Lm317

[ylioily]

i have using the LM317ic and not very sure the circuit design can run properly o not.
Here below is my circuit. The battery is design to overnight charge.
any1 can gives me suggestion and comment to my circuit whether is wrong o needed to improve.
Thx alot.

 

[qrper]

Hummm...

that's not going to work.

what are the fets doing anyway? i see no way to turn then on or off, since the gate of both are being turned on by the output of the two regulators. And while i did not look up the specs for the fets you're using, the 6 volt output is not enough to turn on the fet anyway (unless it's a logic level fet)

and the negative side of your battery pack is connected to the input of the third regulator. It's not going to be happy with a neg input voltage.

and what's the load resistor for?

mike

 

[ylioily]

Hummm...

that's not going to work.

what are the fets doing anyway? i see no way to turn then on or off, since the gate of both are being turned on by the output of the two regulators. And while i did not look up the specs for the fets you're using, the 6 volt output is not enough to turn on the fet anyway (unless it's a logic level fet)

and the negative side of your battery pack is connected to the input of the third regulator. It's not going to be happy with a neg input voltage.

and what's the load resistor for?

mike

The actual MOSFET spec i haven't decide yet, but the only things is Vg(M1)-Vs(M1)<=0V will turn off while the Vg(M2)-Vs(M2)>0V will turn on. The Vin is changing in between 0-20V. The output voltage regulator will be starting to drop when the Vin-Vout(U2)<3V (actually i have measured about around 2.3V below the Vout(U2) it starting to drop).

When the U2 turn off the output load (load resistor) will be directly supply by Vout(U2) through forward bias schottky diode of D3.

And about the negative side of battery pack, I had found out there will be current flow from Vout(U3) to Vin(U3) after u mention, thx. Any solution or method to prevent this occur?

And here i had modified the circuit a bit since i make sense the U1 is useless and can be eliminate.

Thx again

 

[qrper]

There's still trouble.

How are the gates being switched?

unless the solar panel is in the dark, it's output will always have some voltage to it as a solar panel is a constant voltage source.

the input to the second lm317 with the negative side to the battery is not going to fly either. the adjust line is at ground.

Grab a hunk of perf-board and build just the battery and u2 connected to each other, then apply voltage with your power supply to the + side of the battery. see what happens in the real world.

if you are only after a charge controller, use one lm317 set up for a constant voltage source. set the output of the lm317 to whatever the full charge voltage of the battery back is. Let's say your charging a small lead-acid battery. A 12 volt pack is fully charged at 14.3 volts (depends on the company) so set the output of the regulator to 14.7 volts and then insert a diode between the output and the battery to prevent discharge at night.

when the battery voltage reaches the voltage set by the lm317, charging current stops or drop to a very small amount to keep the battery fully charged.

 

[ylioily]

There's still trouble.

How are the gates being switched?

unless the solar panel is in the dark, it's output will always have some voltage to it as a solar panel is a constant voltage source.

the input to the second lm317 with the negative side to the battery is not going to fly either. the adjust line is at ground.

Grab a hunk of perf-board and build just the battery and u2 connected to each other, then apply voltage with your power supply to the + side of the battery. see what happens in the real world.

if you are only after a charge controller, use one lm317 set up for a constant voltage source. set the output of the lm317 to whatever the full charge voltage of the battery back is. Let's say your charging a small lead-acid battery. A 12 volt pack is fully charged at 14.3 volts (depends on the company) so set the output of the regulator to 14.7 volts and then insert a diode between the output and the battery to prevent discharge at night.

when the battery voltage reaches the voltage set by the lm317, charging current stops or drop to a very small amount to keep the battery fully charged.

i juz know solar panel is act as current source and not sense it is a constant voltage sources @@ll. i tot the voltage of solar panel will drop proportional linear with current. Thats y i made a mistake from beginning. Thx for ur shared info here...

If in case the solar panel can provide sufficient enough current to my output load and battery pack at daytime. The battery pack was to design taking charging of 0.1C-rate. The battery pack will provide the power to the load at nighttime when the Vinput starting drop below to voltage battery.

Below circuit was modified again, isn't any overall improvement can be add?
(The spec i haven't go through yet, so can be ignore 1st)
Thx again ur reply

 

[qrper]

why do you want to complicate such a simple process?

your battery will NOT charge when the negative side is connected to the + side of a regulator.

the regulator you have with the negative side connected to the regulator + input simply won't work!!!!

THere no way you can turn the fets on or off.

"Vinput starting drop below to voltage battery"

the only thing I see here is you're going to cook the fet as you have it in an not linear region instead of either fully off or fully on.

you need a clean sheet of paper.

Get rid of the fets unless you have some logic to turn them off and on as the battery requires. Depending on solar panel voltage to switch fets is not going to work.

There's no need for a load resistor either.

have you actually sat down and built any of this?

 

[ylioily]

why do you want to complicate such a simple process?

your battery will NOT charge when the negative side is connected to the + side of a regulator.

the regulator you have with the negative side connected to the regulator + input simply won't work!!!!

THere no way you can turn the fets on or off.

"Vinput starting drop below to voltage battery"

the only thing I see here is you're going to cook the fet as you have it in an not linear region instead of either fully off or fully on.

you need a clean sheet of paper.

Get rid of the fets unless you have some logic to turn them off and on as the battery requires. Depending on solar panel voltage to switch fets is not going to work.

There's no need for a load resistor either.

have you actually sat down and built any of this?

The -side of battery connected with U1 can be charging work by adjust the value of R1 to control the charging current which u want. Since the Vout(U1)-Vadj(U1)=1.25V at condition when Vin(U1)-Vout(U1)>3V. The path of current go in U1 will be equal to current go out U1. The Iadj(U1) is small and can be ignore.

i had build and test the circuit as below with input power supply and it can be work exactly almost same as the result of multisim.

Here below are the result of 3 condition the multisim show out.
The problem is the 0.5Amax solar panel tat i need to deal with, since it act as constant volt source.

 

[qrper]

Go for it!!

But I have a good supply of LM317 on hand

 

[qrper]

How's this for simple?

input from the solar panel is sent to the lm317. Its output is set to the final fully charged battery voltage with the pot. The diode between the battery and reg prevents the battery from discharging at night.

when the battery is fully charged, current flow is drop to a very low level.

if you panel has enough grunt, the led will light when the solar panel is producing power.

if the solar panel has more power than you need, or higher input voltage than your battery, you may need to heatsink the lm317.

the 1n5408 diode is over kill times three. a 1n4000 series will work too

it really simple, always works and is cheap and easy to build.

 

[ylioily]

what i trying to design the circuit is something like behave as Uninterupt power supply. my output will drive to MCU to coordinate the 125 LED. when the solar power is low, the battery should be provide power to my LED. if the solar power is high enough thn should be direct provide to my output and also store the power(by charging battery).

 

[qrper]

the circuit I posted will do exactly what you want it to do.

run your loads from the battery

if the solar panel is producing power, then the power required by the loads will be powered by the panel. The battery floats.

when the panel is not producing, then the battery supplies the power. there's no switching, no diodes no nothing.

this is known in the ups field as a floating power system. Large cell phone towers and most communications systems use this exact method, only on a larger scale.

 

[ylioily]

ok, thx a lot, i will try it...
any problem i will post and ask it here again

 

[audioguru]

How's this for simple?
Input from the solar panel is sent to the lm317. Its output is set to the final fully charged battery voltage with the pot.

Wrong.
The 240 ohm resistor is for the more expensive LM117. The LM317 uses a 120 ohm resistor then the value of the other resistor must be halved.

 

[qrper]

Well not according to this data sheet from national. as you can see the device shown as LM117, LM217 and the LM317.

 

[qrper]

lets try this again!

can't seem to get the attached file to come up.

 

[qrper]

third times the charm

 

[audioguru]

National never made an LM217. Just the expensive LM117 that can use a 240 ohm resistor and the LM317 that needs a 120 ohm resistor.
ST Micro copied the LM117 and made the LM217. They "no speeky zee English" (I think they are Italian) so their datasheet is very confusing.

 

[ylioily]

input from the solar panel is sent to the lm317. Its output is set to the final fully charged battery voltage with the pot. The diode between the battery and reg prevents the battery from discharging at night.

when the battery is fully charged, current flow is drop to a very low level.

i think in ur such condition mayb it is float charge considered. but it cannot charge the battery from low power to fully charge. Since the charging current uncontrollable and also sufficient high.

 

[qrper]

Well not to get into a pissing contest with someone over a single resistor.

However that being said, I reference:

Tormets electronic bench reference
Forest Minns electronic bench handbook
The arrl radio handbook 2005 pages 17.17
emergency power for radio communications page 8-5
The modern power supply and battery charger circuit encyclopedia page 6 figure 1-7

and the electronic projects ready-reference by john markus

I can supply others...

and ylioily...

The charging current will be determined by the state of charge of the battery. the more discharged, the more current will flow, up to the capacity of the solar panel. Of course the regulator will current limit at one amp, but you told me the solar panel is quite small.

when the battery voltage reaches the set point of the regulator, for example a 12 volt gel cell state of charge is 14.3 volts (reference Yausa GB7A 7 amp/hr) when the battery reaches the 14.3 volts set by the pot on the lm317, current will drop to a very low float value. of course, you need to take into account the voltage drop across the blocking diode and adjust according.

If you want I can supply you photographs of several commercial versions of the very same circuit. All are used as battery chargers in the form of a LM317. One is particular is the hickroy CH-12 charger. Other than some current limiting resistors on the input (the charger runs on ac) the circuit is identical to the one I posted.

bobier electronics US Patent "solar panel output circuit" uses this same idea.

In higher end outdoor solar lights (not the cheapies from wallmart) use a lm317 to control charging of the two ni-cad batteries. Reference USA patent: Siemens Solar industries " Solar cell packaging assembly for self contained light"

build the circuit, it's just a handful of parts. I don't put much faith in multisim. I downloaded the 30 day version and hell I can't even get a led to flash..

 

[ylioily]

If you want I can supply you photographs of several commercial versions of the very same circuit. All are used as battery chargers in the form of a LM317.



I don't put much faith in multisim. I downloaded the 30 day version and hell I can't even get a led to flash..

Hi,
I m appreciate and interest if u can supply me some schematic familiar of tis kind of circuit. I m try to learning more...

The LED can't light mayb u r not enough current to turn it on.Try to set the on value of LED.