I'm still in the midst of factoring and refactoring my carpc setup. I'd like to power the lcd, line driver and a few other things, only after I've started my car. So what I'd like is a NO relay that is powered by an ACC circuit, that only turns on when the voltage reaches a certain threshold (alternator going) and then stays on until the key is removed and the ACC line is disconnected. (I also do not want to go through the firewall again or touch the alternator ) My battery is usually a little more than 12v and around 13.5-14v when the car is running. My first thought was a PIC, using the comparators (the regulated 5v as vref) and a voltage divider to detect it. I've got a PIC or two going planned to drive certain things in the car as is, but some of it is still changing so I until that's finalized, figured I'd try an analog approach. (My dash is getting crowded as is) My second thought was to use a divider, 12v zener and transistor, but then I thought, maybe I'm overcomplicating it. My grasp on electronics is still pretty "iffy", but figured I'd give it a shot. So this came to mind. I figured I could just tweak the pot until it can't close the relay when the alternator isn't running. Does this sound, um, sound? Anything I'm not considering? Thanks. -Lee
I think with such a gradual approach to activating the relay coil, the slightest vibration might trigger it before the engine is actually running. 2nd thought seems OK, but all you need is a 12V zener and 1k resistor in the base-to-supply line. Should turn things on if supply is over 12.7V, the resistor will limit zener heat to 150mW. Might be an idea to make the relay latch on so any minor dip in voltage (flashing the lights, pipping the horn) doesn't turn your system off.
Thanks. I have a bunch of 12v zener's on hand if that'll work. You mean like this? If this is good, is the shunt still required? Both of these circuits should latch right? That's what I intended to avoid the problem you described.
To latch you need a double-pole relay, with the extra pole shorting out the switcher transistor once it's operated, and a push-to-break momentary switch to reset the system (although it will reset anyway if the supply falls below about 8V and the relay won't hold in). Your circuit runs off the LH switch. And your zener's wrong way round.
Duh! Thanks for pointing the zener out (and the revised schematic). One thing I'm curious about, most circuits I've seen for latching are DPST, but not sure I understand why a SPST can't latch as well. Not that I don't take your word for it, I do, just wondering what I'm missing. (I wish I was smarter about this stuff ) In the schematic I posted, when the NO contact closes, seems like the 12v will be fed to the relay coil and the transistor collector. So why would it close? Is it because if it dips the zener will cut off the base opening the path to ground? And if so, couldn't you just connect NO after the zener? I have both DPST,SPST on hand, more just an academic question Thanks again, as always, much appreciated.
Don't know how I missed that. But what about the second question, connecting NO between the zener and the transistor or am I missing another obvious short circuit?
I would have the transistor connected to the 'high side' of the relay coil, then it's very easy to have the SPST relay latching - just connect the relay contacts to the emitter of the transistor and the 12V rail, and you'd basically just have your load in parallel with the relay coil.
That works OK AFAICS, so you can use a SP relay in this case. One possible snag, there'll be maybe 10-15mS when the zener is conducting but the relay hasn't had time to take over (they're slow compared to solid-state). The load will only see about 2V but any high current may blow the zener. I'm not sure on their surge ratings. But then you've got the transistor as an emitter-follower, and it's following the (Alternator - Zener) voltage, so the relay coil gets under 2V max.
Isn't there a signal from the alternator that operates the alternator check light circuit? Couldn't you use that signal to latch on the relay? Or use your zener idea, but use a second resistor (and diode) from the load side to the base of the transistor to force it on.
cpemma Thanks for the explaination. Given the possibility of frying the zener, I'll go with the DPST. Though wondering about the BZX55C,9V1. Is that a 9v zener? star882 There's a battery indicator that comes on if they key is in but the car not started, but I don't want to tap into those harnesses or the flexi circuit board. Also, it's also low voltage and driven through an ECU, not the alternator directly. Hate to fry that as they cost $1000. Plus, a bit is just hubris. I was trying to think of an analog way, simply to see how it could be done that way.
Yeah, you want a 12V Zener, but the diagram / SPICE model that cpemma drew up worked off a lower voltage from the looks of things, but the idea is the same.
Sorry about that, Livewire only has zeners to 9.1V and I didn't want to cause extra confusion by showing two adding two up to something near 12. As g0th said, just showing the principle worked. First diagram I just hid the labels.
Thanks. Figured it was something like that. You also answered my next question which was "What did you use to make that?"
That works, you don't even need the extra resistor. It's a neat idea for a 2-button latch if you only have a SP relay, like a lot of the car ones. Traditional DPDT way is here. Trad way you can use the other (load) switch for any voltage it will take, not just same as the coil voltage.
