My own simplest discrete 555 timer representation design

[Edin]

While making a custom-tailored circuit with an NE555 timer, I kept coming across issues which I had to resolve one-by-one, and at the same time get a more complete understanding of the 555 IC.
In order to more easily work the problems out, and out of creative and intellectual curiosity, I searched the Internet for a greatly simplified/simplest discrete schematic representing the 555 timer IC, but I didn't find it.
Grudgingly, I decided to come up with one, and after a few hours it was done.
 The attached image is the simplest discrete schematic of the 555 timer I could come up with, outlined with a green box. It includes a few "external components" to make an astable multivibrator, to make it easier to visualize how it works.
 I have actually BUILT this circuit to make sure it works and that I am not missing something before I share it with the world, and the second attached image is the photo of that built and tested circuit.

SR latch
The SR latch section was built and tested first, as I had never built one before and it was critical that it works properly.
I wanted to default its output to "high" at turn-on because that's also how the 555 actually works in an astable multivibrator when the timing capacitor is at zero volts or below the upper, 2/3 of the supply voltage, threshold.
To that end, I had to make sure the two transistors are unbalanced and that one is certain to turn on before the other, and this is done by using a smaller value base resistor for that transistor. I tried using 1.5:1 ratio but that was not enough, so I went with 2:1 ratio and that worked.

Comparator and voltage divider
For the divider, I simply went with the classic 5k resistors for simplicity and stiffness of the voltage thresholds when loaded through the input transistors Q1 and Q2.
Even though the established thresholds are 1/3 and 2/3 of the supply voltage, the actual thresholds at which the input transistors switch are about 0.6V below and above, respectively, due to their base-emitter voltage drops at their turn-on.
In case of a 3V supply, established thresholds are 1V and 2V, but actual voltages at which the lower and upper transistors switch are about 0.4V and 2.6V, so not exactly ideal, but it works.
This may be fixed by inserting diodes in series with the voltage divider resistors to counteract the base-emitter drops, but I decided to keep it simple, nice and classic looking for easier reference to, and understanding of the actual 555 circuit.

I used high value resistors at the input transistor collectors to reduce loading on the voltage divider, but the lower transistor couldn't provide sufficient current to the SR latch through a 100k resistor due to only about 0.4V between lower (trigger) threshold and the base of the SR latch transistor Q6 (about 4uA base current), so I reduced it to 47k and it worked.

Testing and conclusion
The testing of an actual breadboard circuit showed that it works as expected from 3V to 18V, stopping its switching action just below 2.80V and the output was stuck at low at about 20V.
The switching frequency in my case changed from below 10Hz to about 25Hz when supplying from 3V to 18V.
At 3V of supply voltage, 100k and 10k resistors with a 1uF capacitor gave about 100ms Ct charge time (output high) and 10ms Ct discharge time (output low).
The transition times were fairly long at the output (loaded only with a red LED with a 10k resistor towards Vcc and a 10M and 20pF oscilloscope probe) : rise time was about 150us and the fall time was about 5us, which was to be expected due to high resistances involved and no focus on speed or stiffness of the output.
Additional note is that though I tried to make the indicated "IC pins" behave just like the actual NE555, the RESET pin works differently: while on the actual IC it pulls the OUTPUT pin low when less than about 0.8V is applied to it, the RESET pin in my circuit does that when its voltage is at least about 0.5V below Vcc, but it still needs to be connected to Vcc to keep circuit stable like in the actual IC. This circuit was meant merely as the simplest proof-of-concept for the classic 555 timer IC.
To my knowledge, most of the typical 555 circuits should work with this one as a drop-in for the NE555, provided they don't exceed its known limitations which I have mentioned here.

P.S.: I know that much can be changed or improved here, and you are welcome to pitch in, but the most welcome comments/answers would be those which either make my circuit simpler or better-performing without adding more components.

[moffy]

While working with an NE555 timer to give me short, 6ms pulses every 1.5 seconds, along with a few other peculiar needs, like greatly reduced quiescent current and driving a MOSFET, I kept coming across issues which I had to resolve one-by-one, and at the same time get a more complete understanding of the 555 IC.
In order to more easily work the problems out, and out of creative and intellectual curiosity, I searched the Internet for a simplified discrete schematic representing the 555 timer IC, but I didn't find it.
Grudgingly, I decided to come up with one, and after a few hours it was done.
 

The LM555 datasheet: https://www.ti.com/lit/ds/symlink/lm555.pdf
gives a discreet equivalent of the LM555 timer on the first page. The 'trigger' and 'threshold' should at least be differential pairs to compensate for Vbe drops and temperature effects, you could get away without the darlington differential pairs but be aware that the extra base current will produce increased offsets.

[Edin]

I am definitely aware of the FULL schematic, but my goal was the SIMPLEST schematic which would make it imitate the actual IC.
The datasheet schematic has 28 transistors, 31 if you consider the fact that 3 of them can't be replicated in a discrete circuit because of their multiple collectors per single transistor which are only done in integrated circuits. Additionally, it is harder to trace and visualize the signal paths in it.
My circuit uses only 6 transistors in the most basic ways so it is significantly easier to make it and figure out how it works.
And yes, I am aware of the benefits of using differential configurations, but that would increase both the number of components and the complexity of the circuit.
My main goal was simplicity and ease of understanding it even for the beginners in electronics who only know the most basic concepts of how transistors work, and I think I have achieved that.
I was merely sharing this with the world so that others (students, beginners or those who struggle with understanding basic circuits like the classic 555) can be encouraged to go deeper into electronics and to actually build these circuits and apply them without fear and with a more full understanding.
If some people have suggestions which make this circuit even simpler while having the same functions, I welcome them.

[ledtester]

The datasheet schematic has 28 transistors, 31 if you consider the fact that some of them can't be replicated in a discrete circuit because multiple collectors and emitters per single transistor are only done in integrated circuits.

It might be possible to emulate, say, a multiple emitter NPN by parallelizing the collectors and bases of two discrete NPNs, e.g.:

https://www.electricaltechnology.org/2021/08/transistor-types-of-transistors.html#multiple-emitter-transistor

You might be interested in the design of the "Chip 222" - a theoretical proposed building block like the 555 for timer applications:

https://www.researchgate.net/publication/354694789_Shustov_MA_Chip_222_-_alternative_555_PWM_generator_with_independent_frequency_control_International_Journal_of_Circuits_and_Electronics_-_2021_-_V_6_-_P_23-31

(PDF attached below)

Internally it only uses two comparators.

It's also mentioned in this EDN Design Idea:

https://www.edn.com/single-button-load-switches-on-the-chip-222/

[Edin]

The datasheet schematic has 28 transistors, 31 if you consider the fact that some of them can't be replicated in a discrete circuit because multiple collectors and emitters per single transistor are only done in integrated circuits.

It might be possible to emulate, say, a multiple emitter NPN by parallelizing the collectors and bases of two discrete NPNs

 If you carefully read my text you have quoted, you will notice I said 28 transistors are in the IC, but 31 if they had to be replicated in a discrete circuit by doing exactly that, i.e. paralleling multiple transistors instead.
 I was not aware of the 222 IC, thank you for sharing that information. It has some interesting properties.