SCR/Triac - timer circuit.

[Diabolical Artificer]

I have a load of relays that need 50V AC to switch the contacts, no problem powering these as I have some little transformers that have 12V and 50V secondaries. I want to make a time delay soft start circuit to shunt a 100 ohm resistor, so I need to switch the 50V AC.

I'm pretty sure on the timer part, I can use an RC timer circuit, a 15k and 470uF cap should give me about 7 seconds. What I'm unsure about is if I can use an SCR to switch the AC or do I need a triac?

I'm also a bit misty on the difference between an SCR and Triac, but have a suspicion that a triac can conduct both ways, IE AC, whereas an SCR conducts one way, so is only good for switching DC.

I'm going to have a 12V DC supply so don't need to rectify the 50V. Saw this circuit below that it may be possible to modify.

Any help much appreciated, Andy.

[Al (astral highway)]

Hi Andy,

1) this circuit only works from AC the power supply you are switching, not from an independent DC supply. You will need lower voltage-rated electrolytics there accrordingly.

2) Your Time Constant calcs are correct.

3)Triacs are for AC, SCR's are DC switches since once latched , they pass current in one direction only.

Cheers!

[Diabolical Artificer]

Thanks Al. so I can use my 12v DC supply to feed an RC time circuit, which can turn on a transistor, which in turn can switch the triac which switches the 50v AC needed for the relay.

Time to get the breadboard out.

A.

[Al (astral highway)]

Originally Posted by Dr Wobble

..so I can use my 12v DC supply to feed an RC time circuit, which can turn on a transistor, which in turn can switch the triac which switches the 50v AC needed for the relay.

Hey Andy, I think there's a general feeling out there that simplicity is always best. You don't need to be breadboarding with a transistor, but of course go ahead if you feel like having a play!

If you're intent on using your DC supply instead of the shown configuration, simply leave out the rectifier diode and series resistor and apply your 12V+ where the cathode of the diode is shown instead. The time constant will work just the same.

But arguably better, (since you're not using your DC supply as an auxiliary role that will now need decoupling - can't remember why you have a 12VDC supply?)...
...just use the circuit shown but substitute a 4.7K resistor - which will pass the same current (10.45mA) at 50 volts, as the one shown passes at 230V - and a silicon diode. The resistor will dissipate just over half a watt so still use one rated 1W for good measure.

That would be my preference to getting involved with DC or anything more elaborate!

Oh, you should add a snubber. Triacs are noisy and you don't want that noise appearing somewhere where there's gain going on!
Cheers

[Al (astral highway)]

Hey Andy,

If you want to go the transistor route (which avoids introducing noise and all the associated problems with a triac), I can recommend this circuit (attached.)

I am using it successfully in a similar application to soft-start two giant transformers. Your time constant components can be subbed-in, obviously.

The first transistor is shown correctly with the emitter-base junction connected and the collector disconnected. After an interval determined by your time constant, the biased emitter-base junction reaches its avalanche voltage and passes enough current to switch on the Darlington transistor array.

You can also one higher powered transistor for the Darlington.

[trsomian]

A snubber will not do much if the load is resistive, but can be useful if the load is reactive as it will reduce the dv/dt at switch off, thereby reducing EMI and reducing the chance of spurious firing of the triac

[Diabolical Artificer]

Thanks Al. I had this idea in my head that an RC time delay circuit wouldn't work with AC, hence the idea of using the 12v DC for the timer bit to switch the 50v AC. I'll be fine when I knock the circuit up and can see what's what.

As the load in this case a relay coil is inductive, a snubber would probably be in order. where talking an RC filter yes? So I'd need to calculate R and C value's for a low pass filter?

Thanks again, Andy.

[Herald1360]

You can use a SCR to switch A.C. by connecting its anode to the + and cathode to the - terminals of a bridge rectifier. Connect the ~ terminals of the bridge in series with the supply to be switched.

The timer on the gate drive will need to be floating since its negative rail will be referenced to the SCR cathode.

[Al (astral highway)]

Andy, 39R is a typical snubber value for this application with 0.01uf in series, save you from calcs!

[Diabolical Artificer]

Interesting idea Chris, but probably no good for switching a relay with an AC coil as it would rectify the AC, or would it?

Cheers Al, that's saved me some time and brain power. I was researching snubber's yesterday, have a small pile of papers to digest.

Ordered some little 2w triac's in TO92 package, will save on HS's, and will have a play tomorrow.

Andy.

[TrevorG3VLF]

This all seems complicated to me.
You have a relay designed for 50V AC.
You can run this on DC so that an electrolyic capacitor can be used for timing. The relay contacts will shunt the 100ohm resistor.

Some experiment will be needed to determine the necessary DC voltage.

I showed elsewhere that an AC relay can be used on modified AC and DC. Here it is again.

Norslo 240V AC 2 pole change over relay. resistance 7k5.
AC pull in 130V, drop out 130V

Diode in series with coil
pull in 110V drop out 75V

Diode in supply and diode across the coil
pull in 90V drop out 20V

Bridge rectifier
pull in 65V drop out 10V

DC
pull in 60V drop out 7V

Try it as if it were a 20volt relay across the capacitor. No need for diacs, triacs or scrs. Watch for power consumption, there will be no impedance effect at DC but current consumption should be very low.

[McMurdo]

you could always use a solid state relay. Full isolation and gate firing logic built onto one little DIL or SIL block!

[Diabolical Artificer]

Trevor the relays in question will run on DC and pull in at about 24v, however I thought running AC coils on DC was a no no. See here - https://www.vintage-radio.net/forum/...d.php?t=133647 - post #5 ... "You need to be careful if you apply a DC voltage to the coil of a relay designed for AC.

AC energised coils rely on the resistance and the inductance of the winding to limit the current and avoid overheating.

As an example I have an octal based 6V AC relay which has a coil resistance of 5 Ohms. If you used that on 6V DC the coil would dissipate over 7 Watts as heat."

I could Kevin but I have all these big relays knocking about. These are going to be switching 450v at nearly an amp, a similar spec'd solid state jobbie I think would be expensive.

Andy.

[TrevorG3VLF]

The shaded ring on an AC relay dissipates heat and kills a lot of the inductance. I could not measure the inductance of my relay with a Peak meter, it interpreted it as a resistance. The resistance of my relays is 7k5 so would dissipate 0.7W on 70V DC. It would be interesting to find out the dissipation with the rated 240V AC.

The relay will be much more sensitive on DC and will run cooler on the correct reduced voltage.

Trevor

[Diabolical Artificer]

I've tried a few circuits out to get this relay to switch, but it's vibrating. I've attached a datasheet for the triac I'm using. It's very sensitive and if I read the datasheet right it needs 1v at 10mA on it's gate.

I've discounted the original circuit I posted as it needs an SCR. I've bought a lot of bits recently and don't want to getting any more bits.

So I connected 50v AC from a tfmr to M1 and the other half to the relay with M2 going to the other side of the relay coil. By using a red LED and shorting this to M2, the relay clicks. But all attempts to put a delaying circuit in between the LED results in a vibrating relay. I'm using the LED to fix the gate voltage at around 1v.

I could do what Trevor suggests and power the coil with DC. Which means rectifying and smoothing the 50v AC, then into an RC delay circuit that then switches a BC337 say that's connected in Darlington config to a BD139 that switches on the relay.

Apart from using quite a few components this seems overly complex. Is there a way I can make a simple delay circuit without using an SCR to trigger the triac? I tried a transistor but the relay vibrates.

Bit stuck here, any help/schematics on a fag packet much appreciated.

Andy. : )

[TrevorG3VLF]

What is the resistance of the relay coil?
I would think that all you need is a diode, resistor and capacitor with the relay across the capacitor. No smoothing is required.
If the capacitor is too large, then it can be made smaller with higher value resistors and a transistor to amplify the diddy current.

Sorry that I cannot provide a fag packet, they are all in use storing components and my supplier has left this world.

[McMurdo]

No, I mean use a little solid state relay to switch the big one. Saves messing with triacs and diacs and firing circuits etc

[Herald1360]

Originally Posted by Dr Wobble

Interesting idea Chris, but probably no good for switching a relay with an AC coil as it would rectify the AC, or would it?

No, the only "DC" is the full wave pulses through the SCR. The bridge acts as diode switches to route the AC the right way through the SCR irrespective of which way they go through the bridge.

The technique was used for heavy duty AC phase controlled dimmer circuits before suitably beefy triacs were available.

[Diabolical Artificer]

Got you Trevor, will have another try today.

Also with you Kevin, I have some little SS relay's. I read that Diac's can be used to tame triac's, but not familiar with using them, a SS relay sounds a better bet.

Ditto Chris, light dawns.

Thanks all, Andy.

[Al (astral highway)]

Andy,
I'm surprised you didn't try the full transistor version I posted up there...it takes about 10mins to do a great job, so low is the component count. Wanna try?