Author Topic: LDO bypass capacitors  (Read 3303 times)

0 Members and 1 Guest are viewing this topic.

Offline Nikos A.Topic starter

  • Regular Contributor
  • *
  • Posts: 240
  • Country: cy
LDO bypass capacitors
« on: April 22, 2019, 01:50:50 pm »
Hi everyone,

I want to use the TPS79733QDCKRQ1 LDO in one of my projects. According to the datasheet two decoupling capacitors are required in both input and output.

The datasheet indicates electrolytic capacitors. Can I use ceramic instead?

Thanks in advance!!

 

Online wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: LDO bypass capacitors
« Reply #1 on: April 22, 2019, 01:58:22 pm »
It says any capacitor, so MLCC supposedly is OK.
 

Online wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: LDO bypass capacitors
« Reply #2 on: April 22, 2019, 02:03:47 pm »
But I would suggest something more modern and cheaper. For example TPS7A0533PDBVR is explicitly said to be fine with MLCC.
 

Online wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: LDO bypass capacitors
« Reply #3 on: April 22, 2019, 02:08:06 pm »
MCP1811 or MCP1812 is also multiple times cheaper.
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: LDO bypass capacitors
« Reply #4 on: April 22, 2019, 02:19:16 pm »
That 0.47 ufd provides bulk capacitance to the load.  It can be larger and I would move toward 10 ufd electrolytic.  Then I would add a parallel 0.1 ufd ceramic to handle high frequency noise.  The input capacitor is still 0.1 ufd ceramic.
 

Offline Nikos A.Topic starter

  • Regular Contributor
  • *
  • Posts: 240
  • Country: cy
Re: LDO bypass capacitors
« Reply #5 on: April 22, 2019, 02:51:45 pm »
Thank you for your advice!!
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: LDO bypass capacitors
« Reply #6 on: April 22, 2019, 08:22:57 pm »
That 0.47 ufd provides bulk capacitance to the load.  It can be larger and I would move toward 10 ufd electrolytic.  Then I would add a parallel 0.1 ufd ceramic to handle high frequency noise.

Adding a parallel 0.1 microfarad ceramic capacitor to the existing bulk output capacitor does nothing useful.  High frequency decoupling capacitors must be close to the load to be effective.  For the same reason, the output of the regulator cannot "see" the load at high frequencies.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: LDO bypass capacitors
« Reply #7 on: April 23, 2019, 01:56:36 pm »
Adding a parallel 0.1 microfarad ceramic capacitor to the existing bulk output capacitor does nothing useful.

It may be needed for stability if feedback loop bandwidth is quite high. At least that was a suggestion from one application note which I don't remember. The idea was that bulk electrolites may be too "slow" for high-speed regulation.

I personally always include a 100nF on the output, but never measured the impact of this. Although, when I do measurements that can be affected by the power supply, I try to make power cable as short as possible, often connecting the test fixture directly to output bananas, in addition to a capacitor a the point of load. But the thing is, because of parallel resonanse between decoupling caps (and their parasitic inductances), it's impossible for me to predict if adding more capacitors helps or not. I often try to measure "before" and "after" and compare if there is any change and if I have enough time. Otherwise there is some uncertainty. Fortunately, I work with hundreds of kHz, so this is not that big issue.

So I what to suggest: do measurement to see what actually makes sense, and what is not. Dave has two good videos on the subject: 1) showing impact of mlcc and electrolyte caps as decoupling caps 2) in red pitaya video he showed that decoupling caps in parallel may cause problems.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: LDO bypass capacitors
« Reply #8 on: April 24, 2019, 06:58:28 pm »
Adding a parallel 0.1 microfarad ceramic capacitor to the existing bulk output capacitor does nothing useful.

It may be needed for stability if feedback loop bandwidth is quite high. At least that was a suggestion from one application note which I don't remember. The idea was that bulk electrolites may be too "slow" for high-speed regulation.

