I think that NPN transistor has emitter and collector reversed. It's the classic "programmable unijunction" circuit.

 

THanks for the reply, I'll be reading it with an IR receiver so not worried about seeing it, just about saving power by keeping the pulse short.

OK, so the reason for keeping the pulse short is battery conservation, not some precise time requirement. That makes the solution simpler and much less critical.

 

OK, so the reason for keeping the pulse short is battery conservation, not some precise time requirement. That makes the solution simpler and much less critical.

In which case, you need Danko’s circuit. You will be surprised how much current HC14s use!

 

In which case, you need Danko’s circuit. You will be surprised how much current HC14s use!

That is why I suggested the CD4049 circuit. The current is what the load draws.

 

This is how a 4069 behaves when the input voltage approaches mid-supply.
There's no equivalent graph in the 4049 datasheet. Do you think it behaves differently?
I'm assuming you're using the classic 2-gate CMOS oscillator, as a single 4049 won't oscillate, having no hysteresis.