Author Topic: Indicator LED for valve controller  (Read 1385 times)

0 Members and 1 Guest are viewing this topic.

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Indicator LED for valve controller
« on: June 19, 2020, 05:52:53 pm »
I have a controller for 24 VAC irrigation valves in a pump house and would like to have a flashing LED on an outside the pump house wall that indicates when at least one valve has power.  To be clear, there will be only one LED,  not one for each valve.  Each valve already has an LED, but the valves are located in boxes or cabinets 5-100 m away from the pump house.   

There is plenty of 24 VAC and 9 VDC power available.  The 9 VDC is used inside the pump house to power indicator LEDs which get switched via DPDT relays or DPDT switches.  Other pole of the relays and switches controls 24 VAC to valves.

Simplest and cheapest solution I have devised is to have a small SPST (or SPDT) relay for each valve control  "module" ... relays to have 24 VAC coil. Cost of relays is less than US$ 4.  See the pic.

Other ideas?

Mike in California

Edit:  Yes, I have considered a rectifier for each valve unit to send power to the LED and may go that way.
« Last Edit: June 19, 2020, 06:20:39 pm by calzap »
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
Re: Indicator LED for valve controller
« Reply #1 on: June 19, 2020, 07:15:11 pm »
We can discuss circuit details later once you've picked a way to go:

A diode for each AC valve output, cathodes wire ORed together would provide half-wave rectified DC when any valve is active.  Smooth it and regulate down to an appropriate voltage for a commercially available exterior LED security strobe.

A bunch of AC input optocouplers + a dropper resistor for each to monitor all the AC outputs, with their photo transistor  outputs wired in parallel . . . .

A current sensor on the 24V feed to the switches and relays - if there's negligible current all are off, if there's one or more on the current will be detectable.  Add a comparator for thresholding and low pass filter its output to eliminate line frequency 'chatter' to drive a transistor to control a LED strobe.
 

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Re: Indicator LED for valve controller
« Reply #2 on: June 20, 2020, 12:36:00 am »
Ian,

Thanks for your response.  I've been leaning toward electromechanical relays because it's hard to beat the isolation and environmental toughness they provide.  However, I did consider the diode/rectifier solution, which is made easier by the fact that the valves share a common 24 VAC line.  See the first pic.  It's hard to beat this solution for price.  Do you see any downside in it?

I thought of optocouplers/SSRs but decided they wouldn't be easier or cheaper than electromechanical relays.

Didn't think of current sensing.   The 24 VAC is used for several purposes in the pump house.  And there is are several multiwire cables leaving for the valves.  Not sure there is a single spot where there is just one wire carrying the  common  24 VAC to the valves, but this could be created.  Still doubt I would go this route.

Second pic is the LED circuit I've used for years, powered by 24 VAC and used at the valves.  No smoothing has been necessary (but haven't tried it with flashing LEDs).  No electrical component failures.  Biggest problem was encapsulation.  Tried clear heat-shrink, but it didn't hold up.  Went to polycarbonate tubing and haven't had a problem.

Mike in California


 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
Re: Indicator LED for valve controller
« Reply #3 on: June 20, 2020, 01:27:57 am »
Yes, that would work for plain LEDs if you don't mind the 60Hz flicker.   However if you want to be able to use an off-the-shelf LED strobe intended for exterior mounting or even a self-flashing LED, you'll need to smooth the half-wave rectified DC and limit it to approx. 12V.

The circuit lineup after the diodes would probably be something like a 15R, 3W wirewound resistor to limit the inrush current, charging a 50V 1000uF electrolytic capacitor, feeding a LM317 regulator with a 120R upper feedback divider resistor and a 1K2 lower resistor for approx 13.5V out, near enough the same as the float voltage of a 12V SLA battery, which most security system accessories (e.g. a LED strobe) expect as their supply voltage, and with 100uF,16V decoupling on its output.   That would be good for up to 200mA to power the strobe.  If it draws over 100mA the LM317 will need a small heatsink.   The diodes should be min. 1A, 100V.  N.B. The 1000uF capacitor *MUST have a 1/3A or greater ripple current rating for 200mA load current - if nessercery use a higher value capacitor or two in parallel to get a high enough ripple current rating.  If you need more than 200mA, decrease R1 to 10R, replace the regulator with a nominal 3A Dc-Dc buck module from EBAY set to 13.5V and omit C2, R2, R3.  However its not a good circuit for high load currents - if you need that you'd be better off minimally smoothing the voltage from the combined diodes and using it to switch a SSR controlling a separate 12V PSU.
« Last Edit: June 20, 2020, 01:52:14 am by Ian.M »
 
The following users thanked this post: calzap

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Re: Indicator LED for valve controller
« Reply #4 on: June 21, 2020, 01:59:45 am »
Ian,

Again thanks.  You're very helpful.  So far I've used self-flashing LEDs.  A minor hobby is collecting as many different 5 mm types as I can find  ... 14 distinct types so far.  However, I've always powered these with DC.  For AC powering, I've only used non-flashing LEDs powered with the simple circuit in my previous post.

