Thank you all for your fantastic feedback - I already ordered some 555s to play with and will try to go down the P-channel route again, when time permits. Yesterday I did find the cause of my original circuit shutting down - it's always the thing you take for granted: The Shottky diodes weren't up to the amperage of an incandescence bulb.

I replaced those with some SB5100 (https://www.mouser.at/datasheet/2/308/SB580-D-1814696.pdf) I had lying around and everything worked fine - though the diodes still became hot to the touch when feeding them with 2-3A.
I've looked around a bit; the SB520 has lower voltage drop so should be cooler and I don't need the 100V from the SB5100.
Any other Schottky diodes you could recommend for this? Ideally they'd take 3A at 14V in 50C ambient with no mentionable heating.

 

Any other Schottky diodes you could recommend for this? Ideally they'd take 3A at 14V in 50C ambient with no mentionable heating.

It's not 14v but 0.48v max drop @ 1.5A each diode which is ~0.75W. The SB520's main thermal path is through the leads so keep these as long as possible and solder them to the largest area of copper you can, preferably in a vertical position. They should be OK for this.

A more interesting choice would be something like the SBR10U40CT which has two diodes with a common cathode in a thermally more efficient TO-220 case. They are lower voltage drop diodes as well.

 

It's not 14v but 0.48v max drop @ 1.5A each diode which is ~0.75W. The SB520's main thermal path is through the leads so keep these as long as possible and solder them to the largest area of copper you can, preferably in a vertical position. They should be OK for this.

They'll have to deal with up to 3A (2 blinkers at 21W each at around 14V) - realistically it'll be lower amperage as they'll be dimmed. The drop on the SB5100 I used is about 0.6V at 3A which is about 1.8W - enough to get it warm.

Great tip about theSBR10U40CT - amazing the things that exist! That one really seems to fit my needs!

Thank you so much for your time & energy. I feel very welcomed and supported in my beginners-endeavours!

 

They'll have to deal with up to 3A (2 blinkers at 21W each at around 14V) - realistically it'll be lower amperage as they'll be dimmed. The drop on the SB5100 I used is about 0.6V at 3A which is about 1.8W - enough to get it warm.

With a 25C/W thermal resistance to ambient, at 50degC ambient and 1.8W the junction temp will only be 95degC, well below the 150degC limit for those devices. The lead temperature will be hot to the touch at around 81degC.

 

The diode's leads act as a heatsink. Within reason for whatever the circuit packaging is, longer leads are better than shorter.

ak

 

The diode's leads act as a heatsink. Within reason for whatever the circuit packaging is, longer leads are better than shorter.

ak

Agreed. That's true for BJT/MOSFETs too. Its little understood but upwards of 10% of cooling on TO-220 etc packaging comes from the collector/drain pin. That's why, ideally, you should always have a PCB attached to the device rather than just a wire to some remote board.

edit:
Just found the CFD simulation I did recently on a PLUS247-cased IXYS MOSFET running at 1W in free air on a PCB as spec'd by JEDEC to assess thermal resistance junction-ambient by natural convection. Red is hottest (90+C), blue coldest (25C, ambient), yellow then green are 70 - 45C. See how the heat also travels down the drain pin and heats up the PCB to around 35-40C at the solder joint..