Author Topic: Switch Current calculations for PMIC topologies  (Read 1783 times)

0 Members and 1 Guest are viewing this topic.

Offline smoothVTerTopic starter

  • Regular Contributor
  • *
  • Posts: 151
  • Country: us
Switch Current calculations for PMIC topologies
« on: March 12, 2019, 07:42:44 pm »
Hello All,

Searching on Digikey for DC/DC converters, one often encounters the Current Output parameter listed as (Switch Current) and not just the max current output.

Why is this so?

Also:   for buck, boost, and buck-boost converters, how does one convert from the "Switch Current Limit" to the implemented real-world DC current output capability? 

Is there a resource anywhere that derives this Switch Current Limit for all topologies in one place? It seems that different manufacturers all have a different approach to handling this parameter in their datasheets & the calculations involved, so I am rather confused as to what the base-line theory behind this "switch current" or "peak current" parameter is.


 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Switch Current calculations for PMIC topologies
« Reply #1 on: March 12, 2019, 11:14:38 pm »
Searching on Digikey for DC/DC converters, one often encounters the Current Output parameter listed as (Switch Current) and not just the max current output.

Why is this so?

Also:   for buck, boost, and buck-boost converters, how does one convert from the "Switch Current Limit" to the implemented real-world DC current output capability?

The switching topology has a major effect on output current.  For instance a boosted buck converter using a tapped inductor can multiply the switch current by several times.

Quote
Is there a resource anywhere that derives this Switch Current Limit for all topologies in one place? It seems that different manufacturers all have a different approach to handling this parameter in their datasheets & the calculations involved, so I am rather confused as to what the base-line theory behind this "switch current" or "peak current" parameter is.

Linear Technology published application notes which discuss all of the characteristics of different switching topologies.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Switch Current calculations for PMIC topologies
« Reply #2 on: March 13, 2019, 12:15:14 am »
Because peak current mode controllers are very popular (and with good reason!), and performance is therefore tied to switch current.

It would be disingenuous to rate these by, say, input or output (DC) current, because that current varies with condition.  For example, a boost/flyback/SEPIC delivers (almost proportionally) more output current at higher input voltage.

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

Offline smoothVTerTopic starter

  • Regular Contributor
  • *
  • Posts: 151
  • Country: us
Re: Switch Current calculations for PMIC topologies
« Reply #3 on: March 13, 2019, 02:41:00 pm »
I see.

After reading Linear Tech's note and also this series of notes from TI ...

http://www.ti.com/lit/an/slva477b/slva477b.pdf
http://www.ti.com/lit/an/slva372c/slva372c.pdf
http://www.ti.com/lit/an/slva535b/slva535b.pdf


... I think the gist of what I am getting is ( feel free to refute/comment ) ...


>> You can't predict the max DC output current of any PMIC based solely on the switch current limit in the datasheet. You have to take into account your Vin/Vout, min/max duty cycle, and switching frequency and inductor values.

>> You have to run through all the ideal calculations/equations first to determine your ideal duty cycle, fsw, L value, then come up with a ballpark figure of your max switch current.   Then, the values in a parametric Digikey search for "Switch Current" can be useful.

 I am utilizing DC/DC converters in very tight, space-limited systems so I always look for a synchronous rectifier and integrated, internal switch.   Might be why I am seeing such limitations in switch current.   However I found a very nice regulator, the LT1935, which claims an "unprecedented 2A, 40V internal switch"    for one of my applications.     Why is it so difficult to manufacture a 2A/40V switch when there are many many mosfets that can do that easily as discrete transistors?  Also why are NPN switches used in a lot of LT's converters?   Wouldn't it make more sense to use MOSFETs due to the lack of Vcesat? 

Thanks for all the replies, everyone.




 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Switch Current calculations for PMIC topologies
« Reply #4 on: March 15, 2019, 04:11:23 am »
Might've been "unprecedented" at some point... MC34063 is still disturbingly popular, and was probably the only thing that did an ampere (by itself) for a long time?  No idea.

Datasheets rarely carry dates, so it's hard to say what the timeline is when that's what you're looking at most of the day... :-\

At any rate, 2-3A internal switch (regulator) devices are pretty common nowadays, even in very small (SOT-23-6, DFN/QFN..) sizes.

I suspect LT preferred BJTs partly because of their captive foundry capability, and partly for performance, at least at low supply voltages (good luck making a 40V MOS switch that runs on 2 or 3V).

Vce(sat) isn't a constant, it's largely resistive too.  In fact there are low-Vce(sat) and audio muting type BJTs with specs any FET could only dream about.  A jellybean that's 2A, 80V, 20mohm and only 50pF?  You bet! :)

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: smoothVTer

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Switch Current calculations for PMIC topologies
« Reply #5 on: March 17, 2019, 12:02:17 am »
Might've been "unprecedented" at some point... MC34063 is still disturbingly popular, and was probably the only thing that did an ampere (by itself) for a long time?  No idea.

The MC34063 is a cut down 78S40 so at least going back to 1982 with Fairchild.  They dropped the uncommitted operational amplifier, power diode, and some flexibility allowing an 8 pin instead of 16 pin package.

Quote
I suspect LT preferred BJTs partly because of their captive foundry capability, and partly for performance, at least at low supply voltages (good luck making a 40V MOS switch that runs on 2 or 3V).

For a given current they are also more economical because they require a smaller area.

