Protecting a SSR (Solid State Relay) From Overcurrent/Short Circuit - With PTC Resettable Fuse?

Thread Starter

Mahonroy

Joined Oct 21, 2014
406
Hello, I am using a solid state relay (SSR) to drive a 24VAC load of around 0.5 amps. This is using an external 24VAC (40VA) transformer - similar to those rain bird transformers from home depot.
This is the SSR Part #: "TLP3546A(TP1,F"
The SSR is 3.5A, but can support 10.5A pulse for 100ms.

I am planning on limiting this to around 1.0 amp, and I would like to use a PTC resettable fuse. This one for instance:
PTC Fuse Part #: "2920L110/60MR"
The PTC has a hold current of 1.1 amps, trip current of 2.2 amps, and a maximum amp rating of 20 amps before breaking.

The main thing I want to protect against is an accidental short circuit.
PTCs are not very fast - in this case it will take 500mS to start heating up, increase in resistance, and finally enter its "tripped" state.

So it seems the SSR will become damaged before the PTC trips? And the PTC fuse also states that it can only withstand a maximum current draw of 20 amps before breaking.

Do I need to insert a series resistor to help limit current draw? For instance, if I placed a 1.8 ohm resistor in series with the 24VAC transformer, and the load, this would limit the current to 18.8 amps - and buy just enough time to protect the SSR while the PTC starts to heat up and reduce current draw itself? Not sure if its supposed to work like that. Also, I don't think these 24VAC transformers are actually capable of sourcing 18.8 amps anyways. And they do have PTCs that have a breaking amp rating of 100, but I thought this was still a good example.

I'm just a bit confused how to go about this. I am not sure if I am making this overly complicated or not.

Any help or advice is greatly appreciated, thanks!
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
You don't want a series resitor, especially when using AC.
What is the nature of the load?
Why are you anticipating a sudden short circuit?
Use a stand fast blo fuse at the SSR rating.
 

sagor

Joined Mar 10, 2019
903
Depending on your load, a 1A fast acting fuse would be better than any PTC. A fast acting fuse should blow within 100ms when current exceeds rating by x2.5. Thus,a typical 1A fast acting fuse will blow in 100ms when current reaches about 2.5A
All fuses have a current vs time chart (manufacturer). Look those up if concerned.
 

Thread Starter

Mahonroy

Joined Oct 21, 2014
406
You don't want a series resitor, especially when using AC.
What is the nature of the load?
Why are you anticipating a sudden short circuit?
Use a stand fast blo fuse at the SSR rating.
This will typically control solenoid valves. An accidental short circuit if someone wires something wrong, pinches a wire, etc. Or when a solenoid valve fails (when water solenoid valves fail, they typically pull high current).
My old design used a standard fast blow fuse, and it worked just fine. The new requirement is it now has to be auto-resettable... I cannot use replaceable fuses any more. And when I say worked fine - on the initial prototype with a standard fast blow fuse, the SSR would still get damaged with a short circuit. I had to put a 0.1 ohm series resistor inline with the SSR and this did the trick. I have to use an auto-resettable fuse now.

Depending on your load, a 1A fast acting fuse would be better than any PTC. A fast acting fuse should blow within 100ms when current exceeds rating by x2.5. Thus,a typical 1A fast acting fuse will blow in 100ms when current reaches about 2.5A
All fuses have a current vs time chart (manufacturer). Look those up if concerned.
My old design used a standard fast blow fuse, and it worked just fine. The new requirement is it now has to be auto-resettable... I cannot use replaceable fuses any more. And when I say worked fine - on the initial prototype with a standard fast blow fuse, the SSR would still get damaged with a short circuit. I had to put a 0.1 ohm series resistor inline with the SSR and this did the trick. I have to use an auto-resettable fuse now.
 

Thread Starter

Mahonroy

Joined Oct 21, 2014
406
How about a small P.B. style circuit breaker.?
They hold in during the solenoid pick up time.
Those are way too big. This is for a small PCB that I am making. Everything has to be reasonable sized for a small enclosure.

Why can't this work with a PTC?
 

Ian0

Joined Aug 7, 2020
9,667
Those are way too big. This is for a small PCB that I am making. Everything has to be reasonable sized for a small enclosure.

Why can't this work with a PTC?
If you use a triac, you could buy one that will run continuously at 16A and take a surge of 160A for £1.50. and that would give ample time for your PPTC fuse to trip.
 

