SCR/Triac - timer circuit.

[Diabolical Artificer]

Havn't got the correct transistors Al and all efforts to use transistors today have resulted in failure. Even if I can get a 12v relay with delayed turn on, I still have to use that to switch 50v AC. which means using a bridge rectifier, 3 transistors and a 12v relay to turn on one relay. there must be an easier way. Sorry Al, don't wish to sound ungrateful.

I tried Trevors suggestion of 50v AC in through a 1N1004, into a R, relay across the cap. Result relay turns on straight away, no delay. So used bigger value R - relay won't switch. In the end I found I had to use a very low value of resistor to get the relay to click, but no delay.

The relay I want to use is an Omron MK3P-5 which has a coil DC res of around 400 ohm.

For example I got a NPN tranny, used first a BC337, then a BC549. Connected 12v across a 1000u cap, 22k pot in between positive end of cap to base of Q. Emitter to 0v. So with voltage on the base the Q should conduct, but I can only get less than 2v on the collector ( red probe to 12v +). The Q should be in saturation.

I,ve had a few timer circuits work, but as soon as I put a load on, IE a little relay the voltage on the emitter drops.

So I'm either getting relay not on, or relay turns on straight away. It's driving me a bit bonkers to say the least. The one circuit I havn't tried is the one first posted. Sorry, rambling.

Andy.

[Oldcodger]

Andy - many years ago, early 70's I think, where two small transistors were used in an RC timer to switch a 12v relay . I think I got it to work with a PO relay and 6v source as the car I had at time was an old Renault 4. Circuit will be in an old copy of one of the mags under heading of windscreen delay.
Can't remember all the circuitry, but it used a RC circuit ( pot + electrolytic cap) to bias a Darlington pair made up of transistors.

[Diabolical Artificer]

Thanks, I think I have that in an ETI somewhere. Will dig it out and have another bash tomorrow. Will also try and be a bit more methodical.

A.

[MotorBikeLes]

Andy, you know how to calculate your delay time, but you also know that you can double the R and halve the C for the same time. ie, you have plenty of flexibility of component choice.
But do you? You MUST have sufficient current passing to both open and hold the relay open. You found that reducing your R kept it ON (that provides the current), now logically you need to INCREASE your C to get your delay time. The holding current is much lower than the closing current for the (any) relay, so if you get your C large enough for the 7 secs (I think you said), it WILL hold in.
Les.

[TrevorG3VLF]

Chose a resistor as large as possible which will pull in the relay without the capacitor, say 470ohm.
Add a capacitor, say 100,000uF (0.1F, 100nF), this should give plenty of delay.

I was hoping to simulate this using LTspice but my computer died and its temporary replacement uses XP and LT no longer supply LTspice for XP.

[Diabolical Artificer]

I'd come to that conclusion, not enough current and tried bigger caps, but there's the rub 100.000u 63v caps arn't small, cheap or plentiful in the junk box.

Having looked at a few delaying circuits last night in old Elektor's and ETI's there's more to these circuits than meet the eye. There are some subtleties at play. I'm slowly getting my head round the problem.

Thanks for your advice, sorry to hear about your PC problems Trevor..

Andy.

[Al (astral highway)]

Andy, your relay is a 24VDC device
http://http://www.mouser.co.uk/Searc...gqfbuxFw%3D%3D

The transistor circuit I showed has one diode-connected device (emitter-base junction only , like a diode ) and works with any general purpose device as it just forwards biases the junction when it gets to about 0.7V.

You can make up the Darlington pair from any general medium current transistors like 2N2222 or just buy a suitable Darlington. But crucially you need 24VDC to switch this relay (I checked three sources). Good luck !!

[TrevorG3VLF]

The circuit shown by Al #5 will use a much smaller capacitor at the expense of a transistor.
The transistor used as a diode is biassed in the reverse direction so is used as a Zener diode of about 7V. It is thought that doing this is not a good thing but I have one in my Heathkit HW100 to stabilise the VFO. I would use a proper Zener, you will need one with a sharp knee particularly if a Darlington is used.

It should be posible to run the relay on unsmoothed voltage so not needing a reservoir capacitor.

[Al (astral highway)]

Originally Posted by TrevorG3VLF

The circuit shown by Al #5 will use a much smaller capacitor at the expense of a transistor.
The transistor used as a diode is biassed in the reverse direction so is used as a Zener diode of about 7V

Thanks for clarifying, Trevor. It certainly works just as well as a 6.8V zener diode as we've both proved. I've run my version of this in a relay control application for hours. It's such a trivial circuit, definitely worth using in Andy's application too.

Best

[Diabolical Artificer]

Finally got a circuit to work yesterday with a delay of around 4 seconds. The circuit is out of an old Elektor, I had to change some values, see attached.

Originaly the RC part was a 47u and 100k pot, I found those values in the schematic work. This is only switching a little MR62 24S relay on so far though, which I can use to turn on the 50v relay which is nuts. What I'll probably do, is adapt the circuit for 50v. (My relays are definately marked 50vAC Al, but may work as Trevor says on DC.)

I've had a frustrating 3 days trying out lot's of different circuits. I started out using a circuit the same as Al's, IE RC circuit, zener from the cap to a Darlington pair. Simple yes? Except it didn't work. Either the relay came on straight away or nothing happened, with the base on Q1 sitting at less than a volt.

I've tried lots of circuits found online using mosfets, PNP transistors, triacs etc, non had a decent delay.
I,ve found two things out, at the cost of a pile of dead semiconductors. Swapping a Q out of the parts draw for the one in a circuit doesn't work, even ones with similar specs.

The RC part of the circuit isn't straight forward either. I tried 10meg R's at one point to get a decent delay. The tiny current through this still turned on transistors and mosfet's straight away. The only thing that has increased the delay is a bigger cap.

So I think though a delay circuit looks deceptively simple, I think it's finely balanced depending a lot of the Q your using hfe and the relay. I find transistors hard to use compared to valves and need to use them a lot more before I understand what's going off.

Cheers, Andy.

[Diabolical Artificer]

Right, sorry, you all must have been tearing your hair out and banging the mouse in frustration, but after chasing red herrings down rabbit holes, I've built a circuit that works with a delay of about 8 secs.

After more mucking about I got a 6k8 R into a 4700u cap, 4v7 zener off the junction of the two into the base of a TIP122. Collector to ground emitter to relay. Power on, no delay. Swopped C and E, works like a good un. This is with a 22v supply, using the 50v AC relay's. So it looks like a lot of the time I've had the Q's wrongly connected, which explains a lot.

Regarding Q's as switches. Am I right in thinking we use a NPN to switch a negative voltage and a PNP to switch a positive one?

Another thing I've found is you get more delay if you put a bleeder R across the cap, obviously. You can fine tune the timing. Al's circuit used one. It will also discharge the cap quicker, turning off the relay. I put a reverse biased diode across the R of the RC circuit so the relay latches off quicker. BTW, what is the purpose of a reverse biased diode across the relay coil? To stop it vibrating?

Are there any other tweaks I should include? I'll leave this circuit on all day to see if the relay coil gets hot. I'll probably use different RC values to reduce the physical size of the cap.

Thanks for sticking with me, apologies for being dense at times. Andy.

[Al (astral highway)]

Andy, when the current through your relay's coil stops flowing -,when it's de-energised,, the magnetic field around its coil rapidly collapses and a high voltage back EMF is produced. ('Flywheel effect') The diode becomes forward-biased and conducts , clamping that voltage which would otherwise exceed what your transistor can tolerate, destroying it. It also reduces RFI but that's of secondary importance.

Unrelated, the relay coil would overheat to a noticeable degree in seconds to minutes if that was going to happen, I'd say.

[TrevorG3VLF]

A diode across an inductance (relay) is used to eliminate a large voltage since the current will contiue through something for a while. You do not want it to go through the switch when the switch is off. Mosfets are often specified to accomodate this but a diode is cheap.
An AC relay has a shaded pole which will absorb some of this energy but a diode will do no harm.

I see that you have two diodes in series across the input of the transistor. Since the voltage will be under 10V and the current low, then one diode, even a 1N4148 would do.

The circuit is just an emitter follower and doesn't turn on fully so the transistor will dissipate some heat. Check the temperature, it is likely to be warmer than the relay.

Trevor