Logic latch

[Wendymott]

Hi peeps.
I need a simple logic latch.... fool proof. I need to latch a relay or solid state switch to be On / Off with a single "push" operation of a "make" button. C Mos or TTL. Any one out there got a circuit in mind please.

NO Pics please as I already have one in the item and it cannot be re jigged.
TTL or Cmos will be fine.
Thanks

Sounds like a D type flip flop configured as a divide by two with the input well debounced. You can get eight in one box, paralleled it might give enough drive for a real relay, no need for the ubiquitous diode, push pull outputs supress the spike.

[Cruisin Marine]

You need something like a bistable flip-flop, use two NOR gates or two NAND gates.
You can even use a 555 I recall, if you do a search there must be loads of circuits out there.

http://www.easyelectronics.byethost1...ble-latch/?i=1

Or you could use a SMALL thyristor, reset by removing the supply line momentarily, that is how the old GEC selcall units stayed open from distant memory.

[Lloyd 1985]

I’m sure I made one with just a couple of transistors in it, can’t remember where I got the circuit now though! It works very well, I used it in a modulator as a power switch, and a second one to switch it between test pattern and video input, I’ll have a dig around see if I can find it..

Regards
Lloyd

[TonyDuell]

Do you want separate set and reset signals, or a single signal which toggles the state?

The former is an SR latch, which is a pait of cross-coupled NAND or NOR gates. You can also cross-couple a pair of transistors, these are effectively RTL NAND gates if you think about it. It's the classic Eccles-Jordan circuit.

The latter can be made using a D type flip flop (74x74 or 4013) with the D input driven from the Q/ output. Or a JK flip-flop (74x76, 74x112, etc or 4027) with J and K both asserted. Note that some TTL flip-flops have J and K/ inputs so you need to tie J high and K/ low to assert them. In all cases you need to connect the Set and Reset (direct) inputs to the non-asserted state and apply the input signal to the clock input of the flip-flop.

Of course these flip-flops will respond to contact bounce, so you will have to debounce the switch. If it's just a make contact switch, try a 0.1uF capacitor across it. If you have a changeover switch you can debounce using an SR flip-flop, the other half of the 74x74 could be used for this.

[74x74 means 7474, 74LS74, 74HC74, 74HCT74, etc]

Take caution with caps directly across switches, the instantaneous current can be very high and wreck the switch, this is one of the reasons why suppressors are a cap and resistor in series.

[TonyDuell]

I would be surprised if a 0.1uF capacitor charged to 5V held enough energy to damage a normal switch contact.

I can think of a once-popular application that can't be discussed here where a capacitor of about 1uF charged to 12V was connected across a switch contact that operated many, many thousands of times.

[Radio Wrangler]

Something like a 7474 in any flavour.

Debounced switch pulse to clock input (else it'll drive you insane) Q bar to D. Fit an R-C delay pullup (series R, shunt C) to either set bar or reset bar to choose the power up state, with just a plain pullup to the other.

For bonus score, use S bar and R bar on the spare flipflop as the debouncer.

David

I found this https://www.electronicdesign.com/tec...e-pdf-download

[Mark1960]

An alternate way to create a bistable latch with a single non inverting gate, for example 74hct08. Approx 4k7 resistor from the output to the input, then drive the input with either an spdt switch from 0v or 5v, or drive it with a tristate output. This can also be used as a bus hold circuit to avoid floating inputs.

Similar circuit is possible by adding positive feedback to a comparator or op amp.

[Cruisin Marine]

Merlin, that latching relay cct. is very clever- thanks, I may use it.

[MotorBikeLes]

You did include mechanical relay in your original post, so surely all you need is a relay with an extra NO contact, and take an extra feed to that contact back to the coil. Switch on, relay latches. Every common work space with motors is full on them.
Les.

[emeritus]

That would certainly provide 'latch on' operation, but I read the original request as needing 'Turn On' and 'Turn Off' on sucessive presses of a single push-button switch. Using a second set of contacts of a conventional relay to latch it on would require a second NC switch to interrupt the coil current to turn it off. There is a special design of relay (Carpenter relay) that changes state with successive momentary energisations. They seem to be almost unknown in the UK but are commonly used in France for controlling lights from more than two positions using momentary operation push button switches. The 4 terminal intermediate mains switch for operating lights from three or more positions using three core cables is unknown in France.

[Wendymott]

Hi Peeps..... Lots to go at..... Thanks.... I will get the test bed out tomorrow and give the "D" bistables a whirl. Tony.. It is a single switch that will toggle.. On and Off as in 2 pushes.

[TonyDuell]

Is the switch a simple normally-open contact or can it be a changeover contact?

[Wendymott]

Hi Tony. Its a simple "Make" only not 2 way. I made a pcb with 4 make only to fit in the space where mechanical latch switches were previously in the Eddy EC10. Thanks for asking

[MotorBikeLes]

Regarding #13, I recall the "Elliot-Satchwell" control system on our intermittent gas kilns installed in 1967 included Carpenter relays. From memory, a negative supply gave one direction, a positive supply the opposite. No supply, centre position with no output. The system involved a proportional valve output which positioned itself somewhere between 1 (low gas) and 10 (high gas). The controller itself was two term (proportional plus integral), suitably modified from a Satchwell air conditioning system.
Les.

[Wendymott]

Hi peeps....I have decided to use a PIC as there are now 3 channels required... saves logic chips.. if I use a 16F84..... not much programme memory needed.. Thanks anyways for your inputs.

[Herald1360]

Originally Posted by TonyDuell

I would be surprised if a 0.1uF capacitor charged to 5V held enough energy to damage a normal switch contact.

I can think of a once-popular application that can't be discussed here where a capacitor of about 1uF charged to 12V was connected across a switch contact that operated many, many thousands of times.

Yes, but the contacts in those switches were pretty meaty compared to those in a little pushbutton, and they still needed to be replaced after about 400 operating hours.