Sunday, April 19, 2015

Simple Current Limits Compared (Updated!)

Type.1

Type.2

Update !!!
► emerged by the predecessor of the second from the right - a "constant current shunt" - that was the best performing such in one of the experimental linear regulators (now revisited)see also :: yet another constant current source

Friday, April 17, 2015

Regulator Redesign (/Lab -- as always . . .)

________________________________________________
2015-04-15 ::
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
 . . . trying to "de(/un)-bug" the last design from prev. post -- normalize power range , minimize the error . . .

+ a modified v.'s foldback diagram


Customizing/"improving" the 78Lxx for 1.2V

i only guess it's 3.0V
________________________________________________
2015-04-16 ::
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
"Minimizing Compile" , exploring the switcher mode v. , experimenting with foldback . . .

thought to use the LT's µ-P 2.5V LDO -- but the 3uA is likely causing it's instability in high loads env.


________________________________________________
2015-04-17 ::
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Optimizing "Design"

Update! :: the foldback re-visited . . .
Update+ :: ±10µV version ::
Update+ :: ±12µV version with foldback updated to support high load capacities (no frequency test - what so ever- carried out + i donno how to thermal compensate - /!\ so i never do /!\) e.g. you have to recalculate the design before any attempt to assemble it

________________________________________________
[EoP]
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Saturday, April 11, 2015

a low control current capable voltage regulator

Here's the grid - simulated fine (actually excellent -- that's why the design's revisited . . .) in prev. (a non TEST) v. but it didn't regulate "at all" in a real build -- here "we" possibly identified why -- if the R1 is too high (too little load) the VZ may go high thus also VT climbs out of the range and - bam! - we have a multiples of higher voltage on output than it's expected (the RED outlines the voltage regulation logic) ::
Next - omitting the CC module + startup chain - simplifies the design but we loss a magnitude of precision ::
Here follows a very poor example of a step-dn swithing regulator - asctually i didn't preserve the original v. of it (this is a fast feature restoration) coz i could not clear/filter the unwanted ringing and simulates extreamly slow . . . however the feature is a "built in" PPM (a pulse position modulation) capability (extra to low control current capability) -- needs a lot more testing to get it near right . . . (notice the varying supply internal resistance -- is why the efficiency graph wont link up exactly) ::
Omitting the VT reference point results in additional precision loss . . . but it revealed some specific behaviours of this type of regulation scheme (some of which're mentioned on o figure) ::
Here we've a near max low control current scheme - it starts auto-limit it's output current (depending on input supply) @ 150...200mA (output current max 235mA -- SC current 238mA) . . . and it seems to use some 110µA for regulation . . .
. . . a big deal considering the Solar battery test design with avg. 10.5µA TOT.supply draw - 8.8µA load current = 1.7µA (16%) regulation current . . . ::
. . . while the "near max low" v. ↑↑ has 108uA / 18mA avg. = 0.59% ( 120µA / 83mA peak = 0.15%) regulation currents

Sunday, April 5, 2015

Adjusting the adjustable voltage regulator

The goal as in trivia is to gain more stability and precision while keeping the design robust . . .


9V ± 1.8µV , 300/646mA (max.load/SC.supply current)
-- the stability (ringing(OpAmp-s e.c.)) + SC condition are a question marks here


((Simple with SC don't care /!\)) 9V + 0.6 ± 2.6mV , 300/687mA


((Simple SIPMOS with SC handling)) 9V + 4 ± 475 µV , 300/687mA


((BJT alternate of the above)) 9V - 48 ± 255.6µV , 300/56.7(398)mA (max.load/SC.supply current.average(.peak))


((feature test for FET switching)) 9V - 7 ± 13mV , 346/706(CuLm+)mA , 23.7kHz (here for err. compare)
-- the main targets were the inductor&frequency versus load&error (the current limit is a quick fit -- feature test)

[EoP] (XT kogu aeg on mingi vitustus selle küljendamisega -- ma korjan mittevajalikud reavahed maha - aga voila! - peale save´imist on nad kuskilt tagasi tekkinud (teksti ja pildi vahele topitakse <BR> tag´id - et kui on huvi lõpmatuseni oma tekstiga jantida ÕööÖ×ØÛÜÚÝ♒☣☠☦☯ ))

Rebuilt two transformers

There seems to be a dependency for 50Hz Sine transformers - such that the less ampere turns on secondary the less actual output power . . . no matter the copper diameter !

