Talking of capacitors, I remember UK TV news back in around 1990 give or take, during the time of the Iraq "Super Gun". Apparently some capacitors had been intercepted by the authorities in London, whilst on their way east. They were claimed by BBC/ITN to be initiators for an atomic bomb, but the ones displayed looked very similar to some I had bought from likely RS Components for a low voltage DC power supply.  Whether the BBC had just grabbed some surplus from an Edgeware Road electronics shop, I don't know.....

Clive

These are both horrible designs. Mikes depends on the transistors being mismatched to do anything, and if the power supply is raised slowly enough would still do nothing. Probably it would work 90% of the time and do something! The first is very odd, in that VR2 is not required and does nothing useful, and this circuit was probably thought up by a 1st year undergraduate. Good for MUGS I suppose and might cause a little discussion over coffee, which might be the idea. Consider the temperature stability of a regulator against a voltage reference, .some are good, some are not! The hysterysis of the comparator circuit would probably stop most of the relay chattering I suppose.

The relay coil will be rated for a specific DC operating voltage which probably explains the use of the regulator. We do not know what the unregulated voltage range is from the diagram and it could vary considerably even with transforming of the mains AC down to a lower voltage.

You may not like it Dave, but even in the perfect world of computer simulation, the astable flip flop does not  require the devices to be imbalanced - it will start on noise - I've never yet built one that does not start, (*) and they are in all sorts of things I have done,

The very first flashing lights for my toy train (with two germanium, transistors and small torch bulbs), a very bad electronic music generator (factory reject BC107s from a magazine advertiser of the late 1970s), the saving grace being it only drove headphones, and a device for measuring light levels on forest floors using more of the BC107's, but with the top cut  off and blooped with clear resin to make photocells in series with the timing resistance) Later as the oscillator to generate the programming voltages for an Eprom programmer (2n709s I think - I was nearly 20 by then)  and many since some professionally, and  with scouts for flashing LEDs and so forth.

And full circle, with my kids, as lights for a toy railway, but with LEDs not filament lamps..



Mike

(* ) you do need HFE >> Rbase/Rload and electrolytic caps need to be low leakage, and RC to be long compared to ft/HFE, so not much faster than audio only.


I grant you that the three and 4 stage ring oscillator versions can be unreliable..

Why is the load connected to L and switched via the relay to N ?

anti-meddling config?