I have plenty of self-flashers, so I'll try one AC-powered with the simple circuit.  Should be able to do it on the bench next week.  Then if it dies (likely), I'll move on to a 24 VAC to 12 VDC converter (less than US$10 on eBay) with appropriate series resistor.

I really don't want anything brighter.  We're in the country, with homes well spaced, but brighter lights might alarm or annoy the neighbors, especially at night.

The pic is a shot of the pump house with some indicator/trouble LEDs on the outside wall.  There are more on another side of the building.  They are all flashers of different single colors or alternating colors.  Even in daytime, they can be seen 50 m away and much further at night.

The housing is a short, threaded PVC plumbing nipple with a PVC conduit bushing on each end.  Between the bushing and nipple on the outside end is a rubber disc with a small slit.  The dome of the LED is pushed through the slit up to the flange.  These have been in use for 15 years with one failure.

Mike in California


« Last Edit: June 21, 2020, 02:17:00 am by calzap »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Indicator LED for valve controller
« Reply #5 on: June 21, 2020, 07:20:49 pm »
I might use a pair of 24 volt transformers to step the power to each valve to 120 volts AC to drive a neon bulb relaxation oscillator for the flashing.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20363
  • Country: gb
  • 0999
Re: Indicator LED for valve controller
« Reply #6 on: June 21, 2020, 08:30:47 pm »
If you want a big, weather proof, flashing LED, then buy one of those LED strobes. They cost a bit more than a PIC LED flasher, but at least they're properly sealed and will last.
https://www.grainger.com/category/safety/safety-alarms-and-warnings/strobe-and-flashing-lights?attrs=Power+Source%7C24VAC%2FDC&filters=attrs
 

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Re: Indicator LED for valve controller
« Reply #7 on: June 23, 2020, 02:48:13 am »
Thanks for the additional responses. They're truly appreciated.

For safety reasons, I wouldn't step up the voltage.  Much prefer to keep it at 24 VAC or convert and step down to 12 VDC or lower.  Have used LED indicators for years.  They're reliable and having a variety of colors is crucial.

I don't want "big" flashing alarms.  I want small, subtle ones that won't disturb the neighbors.  It's a balance between what can be seen by personnel on our property versus not bothering folks off the property.  Not all of the indicators are alarms in any case; some are status indicators.  The alarms, which are for  low flows or low pressures, aren't immediate emergencies.  A delay of a few hours in fixing the problem would cause no harm.

The self-flashing LEDs have been in place 15 years with one LED failure and zero rain leaks.  Pretty good record.

Mike in California
 

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Re: Indicator LED for valve controller
« Reply #8 on: June 25, 2020, 11:08:00 pm »
I tried a self-flashing red LED with the circuit that I have used for non-flashing LEDs supplied with 24 VAC (see 1st pic).  The LED did not work at all, but wasn't killed.  Guess the timing IC in it couldn't cope with supply that was alternating 1/60 sec on and 1/60 sec off.  So I tried a circuit with a rectifier (DF04M) and smoothing capacitor (see 2nd pic).  This worked well when tried with a variety of self-flashing LEDs (simple flashers, flickerers, alternating 2 colors, alternating 3 colors, random color, fade-in/out, etc.) ... all were happy.

Max current draw by an LED is 30 mA, frequency is 60 Hz, and capacitance is 47 μF.  So Vr = 0.03 / (2*60*47/1000000) = 5.3 V.   Rectified voltage would be about 35, and 5 V ripple is more than tolerable for this application.  22 μF or even 10 μF would probably be OK.   I really don't see a need for a regulator.  This circuit is cheap and compact enough to be put in a heat-shrink wrapper.

Just for fun, I tried a red flasher with no capacitor in the circuit.  Interesting result.  Worked pretty well but with occasional partial flash.  Again guessing, but probably the pulsing, inverted-sin-wave DC voltage was occasionally too low when the timing IC needed more.  There was probably some capacitance in the wiring and circuit elements that allowed it to work at all.  The setup was inherently unstable. When I connected a multimeter to read DC voltage, the spasticity grew much worse.   Definitely need the cap.

Mike in California


 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20363
  • Country: gb
  • 0999
Re: Indicator LED for valve controller
« Reply #9 on: June 28, 2020, 08:26:42 am »
Interesting. I thought flashing LEDs have a maximum voltage ratings of under 15V or so, because when it turns off, the voltage across it rises to the open circuit voltage. They appear not to limit the current and require an external resistor, which is news to me, as I'm sure I used one without a resistor, but it was a long time ago. I suppose the resistance of the battery might've been high enough to limit the current to a safe level.

Some data sheets:
http://www.farnell.com/datasheets/2047559.pdf?_ga=2.209356946.10875200.1593257277-1107606318.1553553996&_gac=1.238357812.1593257277.EAIaIQobChMIl7LH6fGh6gIVTbDtCh2wUAUHEAYYAyABEgLRqfD_BwE
http://www.image.micros.com.pl/_dane_techniczne_auto/oly.3m0010m.pdf

I wonder if the one you're using has a built-in zener diode to limit the voltage?

I've not being able to find a schematic for the internal workings of a flashing LED and most data sheets I've found are quite vague.
 

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Re: Indicator LED for valve controller
« Reply #10 on: June 28, 2020, 08:48:53 pm »
Zero999,

Thanks for your reply and the data sheet links.  Most of the self-flashers in my collection came from eBay with little or no info on electrical characteristics.  I really haven't tried to find data sheets for them, which might have been impossible as no manufacturer or part numbers were given.

The data sheets for the Wenrun and Kingbright LEDs don't  specifically address the question of higher voltages and need for current limiting resistors.  Reading the Wenrun data sheet, I get the impression that it would be OK to connect it to 3-8 VDC without a current limiting resistor.   The Kingbright data sheet is ambiguous. It states operation between 3.5 and 14 VDC.  But in a footnote, mentions limiting driving current to avoid killing it.

I have some single-color and alternating-color self-flashers installed for 5-15 years supplied with 9 VDC with a current-limiting resistor.  They have had a pretty high percent "on" time.  I can recall one failure.

I assume there is a simple timing IC in them.  Not sure whether it would be a separate die from the LED die.  I've seen a few self-flashers advertised for direct connection to a 12 VDC source w/o a limiting resistor, but they always had pre-attached leads with heat shrink over the solder joints ... could have had a tiny resistor under the heat shrink.

The only self-flashing LED I tested for hours with the 24 VAC circuit in my previous post was a regular red type   ... no problems with it.  The more exotic ones like fade-in/out, color alternators, etc. were tested for less than a minute ... so I'll try to do some 24-hour tests real soon.  Just for fun I may then try a few at 60 or 120 VAC with a correspondingly higher value resistor.

Mike in California


 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20363
  • Country: gb
  • 0999
Re: Indicator LED for valve controller
« Reply #11 on: June 28, 2020, 09:40:31 pm »
The flasher IC is definitely separate to the LED die. It's the black dot, in a diffused package, when viewed from the side.

If you're going to run the LED off the mains, make sure the resistor can handle the power & voltage. You might also want to try a capacitive dropper.
 

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Re: Indicator LED for valve controller
« Reply #12 on: June 29, 2020, 12:16:09 am »
I have no plans to power the LEDs directly from mains.  If I do any testing at more than 24 VAC, it will be short term from an isolated supply.   I'll use an appropriate resistor, both for resistance and power, for the tests.    For an actual application at mains voltages, using a resistor to limit current becomes more than a little wasteful of energy.

Mike in California
 

Offline wizard69

  • Super Contributor
  • ***
  • Posts: 1184
  • Country: us
Re: Indicator LED for valve controller
« Reply #13 on: June 29, 2020, 05:45:58 am »
There are a number of ways to solve a problem like this.   Here is a suggestion that might do it for you, use current sensors to detect power going to any one valve.   Something like this guy: http://www.phenixcontrols.com/Download/245/999300D8.pdf    There is one big advantage to detecting current instead of voltage, you can have a voltage present with no current flow (an open line).   In this  particular case you have relay outputs with would allow OR-ing together all of the transducers to drive your one indicator.  This particular guy is only one of many types available in the industrial market, some that get very expensive, others on PCBs.   

These sorts of transducers are almost zero engineering time compared to many other solutions.  Installation is fairly easy, though the one referenced is likely less than ideal for panel mounting, you just turn a screw to fix the set point.   The only thing I might consider is having them drive a relay instead of the lamp directly.   The reason there is that you may need a relatively bright LED or even a higher wattage "lamp" solution for it to be seen outside.
 
The following users thanked this post: calzap

Offline calzapTopic starter

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: us
Re: Indicator LED for valve controller
« Reply #14 on: June 29, 2020, 11:07:25 pm »
Wizard69,

Good point on sensing current rather than voltage.  And it would be hard to beat the isolation.

 Most of the valves are pilot-operated solenoid types and have a holding current of 190 mA.  So the Phenix sensor should work, although that's near the bottom of its range, which is a concern. A couple of valves are direct-acting solenoid types and draw more current.

A couple are motorized ball valves  ... these could be a problem for current sensing because they draw very little except when the motor runs.  The type in use now supposedly draws as little as 1 mA when in the closed position (that's from specs; I haven't measured it).  These are scheduled to be replaced eventually with ones that supposedly draw 40 mA in the closed position.  Simplest solution would be to wire a resistor parallel to the valve terminals to increase current draw to 200 mA total.

Each valve has at its location, a LED indicator assembly wired parallel that draws an additional 12 mA. 

I would only need one current sensor per 8 valves.  That's because the manual override and enable switches are clustered in modules of 8 with a common wire for all eight valves.  Could run the common through the current sensor.  A couple of minor problems are where to put the current sensors ... they won't fit inside the switchboard enclosures, which are full already ... and a power source ... they need 5 or 12 VDC, which aren't currently available.

I'll investigate what's available, order one, test it and go from there.

Mike in California

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf