PDA

View Full Version : LCD backlight overheats 7805



RYTECH
- 7th September 2006, 16:36
hi i have my lcd backlight plugged into the 5v rail via a 11ohm resistor... it seems to cause the 7805 regulator to over heat. Why is this.. what value resistor should i be using with my 16x1 lcd display backlight?...

keithdoxey
- 7th September 2006, 17:05
What does the LCD datasheet give as the forward voltage for the LED ?

How much current is flowing through the backlight with your 11 ohm resistor?

What else is being powered from the 5V rail.

What is the current rating of your 7805 ? 100mA 1A 2A

Does the 7805 have a heatsink an what is its input voltage ?

Has the smoke escaped yet because when it does it will stop working !!!

RYTECH
- 7th September 2006, 17:28
The lcd datasheet doesnt say much about anything.

i metered the lcd backlight, it seems to only pull .05 amps or 50 milliamps via the 11 ohm resistor at 4.9v from the 7805.

The only other things powered off the 7805 is the LCD itself, a pic 16F84a.

The current rating is 1amp on the 7805, no heatsink is attached.

Input voltage is between 12 and 18v. When the LCD backlight is unattached, the 7805 remains cool, the circuit and lcd function normally.

When the backlight is attached to +5 via the 11ohm resistor, the 7805 heats up quickly and i shut down the circuit as soon as it's too hot to touch.

no smoke yet... but i realized the 7805 was overheating when it smoked an electrolitic cap on teh 5v output. (the cap was almost touching the 7805)

lemme know what u think.

keithdoxey
- 7th September 2006, 18:05
The lcd datasheet doesnt say much about anything.

i metered the lcd backlight - 50 milliamps

The only other things powered off the 7805 is the LCD itself, a pic 16F84a.

The current rating is 1amp on the 7805, no heatsink is attached.

Input voltage is between 12 and 18v. When the LCD backlight is unattached, the 7805 remains cool, the circuit and lcd function normally.



I think you need to add a heatsink. Your regulator is dropping upto 13V with worst case supply voltage. At 50mA the LCD will be causing it to disipate an additional 0.65W with only the tab to get rid of the heat.

The 7805 datasheet I have just looked at says that maximum power dissipation is "Internally Limited" so it *shouldnt* self destruct but best not risk it :)

What is the TOTAL current you are drawing from the supply.

RYTECH
- 7th September 2006, 18:58
LCD datasheet

mister_e
- 7th September 2006, 20:55
just use a heat sink could help... but i would check what's happen at the 7805 input first with a scope. Is this really clean? is this really between the voltage you mentionned OR it have also up to 20 volts peak...

My best suggestion will be to change your PIC to have a PWM out and use it. Take the output, send it to a transistor. The LCD voltage will be taken at the input of the voltage regulator. This way, you will reduce heat.

a 16F628 is just perfect, cheaper and also give few free i/o if the internal OSC is enough. 1 more if you don't need the MCLR pin.

Melanie
- 7th September 2006, 22:27
I find that even with quite low current consumption the 7805 dissipates sufficient heat to require a heatsink and to that extent I avoid that part. On the other hand, the 78L05 is great. Small TO92 footprint, low heat dissipation... yes it's spec'd at 100mA, but I seldom draw more than 50mA from one, and I'd fit two... one for the +5v for your microprocessor and logic, and the other for the Backlight. Two or three 78L05's take up less real-estate as a single 7805 horizontal with a heatsink and in total cost less. Actually, most backlights can happilly run from an unregulated supply provided you suitably current limit them.

RYTECH
- 7th September 2006, 22:49
Thx for the reply's everybody.

I chose the 7805 part for my project because i had one on hand (many on hand), and also because the end goal for my little project involves around 12 IC's, and a backlight.

The project involves 1 pic 16F767 running@20mhz which aquires data from several 74LS151 8-1 data selector IC's.

The data is "line status" information coming over 2 wires and 1 common for each "phone line". The pic runs the address sequencing and aquires the bit information and then transmits it to PC software.

The inputs on the 74LS151 IC's are tied high and are "commoned" to ground. Active Low.

I decided to Add on an LCD connected 16F84a. It is to receive serial output from the 16F767 PIC and print certain information to the LCD display (totals for each of the 4 states of the phone line).

So lets say I move the lcd backlightpower to the source before my 7805 regulator which is powered by a 12v wall wart adapter. Do i need a voltage divider before feeding the power to the LCD backlight?

I think the ultimate solution would be to have a button on the 16F84a that turns on the backlight. So i imagine this would involve an interupt? But lets start off by powering the backlight..

keithdoxey
- 7th September 2006, 23:57
So lets say I move the lcd backlightpower to the source before my 7805 regulator which is powered by a 12v wall wart adapter. Do i need a voltage divider before feeding the power to the LCD backlight?


