automatic duty ratio control experiment

a definition correction :: the 555 acts as the 555 e.g. an inverting comparator with the high accuracy hysteresis relative to supply rails (however it's internal hw setup is similar to a window comparator namely an "outside the window comparator" - only it's comparators´outputs are not bound together by simple combinatorical (read: boolean algebraic) operation - which for outside the window comparator would be logical sum) - but instead they drive an RS-trigger which inverts the upper comparator output and remembers that output until the lower comparator triggers - then remembers it's output until the upper triggers - which in a sense of a window comparator would be - depending on input slope - an unidirectionally active operation of the pair of inverting and non-inverting comparators . . . the least however reduces to the case of a single comparator with relatively large hysteresis and inverting output

Low Voltage Op-Amps -- in-simulation TEST

(1) -- verifying "std. responce" + "model ok for simulation"

(2) -- relative "performance"

(3) -- Exactness/Speed // Design/Setup

(4) -- capability for specific operation in low-/limited- supply voltage application

---

[Eop]

Linear Supply with LM113 - experiment

null

null

LM113 quest for practical use

what it seems by now is that utilizing the potential power of the chip requires the target oriented , also a clever and sophisticated "trick" -- not a straight forward block level design but the one that takes into account the specific nature of the LM113 in this case . . . i yet am to figure out both - the trick and the nature . . .

experiments ::

used as shunting regulator

d/s version
PS! altering the ellipsed values causes the circuit to become unstable/oscillate

combined trivia

 

used as zener

shared function in feedback chain
not used at all / instead we use positive error feedback here ▼

fast opening non-linear/log. bridge/circuit elements

 

used as zener -- OR -- if we consider the j-Fet´s V.gate = fn.(I.source) then seeing it upside-down it acts as shunting ref. functionally shunted to Op Amp OUTP or V.gate

j-Fet bridge

---

[Eop]

FW - TL431 LM113 Spice model variants´ TEST

it looks like i already forgot what i graphed
Z=V(voltage point)*normalizer -- sets all references to a default unit value (1V)
relative scaler log10(Z)*relativeNormalizer -- resizes plots so they fit to same scale . . .
. . . so the smaller the "relativeNormalizer" is the worse the device performs (has greater fluctuation range)

so the "Z" re-plotted (for stepped current) ::

 . . . the 1.67V µm 431 variant seems to perform best here in both current ranges (x431/x113)
- - - -

---

[Eop]

TL431 Spice model variants´ TEST

presented as we go
► 1-st

► 2-nd

► 2-nd - in detail

wider range

component level models (sub circuits compared)

the best - upper two are the attempts to achieve better precision×frequency-rejection //// the reverse Vref´s emitter (EB) diode in Rank#(7) is a contemporary relic from an attempt to force the "Hot Start" for U4,U5 . . .

pulse e.g. stepped load response

overall ranking so far ...

? best CLM-s !!! for static operation . . .
. . . for undefined input the U5 e.g. X4 is likely better choice /// might be quite expensive regulator as the RF transistors cost $ 0.2÷1.0 per pcs.  /// (i can't comment nor recommend nor reject the macro models - but usually they are too optimistic !!!)

hi, hi - UPS! seems i used Amps insted of milli-Amps

that re-ranks the rank back to "Static" -- the red box is "for super cool experiments only!!" . . .
. . . so U2 // X1 the best CLM !!!

---

[Eop]