[MAMS] relay driver

[Dr. Gerald N. Johnson]

Yes, that should work, rather hammer and tongs though.

You want the capacitor to not really get much charge in the first 20 
milliseconds. So the relay coil resistance and inductance  with the 
capacitance want a time constat about 40 to 60 milliseconds. The bleed 
off resistor and capacitor needs a much longer time constant. Probably 
two RC time constants for the shortest period you expect to stay in RX 
or TX. Like a second or two.

While the relay takes 15 ms to switch, what length pulse do they 
require? It might be a hair longer. One spec shet I saw said 20 ms pulse.

A hair more complex and probobly a lot smaller space could use a pair of 
555 timer chips or a 556 dual timer. Wire each to be a 20 ms one shot 
pulse generator, one to trigger on rise, one to trigger on fall.

Or getting fancier, a 74hc221 one shot, and a dual interface driver chip 
in the 75460 family. The 75460 have a dual input logic gate input, so 
you could use one NAND and one NOR with the 221 triggered by rise or 
fall and that same trigger signal going to the two drivers, one NAND and 
one NOR so the output of one 460 would be enabled on a high and the on a 
low. Providing these chips can be found these days.

Or you could pick a PIC chip with a couple sturdy power outs, and use an 
input to sense the relay control line, high or low logic level (3.3 or 5 
volts depending on the PIC chip supply) and program it to produce a 20 
ms pulse on the appropriate output. Or the PIC chip could drive some 
power transistors, bipolar or MOSFET. The PIC chip could probably be one 
with 8 leads, maybe 6 using an internal oscillator and if in a surface 
mount package would be tiny. I'd tend to use a DIP package because its 
easier to use on a RS universal PC breadboard. The breadboard will cost 
more than the small PIC chip, which might be as costly as 79 cents in 
single lots. And if you buy enough I make money because I've invested in 
that company's stock which has risen at least 15% since I made my 
investment. Where it may be time to start taking profit by selling the 
stock, soon as this month's dividend arrives.

Op amps can also be made into level triggered one shots, and an audio 
power output chip with lots of gain and usually operated with lots of 
feed back is simply a power op amp and ought to be able to drive the 
relay coil directly.

Look up indices to CSVHF and other VHS society proceedings as well as 
MUD and I'd expect circuits for driving pulsed relays have been 
published in the past decade or so. Not having any such relays yet, I've 
not bothered to remember such circuits in detail.

Your link to dowkey didn't work for me until I appended the hp.

If you were to add enough capacitance in parallel with each coil to have 
a discharge time constant of 10 or 15 milliseconds, you might be able to 
use the indicator contacts for steering, so you apply power once to 
change to TX and again to change to RX with the indicator contacts 
selecting the coil to pulse and disconnecting power as the contacts 
change position. It probably will disconnect before 10 or 15 
milliseconds so the capacitor on each coil is needed to supply power to 
the coil just long enough after the indicator contacts disconnect.

8 ways to build a clock, beats RIW's 6 ways.

73, Jerry, K0CQ

On 9/7/2012 4:44 PM, Danny Pease wrote:
> I wonder if I can simply put a rather large capacitor on series with each
> coil and switch the DC from one to the other when going from TX to RX? The
> discharged cap should look like a short for a few milli-seconds, allowing
> the pulse to activate the coil and then when the cap is charged, no more
> current. Of course there would have to be a discharge path to bleed the cap
> off. The coils are about 60 ohms and the specs say the relay takes 15
> milliseconds to switch, so maybe a 470 uF cap with a 610 ohm resistor in
> parallel?
>
> Danny NG9R
>
>
>