Some days ago I was asked by a visitor about the 5v power....
So here is just a simple drawing for anyone who want to build an external 5V power supply for LCD Screens.
The circuit converts any 9 to 16 V DC power to 5V(1A max) Voltage higher than 16 volts isn't recommended! may destroy the regulator!
Its very simple and you can use it to power your LCD screen for testing with a 9V battery!
You can also connect a 9V transformer with a bridge and a capacitor of 2200uF and make a complete power supply unit!
You can also add an indicator (e.g a LED), a switch to turn of the whole power supply and a fuse to avoid destroy your PSU by a short circuit!
WARNINGS You must be sure about the output voltage before tou connect any device to this circuit. I advice You to check the output to be 5V with a Voltmeter before you try to connect any LCD Screen. Using transformers from 110 V or 220 V AC may be dangerous! You must be sure about what are you doing! If the voltage regulator temp is too high you must add a heatsink
The circuit is working for sure!
Last edited by limbo on May 1st, 2008, 1:27 pm, edited 4 times in total.
Limbo, all of the caps arent really needed. They do clean up the power but I have been using just the regulator from a wall transformer for a while now with no problems.
Although I bet I will get corrected on this, I found it pretty useless to have an external power supply because
* You can't (most people anyway) use it without it connecting to a computer
* You have to connect together the two grounds together, and I know this will trip many up
* It requires another wall socket on the already full computer powerboard
* Although extreme, if you stuff up the wiring of this or your lcd you can seriously fry your computer
I honestly don't even use molex, as they are inside the computer, and I couldn't be bothered to drill holes in every computer that I plug my lcd into. Instead I have made a set of adapters for it. The socket coming off the lcd is a female molex connector that I stole from a splitter, so I can use it on molex. I then ripped apart a broken usb mouse, but one could use an extension cable, and wired the power from the cable to a male molex connector. Another bonus of usb is that it can cut off power to the bus if there is a significant malfunction. I then did the same thing with a 15 pin (game socket). So I now have a portable lcd screen that I can lug to anyones house just to show off
My next plan is to get a usb hub, usb->parallel emulator and a small usb key. I can then ram the whole thing into a box and have only one cable coming out, with smartie on the usb key. Just one question though, can smartie send data out of these emulated ports? (they appear as parallel ports under windows)
I found it pretty useless to have an external power supply because
* You can't (most people anyway) use it without it connecting to a computer
* You have to connect together the two grounds together, and I know this will trip many up
* It requires another wall socket on the already full computer powerboard
* Although extreme, if you stuff up the wiring of this or your lcd you can seriously fry your computer
My post was a "Simple 5V power supply".
There is NO need to connect "grounds together" as you said. Anyone with the basic knowledge in electronics can build it without risk to "fry the computer". "another wall socket"I have one free!
So I now have a portable lcd screen that I can lug to anyones house just to show off
Using this circuit with a 9V battery, anyone can have a portable power supply (in the size of the lcdscreen or smaller).
I admit that your idea (using Usb power) is bright.
But, I personally, can't suggest a "newbie" to strip out USB cables.
Finally: Useless or not this circuit may be helpful for someone.
Last edited by limbo on October 19th, 2005, 9:43 pm, edited 2 times in total.
sorry my post came out a bit annoying or agrivating, i was writing it while i was angry at someone else . its ovbiously not the best for my situation, but im sure that many others will find this wall socket a better option. Also, i would have had to connect the grounds together on my setup, as i did the wiring to my own instructions, and probably not correctly
It seems like anything over a few hundred milliamps really heats up these regulators. I am driving a 4x40 lcd backlight with a 7805 and it is getting pretty hot even with a mid-size heat sink on it.
You can get 1A out of these with the proper cooling.
You can get 1A out of these with the proper cooling.
I agree with it _X7JAY7X_
I use a power supply like this to power two lcd's one "HD44780" and one "MO" together, with proper heatsink attached to regulator!
It's pretty stable (and pretty hot too).
But it's working.... For Sure
Last edited by limbo on October 19th, 2005, 9:44 pm, edited 1 time in total.
