above.

When I was originally doing this work, I was using a 68hc11 from Wheatstone Microsystems, Inc, that had the zero point crossing and the interupt routines (I was coding in assembler) but sadly they seem to be out of business now.

So is it possible with a PLC/ladder logic to have a routine that runs 'only' when an input goes 0 or 1, AND have it do this within say a few microseconds, potentially 60 times a second?

== John ==

[barebottoms]

Originally posted by jgwinner

When I was originally doing this work, I was using a 68hc11 from Wheatstone Microsystems, Inc, that had the zero point crossing and the interupt routines (I was coding in assembler) but sadly they seem to be out of business now.

Was that just 68hc11 based prototype board? Or was there any special software? Sounds like the freebie MIT Handy Board if it was just a uc based board:

http://handyboard.com/

[sailfinreef]

got ya, I was definatly looking at that wrong. its all coming clear now, haha.

[jgwinner]

It was an SBC, about 8" wide by 4 or something, not any specific form factor. The nice thing was that it routed all the A/D signals and the digital out's directly to headers so you could do whatever the chip was capable of.

The handy board looks nice - motor controls even, although in this case we really just need a digital out as I would always use some kind of opto isolated outputs.

The Wheatstone board had the optoisolators on every digital out pin but they weren't big switches, really just protection for the controller itself. I think there were 24 of them. With all the sensors I was planning I think I can exhaust 8 i/o's pretty fast.

I bought it at the time as I couldn't find any proto board 52 pin sockets for the blasted 68hc11's.

The only reason I went with the hc11 was that it had A/D built in, and I had some freeware C compilers; simple, not C++ but I could live with that for what we are doing. Due to the goofed up architecture of the 8051 series, a C compiler is real money, or was at the time. I did most of this work more than 6 years ago. Man I can't believe it's been that long.

So are interupt routines available for PLC / Ladder logic devices?

[Stbringer]

Some models do have STI or Selectable Timed Interrupt or IO based, PII Programmable Input Interrupt. PLC's are still WAY slower than the SBC. For rapid motion control apps, high speed operations, etc other devices are generally needed that the plc simply communicates with.

Nice sounding SBC!