78L05 troubles

LoopyByteloose · 2010-07-18 14:23

For years I've used the larger 1amp 7805 regulators, but I began to think they were a waste of space for many projects, so I bought some little 100ma 78L05s.

But I can't seem to make them work well.. At first I figured that I got a defective one, but replacement seemed not to solve the problem. I had used two 0.1mfd tantalum caps with them.

Finally due to voltage seeming to wander, I removed the tantalum caps and the readings have gotten more consistent. Any comments? At this point, I am thinking the tantalum caps were doing something odd.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ain't gadetry a wonderful thing?

aka G. Herzog [noparse][[/noparse] 黃鶴 ] in Taiwan

Mike Green · 2010-07-18 14:52

You should try a 0.1uF ceramic capacitor on the output of the regulator rather than the tantalum capacitors since its purpose is to limit high frequency noise on the output rather than acting as a filter capacitor for power demand surges. The input capacitor could be tantalum, but should be bigger if supplied since its purpose is to provide a local source of short-term power if the regulator is some distance from the power supply. I'd use a 1uF tantalum, but it is optional.

I tend to use a 0.1uF ceramic capacitor in parallel with a 1uF to 10uF tantalum capacitor on the output of these regulators since that's what I would use at the input supply terminals of a small board without a regulator on-board anyway.

Leon · 2010-07-18 15:12

Use the capacitors suggested in the data sheet (ceramic 330 nF input and 10 nF or more output) close to the pins, and you shouldn't have any problems. I use them a lot and have never had any problems with them.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

Phil Pilgrim (PhiPi) · 2010-07-18 16:03

Also, don't forget that the pinout (viewed with the legend side facing you and the pins pointing down) of a 78L05 (Output, Gnd, Input) is the reverse from that of the 7805 (Input, Gnd, Output).

-Phil

LoopyByteloose · 2010-07-19 08:08

Thanks, I suspect that I damaged the tantalum caps by reverse polarity at some point, but I can't be 100% sure as I set aside the project for several weeks due to frustration.

Having let the unpopulated board 'burn in' for 24 hours without caps, the voltage regulation now appears stable.

I still like the 78L05s, after all it seems the BasicStamp does fine with them in an SMD form.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ain't gadetry a wonderful thing?

aka G. Herzog [noparse][[/noparse] 黃鶴 ] in Taiwan

Leon · 2010-07-19 11:48

You do need them, especially the one on the output. The 78L05 might oscillate when you put a load on it. I've seen it happen with a 7812 that a friend wired up, the result was no output. I put my finger on the chip and it started working!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

LoopyByteloose · 2010-07-19 12:50

I did have it reading 0 volts for a while. The little 100ma devices do seem to be more sensitive to which size caps you use. I have rebuilt the board with ceramic ones on the output and an electrolytic on the input. This seems to be stable whereas before figures were wandering both high and low in weird way.

As always, it is best to power up your bare board with regulator before you populate. Solder bridges and defective parts occasionally happen. Right now I have it loaded with an LCD display and it has settled down to a perfect 5.00VDC. I suppose part of the problem is that there is no heat sink included.

Now I am ready to insert an SX18 without risk of damage.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ain't gadetry a wonderful thing?

aka G. Herzog [noparse][[/noparse] 黃鶴 ] in Taiwan

lardom · 2010-07-19 16:41

What does 'burn in' mean?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

Leon · 2010-07-19 21:32

Basically just powering it and leaving it for a few hours. Technically, it's a term often used for hi-rel devices and systems - they are powered up and run at at an elevated temperature for long periods. Units that are still working are then sold, and should last a lot longer than standard ones under harsh conditions. They cost a lot more, of course.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

LoopyByteloose · 2010-07-20 17:24

I build and test in stages as I have found that I need to catch mistakes before I commit to expensive parts.

And so, I have learned to make sure that I have stable +5 VDC (or whatever required) before I bother with fill IC sockets with chips. Quite often the first attempt gets bad results due to a solder bridge somewhere. Some voltage regulators - like the 2940 series - are quite sturdy and will tolerate momentary direct shorts and reverse polarity. The 78xx series do okay too.

I've given up on the 78L05 for this project as the voltage drops to 4.6VDC when I insert the SX18 and the 2X16 LCD. I switched to a regular 1 amp 7805. I presume that I am over my 100ma limit, though I don't quite see how ( LCD without back light is supposed to be 4ma, the SX18 maybe 50ma, and then 3 or 4 pull up and pull down resistors).

Anyway, when I really have abused what I am building, I let the device run without ICs for 24 hours to assure that it is stable. Electrolytic capacitors may suffer the most and take time to fail.

I also check all my IC connections for voltage, circuit continuity and leaks to adjacent pins. Sometimes I goof and +5 shows up where it isn't supposed to, other times a solder bridge will tie two pins together or I will have a pull up instead of a pull down.

I know this all sounds rather tedious, but it is even more tedious to try to figure out what has gone wrong after everything is assembled and something burns up.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ain't gadetry a wonderful thing?

aka G. Herzog [noparse][[/noparse] 黃鶴 ] in Taiwan

78L05 troubles

Comments