Car Alarm fitting

A decent one will have a delay built in so they don't sense for a couple of minutes after setting. After this it should still sense the interior lights coming on though regardless of a dimmer function.

Plus they don't trigger on the lights *being* on, they trigger on the transition of the lights *turning* on (i.e. they trigger on the event of the lights going on, not the state of the lights being on, if that makes sense) and the subsequent small voltage drop in the wiring caused by the resistance of the wiring and internal resistance of the battery, plus possibly any inductance in the circuit (he says, speculating wildly).

Depends on the alarm. Mine is connected directly to the switch side of the courtesy lights circuit, and responds to the voltage being at ground when one of the switches closes. If the alarm is armed while a door is open, it will go off as soon as the ~40sec arming delay is complete.

That said, it's a pretty high-spec cheapo alarm (you can get some good deals on ebay, if you're prepared to spend an evening with a bench power supply and voltmeter making up for the inadequacies of the installation instructions' Engrish), and needs to know the state of the doors for the auto-rearming and central locking control to Do The Right Thing.

My previous banger had a voltage-drop-sensing alarm as you describe, which was very good at warning you that it might be difficult to start the following morning (the battery was somewhat knackered), and not a lot else.

Kim.

Hence my wording - 'sense the interior lights coming on' ;-)

Voltage drop.

That's not how they work - or at least none I've come across. They stabilise at battery volts then look for a *quick* but tiny change in voltage as happens when a load is applied. They shouldn't react to a gradual drop in volts.

I know, but any inductance in the loom might help, V = - L.(dI/dt) n'all that.

Yar, but that's one that you've got to wire in specifically to that point, most cheapos just wire straight in to +12V, 0V and maybe a connection to the fan so the unit knows when a voltage drop is caused by the fan kicking in, not a courtesy light coming on (note to burglars: hang around the car until the fan kicks in, then quick, nick the stereo. This is a joke. :-)

Top tip is to wire the alarm in the actual supply feed to the courtesy light and as far along the feed from the battery as possible, preferably between the fuse box and the battery, although on most vehicles this would be a bugger to wire in. That way you get the voltage drop across the battery internal reistance *and* across the wiring, hence better chance of the unit detecting the courtesy light going on. Alternatively, bodge a higher wattage bulb in somewhere in parallel with the existing courtesy light bulb - the more amps, the bigger the voltage drop (sez Mr. Ohm :-)

Oh bollocks, ignore that, I meant between the fuse box and the courtesy light. (So you also get the advantage of the voltage drop across the fuse.) But it is a bugger to wire in if you want to put the alarm in the engine compartment.

Emphasis on the 'shouldn't'. This was a particularly cruddy example of the species, so I didn't waste much time on it. The only other sensor was an ultrasonic wossname, and disconnecting it made no difference to the random false alarms.

Of course, the banger in question was a Fiat, so it could have been those Italian electrical gremlins...

Kim.

The one I've got on my SD1 is over 15 years old, but happily senses the boot or under bonnet light. The boot one is just 6 watts. Few interior lights are less. Sensing anywhere on the loom really shouldn't make any difference - the sensor will be very high impedance and the wiring only a few ohms. So altering the 'few ohms' part of a voltage divider where the other part is high resistance makes bugger all difference to the voltage across that.

The sensor impedance is as you say very high but that's neither here nor there really. You want to be at the point where the voltage drop caused by the light coming on is at the max and that's as near as possible to the light, so that you benefit from maximum ohmic drop through the wiring loom - and especially through the fuse as basically a fuse is a low-value resistor (I^2 x R power dissipation of which causes it to melt or oxidise away when the current gets too much).

Draw out the circuit with the following elements and play around with the order and you'll see: voltage source, then internal battery impedance, then wiring loom impedance to the relevant fuse, then fuse impedance, then impedance of the loom from fuse to light, then the light, then the loom to the doorswitch, then ground. Putting the sensor at the light gives you a better point in the voltage divider for a bigger voltage drop than putting it up near the battery. Basically because of the ohmic drop through all the current carrying elements up to that point. Put it at the light and you've got an ohmic drop of Ix(Rbatt+Rwire1+Rfuse+Rwire2). Put it at the battery and all you've got is IxRbatt and I is not much and Rbatt is bugger all, and the degree of buggeration increases the higher the cranking rating of the battery is. Assume the sensor is infinite impedance because it is compared to the other elements, it's not really part of the equation. Yes the impedances of wires and fuse are bugger all but so is the change in voltage you're trying to measure, so every milliohm helps.

Why couldn't it be wired directly to the courtesy light's switch inside the door casement?

When the door opened it would operate then wouldn't it?

I'll shut up!!

We were somewhere around Barstow, on the edge of the desert, when the drugs began to take hold. I remember "Vim Fuego" saying something like:

Here's a useful tip, from an old post...

--------------------------------------

Some of the Moss alarms loose sensitivity on the voltage drop after a while.

This can be overcome by placing a small capacitor (say a few uF, rated over

12v) across the courtesy light bulb observing correct polarity. This causes a momentary larger current draw when the door is open and helps trigger the alarm. Also ensuring the alarms +ve is connected to a permenent live on the fusebox, preferably the same circuit powering the courtesy light helps. It should not be wired direct to the battery - less chance of a voltage drop.

Andy Evans

--------------------------

Never thought of that. Good advice. Especially the bit about observing correct polarity, have you ever seen the mess an electrolytic capacitor makes if you reverse-voltage it? RC rise time shouldn't be a problem given the relatively low resistance

I think geek mode began and ended before and after your HTML input ;)

In news:4550bec9$0$8729$ snipped-for-privacy@ptn-nntp-reader02.plus.net, Billy H wittered on forthwith;

I always wire 'em to the door switch circuits, works for me!

"Pete M" wrote in message news:eisr0p$79t$ snipped-for-privacy@registered.motzarella.org...

Hence a voltage drop of 12V rather than 12mV when the door switch closes. Certainly a bigger voltage but not necessarily what the alarm designers had in mind. :-)

It's not exactly rocket science to measure the voltage drop produced by even a 3 watt bulb anywhere in the car - a tuppenny op-amp will do it.