This is my first post on this forum. I have had several successful circuits with other regulator types, but with the ZLDO1117 I had it fail letting 8 volts through to a picaxe 20x2. This naturally caused the 20x2 to fail. It was far from overloaded, rated for 1 amp and only supplying power for the 20x2 and an ACS714 current sensor. (circuit in question is a battery charger) Im wondering if anyone has used this regulator type and had this problem, and if using capacitors would prevent the problem.
ZLDO1117 Regulator problems
- Thread starter sniper887
- Start date
How did it "fail" and how did you measure the 8 volts (i.e. with what?)
So you didn't have any capacitors in your circuit at all?...despite them being shown in the http://www.diodes.com/datasheets/ZLDO1117.pdf datasheet "typical circuit" and being mentioned as being required in the "Stability" section?
I've not used that particular regulator before but plenty others and be assured there will be no generic problem with it - but it's not uncommon for un-capacitor'ed regulators to act as high frequency high voltage oscillators without the recommended capacitors in place...
Try again with some in the circuit and report back.
PS Welcome to the Forum
So you didn't have any capacitors in your circuit at all?...despite them being shown in the http://www.diodes.com/datasheets/ZLDO1117.pdf datasheet "typical circuit" and being mentioned as being required in the "Stability" section?
I've not used that particular regulator before but plenty others and be assured there will be no generic problem with it - but it's not uncommon for un-capacitor'ed regulators to act as high frequency high voltage oscillators without the recommended capacitors in place...
Try again with some in the circuit and report back.
PS Welcome to the Forum
Thanks for the reply and welcome. The failure was specifically it passing too high a voltage, although your mentioning the capacitors makes me think the reg is fine. I've ran other regs without capacitors but I'm thinking this one is particularly susceptible to not having them. The voltage was measured with a multimeter on the output lead (tab) on the reg. I also noticed the 20x2 getting warm. I will try it with caps next time and see how it goes. I have been playing with SMD parts so I'll have to find assortments of caps. Do you happen to know good places to get specifically assortments of SMD parts?
SMD assortments
Sniper 887.
I have had good luck finding SMD assortments at http://www.goldmine-elec.com/
alternatively you can check http://www.allelectronics.com
Regarding bypass and filter caps in regulator circuits, they are mandatory.
As are current limit resistors for active devices. I like to use the LEDS that are resistorized, less messy. Get those at http://www.jrhackett.net/LEDs.shtml
of course a google search will turn up all sorts of vendors.
PT
Sniper 887.
I have had good luck finding SMD assortments at http://www.goldmine-elec.com/
alternatively you can check http://www.allelectronics.com
Regarding bypass and filter caps in regulator circuits, they are mandatory.
As are current limit resistors for active devices. I like to use the LEDS that are resistorized, less messy. Get those at http://www.jrhackett.net/LEDs.shtml
of course a google search will turn up all sorts of vendors.
PT
lbenson
Senior Member
>Do you happen to know good places to get specifically assortments of SMD parts?
I don't know if you mean an assortment as in grab-bag, but www.digikey.com is likely to have what you want in specific values--for instance, 1206 resistors at $1.40 for 100, 0805 resistors at $.88 for 100, 1uF 1206 ceramic caps at $.186 each, quantity 20. I happen to know this because I placed my order today.
I don't know if you mean an assortment as in grab-bag, but www.digikey.com is likely to have what you want in specific values--for instance, 1206 resistors at $1.40 for 100, 0805 resistors at $.88 for 100, 1uF 1206 ceramic caps at $.186 each, quantity 20. I happen to know this because I placed my order today.
Grab bag is one idea, and I have actually done that from electronics goldmine. It would be better to have an assortment with a few each of common values, like you can get at Radio Shack for resistors at least (there isn't one locally where I live though) I do like both Digikey and Mouser. They have tons of stuff there, both through hole and SMD.
inglewoodpete
Senior Member
What is the source of your power? I had a 230v AC to 12v DC (100w) slimline swtchmode supply do damage to my PICAXEs, after damaging a regulator a few months ago. The 7805 regulator developed some sort of (thermal?) runaway. The output voltage would creep up from 5 to 8 volts. The PICAXEs started to run funny before their outputs failed when the supply was between 7 and 8v. Not surprising, of course.
