Hello, I made this circuit so I can control a 24VAC sprinkler solenoid valve (rain bird ones from home depot), using a microcontroller. I also want to mention that I am new to using triacs by the way.
When I activate the circuit by supplying 3.3VDC to the Q1 mosfet, the solenoid activates, but I noticed its only getting between 19VAC to 21VAC. I expected to get 24VAC since this is what I am supplying power with. I also notice that if I have no solenoid connected, I get an output of 9VAC. I believe the 9VAC is normal since there is no load? Should I be concerned that I'm only getting 19VAC to 21VAC with a solenoid connected? This seems to be a problem.

For reference, the rainbird solenoid valves that I am using draw 0.3 amps when connected directly to a 24VAC source.

Here are the components I'm using for reference:
MOC3063: https://www.digikey.com/en/products...=N4IgTCBcDaILIHkDCBmADANhQDQMpwBUB6AJRAF0BfIA
T1620T-8G: https://www.digikey.com/en/products/detail/stmicroelectronics/T1620T-8G/10414742
PTC1 Fuse: https://www.digikey.com/en/products...56294?s=N4IgTCBcDaIAwC0DCAxOcDsBWAgisSIAugL5A
Q1 Mosfet: https://www.digikey.com/en/products...17723?s=N4IgTCBcDaIEIGUEEYDMAOAtAdkwMRAF0BfIA

Thanks and any help or advice is greatly appreciated!

 

What is your PTC dropping?
Any reason you are not using a simple wired fuse?

 

What is your PTC dropping?
Any reason you are not using a simple wired fuse?

If I am understanding the datasheet correctly on the PTC, it looks like it has an initial resistance of 0.18 ohms at room temperature, and it could have a resistance of 1 ohms at room temperature 1 hour after a trip (since they heat up when they trip and need to cool down to lower ohms).

I can't use a wired fuse. Its inaccessible and must be auto resetting. Otherwise I would love to use a wired fuse but I can't do it in this design.

 

Triacs stink.
Use FETs instead ...............
Virtually Zero Voltage-Drop.
( Just select more appropriate FETs for your application, the ones shown are High-Voltage ).
.
.
.

 

yes...

what is in the circuit when you are doing voltage measurements? one thing to keep in mind is that triacs have significant leakage current. depending on connected load and instrument use to measure voltage, you may see pretty much any value 0-24V. you

 

Triacs stink.
Use FETs instead ...............
Virtually Zero Voltage-Drop.
( Just select more appropriate FETs for your application, the ones shown are High-Voltage ).
.
.
.
View attachment 258093

yes...

what is in the circuit when you are doing voltage measurements? one thing to keep in mind is that triacs have significant leakage current. depending on connected load and instrument use to measure voltage, you may see pretty much any value 0-24V. you

I was originally using solid state relays (SSR's), but had a problem with protecting them from over currents / short circuits. I had a thread about it here, and the general concensus was that I needed to switch to triacs, so that is what I did.
Here was the thread for reference:
https://forum.allaboutcircuits.com/...hort-circuit-with-ptc-resettable-fuse.182415/

 

The Resetable-Fuse seems like a good idea at first glance,
but they are only good for a finite number of protection events,
then they fail.

You need to have your Controller monitor the Current,
and shut-down for X-number of seconds or minutes when an over-Current-Event occurs.
.
.
.

 

What is the nature of your AC supply, usually they are a little above 24vac (unregulated! )

 

Where exactly is the voltage probe? With or without load?
From what I see, ohms law likely busting in here.

Start measuring voltage across components to see where the Vdrops are.

 

if you don't mind audible click, just use a small relay and make your life easier.

btw. one issue with this is that you are using zero crossing version of the optocoupler. according to the datasheet, this may not even fire until AV voltage reaches 20V. in normal 120VAC or higher applications this is insignificant but in your case, voltage is low. as a result your load current is not a good sine wave, it is very distorted and can be throwing off your multimeter.

 

The given schematic is way too complex for a on/off switch.
They do make AC FET relays, and the gate can be cmos level to turn on/off. Literally would be one component.
I use Omron FET AC relays for AC and DC motors.
"KISS" when you can, build it when you can't find it.

 

https://octopart.com/search?q=solid...throwconfiguration&sort-dir=asc&loadcurrent=1

 

Why not use a simple Relay?

 

Where exactly is the voltage probe? With or without load?
From what I see, ohms law likely busting in here.

Start measuring voltage across components to see where the Vdrops are.

DC kid has a very good point, in that until you know what is dropping the voltage all actions are guesses. That means that even interconnecting wires can be the source of voltage drops. And certainly using a 120 volt triac on 24 volts may not perform adequately. look at the data sheet for the triac minimum voltage limits. You may need to use a different device intended for 24 volt operation. And it might even be that the transformer is inadequate. One option would be to use a 28 volts out transformer.

 

What is the nature of your AC supply, usually they are a little above 24vac (unregulated! )

It is a plug in transformer (40VA). And yes you are correct, with no load the transformer outputs closer to 27VAC. Once there is a small load on the transformer its closer to 24VAC. I think 25VAC is typically where it sits.

Where exactly is the voltage probe? With or without load?
From what I see, ohms law likely busting in here.

Start measuring voltage across components to see where the Vdrops are.

The solenoid hooks up to the terminal on the left of the schematic (24VAC Output, and 24VAC Common). That is also where I have the probes for the volt meter.

The given schematic is way too complex for a on/off switch.
They do make AC FET relays, and the gate can be cmos level to turn on/off. Literally would be one component.
I use Omron FET AC relays for AC and DC motors.
"KISS" when you can, build it when you can't find it.

https://octopart.com/search?q=solid...throwconfiguration&sort-dir=asc&loadcurrent=1

I already explained earlier that this does not work because I cannot protect the SSR from over current scenarios (e.g. accidental short circuits). When solenoids fail (not if, but when), they typically fail in a short circuit scenario, and I need to be able to protect the relay. Auto resetting PTC fuses are not fast enough to protect the SSR, and the SSR gets damaged by the time the auto resetting PTC fuse trips. This is why we switched to the triac since it can handle a lot more current draw. I also explained that I cannot use a cartridge style wire fuse because its inaccessible - it needs to auto reset.

Why not use a simple Relay?

I would like to use a mechanical relay, but I can't use them for this project for these reasons:
1. It needs to be surface mount.
2. They don't like the clicking sound, and claim other devices like this in the industry don't make noise.
3. The relays are too big to fit in the enclosure properly.
So it has to be some kind of solid state / semiconductor based switch.

 

The Circuit I provided will do all that You ask for.
The only addition would be a Current-Sensor-Chip to signal your Controller to
interrupt the Switch during Over-Current-Events, and provide a delayed reset.
And beside that, SMD-FETs are available in ratings well over ~50-Amps,
your Transformer will smoke way before the FETs even start getting hot.
.
.
.

 

LowQCab makes a good point about the xfrmr, too small to deliver lots of amps. I believe those controllers also have short protection.

Valves are typically 20-60ohm, so 1A down to about 1/3A. Inrush amps to a coil via fairly thin wire at length will be dampened. A 10A(rated) SSR with a 3A or 5A blade fuse in front of the SSR , the SSR should survive a-ok time and time again.

But as example, 24vac via say 30ft or 100ft of buried wire, out to a ~25ohm(Z) coil, probably will see around 800-900mA of current. 1A(rated) seems to be normal operating design amps per channel (valve).

I would also place a SPD across the lines aft of the SSR (near SSR). When the solenoid turns off the existing mag field will cause a small EMF to build, so to protect the SSR silicon, a 36v SPD (with ~100ohm resistor) across the 24vac lines will help protect the SSR from induced EMF overvoltage.

I don't really build stuff without testing, so, test it, install a SSR and 3A fuse (blade, PTC, whatever), then short the far end, see what happens.

And just a side note, measuring voltage on a solenoid is not as important as measuring the current, because amps is what makes the solenoid work, not voltage.

 

A transformer rated 40 watts at 24 volts is a bit over an amp and a half. And if it drop-s from 27 volts to 24 volts with a "light load" then it is quite likely a big portion of that additional drop. So the first thing to check , at the transformer connection to the controller, is the transformer voltage with the solenoid valve operating. That may be the cause of the problem, or at least a good part of the problem. In normal sprinkler control installations the valves are quite far away from the controller and so there may be 50 feet of #22 copper two conductor cable. And I am sure that I have opperated one of those sprinkler valves from 12 volts as part of a non-sprinkler project.
And given the overload protection device plus the transformer characteristics iit does not seem like an SSR would have a problem.

 

A transformer rated 40 watts at 24 volts is a bit over an amp and a half.

The typical Rainbird controller box, and many like it, have 4-6 channels, which can all be on at the same time. The Rainbird solenoids are typically 20-60 ohm(DC). My mistake earlier, I noted solenoid as (Z), but online specs show DC ohms for those solenoids. That said, Z is probably a bit higher, probably running each channel near 1/4-1/2A. If it's a 40w xfrmr then I speculate 10w x 4ch controller.

Agreed, an SSR will be a-ok, but adding a 3A fuse and 36v SPD is not a bad idea.

 

The typical Rainbird controller box, and many like it, have 4-6 channels, which can all be on at the same time. The Rainbird solenoids are typically 20-60 ohm(DC). My mistake earlier, I noted solenoid as (Z), but online specs show DC ohms for those solenoids. That said, Z is probably a bit higher, probably running each channel near 1/4-1/2A. If it's a 40w xfrmr then I speculate 10w x 4ch controller.

Agreed, an SSR will be a-ok, but adding a 3A fuse and 36v SPD is not a bad idea.

Really, a spike prevention diode on an AC solenoid is NOT a good idea. And there is already that PTC protection device as part of the design.