Re: Constructing a "Power on time delay relay" with 5 sec. delay. max 6V

<david_gillberg@xxxxxxxx> wrote in message
news:1121006636.608186.31480@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
[...]
> I have found a circuit-drawing of a time delay relay, but it demands
> 12V and has a delay of roughly 7 sec.
>
> http://ourworld.compuserve.com/homepages/Bill_Bowden/page2.htm#delay.gif
>
> Since my motors run on 6V I would like the relay to do that as well.

> QUESTIONS:
> 1. How do I calculate the time delay? In the text the author states
> that the delay can be reduced by reducing the R or C values (R=47K,
> C=100uF)but is there a formula that can be used?

The time delay will depend on the reverse breakdown voltage of the
base-emitter junction in the '3904 used as a Zener. You can figure that the
2N3904+2N3053 pair (in a "darlington" configuration) will be fully 'ON' when
the voltage at the base of the 2N3904 (of the pair) is 1.2V. So, as the
100uF cap charges through the 47K resistor, its voltage rises. This voltage
is reduced by (according to the web page's author) 6-9V due to the "Zener"
and then applied to the base of the "darlington".

> 2. Will this solution turn the robot on at "full power" after the delay
> or will it start softly (when I push the button or when the time is
> up?)?

If you use a relay to switch the robot power, it will be ON or OFF. I
wouldn't use this circuit to control the motor directly, as the motor
current will "ramp up" somewhat during turn-on. Stick with a relay to
switch the motor current ON or OFF.

> 3. How do I have to change the circuit to reduce the voltage to 6V?

a) Use a relay with a 6VDC coil, making sure that 'pull-in' votlage is no
more than 5V.

b) Replace the "zener" 2N3904 with something that will drop less voltage.
The 6-9V dropped by the '3904 clearly won't be compatible with a 6V supply
rail. Some people use an LED which will drop between approx 2V and 4V
depending on its color, others use several 1N4148 or 1N4001 diodes in series
to obtain the desired drop. If you were to use 3x 1N4148 diodes (very
common "switching" diodes) you'll end up with the darlington pair saturated
(ON) when the cap voltage is approximately 1.2V + 0.6 + 0.6 + 0.6 = 3V.

c) Adjust the 47K resistor to change the delay time, keep it above 1K ohm
to keep excessive current from flowing into the base of the darlington.

> 4. Is there a better alternative that wouldnt cost me more than perhaps
> 20$?

This shouldn't cost anywhere even close to $20.

This circuit that you found will not behave so well over large temperature
extremes, or if the 6V supply voltage is going to vary quite a bit (the
threshold of the transistors doesn't track the supply votlage). Nor will it
be that repeatable from unit-to-unit -- so if you're going to build a whole
army of sumo robots, they'll all have slightly different turn on delay
times. For an army of one that only has to work indoors, however, it'll
probably work just fine.

I would recommend a means to help discharge the 100uF cap when the power is
interrupted - the web page author put a 10K from V+ to ground, but the 100uF
cap will have to discharge through the 47K resistor in SERIES with the 10K,
that's a pretty long discharge time. What this means is if the power were
applied briefly, then disconnected, then reapplied briefly, then
disconnected, &c, the cap will charge up a little bit each time, and
eventually the relay could turn ON before 5s elapsed when the power was
applied "for good" If you place a diode across the 47K resistor with the
anode at the 100uF + terminal, and the cathode at the +6V point, the
capacitor will "discharge" across the power supply rail when power is
removed.

A more precise turn on delay could be generated with the venerable LM555
timer circuit - a Google search will reveal lots of references to this very
common (and cheap) IC. While you might not find the circuit you need
exactly, it ought to give you lots of good ideas.

> As you can see I have'nt got much experience of electronics, so pelase
> be patient with me :)

Have fun!

TJL


.