I think the 7805 can run in excess of 200+ deg farenheit. Cant check the datasheet right now for the exact figure. Most have thermal shutdown. With a cheap heat sink mine runs at about 120 deg farenheit, which feels hot to the touch.
Damn, _X7JAY7X_ you are absolutely right
The "official" Operating Temperature Range is 0?C to +70?C (32?F to 158?F) according to component specs.
But in real world you can see higher temprature: up to 150?C (302?F) which is the Maximum Junction Temperature (of course the build in thermal protection decreases the output voltage about 20% in temps higher of 110?C (230?F).
A bit of information on this from an electrical engineer (I know this topic is old, but still important):
The caps are VERY important, especially the one on the output of the regulator. You "can" do without them sometimes, but the biggest reason they are there is to keep the regulator from oscillating internally. This results in very large ripple in the output voltage, and will contribute greatly to the heat output.
The maximum junction temperature is the temperature of the actual chip inside; due to the laws of heat transfer the inside of the chip will always be considerably hotter than the outside of the case (especially because plastic is such a good insulator). Even for simple projects requiring little power you should use a heatsink. If you plan on driving anything near or above the 1A limit, there are simple circuits you can use to increase the load handling to 10A or better (using a bypass transistor).
The biggest reason you will see heat being output is because this is a linear regulator, meaning that it drops voltage by acting as the "perfect resistor" which will absorb the extra energy. According to Ohm's Law, the drop will be P=VI, where V is the voltage drop and I is the current supplied to the circuit. The amount the temperature of the chip will increase above room temperature can be found from the Thermal Resistance. The LM7805 datasheet lists two; one for Junction-Air and one for Junction-Case. For a maximum operating temperature of 125C and an ambient temperature of 25C (typical), it can stand an increase of 100C. With no heatsink, we must use the Junction-Air resistance, which is 65C/W. 100/65 gives us 1.54W of power which can be absorbed. Given a 9V supply, we have a 9-5=4V drop. Sinve P=VI, I=P/V, so the maximum current is 1.54W/4V=308mA. As you can see, even just powering your LED backlight (mine on my 2x40 was rated 400mA) will push this to its limit quickly. However, with the heatsink (5C/W to the edge of the case, 24C/W for a small heatsink) we can handle 100/29=3.45W, which is 3.45W/4V=863mA; not even the full 1A it's rated for.
However, this is all assuming we use a 9V supply. It can actually handle over 25V input, but obviously the heat will become considerably higher. Even just at 12V, pulling 500mA, our heat output will be 7V*0.5A = 3.5W, which is the theoretical limit of our regulator. You can see how quickly this causes problems.
One good solution is to use a low dropout (LDO) regulator. These don't handle as high of a current, but because they have a smaller voltage drop (usually as little as 0.5V), they can dissipate far less heat. Most manufacturers have the 78R05, which is the same regulator but with a much lower voltage drop. For example, you could use a 6V input instead of 9V, which would only be a 1V drop. Then we could use the full 1A easily with a heatsink (3.5W/1V=3.5A, considering only overheating), and even without a heatsink, we would still be okay (1.54W/1V=1.54A). These numbers are all approximations, ignoring things like the quiescent current of the regulator (which it draws even without a load current), but it shows the most important part of using a regulator in a power supply.
Now before you go out and grab 4 AA batteries and expect to make a portable 5V supply, let me tell you something else. A brand new AA battery doesn't have a voltage of 1.5V, but rather somewhere between 1.2V (for NiCads) and 1.4V (good alkalines, like Duracell Ultra or Energizer Max). They also usually drop to about 0.9V by the time they die. This is nearly a linear drop for alkalines, but stays much closer to 1.2V throughout almost the entire lifetime for NiCd and NiMH. 4x1.2V=4.8V, which obviously isn't enough to make a 5V supply, and even 4 brand new alkalines will just barely beat the dropout of the regulator (4x1.4V=5.6V, which will drop below usable long before the batteries are actually dead).
So all in all, either use a solid 6V supply (perhaps a 120V:6.3V transformer into a diode bridge, which will be about 5.6V) into a LDO regulator (in which case you can avoid the heatsink, but might as well use one anyway), or use a 9V supply with a regular regulator and a very good heatsink (not those little fins you attach to the tab of the regulator, but a large brick with lots of fins). If you want to use batteries, be prepared to use 5 or 6 with a LDO regulator, or else you'll have annoyingly short lifetime.
