Forums

Boost Converter Tutorial?

Started by Jim Thompson February 11, 2012
"Jim Thompson" <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote in 
message news:i5fdj71se8ju71to9r931quld5qlig3rrd@4ax.com...
> On Sat, 11 Feb 2012 13:34:08 -0500, Phil Hobbs > <pcdhSpamMeSenseless@electrooptical.net> wrote: > >>Jim Thompson wrote: >>> >>> On Sat, 11 Feb 2012 13:25:00 -0500, