I need some advice

[BabyOxide]

Okie, heres the main story,

my cable modem occasionally stops responding and requires a power off- power-on reset, no other ways, (does not happens often, but it does)

I'm heading to aussy for study 3 yrs, but i intend to keep my home system, online for this period, so i can access it and webcam to my folks at home who are computer idiots..,

but this requires my net connection to be stable
so i came up with this idea to reset my modem

i bought a digital timer switch - programable
it has a back up battery (but since this entire unit is kinda of "MADE in CHINA"), i doubt the button battery is even rechargeble, thus it requires replacing say every 100 hours..( this is only if the battery is being utilised)

the guy sold it says thats roughly a replacement every 1 yr or so..
but i'm not about to find out the hard way,

the battery is 1.2v button battery
i bought a variable output transformer from 1.5 v to 12 v
since its another stupid china brand, i measured and found out, that
it does not provide 1.5 instead it gives from 3.5v to 17 v (WTH ?!)
anyway, i switched it to the lowest possible, its still 3.5v

so i created and design a simple circuit which now i may need some help
a simple diagram first for those who understand it,

VCC 3.5 --- R1 --- R2 --- GND
since i rumaged thru what ever resistors i had left from other projects
i manage to find,
4x 1k ohm resistors
2x 2 k ohm resisters

connected them pareller in this order

R1 = 3x 1k ohm in parallel, this gives me about 333.33 ohms
R2 = 2x 2k ohm and 1x 1k ohm in parallel, this gives me about 500 to 600 ohms

i tested the voltage accross this circuit
@ R1 i get 1.3 Vs (just nice)
@ R2 i get 2.0 Vs (useless at the moment but may come in handy to charge my batteries)

now after i designed tested this circuit, it seems to work fine and perfectly
until i suddenly went back to square one. WHAT if my power fails ?
my computers are automated and i have a modem attached i can boot it up with a phone call any time, but the modem.. its stucked well

case1, power fails, modem reboots, system reboots (perfect)
case2, power fails, modem reboots, system reboots, timer switch reset (since power was based on a transformer, if power failed, transformer loses power, the programmed settings on the timer switch would be lost (bad stuff) thus if power doesn't fail very often, i'm stucked with a dead modem

so i added a new circuit to this

VCC - R1 - R2 - GND
Between R1 terminals
1 set connects to the timer switch to power it so i dun have to replace the battery,
the other set connects to a normal ni-mh rechargerable battery, lol by sony if you would..

now the question is will this circuit works ?

will 1.3 v shared between the timer switch, also be sufficeient to recharge this battery ?

2ndly, in the event of a power fail, in that configuration, won't the battery be drained faster since it now powers, both R1 circuit (simple power dissipation) and the timer switch itself....

or should i switch to the other side of the circuit and have the R2 parallel the batt and the timer switch, but then this would also up the power dissapation since there are higher resistor values ?

lastly, even tho in this circuit i'm supposed to use lower resistor values as this restricts quite a number of current, thus weaken the entire cicuit and slowed down the normal recharge time of the battery, (but ths produces less heat (safety 1 i guess), and will such constant recharging, casue the battery to leak ? or be damage or worse expldoes....

thanks in advance

 

[audioguru]

Your resistors voltage divider has an output voltage of only 1.3V. All of its current is in the resistors with none available to charge a Ni-MH cell that has a fully charged voltage of 1.4V.
Without having a diode, the rechargable cell will quickly discharge into the lower resistor of the voltage divider when the power fails.

 

[BabyOxide]

what kind of diode should i use and how should it be placed ?

 

[Sceadwian]

You're better off using an electro mechanical appliance timer, you should be able to get them for less than 20 dollars and it's going to be a whole lot more reliable than any hack you can come up with.

 

[BabyOxide]

if i remove the R1 resistors and short in the battery instead, would this be advisable ?, since i tried for a brief time the voltage would stablise at 1.3 to 1.4volts, but this means the batt is constantly charging, would prolong charging in this state casue damage to this battery ?

 

[BabyOxide]

You're better off using an electro mechanical appliance timer, you should be able to get them for less than 20 dollars and it's going to be a whole lot more reliable than any hack you can come up with.

you mean the non digital types? i did thought of that,. but, then my modem would be shutting down every day..and for atleast 30 mins give or take 10mins, the one i had only had about 48 legs..

 

[audioguru]

The datasheet for the battery shows its recommended charging current and for how long. Energizer recommends using an extremely low 1/40C trickle charge current and Panasonic recommends 1/20C to 1/30C.

They say that overcharging a Ni-MH battery shortens their life.

 

[BabyOxide]

so thats like in the best scenario , 0.025 current over a max of 20 hours or "heck care about the battery life" (like those in cordless) it still last for more then 2 yrs under constant charging..

if i reduce the resistance of my circuit to R1 = 100 and R2 = 200
i'd get abt 1/80 C would this be good enough ? even lower then 1/40 but will this be enuff to charge up ?

and one noob question tho ?

circuit 1 > batt Vcc - 100 ohms - gnd
circuit 2 > batt Vcc - 200 ohms - gnd

which battery above discharge faster ?
my guess is circuit 1, am i right?

the diodes you were referring to are those the zener type diodes ? or the one way type ? i know i cant use one way type diode cos this stupid transformer, (or maybe i'm wrong), but i do believe my eyes and the voltmeter, its showes 23 Volts potential when i added a 1 way diode into the circuit connection

 

[BabyOxide]

btw where can i find this datasheet ?
thanks

 

[BabyOxide]

okie i found the lower resistors now,
i have R1 on 50 ohmz and R2 aprosimately 100 ohmz,, will these work better now ?

