Simple Power Supply Filter?

GICU812

I have a LCD touch screen for my Stamp from PDA inc (symmetry.com) Its cool and all, but it doesnt work right

The touch keeps activating on one particular spot every second or so. I figured out that if I connect it to a different power supply it stops, so I think I've narrowed it down to dirty power.

My original power supply was a 12v adapter running into a 5v regulator. Thats it.

I replaced it with a 5v adapter, it solved the problem, but I blew up the screen after a minute. I was afraid that might happen, and they even warned about it too, something about harmonics, I dont remember its been a while.

Anyway, I've got another screen, but still the same problem with the 12 ---> 5v setup. Seems to me I remember being able to build a simple filter out of a cap and a resistor. Can someone suggest a design?

thanks!



Also, any idea what could have blown on the old one, maybe a simple component? Probably cant answer that without the schematics, they're on the website above if someone is feeling particularly bored and helpful.

Adam

Mike Green

It's impossible to tell what's going on without more information like schematics of what you actually did.

As a general rule, regulators require some kind of filter capacitor on their outputs. Usually a 47 or 100uF electrolytic with a voltage rating higher than the regulator voltage (like 6V for a 5V regulator). They usually need a small capacitor across their input (like a 0.1uF ceramic 50V rating).

GICU812

Thats what I thought, and tried, though maybe not both at once.

So here's what I'll try tonight.



As for the schematic,

http://www.simmetry.com/schcs1.asp

All I did was hook a 5V DC transformer to pins 1&2 which are 5v in and ground. There are a couple of diffrent power input pins on the LCD, Id have to check later to see which connector I use, its either J4 or J1, but they both go the same place.

RDL2004

Well, if you used a typical 5 volt AC adapter, that's why it fried the display. Those are usually not regulated and put out considerably more than their specified voltage when not under full load.

Check this article for some good info on AC adapters:

www.glitchbuster.com/wallwart.htm

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Rick

Skeil

In the circuit you have the panel is only getting 5V 30hz DC the only thing keeping it alive is the capacitors currently

Wheres your full wave bridge rectifier?
use the 12 volt brick
full wave bridge rectifier
ceramic capacitor
voltage regulator
eletrolytic capacitor
then your display
I'd use a 1000 to 4700uf for nice clean power

GICU812

Well I tried several different capacitors, nothing helped. I checked the signal with my oscilloscope, and its good and level with only the slightest hint of noise, which could just be the tolerance on the scope.

I just happened to have another regulator on the testing board, so put it in as well, wired the display to it and everything was great. But when I hook the stamp to this regulator as well the problem returns. I had already tried taking the stamp out of the circuit, so my best guess is that the regulators just barely put out enough power to run the display, and the original regulator just didnt put out as much as this one.

This had occurred to me a couple of times, but I dismissed it thinking it was just a little LCD display. Oh well. I'll just have to get a bigger regulator.

Thanks though!

RDL2004

I have one of those displays I bought for a project last year, but ended up not using it. I'm not certain, but during testing I think I just ran it off the Board of Education power supply. I don't remember having any problems. You can use separate power supplies, but the display needs 5 volts regulated, and you need a common ground between it and the Stamp.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Rick

Mike Green

You know, you seem to be an adventurous type, the sort who likes to hook things up without referring to references or datasheets and the like. You might as well enjoy it. It can work sometimes. On the other hand, you might blow up things (literally sometimes). Do wear safety glasses when you work on stuff ... and ... it can get expensive sometimes ... ah well!

There are plenty of articles and tutorials available on power supply construction ... what to do and what not to do, especially when trying to "feed" digital electronics. They're available with a Google search. There's good information on the Wikipedia and Circuit Cellar has run a series of articles in the past on power supply design. The ARRL (Amateur Radio Relay League) publishes a Handbook with lots of basic electronics information. There's no shortage of information.

Specifically in the case you've described here, your power supply setup is dicey. Some regulators might survive the AC input. Some will not ... depends on the internal design of the regulator. That doesn't mean they'll work properly. Neither a Stamp nor most LCD displays draw a lot of current. The backlight on the LCD would draw the most current of anything. Most regulators (in a TO220 package ... the one with the metal tab) will supply on the order of 1/2 to 1A without shutting down although that depends on how hot they get (any voltage beyond their output voltage gets dissipated as heat). A 12V transformer will put out peak voltages well beyond 12V (that's the average output voltage) and that all has to be gotten rid of.

Anyway ... good luck.

GICU812

The regulator is getting 12VDC from the wallwort transformer, the reason I had not looked into it further was I had assumed the wallwort would be doing a decent job of filtering the power, and the regulator was just to get the voltage to the right level. From what I saw last night, its apparent I was correct.

All in all I didnt consider it an adventurous setup, I dont understand why im getting all this flack.

As for the regulator, I dont recall off hand, but I did check the outputs and the draw of the display, and it was within limits. Its obvious the one regulator is not putting out as much as the second so I might assume its bad or at least out of spec, the second should power the display and the stamp handily, IIRC, so it might also be that the display draws more power than claimed.


RDL2004: I dont have a BOE, just a stamp. The BOE might be stronger than my regulator, these are just little ones I picked up so I could feed the stamp off of 6, 9, or 12V. Are you going to use that display for anything? I might need another one. I've been playing with the idea of an active HVAC control system, and I would need a display for the thermostat.