The best solution would actually be to make a switching power supply. One can buy a small inductor, capacitor, and boost regulator which will take 3V (down to even 1V as batteries die) and successfully boost it up to 5V at a decent amount of current (250mA at least). These are VERY efficient; much more so than the typical 7805 linear regulator, because rather than burn off the extra voltage, they instead turn the power on and off very quickly, such that the average voltage (as filtered by the capacitor) becomes a solid 5V. They are from 90-98% efficient, which is much more than even an LDO regulator can provide.
This concludes my "power supply" tutorial. Come back next week when I explain how to make pigs fly using only super glue and chicken feathers! B)
Just a note to anyone else:
I have used a non standard regulator for mine, it generates a _lot_ of heat. Most people don't realise how great heat sink compound is, it makes the heat transfer to my heatsink much much better. If you go to an electronics shop they may recommend that you use it. That said, you must be extremely careful with it. It contains heavy metal oxides (very poisonous) ie don't get it on you. I sell computers for a living, and let me tell you, that stuff is not nice.
A switching (aka switch mode, SMPSU) power supply is a bit more difficult for the hobby constructor than the simple 7805 linear supply shown above. Unless you need lots of current, it's probably not worth the extra expense and effort.
It's best to avoid making a completely switch-mode supply (the sort that rectifies and chops the mains voltage, like a PC PSU does) as that's a lot more complex. The best bet for that sort is to simply buy a wall-plug switching supply that can provide 5 volts.
There are several single-chip circuits for a halfway-house supply, which uses a normal mains transformer (or accepts 9-12V DC) and has a switching regulator. They are the switch-mode equivalent of the 7805 family of regulator chips. The LM2574/LM2575 series is possibly the most common (fixed voltage and 0.5 or 1 amp output). All you need are a few capacitors, a schottky diode and an inductor, plus the transformer and rectifier (or a wall-wart plug-pack unregulated power supply).
As we know, these regulators 7805 etc dont accept when output voltage is bigger than input. Caps in this case prevent these peaks (especially when you turn off the power).
Actually, i wouldnt be too quick to dismiss the idea of using a usb port to power an LCD screen. Since the introduction of the USB battery charging protocol, its become pretty safe to use a USB port to power a device, without the need of anu digital chips or resistors or anything. A USB cable has 4 wires in it, and a lot of insulation. The easist way to get such a system working would be to buy a usb extension cable. Cut it in half, and strip the male end (the one that plugs into the PC) revealing what should be a white, a green, a black, and a red cable. The white and the green cables are the data lies, called D- and D+ respectively. Because you dont want to transfer any data, simply ignore these wires. Its probably best to leave those wires insulates as those have no part here and you dont want them shorting. The red wire, is logically positive, while the black wire is negative, and these can be connected to the screen.
This sort of system has several advantages, such as the fact that USB is by standard supposed to supple 5V (5%) without any data thruput, and where there was previously a problem before the spec was introduced in current not being supplied to non-standard devices, however the standard (althouth technically still referring to those devices that have digital circuitry), stpulates current be granted to non-communicating devices. Although this is not technically what the standard is set down for, many device manufacturer's have used this loop hole to create devices that do not require complicated circuitry. Additionally, given the limits imposed on USB, shorting the pins should not (although cheaper or older motherboards may not provide such functionality) damage any other devices, or the PC. Despite this, i still dont recommend its tried. I implemented this system with my HD44780 and parallel port configuration without any issues, and its running quite nicely, even powering the backlight. Note however, that the screen should not draw more than 500mA, unless you do include digital circuitry to inform the hub that you will be so it can supply it to the device, as per the spec.
You must be sure about the output voltage before tou connect any device to this circuit. I advice You to check the output to be 5V with a Voltmeter before you try to connect any LCD Screen.
. its ovbiously not the best for my situation, but im sure that many others will find this wall socket a better option. Also, i would have had to connect the grounds together on my setup, as i did the wiring to my own instructions, and probably not correctly 
