Relay bounce with fade out signal

Hi guys, I've got a relay in the car powered directly from the battery, and controlled by a Normally +12v ecu pin. For the relay to close, the pin grounds, that works fine, however, when the pin is sent back to +12v, it does so with a sort of fade out. The ambiguous current causes the relay to vibrate a few times rather than just turn straight off. Is there a way I can level this fading signal to an on/off type thing? I'm thinking a small capacitor might do it?

It's probably a capacitor that it causing the problem unless it's some sort of soft turn off. The circuit you describe sounds like a typical ECU/CPU "open collector" aka ground-switched output. IOW the output is not GND (ie not 0V = zero Volts) when off. When on the output grounds the relay coil (85) whose other end (86) is +12V. There should be no capacitance on that EMS output. It should also have a spike protection diode else a diode should be fitted across the relay coil (from 85 to 86, line end towards the +ve end (conventionally 86). Either of those can cause relay chatter tho the lack of latter spike protection usually blows EMSs/CPUs or at least their outputs. If the EMS output is a soft on or off - ie, PWM or slowly changing voltage - then yes, some sort of squaring circuit will need to be fitted - eg, a Schmitt trigger or other hysteresis type circuit. (Or the EMS reprogrammed.)

Yeah its a soft off on the body ecu, I need something that will change that signal to a hard off. This schmitt trigger, that should do it? Thanks

Yes - a Schmitt or comparator circuit. With an RC filter if its PWM.

i am an idiot wrote: https://www.the12volt.com/installbay/forum_posts.asp?tid=114184

Excellent, thankyou :)

A diode & cap may not solve the problem since it can from an LC oscillator (see Wiki) but it's worth a try. That's the other reason I don't like the brute force circuit as per relaydiagram22 - several posters have mentioned its relay chatter.

Thanks for the help all, I also needed some extra features, so I knocked up a quick Diode OR gate as the logic required should help my needs. I have a feeling though that there may be current leakage which could prove the opposite what I need. So, IGN is a standard open/closed circuit, its either 1 or open. ILE is the interior lamp circuit, normally a 1 for lights off, goes to 0 when the light is on and fades back up to 1 again. HL is also normally 1, then grounds when the headlights are on. Originally I simply connected ILE to HL via a single diode that allowed HL to ground through ILE when the interior light came on, but that obviously created the relay chatter when the fade came back on. I added ACC as I don't want this to happen when the keys are in the ignition, only when I'm out of the car. The ground I added as a pull down, so when both ACC is open and ILE is grounded, the from HL will ground through that resistor, until ILE is at full power again. Unfortunately my digital electronics knowledge is limited, so all the reading I've done tells me this should work, but logically I'm sure that HL will ground through the R1 regardless of ILE and ACC's status. Help!

Saying that, looking at what you've just said Oldspark, I think brute forcing it, though dirty, may well work for my application.

Yes, brute forcing may work. If the relay chatters, you'll know it doesn't (with the component values used). What is R3 in your above diagram? If it's the capacitor & relay, fine, but don't assume open state voltage levels supply high current. And beware logic 1 & 0 - you are mixing logic and voltage - eg, logic 1 & floating - IOW logic 1 else 0 (IGN is 0V when not +12V tho its impedance/resistance can vary from mOhms to kOhms etc).   But I'll assume you meant "1" as shorthand for +12V (because "not +12V" if meaning GND/0V has no impact on the above diagram).