 

[audioguru]

The 3.3V from the power supply might drop when it is loaded with the resistor voltage divider. Let us assume that its voltage stays at 3.3V:
1) With a 50 ohm resistor in series with a 100 ohm resistor then the voltage across the 100 ohm resistor is only 1.1V. Any current for the timer circuit will reduce it.
2) A Ni-MH cell needs about 1.4V to fully charge. So the voltage divider doesn't have a high enough output voltage.
3) IF the battery is externally fully charged then the 50 ohm resistor across it will cause a 25mA drain on it during a power failure.

Energizer has datasheets for all their batteries at www.energizer.com . Click on Technical Info at the top of the first page, then select the battery type on the second page.

Do it like this:

 

[BabyOxide]

sweet, i think i understand abit more now ,
my 1.2 altho is enuff to run the timer won't be enugh to chareg the battery.
, the last i checked after i did my version of the circuit i had 1.3 across the 50 ohms , i paralleled the timer and the batt on it, if my calculations were correct i had about 2mA extra for charging the battery, at 1.3v.. ( i think the 5% tolerance band means 10 % lol was a huge differnce in voltage compare to the calculated one..

i did thought of that 1 way diode and even tried it, and it excatly as you drew , only that mine was the other way round for the vcc and gnd,
but with the diode in such manner, i read a +20 v accross my volt meter .....
puzzles me too .. no matter how i punch in the diodes, if IIRC,

VCC R1
(DIODE)
R2
Gnd

diode placed >| , i get + 20 v respect to gnd.....
diode placed <|, i get -16 v respect to gnd....
SOME transformer eh ....

 

[BabyOxide]

wahhahahaa i dun believe it i went to try again this time i used crocodiles clips, its now displaying correctly, lol then i stick my finger on the contact it jumps up again.... SOME conductor i am . wth.......

must i use the said diode or any other diode i can find ?

just pondering, the diode has a v-drop of 0.7
can i like series 3 of them and be left with +-1.4 v ? wihtout using resistors

in your illusatration, in case power fails my battery will not be dischargin to the resistors right

 

[audioguru]

The voltage might jump if your wires are very long or you touch a wire and the wires or you act like an antenna to pick up mains hum or a local radio station.

Any diode will work. A big diode will have a little less voltage across it, it might be 0.6V. The diode stops the battery from discharging into the resistors when the power fails.

If you connect three diodes in series with the 3.3V supply then their output will be 1.2V without any current limiting. A dead battery will begin charging with a very high current that might break it, but the voltage is too low so it will never fully charge.

 

[BabyOxide]

okie got that, time for bed i shall re do the circuit for the 4th time lol,,, thanks alot i see how well it goes tmr..

 

[BabyOxide]

minor problem, which confuses me..
V-drop across diode is 0.4 to 0.5

my circuit is as so
3.125V - 99.7 ohm - 56 ohm - GND
-------- 2 Vs-------1.13v----
i had the diode placed at vcc
gnd* here is the mid point between the two resistors
3.125V - >| - LOAD - GND* ( with battery )
--------------1.283v----------------------
3.125V - >| - LOAD - GND* ( without battery )
--------------1.763v----------------------
battery measures 1.250v

previously when i had more then 1.2 v across the timer switch terminals the timer lcd display grew darker to the point it looks as thought it resets, everything was displayed as in the time were all 88:88, but in my version of the circuit it was lighter and look as tho it was going to die

but now it has 1.7 v across (without battery) and it looks fine and works normally.. wth ???
with battery it works fine too at 1.2 v and when i simulate a power fail, it continues to be backed by the battery

another wierd thing is , my fluke can't mesaure the current without using the min max button which gives bout 4 mA without battery connected, it goes 4mA for max and -4 for min... confuses me..

with the battery it goes from min -4 to max 12

does that mean the battery is charing at 12 mA ?

if you could explain the 1.7 volts thing ?? im really confused lol

 

[BabyOxide]

ah krap 1 hr after i soldered everything the clock went dead
i redid the circuit and follows excatly as you drawn
when i put the battery on and waited a few mins, the clock slowly fades
when i reduce the 56 ohm resistors, the clock didn't come back up ( i did this by parraleling the 56 ohm resistor with a 22 ohm resistor) (if i'm not wrong this will increase the current flow to the clock), when i remove the battery the clock darkens so much, when i remove the parralleling resistors, they clock went back to normal ( at this pt the batt is not connected)

my guess is either the batt is crazy or the diode is crazy?

wat gives?

 

[BabyOxide]

with the added resistor the output A went from 3mA to 28 mA ( chanegd the battery on my fluke if my earlier recrdings were somewhat puzzling)

the battery seems to suck all the power

 

[BabyOxide]

just incase the batterry might be the hge problem, its ni-mh 2100mAh, hr15/51 1.2v hr6 migon AA

 

[audioguru]

The timer's display needs a certain voltage (maybe 1.5V?) for proper contrast. With a different voltage it will be too dark or too light. Your circuit has nothing to regulate its voltage so it is too high when the battery is fully charged and is too low the rest of the time.