I would like to see that application note.  The only examples of this I have seen indicated the opposite and the only reason designers got away with it was because the large aluminum electrolytic or solid tantalum bulk decoupling capacitor damped the response of the regulator enough to tolerate the small low ESR ceramic part.  (1)

It certainly makes a difference if you make a measurement directly at the regulator's output but a remote load never sees that!  Only the probe does.

There are some special regulators designed specifically to work with low value ceramic capacitors but it is not easy.

(1) The same trick can work with an operational amplifier where you add say a 10 microfarad aluminum electrolytic directly from the output to ground.  Now the operational amplifier will be quite happy driving a 0.1 microfarad ceramic capacitor albeit with low gain-bandwidth product.
 

Online wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: LDO bypass capacitors
« Reply #9 on: April 24, 2019, 08:05:43 pm »
LDO in general (such as LM1117-xx) don't like LOW ESR capacitors such as MLCC on the output. So it's completely counterproductive to use such if particular LDO is not designed to work with ceramic capacitors (mostly relatively recent models). They may become unstable and start to oscillate. Don't confuse them with 78xx which are not LDO to begin with and recommended to have 0.1uF cap on the output. You could use them with small resistor in series to increase ESR. But why would you do this when you can buy LDO which are stable with MLCC to begin with, not to say usually they are even cheaper than older designs.
« Last Edit: April 24, 2019, 08:08:54 pm by wraper »
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: LDO bypass capacitors
« Reply #10 on: April 24, 2019, 09:18:25 pm »
Don't confuse them with 78xx which are not LDO to begin with and recommended to have 0.1uF cap on the output.

The emitter follower output of the 78xx series makes them very tolerant of the output capacitance and ESR.  Usually 10 to 100 microfarads of output capacitance per amp was used but often no output capacitor was required.  The input capacitor in one form or another was always required however.

You might still get into trouble with a lower value solo ceramic output capacitor.

The 79xx series are whole different matter because being a negative regulator with an NPN output stage, the collector is used.  They *must* have a "high ESR" output capacitor and typically a 1 microfarad or larger solid tantalum or 10 microfarad or larger aluminum electrolytic was used.

The 317/337 have the same disparate requirements.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: LDO bypass capacitors
« Reply #11 on: April 25, 2019, 05:03:31 am »
I would like to see that application note.  The only examples of this I have seen indicated the opposite and the only reason designers got away with it was because the large aluminum electrolytic or solid tantalum bulk decoupling capacitor damped the response of the regulator enough to tolerate the small low ESR ceramic part.  (1)

I'm afraid I won't find it. I think it was a datasheet for an LDO with high loop bandwidth. It was saying that an electrolyte may have too big impedance at high frequency (some hundreds of kHz, don't remember exact number) and may not meet requirements of maximum ESR for stability (I hope this sentence makes sense, even though I mix ESR and impedance). As I understand, capacitors are useless above their self-resonant frequency because of phase shift. May be that was also the reason. Frankly, I don't remember details.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: LDO bypass capacitors
« Reply #12 on: April 26, 2019, 01:58:32 am »
I'm afraid I won't find it. I think it was a datasheet for an LDO with high loop bandwidth. It was saying that an electrolyte may have too big impedance at high frequency (some hundreds of kHz, don't remember exact number) and may not meet requirements of maximum ESR for stability (I hope this sentence makes sense, even though I mix ESR and impedance). As I understand, capacitors are useless above their self-resonant frequency because of phase shift. May be that was also the reason. Frankly, I don't remember details.

LDOs have different and more stringent stability requirements because their output impedance is higher.

I had a thought about this today; I may have encountered the problem with sticking a ceramic capacitor on the output of a 7805/317 before.

The emitter follower output is low impedance and if driven from a low enough impedance, acts as a negative resistance when driving a low impedance load.  This can cause the output transistor itself to break into oscillation if a small low loss capacitor is used.

The result would not necessarily compromise the regulation but it is a major EMI and power dissipation hazard.  The resulting oscillation can be difficult to detect without a sufficient bandwidth instrument and applying a probe may stop it anyway.

I have encountered regulators which just mysteriously ran hot and heard several similar reports which may have been due to this.  Of course now knowing better, this is one of the things I would immediately check for.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf