accurate 30second period for a clock

[alec_t]

A cheapo clock from a dollar/pound store would give you a pretty accurate 1-pps as a starter.

 

[Les Jones]

Following on from Alec's suggestion (Post #21) if you get a clock with a seconds hand you could use that with an slotted opto sensor to generate a pulse every 30 seconds. You could make a second hand that had fingers 180 Deg. appart or use two sensors spaced 180 deg. appart with the outputs ORed together. You could adjust the pulse length by changing the width of the second hand or you could trigger a monostabe to adjust the pulse width.

Les.

 

[mab2]

Ooh, more ideas! Thanks. Well I've ordered a cd4059 so I can either go with my old 3.2768 Crystal (found another crystal with legs) - that will probably be accurate for the 6months between bst and gmt changes anyway. Or i can get one of those rtc will 1pps outputs.

As for the pulse requirements: it just pulls in at 5.6v, pulls in at 6v but takes over 0.75secs, or at 12v it works in ~ 0.2 secs. Above 12v it seems to be a bit too energetic for my liking. Coil is 44ohms and I've no idea what the design pulse is supposed to be.

 

[crutschow]

it just pulls in at 5.6v, pulls in at 6v but takes over 0.75secs, or at 12v it works in ~ 0.2 secs. Above 12v it seems to be a bit too energetic for my liking.

So it would seem a 12V pulse of about 300ms should work.
The current will be about 273mA so a transistor such as a 2N2219 or 2N2222 or a logic-level MOSFET would work as a driver.
Make sure you put a diode (1N4148 or similar) across the coil (cathode to plus) to protect the driver from the inductive kickback.