Skeil

Wheres your full wave bridge rectifier
without it the meter is only getting 50% duty cycle DC

Or switch to a DC power supply instead of AC

Mike Green

Sorry for the flak, but the diagram you drew of what you were going to try last night showed a 12VAC transformer feeding a regulator of some sort which would have behaved as I described (maybe even explode if the transformer could supply enough current) and you seemed very cavalier about the whole thing, particularly about what kind of regulator you were using and what kind of parts it might require around it. All regulators are not equal. Some will malfunction without the proper capacitors attached to their input and output circuits (and possibly destroy or damage what digital logic is attached to it).

The BOE has a regulator on the board that will supply up to about 1A with an input of up to about 9V. It can withstand an input of up to about 30V, but is not designed to supply anywhere near its specified current output with that kind of input voltage. It has a small heatsink and will shut itself down from the overheating if it has to dissipate too much power (30V - 5V x 1A = 25W, even 12V - 5V x 1A = 7W is too much).

The Stamp itself has its own regulator (if you use the Vin terminal for its power input) that is rated for only 50mA output which includes the Stamp itself and any I/O pin current drain.

RDL2004

Hey, no flack here. You said all along you were using DC not AC. What you did originally seems like it would work. Are you using those little TO-92 size regulators or the bigger TO-220? The little ones are barely good for 100 MA, maybe that's your problem? I wouldn't expect a wallwart to be well filtered, I'd put a 470-1000 uf (use 25 volt rating here) capacitor in front of the regulator, in addition to a 0.1 uf close to the in and out pins, and probably a 100 uf at the output. That should cover any filtering you need. Like this:
V in > 470 > 0.1 > regulator > 0.1 > 100 > V out
·
I'm going to keep the display for now, I have another project in mind, but Sparkfun has something similar you might check into. http://www.sparkfun.com/commerce/product_info.php?products_id=8358





▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Rick

GICU812

Sorry guys, chock that up to the limited selection of graphics in Visio, and a little miscommunication that there was a wallwort and not just a transformer sitting there. I wasnt too concerned with the regulator, thinking all I needed to do was take a clean 12VDC to 5VDC.

I might try to see what I can do with my broken display, Im hoping its something simple, but at the same time I was prepared to part with it.

I'll keep the capacitors in there, but I saw no noticeable noise on the scope, nor any change with the cap in the circuit. But you guys are right, they couldnt hurt.

IIRC the regulator might be an 8506

GICU812

Okay, its a 7805, rated for 5V @ 1A

This display claims to draw 55 ma @ 5v.

But when I connect the display and the stamp, the backlight dims a little, and the touch screen malfunctions.

Id say the stamp was causing the problem, but on the original regulator with just the display, I still have the same problem.

Mike Green

Part of the problem here is we still don't know what kind of display you have, don't know how you've connected it, don't know how you've connected the Stamp, etc. Clearly, if the thing doesn't work when you connect both the Stamp and the display, something's broken.

Stamps are pretty robust and tend to survive a variety of abuse, but they do break eventually. How about testing the Stamp by itself? Hook it up to a PC as described in the Stamp Basic Manual and run it off a 9V battery (using the Vin power connection so the built-in regulator is being used). You should be able to use the Stamp Editor and write little programs that run and display stuff in the DEBUG window in the Stamp Editor.

RDL2004

If I remember right, that display has a demo mode where you can play with the touchscreen and such. Does that work correctly without the Stamp attached? If so, maybe it is a Stamp problem.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Rick

LilDi

The fact that you said the display dims a little when the Stamp is powered up, tells me that the Malmart transformer is not powerful enough. The 7805 is rated at 1 amp max., but the transformer may not be able to supply enough amps to drive both devices.

No flack intended, but I would never use a MalMart power adaptor on anything of mine that cost more than $20. I would include Radio Shack as well. Most Radio Shack products are from the reject bins of other companies. I'll bet MalMart is the same.

GICU812

When malfunctioning it repeatedly "hits" a point on the screen, 104,63. if you touch the screen, it will report that, but shortly after report the 104,63 again. It starts this immediately after power on, in, and doesnt stop for anything unless you remove the connector to the touch screen.

Just clarification, its not a "Walmart" power adapter, rather a WallWart, a generic name for the power adapters that have the transformer built into the plug. I dont know the brand or output off hand, but seems to me, both stamp and display would draw about 100 ma, at 5 v thats .5 watts. Even if the adapter was 600 ma at 12 v, that would still be 6 watts, BUT I will definitely check when I get home, as you're right it certainly does imply a power shortage.

metron9

Why don't you connect a current meter between the wall wart and the input of the regulator to see how much power is being drawn? That is something I do typically on every breadboard circuit. I also make use of the little PTC fuses so I don't blow something up if a wire gets shorted especially when the cats walk around on the circuits and play with the wires.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think Inside the box first and if that doesn't work..
Re-arrange what's inside the box then...
Think outside the BOX!

GICU812

Well doesnt matter now. I wasnt paying attention and connected the +5 feed to the wrong slot in the breaad board, the one that was feeding 12V, it was only a second or so, but long enough.

The LCD still lights and the splash screen shows, but it is unresponsive to touch or serial.

GRRR...