You just need a bigger resistor instead of the 11 ohm you currently have.

With your 11 ohm resistor you have 50mA flowing through the backlight.

11 * 0.05 = 0.55V dropped across the resistor so your LED backlight has a forward voltage of 4.45 volts.

For a 12v supply you need the resistor to drop 7.55v

7.55/0.05 = 151 so 150 ohm 0.5 watt resistor for 12V supply @ 50mA

For the 18v supply you need the resistor to drop 13.55v

13.55 / 0.05 = 271 so 270 ohm 1 watt resistor 18V supply @ 50mA

RYTECH
- 8th September 2006, 02:50
thx for the info

DynamoBen
- 14th September 2006, 23:43
Just for the record, I ran into the same situation recently. I end up tying the backlight to 12V. Works well and keeps the 7805 nice a cool.

mvs_sarma
- 18th November 2006, 08:54
Hi
i saw another application where 47E is used instead of 11E in series to backlit LEDs circuit
However perhaps it is better Rytech cold analyze the current breakup-- may be the current is draining thro another circuit path

mvs_sarma
- 13th January 2007, 11:16
You just need a bigger resistor instead of the 11 ohm you currently have.

With your 11 ohm resistor you have 50mA flowing through the backlight.

11 * 0.05 = 0.55V dropped across the resistor so your LED backlight has a forward voltage of 4.45 volts.

For a 12v supply you need the resistor to drop 7.55v

7.55/0.05 = 151 so 150 ohm 0.5 watt resistor for 12V supply @ 50mA

For the 18v supply you need the resistor to drop 13.55v

13.55 / 0.05 = 271 so 270 ohm 1 watt resistor 18V supply @ 50mA

having regard to RYTECH, i wonder whther 78L05 was used and if so proper termination adopted or Not?
afterall for 50mA with 12V input, the device need not be hot. let RYTECH measure the current at input 12V point and it will really speak.

Acetronics2
- 13th January 2007, 13:11
Hi, Sarma

I just have MEASURED 3v9 @ 50 mA on one of my LCD Backlight ...

so, ...

( 12v - 3.9v ) x .05 = 0.4 W to be dissipated ...


now, as you do, pluggeg in the 5v with 11 Ohms

( 5 - 3.9 ) /11 = 100 mA !!!

so, your 7805 is then overcharged and is close to it's maximun dissipated power before shutdown ( 130°C internal ? ) !!!

Try a 22 Ohm resistor and everything might be Ok !!!

Alain

Ohm's law will still make ink flow ...

mister_e
- 13th January 2007, 13:21
http://www.mister-e.org/Pics/ohmslawchart.gif

Acetronics2
- 13th January 2007, 13:39
really satanic !!!

mvs_sarma
- 15th January 2007, 11:09
Hi,

Acetronics is right in recommending 22ohms and i saw other designs (IK3OIL's frequency counter using LCD display) indicating backlit LCDs draw 60 mA and recommend 7805 but NOT 78L05. No heatsink is noticed.

Hi DynamoBen,

did you use 12V ,7805 and 11ohm resistor for Backlit? Still find No heat?

DynamoBen
- 15th January 2007, 15:51
I have since moved the backlight to my 12V source. That seems to work fine.

mvs_sarma
- 15th January 2007, 19:02
Hi

And would have increased the resistor value to say 47 ohms nearly.. it is a good idea and now even a 78L05 will do ( ofcourse in and out pins are diffrent.)
GOOD improvisation

Archangel
- 16th January 2007, 01:43
http://www.mister-e.org/Pics/ohmslawchart.gif

Thanks mister_e

Acetronics2
- 16th January 2007, 08:51
Hi

And would have increased the resistor value to say 47 ohms nearly.. it is a good idea and now even a 78L05 will do ( ofcourse in and out pins are diffrent.)
GOOD improvisation

Hi, Sarma

When using 78L05 ... just think to have very good decoupling caps.
the internal impedance of the L series sometimes may cause little surprises with PMW use or with outputs loaded ...

Think to the LM 317 L that has a lower output impedance !!! ... or keep the 7805.

Alain

RussMartin
- 12th July 2007, 23:01
I find that even with quite low current consumption the 7805 dissipates sufficient heat to require a heatsink and to that extent I avoid that part. On the other hand, the 78L05 is great. Small TO92 footprint, low heat dissipation... yes it's spec'd at 100mA, but I seldom draw more than 50mA from one, and I'd fit two... one for the +5v for your microprocessor and logic, and the other for the Backlight. Two or three 78L05's take up less real-estate as a single 7805 horizontal with a heatsink and in total cost less. Actually, most backlights can happily run from an unregulated supply provided you suitably current limit them.

This is such a great idea, and not just for this application!