On Sun, 11 Jan 2004 23:54:29 +0100, "Meindert Sprang" wrote:

wrote in message
.. .
On Sun, 11 Jan 2004 13:55:44 -0500, "Len Krauss"
wrote:

Hi Steve,
Thanks for that alram-out power limit tip and the excellent schematic!
Len

As usual, there's always a ton of ways to do any task.

The schmatic refered to above specifies a relay with as low as 120 Ohms
coil resistance. At 12
Volts, this equates to 100 mills, meaning some heat will be generated,
however small, inside the
GPS. I admit the alarm wouldn't be on very long in most cases, unless no
one was aboard.

If it were me, I'd buffer the GPS output with the trusty 2N2222 to drive
the relay. The advantage
is that it'd be trivial to incorporate a "alarm off" switch... and a
sesitivity contol, just in case
the GPS output was analog (transistor, which I'd find hard to believe),
and didn't have well-defined
logic levels.

The fact that the output switches to ground and can take 100mA max, tells
me, as an electronics designer, that the output IS a switching transistor.
No analog values, just a hard switch to ground, with a maximum so called
saturation voltage of about 0.1V. This means that at 100mA the heat
dissipation inside the GPS is a whopping 10 milliwatt. No need to add an
extra transistor.

Futher, it says that the maximum current is 100mA and therefore the relay
coil may not be less that 120Ohms. There's nothing wrong in using a relay
with a coil of 500 or even 1000 Ohms, to reduce power and 'heat' any
further.

Meindert


I agree with your points. I guess it's up to the alarm device builder to choose
the method.
But, also as an electronic designer, I generally don't like to approach max values.
I use a conservative 25% of ratings as a design goal, based on reliability charts.

Moot point, but don't most silicon transistors have a Vsat of about .2 Volts?
I wouldn't be surprised if the device in question was actually a FET.
Norm