Thread Starter

Mahonroy

Joined Oct 21, 2014
406
I did a test with the transformer just now. So its a 24VAC (40VA) transformer - e.g. it can output 1.66 amps.
I took a 1 ohm resistor and shorted it across the terminals, and monitored it with the oscilliscope. The highest voltage I saw during this period of time was 15 volts. So with ohms law, it sounds like the maximum current this transformer will output during a short circuit is 15 amps?

So that is fine for the PTC since that amperage is under the "breaking current". Now its just a matter of if the PTC will reduce current draw in time to protect the SSR. I am now looking at this PTC:
https://www.digikey.com/en/products/detail/bel-fuse-inc/0ZCF0075AF2C/4156294

It looks like the PTC will most likely take between 100ms to 300ms to fully "trip".

So you are saying I should be using a triac instead of a SSR? Is there no way to make this work with an SSR?

I found this triac:
https://www.mouser.com/ProductDetail/STMicroelectronics/T1620T-8G-TR?qs=l7cgNqFNU1hmtE5oowWcvw==

I've never used them before. Are there any "gotcha"s as far as using this in place of a SSR? It looks like I can still trigger it with a microcontroller, e.g. 3.3V on the gate @ 20mA.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
I found this triac:
I've never used them before. Are there any "gotcha"s as far as using this in place of a SSR? It looks like I can still trigger it with a microcontroller, e.g. 3.3V on the gate @ 20mA.
A SSR uses a triac, typically, see Fairchild App notes, AN-3006 & AN 3003
 

Ian0

Joined Aug 7, 2020
9,667
I found this triac:
https://www.mouser.com/ProductDetail/STMicroelectronics/T1620T-8G-TR?qs=l7cgNqFNU1hmtE5oowWcvw==

I've never used them before. Are there any "gotcha"s as far as using this in place of a SSR? It looks like I can still trigger it with a microcontroller, e.g. 3.3V on the gate @ 20mA.
If you want to trigger the triac directly, then your microprocessor’s positive supply must be common with the triac’s MT1 connection.
if this isn’t the case, use an opto-triac to drive it. From the microprocessor the opto triac looks like any other LED or opto isolator.
One output pin of the opto-triac (either one) connects to MT2 on the triac, the other pin goes to the gate via a resistor, select the resistor as
R=Vpeak/1A.
A quarter watt resistor will do as current only flows for a microsecond or so, but make sure that the resistor has enough voltage rating.
 

Thread Starter

Mahonroy

Joined Oct 21, 2014
406
This is what I came up with. Does this seem like a decent solution? It seems I can use any of these MOCs since the difference seems to just be voltage capabilities, and LED activation mA (5mA or 10mA). These are all zero crossing opto triacs:
MOC3063, MOC3062
MOC3043, MOC3042
MOC3033, MOC3032

I had a difficult time figuring out the value of R40 and especially R41. I am fairly confident R40 is good, but I have no idea about R41.

Should this do the trick?
triac_question_1.jpg
 

Ian0

Joined Aug 7, 2020
9,667
This is what I came up with. Does this seem like a decent solution? It seems I can use any of these MOCs since the difference seems to just be voltage capabilities, and LED activation mA (5mA or 10mA). These are all zero crossing opto triacs:
MOC3063, MOC3062
MOC3043, MOC3042
MOC3033, MOC3032

I had a difficult time figuring out the value of R40 and especially R41. I am fairly confident R40 is good, but I have no idea about R41.

Should this do the trick?
View attachment 250188
You can leave out R41.
My suggestion for R40 would be Vpeak/1A = 33Ω. 100Ω would be a good value for a mains-operated circuit. It will never get 1A because it only triggers around zero-crossing.
 

Thread Starter

Mahonroy

Joined Oct 21, 2014
406
You can leave out R41.
My suggestion for R40 would be Vpeak/1A = 33Ω. 100Ω would be a good value for a mains-operated circuit. It will never get 1A because it only triggers around zero-crossing.
Thanks for the suggestions. I changed R40 to 33 ohms. Is a 1/4 watt resistor ok here?

Regarding R41, how do you know I should leave it out? I had a hard time figureing this part out. I found a bunch of examples that had it, and a bunch of examples that did not have it. No explanation for either cases, and no example for how the resistor value was determined, so I was a bit confused on this part.
 

Ian0

Joined Aug 7, 2020
9,667
Regarding R41, how do you know I should leave it out?
Two reasons:
Having built (not personally) this circuit about 10,000 times, operating on 230V mains, and never had any faults caused by a lack of gate-MT1 resistor.
If you measure the gate-MT1 resistance with an ohmmeter (not a diode tester - a meter with a low output voltage) you will find a low resistance across gate-MT1, most likely a lot lower than the resistance you intend to connect.
 
Top