Author Topic: LDO keeps blowing up but without any magic smoke...  (Read 4810 times)

0 Members and 1 Guest are viewing this topic.

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
LDO keeps blowing up but without any magic smoke...
« on: May 12, 2016, 07:55:42 am »
Hi,

I am trying out the Micrel MIC5205 LDO I bought from ebay (item #311513963103) and they keep blowing up.
No magic smoke though.   :(       

After the failure there is always a short between Vcc and Ground. 
The load is very small at around 14mA and this can handle up to 150mA. 
Enable pin is always pulled down with a 100k resistor and I enable it by connecting the EN pin to Vcc.   
A 1uF tantalum capacitor is connected to Vout.


http://ww1.microchip.com/downloads/en/DeviceDoc/mic5205.pdf


At first power on +8Vcc, the output was ramping up at a random manner.  After some tries switching
the power supply on and off, it went to 3.3V and then it stayed there with a very stable output.
I applied a small load and everything was fine but whenever I tried +10V Vcc the output dropped to
almost zero.  At 9.8V it started to draw excessive current.  According to the datasheet this will work up to 16V.
The second IC had the same behavior but it was damaged immediately at 10V.

At first I thought that I damaged it thermally when soldering, because I used 300 degrees C.
The datasheet suggests a 250 degrees for 5 seconds as an absolute maximum. 
The second time I used 230 degrees and very quickly.  Same results though...

Am I doing something wrong, or did I receive a faulty batch ?
Any help is appreciated.  (I have another 3 left...)
Thank you.

p.s. by the way, this fine pitch is simply evil !   >:D
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: LDO keeps blowing up but without any magic smoke...
« Reply #1 on: May 12, 2016, 07:57:54 am »
No input bypass?
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #2 on: May 12, 2016, 07:59:18 am »
No, I did not use any.  None mentioned in the datasheet.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: LDO keeps blowing up but without any magic smoke...
« Reply #3 on: May 12, 2016, 08:08:19 am »
No, I did not use any.  None mentioned in the datasheet.

Quote
Input Capacitor
A 1µF capacitor should be placed from IN to GND if
there is more than 10 inches of wire between the input
and the ac filter capacitor or if a battery is used as the
input.

;)

I like that they give so many PSRR graphs, but... they're upside down, the "rejection ratio" ones.  Rejection is positive if it's, well, rejecting. So they're apparently claiming >100dB of gain at 10Hz, which is obviously false.  The "ripple rejection" ones are positive dB, though... so it's worse than just being wrong, they're inconsistent too.

But more to the point, there is no spec on GND pin current versus Vin (safe to assume it's notoriously bad unless shown), and no stability graph versus C + ESR for either input or output capacitors.  I would not design-in this part.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: hgg

Online wraper

  • Supporter
  • ****
  • Posts: 16803
  • Country: lv
Re: LDO keeps blowing up but without any magic smoke...
« Reply #4 on: May 12, 2016, 08:22:15 am »
At first I thought that I damaged it thermally when soldering, because I used 300 degrees C.
The datasheet suggests a 250 degrees for 5 seconds as an absolute maximum. 
The second time I used 230 degrees and very quickly.  Same results though...
That's reflow temperature (when all device gets heated), not solder iron temperature.
 
The following users thanked this post: hgg

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #5 on: May 12, 2016, 08:48:43 am »
Quote
Input Capacitor
A 1µF capacitor should be placed from IN to GND if
there is more than 10 inches of wire between the input
and the ac filter capacitor or if a battery is used as the
input.

Indeed the input wire was more than 10 inches, which I forgot, but is this a reason for
the LDO to get damaged ?   Can it be damaged by oscillation ?  (none observed though).

Quote
I would not design-in this part.
Any suggestion of a quality part with an enable pin?

Quote
That's reflow temperature (when all device gets heated), not solder iron temperature.
So what is the max safe temperature with a soldering iron?

Thanks.
 

Online wraper

  • Supporter
  • ****
  • Posts: 16803
  • Country: lv
Re: LDO keeps blowing up but without any magic smoke...
« Reply #6 on: May 12, 2016, 08:55:04 am »
300o C is certainly safe, 350oC should be safe too, if doing the job fast enough. All you should care about is that part itself don't heat up above that max temperature.
 
The following users thanked this post: hgg

Offline djacobow

  • Super Contributor
  • ***
  • Posts: 1151
  • Country: us
  • takin' it apart since the 70's
Re: LDO keeps blowing up but without any magic smoke...
« Reply #7 on: May 12, 2016, 08:56:09 am »
I can't speak for blowing up, but I inherited a design with this very part last year and it oscillated like nobody's business the way they had it in there. That output cap, it really needs to be a tant, or at least put a small resistance in series with an MLCC.

When I revved the board, I designed it out. I don't remember what I put in there, but I actually didn't need the enable.
 
The following users thanked this post: hgg

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #8 on: May 12, 2016, 09:03:59 am »
I used a tantalum capacitor.
So, it looks like that its not one of the best LDOs out there... 

Thats why I like ebay.  You can find the worst parts at the greatest prices.   :)

I will give it one more try fully 'caped'....
 

Offline EPTech

  • Regular Contributor
  • *
  • Posts: 168
  • Country: be
    • EP Technical Services
Re: LDO keeps blowing up but without any magic smoke...
« Reply #9 on: May 12, 2016, 09:37:41 am »
Hi there,

Quote
Indeed the input wire was more than 10 inches, which I forgot, but is this a reason for
the LDO to get damaged ?   Can it be damaged by oscillation ?  (none observed though).

