Making a realy latch

[danrogers]

Hi all.

Ok, im working on something in which i need to be able to switch 12v at max 18A.

The signal i need to control that power with is a 0v pulse of about 1 second (normally 5v).

I want to have the relay activate upon a low signal event, and stay active until the next low signal.

Now i understand that i will probably need a transistor with the 0v signal to base, to switch power to the relay coil, but this part of the circuit will only be active for around 1 second.

The main things I'm unsure on is what relay to use, and how to make it latch/ stay active.

any help greatfully welcomed

Edit..

I would also like to be able to switch 5v to another part of the circuit at the same time in the latching manor if possible?

 

[Boncuk]

Hi Dan,

you can make any relay latch using a HCF4027 (J-K/FlipFlop).

The example shows positive triggering but you might invert it.

Both, pin14 (\Q) and pin15(Q) can be used for output.

Using a bistable (latching) relay you'll need two signals reversing when you want to switch on/off.

With the FlipFlop only one input signal is required keeping the output pins at a set state until the next input arrives.

Boncuk

 

[ljcox]

Attached is a relay Flip Flop that will do what you want.

 

[danrogers]

Hi guys, Im stuggling to understand both of those circuits you have posted up for me.

Bonuk, is ICA & B part of the same IC? Whats IC1 P? How would you invert it to be trigged with a low pulse?

LJCox, Is A & B on the circuit two different relays? I presume the thickened lines on the switch and the FET is GND? Also, where is the ouput from the circuit (relays)?

Thanks very much

Dan

 

[ljcox]

You're welcome.

Yes there are 2 relays. This convention is called the "attached" method. The contacts of the relay and its coil are drawn on the vertical line.

The horizontal lines connected to the switch and the source of the MOS FET are the common or gnd points.

I did not bother to show the output. All you need is a second contact on A, ie. the A relay should have DPDT contacts.

The purpose of the MOS FET is to provide a SPDT function. You could try a SPDT switch instead, but one of the relays may release during its transition period.

 

[ljcox]

Bonuk, is ICA & B part of the same IC? Whats IC1 P? How would you invert it to be trigged with a low pulse?
Dan

Dan, the IC Is a dual JK Flip Flop. IC1A is configured as a Monostable, ie. it provides a pulse when triggered by someone pressing the button.

This pulse toggles IC1B.

So it operates the same as my relay FF.

First press sets Q of IC1B High & it remains high until the button is pressed again.

IC1A

 

[ljcox]

Dan,
To trigger it with a low pulse I would:-

Connect pin 5(k) to Vdd
Connect pin 4(R) to gnd
Connect R1 between pin 3 (CLK) and Vdd
Connect S1 between pin 3 (CLK) and gnd

Thus IC1A will trigger when S1 is pressed as before. It is simply a matter of using the synchronous CLK input rather than the asynchronous R input.

 

[Hero999]

The dual relay solution is interesting but isn't it cheaper to just use a single relay, a transistor and a flip-flop IC?

 

[ljcox]

The dual relay solution is interesting but isn't it cheaper to just use a single relay, a transistor and a flip-flop IC?

THis circuit needs 4 states in order to effect the state transitions. You need the second relay to provide the extra 2 states.

The relay could be operated by the button & then held by its own contact. But how do you release it when the button is pressed again?

 

[ljcox]

The dual relay solution is interesting but isn't it cheaper to just use a single relay, a transistor and a flip-flop IC?

Sorry, I missed the point of your post. But what I said above is still relevant.

Yes, it may be cheaper to use the IC + relay solution. But:-
1. The OP may be more comfortable with less electronics given his limited electronics knowledge.

2. I like to provide options and let the OPs decide for themselves.

3. The FF solution has a weakness in that the button can only be pressed briefly. Otherwise, if it is held until C1 discharges, the circuit may not work (I'd have to study the 4027 data sheet to know)

Dan,
The IC1P means IC1 pins

 

[ljcox]

Dan,
I just re-read your initial post. You said the pulse is about 1 second.

So if you use the IC solution you will need to increase C1 to 1 uF. 220 nF is too small for a 1 s pulse.

 

[danrogers]

thanks for all the info LH, very informative indeed!

I have just been studying the circuits and I think I now under stand how the dual relay solution operates, and how the ic solution would be connected.

With regard to the dual relay idea, I want to be able to switch up to 18A, so Im thinking to use an automotive relay? If it would be relay A that i would connect the load to, would that be the same side as the FETS's output?

To reduce cost, would it be feasable to use a lower rated relay for B? Looking at the schematic I think both of them would be switching the high current - is that correct?

Thanks again!

 

[ljcox]

thanks for all the info LH, very informative indeed!You're welcome

With regard to the dual relay idea, I want to be able to switch up to 18A, so Im thinking to use an automotive relay? If it would be relay A that i would connect the load to, would that be the same side as the FETS's output? You can't use the contact that is connected to the FET. I'll answer this in more detail below.

To reduce cost, would it be feasable to use a lower rated relay for B? Yes Looking at the schematic I think both of them would be switching the high current - is that correct? No, The current for the logic will depend on the resistance of the relay coils you choose.

Thanks again!

You can use either relay to switch the load. It depends on whether you want the load to be switched when you press the button or when you release the button.

You need a DPDT relay for the one that switches the load.
The other one can be a SPDT.

One contact for the load and the other for the logic.

You also need to study the relay data as some relay contacts have a MINIMUM current specified in addition to the max current.

I think you may need to ask more questions, so feel free to do so.

EDIT
I suggest you post the data of the relays you are considering so I can advise further.

I just re-read your original post. I noticed that you wrote "The signal i need to control that power with is a 0v pulse of about 1 second (normally 5v)"

If your input signal switches between 0V & +5V, then you will need a FET or bipolar transistor to interface.

Also, is this signal from a switch or is it from an electronic source? The point is that switch contacts have contact bounce and this is a complication. This is the reason for the Monostable proposed by Boncuk in the electronic solution above.

 

[Boncuk]

Dan,
I just re-read your initial post. You said the pulse is about 1 second.

So if you use the IC solution you will need to increase C1 to 1 uF. 220 nF is too small for a 1 s pulse.

By the time the 1s pulse has ended the stage will have triggered the follow on stage already. I use that value for a light switch with multiple buttons at different locations (e.g. porch light to be switched on and off from the kitchen, the hall or living room).

Switch bouncing will probably be a problem but with a light switch this is no issue.

Boncuk

 

[ljcox]

By the time the 1s pulse has ended the stage will have triggered the follow on stage already. I use that value for a light switch with multiple buttons at different locations (e.g. porch light to be switched on and off from the kitchen, the hall or living room).

Boncuk

Boncuk,
I just had another look at your circuit.

In the conventional configuration, C1 would be connected to S so that it delays the Set.

But in your circuit, it is connected to Q/. Is this correct? Or is it a drawing error?

 

[danrogers]

any updates on this Bonuk? Im looking foward to building it soon

 

[ljcox]

You did not answer the question that I posted some days ago. I quote:-

"I just re-read your original post. I noticed that you wrote "The signal i need to control that power with is a 0v pulse of about 1 second (normally 5v)"

If your input signal switches between 0V & +5V, then the Bonuk circuit may not be necessary. It's possible that all you will need is a FF and a transistor to drive the relay.

What is this signal derived from? If it is coming from an electronic source as opposed to a push button, then there won't be any contact bounce, thus Bonuk's monostable is not necessary (I would not do it the way he did anyway).

Another question. Can you obtain the 5 Volt supply from the source? Or do you want to run the circuit from the 12 Volt supply?

What is the unit to be used in?

If it is to be used in a motor vehicle, any electronic components will have to be protected against the alternator spikes which can be as high as 150 Volt.

Is the 12 Volt supply switched?

If not, the unit will have to be designed to consume minimal current in the quiescent state, otherwise the battery will be discharging while the vehicle is not in use.

If you answer the questions and give me some info on the relay you intend to use, I'll design a solution for you.

 

[danrogers]

Hi Lj, sorry I think I must have overlooked that question. Here goes,

the signal is coming from an encoder ic. I could use 5v from a regulator as the ic and some other bits require 5v, but it will be powered from a 12v source so I guess I could use either?

basically, im making a remote control rgb controller. The decoder could be set to toggle, but 2 of the 4 outputs are going to be used to trigger inputs on the pic im using so the 0v pulses are perfect for that. I want to use the 3rd channel from the decoder to switch the 12v supply to the LED arrays, and possibly the power to the rgb controller but thats not essential as it wont draw much. The FET's that im using to control the RGB leds can switch 6A each, hence the high power requirements.

To be honest I havent picked a specific relay to use yet but from looking the automotive types can switch these high loads and are fairly cheap.

Thanks for the help, and your time

 

[ljcox]

A block diagram of what you're doing would be a big help.

As far as I can gather (you did not answer that question), you are not intending to use it in a vehicle.

If you're using a PIC, why do you need hardware logic?

 

[danrogers]

Sorry Len, no it wont be used in a vehicle.

The PIC im using is pre-programmed, Ive not dabbled with programming yet.

I will get something drawn up