<wpwrak>
hmm. another one goes up on smoke. seems that some current-limiting may be beneficial :)
<wpwrak>
well, another FET could detect if Vout < ~0.7 V and trigger an interrupt. or directly cut power
<wpwrak>
and a 1 Ohm series could further limit inrush current. just in case
pcercuei has quit [Ping timeout: 255 seconds]
atommann has joined #qi-hardware
atommann has quit [Ping timeout: 255 seconds]
<wpwrak>
whitequark: also tried a load step from ~100 Ohm to ~18 Ohm (at 1.8 V). i can't see any voltage change. not sure if this is good or bad ;-)
<wpwrak>
at least the FET, 3rd in this role, didn't go up in smoke :)
atommann has joined #qi-hardware
rjeffries_ has joined #qi-hardware
rjeffries_ has quit [Remote host closed the connection]
archang has quit [Quit: Konversation terminated!]
archang has joined #qi-hardware
archang has quit [Ping timeout: 245 seconds]
archang has joined #qi-hardware
atommann has quit [Ping timeout: 255 seconds]
<DocScrutinizer05>
huhß life in here?
<DocScrutinizer05>
(<wpwrak> i wonder if this simple adjustable regulator could actually work) it would, but pretty poor efficiency
archang has quit [Ping timeout: 265 seconds]
<DocScrutinizer05>
so it's OK for very low loads (except you'd still want a dedicated series resistor in line with D-S of FET), but for real supply it burns exactly same amount of energy like a linear regulator
<DocScrutinizer05>
while introducing nasty noise
<DocScrutinizer05>
actually it doesn't introduce any noise when that's an opamp and not a comparator (with hysteresis), since then your FET does work in linear range and not switched like you implied
<DocScrutinizer05>
you should add some small buffer capacitor on output of opamp, to avoid it starting oscillation
<DocScrutinizer05>
the problem with your design though seems that you want to use a *comparator* in a MCU
<DocScrutinizer05>
this won't fly
<DocScrutinizer05>
unless you intentionally design in some RC elements that determine the oscilation frequency
<DocScrutinizer05>
your design best case is using the output buffer C and the assumed Ron of FET as such determining values, and that means you _will_ get wild oscillation at least in the magnitude of comparator's hysteresis
<DocScrutinizer05>
oooh, and of cours Rload for discharge of the buffer C
<DocScrutinizer05>
in theory the circuit works. In practice it's a nasty critter hard to tame and get stable
wej has quit [Ping timeout: 248 seconds]
<DocScrutinizer05>
wpwrak: you probbaly should add a R-C lowpass/integrator after the comparator output and drive the FET gate from that integrated analog voltage
<DocScrutinizer05>
that might actually work and result in a semi-usable and semi-stable output voltage
wej has joined #qi-hardware
<eintopf>
hi DocScrutinizer05 :-)
<DocScrutinizer05>
hi!
archang has joined #qi-hardware
unclouded has quit [Ping timeout: 256 seconds]
archang has quit [Read error: Connection reset by peer]
fengling has quit [Quit: WeeChat 1.1.1]
unclouded has joined #qi-hardware
<wpwrak>
(RC) naw, that would defeat the idea of operating in switched mode. what i want is to minimize the power burnt in the FET, so i want it either all on or all off
<wpwrak>
so far, the critter seems to work suspiciously well, but with a slight tendency of regulating too high with fast settings
<wpwrak>
what is a little odd is that it drives the FET with bursts of pulses. so yes, there's more noise than strictly necessary
jekhor has joined #qi-hardware
pcercuei has joined #qi-hardware
jekhor_ has joined #qi-hardware
jekhor has quit [Ping timeout: 252 seconds]
jekhor has joined #qi-hardware
jekhor_ has quit [Ping timeout: 250 seconds]
wej has quit [Ping timeout: 246 seconds]
wej has joined #qi-hardware
jekhor_ has joined #qi-hardware
jekhor has quit [Ping timeout: 252 seconds]
rjeffries has quit [Ping timeout: 252 seconds]
jekhor__ has joined #qi-hardware
jekhor_ has quit [Ping timeout: 258 seconds]
pcercuei has quit [Ping timeout: 250 seconds]
<eintopf>
in germany it's hotter than in buenos aires