Quote
Vce(sat) isn't a constant, it's largely resistive too.  In fact there are low-Vce(sat) and audio muting type BJTs with specs any FET could only dream about.  A jellybean that's 2A, 80V, 20mohm and only 50pF?  You bet! :)

Integrated designs can take advantage of saturation control without compromising dropout but are almost always lack power PNPs.
 
The following users thanked this post: smoothVTer

Offline aiq25

  • Regular Contributor
  • *
  • Posts: 243
  • Country: us
Re: Switch Current calculations for PMIC topologies
« Reply #6 on: March 18, 2019, 01:55:19 am »
I see.

After reading Linear Tech's note and also this series of notes from TI ...

http://www.ti.com/lit/an/slva477b/slva477b.pdf
http://www.ti.com/lit/an/slva372c/slva372c.pdf
http://www.ti.com/lit/an/slva535b/slva535b.pdf


... I think the gist of what I am getting is ( feel free to refute/comment ) ...


>> You can't predict the max DC output current of any PMIC based solely on the switch current limit in the datasheet. You have to take into account your Vin/Vout, min/max duty cycle, and switching frequency and inductor values.

>> You have to run through all the ideal calculations/equations first to determine your ideal duty cycle, fsw, L value, then come up with a ballpark figure of your max switch current.   Then, the values in a parametric Digikey search for "Switch Current" can be useful.

 I am utilizing DC/DC converters in very tight, space-limited systems so I always look for a synchronous rectifier and integrated, internal switch.   Might be why I am seeing such limitations in switch current.   However I found a very nice regulator, the LT1935, which claims an "unprecedented 2A, 40V internal switch"    for one of my applications.     Why is it so difficult to manufacture a 2A/40V switch when there are many many mosfets that can do that easily as discrete transistors?  Also why are NPN switches used in a lot of LT's converters?   Wouldn't it make more sense to use MOSFETs due to the lack of Vcesat? 

Thanks for all the replies, everyone.

TI has a tool called Power Stage Designer. You can choose different topologies and circuit parameters and it will calculate all the design parameters (i.e. switch current, inductor current ripple, etc...). It is very handy. For example, you can choose a boost topology and click on the FET, inductor or diode to see calculated current for them at different input voltages, output currents, etc...

Here is an example (I found it on google images):


These are all for reference but it's nice to have this calculator. Plus if you use a TI IC you can use their designer online and even get PCB files, BOM, schematic exports, etc... They used to thermal simulations online but I haven't used it in a while. Obviously these are again for reference but very handy to have.
 

Offline smoothVTerTopic starter

  • Regular Contributor
  • *
  • Posts: 151
  • Country: us
Re: Switch Current calculations for PMIC topologies
« Reply #7 on: March 20, 2019, 01:23:38 pm »
Interesting

I've used TI's per-chip designer applet thingamadoo many times in the past but never Power Designer.  I'll give it a shot ... the basic equations for each topology should be more or less the same within that topology, between manufacturers.     I can see a different equation being used if a synchronous rectifier is used vs. an internal diode or external diode ...
 

Offline smoothVTerTopic starter

  • Regular Contributor
  • *
  • Posts: 151
  • Country: us
Re: Switch Current calculations for PMIC topologies
« Reply #8 on: March 21, 2019, 04:53:51 pm »
I see.

After reading Linear Tech's note and also this series of notes from TI ...

http://www.ti.com/lit/an/slva477b/slva477b.pdf
http://www.ti.com/lit/an/slva372c/slva372c.pdf
http://www.ti.com/lit/an/slva535b/slva535b.pdf


... I think the gist of what I am getting is ( feel free to refute/comment ) ...


>> You can't predict the max DC output current of any PMIC based solely on the switch current limit in the datasheet. You have to take into account your Vin/Vout, min/max duty cycle, and switching frequency and inductor values.

>> You have to run through all the ideal calculations/equations first to determine your ideal duty cycle, fsw, L value, then come up with a ballpark figure of your max switch current.   Then, the values in a parametric Digikey search for "Switch Current" can be useful.

 I am utilizing DC/DC converters in very tight, space-limited systems so I always look for a synchronous rectifier and integrated, internal switch.   Might be why I am seeing such limitations in switch current.   However I found a very nice regulator, the LT1935, which claims an "unprecedented 2A, 40V internal switch"    for one of my applications.     Why is it so difficult to manufacture a 2A/40V switch when there are many many mosfets that can do that easily as discrete transistors?  Also why are NPN switches used in a lot of LT's converters?   Wouldn't it make more sense to use MOSFETs due to the lack of Vcesat? 

Thanks for all the replies, everyone.

TI has a tool called Power Stage Designer. You can choose different topologies and circuit parameters and it will calculate all the design parameters (i.e. switch current, inductor current ripple, etc...). It is very handy. For example, you can choose a boost topology and click on the FET, inductor or diode to see calculated current for them at different input voltages, output currents, etc...

Here is an example (I found it on google images):


These are all for reference but it's nice to have this calculator. Plus if you use a TI IC you can use their designer online and even get PCB files, BOM, schematic exports, etc... They used to thermal simulations online but I haven't used it in a while. Obviously these are again for reference but very handy to have.


I downloaded and installed TI Power Stage Designer.  Awesome little bit of software!   I have one question:  How come only  inverting buck-boost is shown?   Is there a non-inverting buck-boost topology that is known by a different name?   I've used the TPS63020 in a few places, and it is specified as just a buck-boost regulator in the data sheet.  Is the buck-boost characteristic switch current  the composite of the buck mode stage and the boost mode stage, since only one stage can be operating at time? 
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf