Help: 1-Wire Switch --> selenoid

[3v0]

It is time to dig up the yard and put in a sprinkler system. I would like to use a Dallas 1-Wire Switch to control each solenoid. The idea is to run the 1wire net, power and ground with each pipe run. That way the system can be extended at any point.

I have been looking for an appnote on interfacing to the DS2405 PIO pin.
I need to understand how to get the DS2405 Addressable Switch to power the solenoid? A mechanical relay is not an option. I plan to pot the circuit for the obvious reason.

Solenoid Specs

24Vac 60HZ nominal
19Vac 60HZ min
Inrush 0.3 Amps 7.2VA @ 24 Vac 60Hz
Holding 0.2 Amps 4.8VA @ 24 Vac 60Hz

The device controlling the solenoid is the
Dallas DS2405 Addressable Switch

PIO pin sink capability is greater than 4mA at 0.4V

The PIO outputs are configured as open-drain and provide up to 20mA continuous sink capability and off-state operating voltage up to 28V. Control and sensing of the PIO pins is performed with a dedicated device-level command protocol. To provide a high level of fault tolerance in the end application, the 1-Wire IO and PIO pins are all capable of withstanding continuous application of voltages up to 28V max.

If I futz with it I might come up with something that works but I need a solution I can trust.

I would like to keep the cost down if possible.

 

[Brevor]

Have you read this Dallas ap note ?
I did some 1-wire stuff once and found they need to be protected from nearby lightening strikes.
Looking at your solenoid specs I would try using an opto isolated triac to drive it.
Bruce


https://www.electro-tech-online.com/custompdfs/2007/03/AN148.pdf

 

[philba]

Sorry but I think this is a bad idea. I've done 4 sprinkler systems so far and I'd stick with the tried-n-true approach - power control at the controller. You will still need to run power to the valve so it doesn't save much in the way of wire cost. The environment is pretty hostile to electronics - cold, dirt and flooding. The signal wire would run with the power so you would have to deal with noise. Not to mention the occasional gardener abuse.

For expansion capacity, I suggest you just run extra wires when you have the trench dug.

Now, a much more interesting project to me is to interface the controller to the network. My biggest complaint is that I have to tweak each zone as the temperature goes up or down (hot days vs cold days). I'd love to be able sit at my computer an tweak all the zones at once rather than screwing around with 33 zones individually (yes, it's a big system).

 

[3v0]

I went on another net hunt and found this.

This circuit uses a transistor to control a LED from the PIO pin. .


**broken link removed**

Notice from the schematic how the DS2405 addressable switch outputs are connected to 2N3904 NPN transistors. This is because the DS2405 outputs can only provide about 4mA sink [ground] output current. The 2N3904 NPN transistors allow the DS2405 limited capacity outputs to easily control higher current devices.

Since the open-drain outputs of the DS2405 addressable switches cannot produce a logic 1 output, the 10K pull-up resistors are included to pull the NPN transistor base to logic 1, and turn on the LEDs. When the DS2405 output is on, it will apply ground to the base of the NPN transistors, turning the LEDs off.

I am thinking if I replace the LED with a Triac Output optocoupler and a triac I have what I need.

Did I miss anything.

 

[3v0]

Sorry but I think this is a bad idea. I've done 4 sprinkler systems so far and I'd stick with the tried-n-true approach - power control at the controller. You will still need to run power to the valve so it doesn't save much in the way of wire cost. The environment is pretty hostile to electronics - cold, dirt and flooding. The signal wire would run with the power so you would have to deal with noise. Not to mention the occasional gardener abuse.

There is nothing to be sorry about you are entitled to you 2 cents worth. And I appreciate you giving it even if I do choose to ignore it. You may be right but I want to try anyway. I understand the problems with this approach but still would like to pursue it.

I understand this is harsh environment. That is why I intend to pot the electronics. I have 1-wire temperature sensors that run submerged in my ponds. I think I can make this work. I am more worried about the noise generated when the 24V switches.

The cost of the wire should be offset (to some degree) by the reduction in pipes.

For expansion capacity, I suggest you just run extra wires when you have the trench dug.

That does not give me the ability to hang another n heads off the end off a run. With this system each zone will hang off a main feeder trunk which can be extended as needed.

At my last house I used a manifold/system in each major area of the yard. One out front and two in the backyard. I did not care for it.

Now, a much more interesting project to me is to interface the controller to the network. My biggest complaint is that I have to tweak each zone as the temperature goes up or down (hot days vs cold days). I'd love to be able sit at my computer an tweak all the zones at once rather than screwing around with 33 zones individually (yes, it's a big system).

Should not be hard to do. Start by scrapping the commercial controllers. Use a micro controller with some 74x595 shift registers and a bunch of relays. You can have as many zones as need. Some DIY Christmas light controllers do this. You can either put the smarts in the micro on just use a PC.

 

[Brevor]

I have 1-wire temperature sensors that run submerged in my ponds.

3V0, How long have you used these, and how reliable have they been ?

Thanks,
Bruce

 

[3v0]

3V0, How long have you used these, and how reliable have they been ?

Thanks,
Bruce

I have used them for a bit over a year with no device failures. Once in a while I get a flyer (unreasonable data value) which is rejected by the software.

The values are read once a minute. The system will retry 4 times prior to giving up. If it is critical use two sensors on seperate 1-wire runs.

 

[justDIY]

I went on another net hunt and found this.

This circuit uses a transistor to control a LED from the PIO pin. .

I am thinking if I replace the LED with a Triac Output optocoupler and a triac I have what I need.

Did I miss anything.

that looks pretty straight forward!

maybe use a small darlington instead of the single bjt, then you can drive both an led and the opto - handy for diagnostics? or use a solid state relay, I see 'em cheap at surplus sites, typically 3v to 32vdc control voltage.

might be prudent to plan your software to support different 1-Wire IO devices, such as the 2 channel 2406 and the 8 channel 2408?

 

[3v0]

The LED sounds like a good idea. Maybe a 2nd one driven by the 24V that shows when the valve has power. This sort of thing is a godsend when you need to figure out why it does not work.