The TF.2 has 2-le - and the TF.1 has 3-ple in parallel secondary winding from 01..0.08mm wire - e.g. - a lot of amp.-turns.
so here what we got



Transformer-Rating-A.xls



Transformer-Rating-B1.xls










TF.2
5W2DRW
After




W2V33
Before


2x
Z.in
10283.84
(Ω)



Z
9833.044786
32xEI+2E
XL.in
10260.87
(Ω)



XL
9809.347823
17.5
dpth
L.in
32.66136
(H)



L(H)
31.22412389
41
wdth
R.in
687
(Ω)



R1(Ω)
682.25
33
hght








6:8:13:8:6
Z.out
74.103
(Ω)





21

XL.out
70.26131
(Ω)







L.out
223.6487
(mH)



L(mH)
26.7864545


R.out
23.55
(Ω)



R2(Ω)
0.9












TFRT
11.78039
(1)



N(1)
34.086046

~AC










Norm.IN
240
(V)




240


Norm.OUT
20.37283
(V)




7.041004405

~DC










Norm.IN
240
(V)







Norm.OUT
28.33159
(V)






about*





Simulated



*REAL
RL(Ω)
u. RL(V)
p. RL(W)
i.RL(mA)





SCir.Max
0
0
0
310.6169

3xVtg.Mpl
1.34W 25.7V 51.5mA @ 497sz

RL.P.Max
74.103
14.17819
2.712727
191.3309

normal
1.27W 8.48V 150mA @ 56sz


80%
170%
210%
130%














TF.1
NiMH-YR
After




YR
Before


3x
Z.in
14195.2
(Ω)



Z
9320.274036
30xEI+2E
XL.in
14153.2
(Ω)



XL
9215.587833
16
dpth
L.in
45.05103
(H)



L(H)
29.33412714
35
wdth
R.in
1091.25
(Ω)



R1(Ω)
1393
29.5
hght








5:7.75:9.5:7.75:5
Z.out
482.172
(Ω)





19.5

XL.out
477.6543
(Ω)







L.out
1520.421
(mH)



L(mH)
27.65934601


R.out
65.85
(Ω)



R2(Ω)
2.55












TFRT
5.425875
(1)



N(1)
32.08109189

~AC










Norm.IN
240
(V)




240


Norm.OUT
44.23476
(V)




7.481042129

~DC










Norm.IN
240
(V)







Norm.OUT
63.58902
(V)






about*





Simulated



*REAL
RL(Ω)
u. RL(V)
p. RL(W)
i.RL(mA)





SCir.Max
0
0
0
115.654

3xVtg.Mpl
2W 9.6V 208mA @ 46.5sz

RL.P.Max
482.172
36.98703
2.837246
76.70921

normal
2.45W 2.88V 847mA @ 3.4sz


0%
1280%
120%
10%





and a figure

  

note :: for the TF.1 is shown L.1.Before some 29(H) and .After 45(H) -- there's another experiment listing the values for the original NiMH charger's TF as R.i = 1251(Ω) and L.i = 47.5(H) , R.ii.YR (yellow,red wire , center tapped , total) = 2.4(Ω) , 9.11(V) - & - R.ii.BB (blue,blue , total) = 54.3(Ω) , 28.8(V) - but there's no much other data - i guess the ~AC.peak! voltages were required at the time --- so the table has data for TF.1 from the more complete experiment . . .if we multiply the 29.3(H) with Pi/2 we get 46.1(H) -- so i guess there's a current measurement peculiarity involved -- as Z = U/I = U.AVG / I.AVG , I.AVG = I.MAX·(2/π) -- or -- the prev. more precise Z.correct = Z·(π/2) → L.i = 46.4(H) , L.ii.YR = [P.i=P.ii=U².x/Z.x] = 44.546(mH) -- !!! however it does not much affect the DC(direct current) test values that are compared here