Remote starter design

[Chancy99]

Hi All -

I'm thinking about designing and building a remote car starter. I know that there are lots of different ones already out there in the market, but where's the fun and learning in going off-the-shelf ?

Most starters use relays, either external triggered from a smaller one inside, or they have larger ones internal to their box. What about using mosfets instead ? It would be nice to have be completely solid-state, no moving parts to fail.

Dean.

 

[daveward]

Chancy, what a brilliant idea! Put me down on your product wait list. Sorry I don't have the kind of expertise to help out, but I appreciate the out of the box thinking (sorry that's now such an overused phrase)...you're using not-from-the-shelf-thinking!

While you're brainstorming, I've got more and more applications these days for continuous solenoids...so if you could add this feature into the mix, I'd appreciate it. Dave

 

[carnage1]

the relays are much less expensive when you start talking large loads but some nice transistors could sure do the trick the thing I have yet to find on the market is a kit that could be installed by a shop on a standard transmission car. most shops won't do it because someone could accadentaly start the car with it in gear, so maybe a nutral saftey switch for standard trannies?

 

[Chancy99]

There are several remote starters on the market at the moment that will handle manual transmissions. The general technique they use is more or less as follows :

1) Put on handbrake
2) Push button x on remote
3) Remove keys (car keeps running)
4) Get out of car, close door
5) Push button y on remote

Step 2 tells it that you're stopping the car, so turning the keys to off and removing them doesn't actually stop the engine. Step 5 actually stops the engine, and possibly arms the alarm.

Now the system knows that the car is in neutral, with handbrake on, so is safe to remote-start.

I'm thinking about possibly paralleling up several heavy-duty mosfets to handle the current surge of the engine start. Any ideas ?

Dean.

 

[carnage1]

parallel transistors is a bit tricky, if they aren't a tight enough spec one transistor tends to "rob" all the current from the others and burn itsself out.
one technique i have seen used with diodes is a small resistor on each with the resistors forcing the current split, not the best solution in the world but it worked well in that specific application.

does anyone know how much current a car draws, minus starter moter but including starter relay?

 

[JRW261]

Just a note, but make sure there is an OFF/ON switch for your starter. I put one on my truck (cheapo) and a pretty good one on the girlfriend's car.... After doing so, I realized, I wouldn't really want to change a belt with or with out the hood switch. The hood switch is fine, but it grounds the system to disengage it... doesn't take much age to cause a bad ground. So, for both I wired a shut off switch, so when I work on them, I shut them off or disconnect the battery. Her car is a pain to disconnect the battery because of passloc.

Its cool to think out of the box, but I would think a mechanical relay for this type of use would be alot more reliable and easier to change/diagnos. just my 2 cents

 

[gregdachs]

I lack quite a bit of knowledge in electrical engineering, but one thing i do know is that it is essentially impossible for a relay to fail on. It might get stuck on or not activate at all, but chances are it will not engage on its own.

In my experience with some home made (by electrical engineering students that were fairly competent) high current motor drivers i've seen parts of mosfet circuits fail causing a significant amount of current to leak through the transisitor. This could be a bit unfortunate in your case. If you can, however, build a more bulletproof circuit than this or provide some sort of failsafe, it could work.

Greg

 

[MacGyverS2000]

greg,

Big time wrong on that one...they can, and do fail while on. The easiest one is a relay that connects circuit A in one direction, and circuit B in the other...in that case, no matter what direction it fails in, it's an unintended and potentially bad thing.

Another possibility, though, is the contacts arc and weld themselves together...once welded together, it doesn't matter what happens with the coil, it's going to stay closed.

Absolutely do NOT count on mechanics or software to be foolproof every time...the moment you make an item foolproof, someone goes and makes a better fool.

 

[BrianR]

carnage1,

Current splitting with power transistors is usually done by inserting a resistor in each emitter lead, chosen to give about 0.2V drop at running current. As you said, without this one transistor will steal most of the current and overheat.