Anotherpower.com Forum
Renewable Energy Questions/Discussion => Automation, Controls, Inverters, MPPT, etc => Topic started by: David HK on May 22, 2012, 08:13:07 pm
-
Who can help?
The story so far is that I have a Gardena model WT 1030 water sprinkler. The unit splits into two pieces which consists of the electronics and battery, and the solenoid and valve unit. The battery is 9 volts PP3 type.
Whilst making up a 9 volt DC circuit to supply power to the unit instead of the battery I think I accidentally supplied around 13.7 volts which has zapped the SMD 16 pin IC controller. I have checked the solenoid with a low volt circuit incorporating a 1K resistor and LED and that is in working order.
The operation of the solenoid is important and I will give a brief note here. It works on a Negative pulse to turn on and a Positive pulse to turn off. The Negative and Positive pulses cause the solenoid to attract or repel a small plunger valve to move about 1 to 2 mm to cause flooding around a rubber diaphragm. Thereafter water pressure along keeps the diaphragm opened or closed as the case may be. The duration of the pulses is not known except to observe on an analogue meter they seem to be about 1 second and the needle swings to about 9 volts or slightly less.
The cost of a brand new controller is HK$500 so I have set myself a challenge to build a new controller unit to work with the existing solenoid and valve unit for the same value or less.
Locally I can obtain an ANLY weekly programmable weekly timer for HK$360. Details of this unit can be found at:-
http://www.anly.com.cn/english/pdf/APT8(9)S(M)DA.pdf
The ANLY PDF is fairly comprehensive and readers should note that is can provide a pulse output which can be varied in width.
In the next block of electronics I have to find a way of using the ON and OFF pulses to cause the solenoid to change polarity. A flip flop of some sort comes to mind, but which type?
There is an interesting IC 4017 circuit which can be used as a flip flop on positive going pulses which can be found here:-
http://electroschematics.com/5504/toggle-switch/
After the flip flop section I have on order a MAXIM 4427 IC non-inverting type which can be used to change the polarity on the solenoid. A copy of the PDF can be found here:-
http://datasheets.maxim-ic.com/en/ds/MAX4426-MAX4428.pdf
A photograph of a MAXIM 4427 can also be found here:-
http://www.robotroom.com/HBridge.html
Scroll down the page to see a bread board photograph.
Other parameters:-
The timer and all other parts must suit DC 12 volt operation.
My knowledge if electronics is mediocre. I can build circuits from diagrams and design PCB’s from circuits, but I have never been schooled enough in the actual component layout of circuits.
I hope I have supplied readers with sufficient details, if you requite more I shall do my best to assist.
Regards,
David in HK
-
Locally I can obtain an ANLY weekly programmable weekly timer for HK$360. Details of this unit can be found at:-
Thats still fairly expensive for what seems a fairly straightforward device.
Is your intended "fix" going to run from a 9V 216/PP3 battery, or something more substantial?
How many daily or weekly cycles are you anticipating?
In the next block of electronics I have to find a way of using the ON and OFF pulses to cause the solenoid to change polarity. A flip flop of some sort comes to mind, but which type?
David, I appreciate you trying to do this "the right way", but sometimes it only needs to be "good enough".
If you put a decent capacitor in series with the solenoid..... wait, let me draw it...
[attachimg=1]
If the capacitor is lets say, 470uF, when you close SW2, you will get a pulse of current through the coil with + at the top end.
Leaving the switch closed makes little difference to current - it will taper off as the cap charges, eventually ending with virtually nothing.
Release SW2, and close SW1 and the capacitor will now discharge through the coil - with the opposite polarity!
Replace the switches with transistors, and you can get your "open" and "close" polarity pulses without ANY expensive driver chips, H-bridges and mucking about...
Or you could use a single-pole double-throw relay to achieve the same thing.
The timer and all other parts must suit DC 12 volt operation.
I just purchased a timer to replace a failed mechanical one to run my pool pumps. Cost me $8.99 AUD (including delivery). It runs on 220V AC (although they also have a 12V version), has a nice digital display and 16 programmable events over a week. Each event has selectable on and off times and days, including selecting the day, every-day, week-days, weekends, mon-sat etc - so highly selectable. OK, it doesn't have selectable output pulse width - but if you use the capacitor idea, you don't need it to.
Just my thoughts....
-
http://yahee.en.alibaba.com/product/318858932-210265811/7_Day_LCD_Electronic_Plug_in_Digital_Timer_Switch_24hr_YHA_PC092.html (http://yahee.en.alibaba.com/product/318858932-210265811/7_Day_LCD_Electronic_Plug_in_Digital_Timer_Switch_24hr_YHA_PC092.html), a 9v universal wallwart, a relay and the cap as details by rossw should do the lot!!
Here in NZ these 7 day timers are about NZ$8 ( < HK$50).
BTW: These solenoids uses a magnet as the plunger hence being reversible and latching at each end of its travel :)
-
, a 9v universal wallwart, a relay and the cap as details by rossw should do the lot!!
Here in NZ these 7 day timers are about NZ$8 ( < HK$50).
David specifically said it had to run on 12V, so the timer you cited isn't really suitable.
I was thinking something more like http://tinyurl.com/d7r8w9m (http://tinyurl.com/d7r8w9m) - just over $10 AUD (under $80 HKD).
Runs on 12V, has a *changeover* relay for the output.
Ie, using one of these timers and a pre-existing power source, his *ONLY* other component would be the capacitor.
-
Ross and all the other people who have read this article - thank you.
Ross, I note your comments about another timer in your second thread. Do you have access to a PDF or information sheet about it?
Yes. I want to stick to 12v DC because it suits my RE system.
As for watering times – right now I am water-on for three minutes at 7 am, and a repeat cycle 12 hours later at 7pm. In the autumn and early spring I often change this to three cycles at 8 hours each with a run time of 1 minute. It depends on what I need to water.
This thread needs to linger on for a while so that people like Ghurd can give it the once over.
Lets see if any other forum members have wiser thoughts.
Regards,
David in HK
-
Ross, I note your comments about another timer in your second thread. Do you have access to a PDF or information sheet about it?
I found some docs here:
http://www.automaticsolutions.com.au/assets/manuals/index.php?dir=&file=7daytimer-quickstart.pdf
In my previous post, I posted a link to someone selling them for about $10
http://tinyurl.com/d7r8w9m
I have one sitting on my desk here that arrived this week - except that it runs on 240AC, it's otherwise identical, and should be fine for your job.
As for watering times – right now I am water-on for three minutes at 7 am, and a repeat cycle 12 hours later at 7pm. In the autumn and early spring I often change this to three cycles at 8 hours each with a run time of 1 minute. It depends on what I need to water.
The only potential problem with the timer I cite is that it has a MINIMUM operation time of 1 minute.
You have PLENTY of operations though - 16 or 17 program on+off times (cycles), if you do it "every day" then you only need 2 or 3 of them.
For $10 plus lets say $1 for a thousand microfarad capacitor, you'd be done....
-
I like Ross's approach.
KISS principal many time fits the best.
I have similar setup for our garden, but I have never even considered putting it on the offgrid system until this thread, thanks for the idea and links Daivd/Ross
-
Ross,
Could you confirm that the following drawing is a correct version of your earlier drawing above.
David in HK
-
In the above drawing ignore the green line going to pin 9
David
-
Ross,
Could you confirm that the following drawing is a correct version of your earlier drawing above.
That looks fine.
Of course, if you use the timer that I linked to, you don't need the (extra) relay as it already has a changeover (SPDT) output.
If you do choose to use a seperate relay as per your diagram, I'd *STRONGLY* advise a freewheeling diode across its coil to reduce the potential of nasty spikes back into your system.
Also, leave plenty of room around the capacitor. I've no idea what the current requirement of your solenoid is, and suggested a nominal capacitor that I "think" would likely do. You MAY have to use a larger cap. Conversely, you might get away with a smaller one. Take a 12V battery outside, 3 clip-leads and a selection of electrolytic caps from your junkbox. Find the smallest cap that will reliably turn it on and off, then use twice that in your final circuit :)
-
Another piece of guidance needed.
The original Gardena solenoid coil probably operated at 9 volts maximum.
If I put 12 volts through it what will happen?
David in HK
-
The original Gardena solenoid coil probably operated at 9 volts maximum.
If I put 12 volts through it what will happen?
It will draw more current and make a stronger magnetic field.
Since you will have a capacitor in series with it, the duty-cycle will be VERY low.
If you were going to be a purist, you'd measure the DC resistance of the coil, work out what current it would have taken at 9V, then work out the series resistor required to limit the current to that when running on 12V, and put that in series with the cap.
Eg, lets say you measure the coil and it's 90 ohms.
It would take 100mA from a 9V source.
In order for it to draw only 100mA from your 12V supply, the total resistance needs to be 120 ohms, so you'd use a 30 ohm resistor in series. (90+30=120). 30 ohm isn't a common value, so use a 27 ohm resistor.
My friend the bot tells me:
<RossW> !ohmslaw 27r 100ma
<RossBot> 2.700 volts at 100.000 milliamps through 27.000 ohms for 270.000 milliwatts
You would only need a 1/4 watt resistor.
Reality is: it's such a short pulse, its unlikely to be a problem even without the resistor.
-
Ross,
Thank you for the very useful advice and your time.
I have updated the drawing and included a diode across the relay coil - I hope its the correct alignment.
Dave
-
I have updated the drawing and included a diode across the relay coil - I hope its the correct alignment.
Yes, that's fine. But if you're going to make a board, at least add a the option of a resistor in series with the cap.
Put it down as two pads the right spacing apart, run a track between the two so the "default" condition is that it's shorted, but if you choose to put in a resistor (or two back-to-back zeners) you can solder them in and cut the track. No need to re-make the board :)
-
Thanks Ross,
Another refined drawing attached. I had not planned to make a board at the moment, I'm just concentrating on getting the circuit layout correct. Not doing too badly today with a slight hangover - may go for a hair of the dog shortly.
Dave
-
Now I’m onto calculations.
According to my multimeter the solenoid coil resistance is 36 Ohms.
Therefore at 9 volts the current passing through the coil will be 250 milliamps.
Increasing the voltage to 12 volts with a requirement to pass 250 milliamps through the coil means a resistance of 48 Ohms.
Therefore 48 Ohms minus the existing solenoid coil resistance of 36 Ohms leaves a requirement for a 12 Ohm resistor.
In my last drawing shown above, should I have put the ‘possible resistor” position on the negative side of the capacitor?
Dave
-
Now I’m onto calculations.
....
Therefore 48 Ohms minus the existing solenoid coil resistance of 36 Ohms leaves a requirement for a 12 Ohm resistor.
I agree with your maths. The only thing not considered is if the old controller may have run the solenoid on less than 9V (perhaps the circuit had say a 5V regulator?). Would seem unlikely however.
In my last drawing shown above, should I have put the ‘possible resistor” position on the negative side of the capacitor?
Makes no difference. It's a series circuit so the current will be limited either way.
I'm not convinced the resistor is required though - 36R coil, 250mA for even half a second, isn't much power to dissipate.
-
Well it works!
However, I have to keep the power onto the relay continuously, and, as soon as I disconnect, the solenoid shuts off as intended.
I have been thinking that 12 volts produces a very fast reaction and I wonder therefore if 9 volts would produce a slower reaction enabling a pulse effect to keep the water balance on whilst the relay is off (dead).