I replaced the 7805 with a LM2940CT-5v and have not had any further problems. However, I have no way of confirming what the actual cause was or if the 2490 is really the solution.
I replaced the 7805 with a LM2940CT-5v and have not had any further problems. However, I have no way of confirming what the actual cause was or if the 2490 is really the solution.
The source of power I used when encountering the problem was an extra ATX power supply I had laying around. All computer power supplies are switch mode much like yours. I had the circuit hooked up to 12V on it. The reason I didn't hook this particular project to a 5V lead on the power supply is the circuit, a battery charger, needs 12V controlled with a MOSFET to charge a battery, using PWM. A previous try at this project wouldn't program until I put a 220 uF capacitor across the input power leads to filter the power coming from the PSU. That might have had something to do with your problem. Once I get around to making a new board, I'll use capacitors on both sides of the regular as recommended by others who posted on this thread.
premelec
Senior Member
Note that the capacitors should be physically as close as possible to the regulator chip... some times you can get away with other configurations - the PICAXE bypass cap also needs to be right at the chip... I've seen too many mysterious events traced to stray lead inductance to bother with not observing this simple protocol and saving lots of trouble tracing time... have fun.
inglewoodpete
Senior Member
I can confirm that I had all the recommended caps in place, with the 0.1uFs as close as possible to the regulator.
My only assumption for my regulator's failure is that the SM power supply may start up a bit spikey, if that makes sense. It only happened the once but took out 5 PICAXEs in one go. 4 x 28x1 and 1 x 40x1: painful to the pocket
.The design of some (a lot of/all?) linear regulators requires there to be external capacitance - it's not really a susceptibility problem, it's needed to actually make them work properly.
IIRC in the n'ty years I've been doing stuff I got a 7805 to give out around 8v on a multimeter once before I put a capacitor on the output (smooth benchtop supply as well) - a quick check showed the 8v was a near perfect 0-8v sine wave at somewhere around 2Mhz.
Yes, many voltage regulators require caps on the input and output to avoid self-oscillation. In addition, for any that don't require them, it's still a good idea to use them. If they have worked without caps for you in the past, then consider yourself lucky. But don't expect to be so lucky next time.
When using computer switchers as picaxe supplies I ALWAYS use a separate (capacitor regulated) voltage regulator IC between the PC supply and my PICAXE. Unloaded (meaning not drawing an amp or two) PC power supplies may very likely deliver large spikes on power-up until they settle down. They are typically designed as cheaply as possible and are designed for a given minimum load.
Using a PC switcher unloaded (and a picaxe doesn't constitute a "load" to a couple of hundred Watt switcher,) is a risky business.
When using computer switchers as picaxe supplies I ALWAYS use a separate (capacitor regulated) voltage regulator IC between the PC supply and my PICAXE. Unloaded (meaning not drawing an amp or two) PC power supplies may very likely deliver large spikes on power-up until they settle down. They are typically designed as cheaply as possible and are designed for a given minimum load.
Using a PC switcher unloaded (and a picaxe doesn't constitute a "load" to a couple of hundred Watt switcher,) is a risky business.
Check the metal heatsink tab. On some devices this is GND, on some devices like this part it is V+. Therefore accidentally touching the tab against a box etc. can cause issues.
I did some travel so I haven't been messing with PICAXEs in a while,but I finally got to work on them again. Like much of you guys have said, capacitors did the trick. After two fried 20X2s it's lesson learned. Now to finish the board and continue playing with my hot-air rework station. . . I appreciate the help.
fernando_g
Senior Member
"After two fried 20X2s it's lesson learned."
Indeed. The lessons that end up costing £,€,¥ or $, are the ones that usually last a lifetime.
Indeed. The lessons that end up costing £,€,¥ or $, are the ones that usually last a lifetime.