Alternate Switching of Relays

[dougy83]

Ahh I see the debounce now that you have mentioned it, your circuit now works a treat, I changed the CLK cap to 4n7 & it fixed the problem, the opposing LED gives a minute flash when the button is released but I don't think this will be an issue.

I had the CLK cap at 1uF (excuse my handwriting), which should have been sufficient for debounce. There should be no flash from the other LED.

If I disconnect the relay contact from the supply rail I can't get power to the circuit to check it out?

Yeah, sorry I didn't think that through. You'd have to supply power to the LEDs. Maybe with the new relay there will be no problems.

 

[willeng]

Thanks alec_t,

I will have a look at the circuit provided, just got home.


Timescope,

Yes, my bad, I didn't worry to much about drawing the relay or the wiring to it on the schematic properly but I do have it correct on the circuits, seeing as I don't have a LTSpice model for a relay or the 4013 4017 etc I didn't worry to much about it.

I would like to get some models if anyone has them so I can simulate these things?

Dougy,

After I posted that I added 1k resistors across the transistor gates to emitters & it fixed the flashing issue immediately, I tried 10k, 4k7 etc but got to the 1k.

Thanks to everyone for the help, it's really appreciated.

Probably a waste of time & you may laugh but I decided to try to see if the button & the relay contacts were in fact bouncing & causing the issues. I am trying to learn more about things so I thought it was worthwhile.
I tried with the scope but had trouble picking it up so I decided to make a bounce detector circuit out of a 4017 & 10 LEDS.
Basically when you press the button with no contact bounce the Led count should increment from 1to 9 one LED at a time & then another button resets the circuit to zero LED.

So with the circuit Zeroed if you press the button any bounce will show up with say number 6 LED lighting instead of number 1 etc.

It was an eye opener & I guess it was a worthwhile exercise.

With the button tested first every now & again the circuit would pick up contact bounce & would explain the intermittent misfires I was having using the button.

With the Relay connected to the circuit & tested it showed exactly why I have had so much trouble.
With the circuit zeroed & then the relay was activated sometimes LED 3 would light & then LED 6 then 9, basically all over the place & explains a lot.

I know that this is only a basic test & more than likely I should & would be better to have tested a time constant of the worse bounce senario instead of the number of bounces to overcome the bounce issue & the then apply a time constant debounce circuit suitable for this.

But it did show the issue & as yet I am not sure how to set something up to test a time constant & haven't got anything to see it on screen.
Any Idea's?

My relays didn't turn up today, it will be Friday they say now but I did see a solid state relay in the shop so I picked it up to try it out & PRESTO all the circuits work perfectly now I have three & maybe four circuits with alec_t's new modded one.

There is no missfires of any kind with the SSR----PERFECT!

I will now try to work out how these circuits work, Timescope gave me an example of how his works but i'll have a look at the others & see if I can work them out--Don't hold your breath

Thanks Again it's much appreciated.

Cheers

 

[willeng]

Remove double post, I worked it out.

Cheers

 

[alec_t]

Glad you've solved the problem. Your idea to use a 4017 to test contact bounce was good.

I would like to get some models if anyone has them so I can simulate these things?

They're available for download (as a set of zip files) from the Yahoo LTSpice user-group.

 

[willeng]

alec_t,

Thanks, just downloaded the models etc, I can see now how the debounce works, the sim works good.

Cheers

 

[kinarfi]

I haven't built this recently, maybe today, but if you're getting both LEDs on, your flip flop Q & Qnot may not be going to 0, so a diode or 2 in the emitter of the transistors will require a higher voltage to trigger. It won't simulate for me because Q & Qnot are only 1 volt on the simulation. Try measuring Q & Qnot and see how low they go.
Posted this before I read every thing. oh well.