Yes it can because each wire and trace has a parasitic inductance. When you power on, you actually make a step. The voltage on the input will overshoot very shortly but surely and could go way past the rated input voltage.
Kind greetings,

Pascal.
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Re: LDO keeps blowing up but without any magic smoke...
« Reply #10 on: May 12, 2016, 10:28:01 am »
When it oscillates it is dissipating a lot of CxV**2xf power dissipation
because of load cap, so die meltdown real possibility. Keep in mind
power is the rate of change of energy, eg. the flow of energy in and
out of the cap constitutes power, the power to move energy in and out
of the cap.

Regards, Dana.
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #11 on: May 12, 2016, 10:39:04 am »
Interesting.
I never saw oscillation on the scope though.  The voltage was a flat line all the time.

Just soldered another IC with hot air but this time with all the capacitors.
Worked fine for a while and after some on/off cycles and a small load it cannot regulate more than 9V.
At 10V the output voltage drops in half.   Never had this kind of problem with any other LDO.

Am still not sure what is going on.   The first time I thought that I might have damaged it when I checked
for shorts with the multimeter before powering it on.  The UT61E outputs +3V in continuity mode and if
you check with wrong polarity maybe you can damage it with -3V (?).   The UT136C outputs a safer 0.44V.

I am at a loss of what is the real problem, but I am definitely not going to use this LDO even if I find the
real cause.   ...which is...  :-//



 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 9410
  • Country: gb
Re: LDO keeps blowing up but without any magic smoke...
« Reply #12 on: May 12, 2016, 11:17:13 am »
Those LDOs are also very easy to damage by shorting the input... discharge current spike of the output capacitor through the internal parasitic reverse diode.
Best Regards, Chris
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Re: LDO keeps blowing up but without any magic smoke...
« Reply #13 on: May 12, 2016, 11:56:41 am »
Protection of linear regs, see link, page 4......


http://www.ti.com.cn/cn/lit/an/snoa402b/snoa402b.pdf


Regards, Dana.
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #14 on: May 12, 2016, 12:05:12 pm »
Hi,

Maybe I should have used protection diodes.
Thanks for the link, I will read it.

I have two more ICs.  I'll give them a try with all the protection possible.   :)

Do you think that this has anything to do with the problem?
https://youtu.be/9DwytzdpQOc

Thanks.
 

Offline krivx

  • Frequent Contributor
  • **
  • Posts: 765
  • Country: ie
Re: LDO keeps blowing up but without any magic smoke...
« Reply #15 on: May 12, 2016, 12:15:41 pm »
Just because no one has mentioned it yet - eBay parts are always something of a gamble. These things could be seconds, fake or ESD damaged before they get to you.
 

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #16 on: May 12, 2016, 12:20:38 pm »
Indeed, and fake might be more possible because of the constant failure above 9.8V exactly!  (Its rated for 16V)
They arrived in a plain plastic bag... so ESD damage is a possibility as well. 

 

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #17 on: May 12, 2016, 01:44:57 pm »
This looks like a properly designed LDO, what do you think?  Any experience with this part?

http://www.ti.com/lit/ds/symlink/lp2985-n.pdf
 

Offline umagarz

  • Supporter
  • ****
  • Posts: 2
  • Country: dk
Re: LDO keeps blowing up but without any magic smoke...
« Reply #18 on: May 12, 2016, 02:03:59 pm »
Just because no one has mentioned it yet - eBay parts are always something of a gamble. These things could be seconds, fake or ESD damaged before they get to you.

+1

I recently bought some 7seg serial drivers on ebay, and none of them would communicate with SPI...
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9886
  • Country: us
Re: LDO keeps blowing up but without any magic smoke...
« Reply #19 on: May 12, 2016, 02:49:55 pm »
To be clear, there should probably be a reverse biased diode between Vout and Vin.  Many regulators don't like reverse current at shutdown.  This becomes a more serious problem when there is a lot of capacitance on the load.  They have to discharge somewhere.

You may see a flat line on the output but how about the input.  You mentioned a long wire, a perfect tuned circuit for oscillation.  You might try some capacitance at the input.  Something like the one used at the output.

I buy my parts from DigiKey or Mouser.  Absolutely nothing from eBay!  Those chips are floor sweepings at best.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: LDO keeps blowing up but without any magic smoke...
« Reply #20 on: May 12, 2016, 04:28:54 pm »
Most likely seems to me, either reverse conduction, or input overvoltage or reversal.

Long wires exhibit inductance, which causes voltage spikes when connected.  This is exacerbated with ceramic capacitors, but dampened with electrolytics.  (And tantalums, but tants don't appreciate current surges.)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #21 on: May 12, 2016, 04:33:21 pm »
I will try once more with all the protection passives in place.
Maybe a 12V zener as well !

It usually works up to 9.8V but above that the LDO starts to draw more current and fail.
 

Offline hggTopic starter

  • Frequent Contributor
  • **
  • Posts: 960
  • Country: gr
Re: LDO keeps blowing up but without any magic smoke...
« Reply #22 on: May 13, 2016, 07:18:20 pm »
Ok, I soldered another one, with a protection diode and input/output capacitors.
Proper soldering temperature and dwell time.   Again, it can regulate only up to 9.8 Volts.
Put 9.9 Volts and the output drops rapidly to 1.36V also drawing a lot of current from the
power supply.

As djacobow said, the original ones oscillate very easily.  These ones oscillate only if I
leave the enable pin floating.  They are stable even without any capacitors but they can
regulate only up to 9.8V

They must be %100 fake.
Cannot think of something else.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf