LED Flasher

[Alistair Ballantyne]

Hello,

I’m struggling to build the following circuit.

I have a reasonable grasp of why it should flash – but I just cannot make it work!

The LED remains constantly lit – and that’s it.

The breadboard is not particularly elegant as I’ve spaced things out to show connections. I’ve used two resistors from the NPN base as I don’t have an exact 330k.

I was hoping someone more experienced could shine some light?

Thank you very much.

 

[danadakk]

Sim shows it sorta works, poorly. Notice the spikes on Q1 emitter LED current, when it turns on,
as you can see for only a very short amount of time.

Google "2 transistor led flasher", there are some other circuits that you could try.

Regards, Dana.

 

[barry]

I don't think that circuit is an LED flasher; I don't know what it is. The way I see it (and I've been wrong before) is that the NPN is always on, which means the PNP is also always on.

At power on, NPN is off and the base of PNP is floating (this already looks like a problem). The 10uF charges up until NPN turns on, turning PNP on. When the PNP (and LED) turn on, the voltage on the (+) side of the 10uF will rise, driving the base of NPN even higher. There's no way the base of NPN will ever go low enough to turn it off.

--- Updated Jan 14, 2021 ---

See, I told you I could be wrong. I also just did a simulation, and I have to admit, I still don't quite understand how this works.

 

[c_mitra]

I guess the 10uF cap is connected wrongly; try reversing it.

 

[dick_freebird]

The human eye can't distinguish steady from
high rate flashing, other than by intensity. Your
eye can't follow flicker above maybe 30Hz. It
could be that the passive component values
set up a flash rate that looks to you like always
on.

Presuming it's working as planned, otherwise.
I know the old LM3909 would work to flash-rates
well above visible.

But that would be a matter for a 'scope, or for
a poor-boy audio tone sniffer like a cast-off
headphone speaker and a series resistor that
you could put across the diode and hear a
range that your eye can't follow (at least some
decades above 30Hz).

 

[c_mitra]

Sim shows it sorta works, poorly. Notice the spikes on Q1 emitter LED current, when it turns on,
as you can see for only a very short amount of time.

Something surely looks fishy. See the simulation graphs: the top one. Q1/Base potential.

The voltage goes above 300V. The power source is a 6V battery and there are no inductors in the circuit.

I do not think the simulation is correct. You need to check and see that it works as expected.

 

[Alistair Ballantyne]

Thanks a lot for your input.
I have previously made the 'standard' flip-flop flasher and worked well.
Just thought it would be interesting to use both PNP & NPN transistors in a circuit.
Your views clearly tell me the circuit is not fit for purpose. No matter - at least I have learned!
I'm intrigued by the simulation model. Is this something you experts can 'manufacture' by computer modeling to analyse the circuit without having to build it? Impressive!

 

[danadakk]

I concur something is fishy with the sim. Looking into it now.

Yes sim can be a useful tool to check out ideas. One has to be
careful, its not foolproof, as manufacturers spice models for their
parts not necessarily complete characterization of part.

A lot of people use LTC Spice, OrCad, Tina TI, SIMetrix.....most are free.



Regards, Dana.