Gardena must have had a reason for using a 9 volt battery, and a solenoid coil of this exact type.
I don't have any 9 volt relays but I can re-jig the solenoid power via the relay to a 9 volt battery and just use 12 volts for activating the relay. Will give it a try
Dave
-
Well it works!
What, you doubted me?? :)
However, I have to keep the power onto the relay continuously, and, as soon as I disconnect, the solenoid shuts off as intended.
Exactly. Thats why my original diagram showed two discrete switches. One for "ON" and one for "OFF". Using a single SPDT switch means it's always either ON or OFF, with no "STAY AS YOU ARE" option.
I have been thinking that 12 volts produces a very fast reaction and I wonder therefore if 9 volts would produce a slower reaction enabling a pulse effect to keep the water balance on whilst the relay is off (dead).
There is no "dead" using a double-throw switch or relay.
Again, using the timer I suggested (which has a changeover relay for its output) would indeed keep its internal relay activated during the "ON" cycle (for 1 to 8 minutes per your earlier post?), but the solenoid itself would only get a short burst of power.
Gardena must have had a reason for using a 9 volt battery, and a solenoid coil of this exact type.
small, cheap, easy to get.
I don't have any 9 volt relays but I can re-jig the solenoid power via the relay to a 9 volt battery and just use 12 volts for activating the relay. Will give it a try
Don't expect a different outcome, because it won't be.
Let me re-explain the theory of operation of my original circuit.
A capacitor, on DC, will charge up to the supply voltage. The amount of current it draws, and how long it draws it for, is a function of the resistance of the circuit, and the capacitance of the capacitor itself. When you first apply power to the circuit and operate the "on" switch, peak current will flow through the coil and capacitor. The capacitor will charge up at a rate determined by the resistance of the circuit and the voltage. The charging current will rapidly reduce as the capacitor aproaches full charge, end eventually will be virtually zero (just leakage current).
Thus, your solenoid has had a brief pulse of current even though the power could remain turned on for hours.
When you disconnect the power, the capacitor will hold the charge - it has no discharge path. The charge would (in theory) hold forever - except of course, capacitors have some internal leakage and the charge will eventually dissipate. (I've had 100,000uF capacitors stored for years that took a chunk out of a screwdriver with the charge still in them after that time!)
Finally, when you close the "off" switch, the charge in the capacitor has a discharge path back through the coil - and will again produce an initial peak current which tapers off as the charge in the capacitor is discharged, ending up with the cap discharged and there being no current through the coil - in effect, "reset" ready for the next cycle.
If you use a changeover (double pole) relay, it is only ever in one of two states - charging the capacitor ("on"), or discharging the capacitor ("off").
-
Ross,
Thanks for clarifying my thoughts. Well what a day! It's so nice when one can get some expert help to make a very simple device to work
Next step is the weekly timer and i will keep the forum posted when that is sorted out.
Its getting late in the afternoon here in HK so to celebrate in your absence i am going for short bike ride and will consume a few tins along the way.
Dave
-
Not that this needs to be "bumped" or anything - but one thing came to mind...
The charge would (in theory) hold forever - except of course, capacitors have some internal leakage and the charge will eventually dissipate.
This is most definitely true... however in the way that the cap is being used in this particular circuit, there's no "risk" of it ever not being a momentary low impedance when the relay changes states. As you mentioned Ross, there's no "in between" state here, it's all one way or the other.
When the relay is "off", the cap is connected across the solenoid's coil, preventing any residual charge from surfacing and increasing the voltage at the terminals. The cap stays at zero.
When the relay is on, it is holding the cap at the 12V rail via the solenoid's coil, so the minimal leakage current that would normally allow for the slow self discharge is counteracted. As a result, there will always be sufficient energy to toggle the solenoid back to it's off position when the relay cuts off.
Just an observation for future readers that might stumble across the concept. Carry on then. :)
Steve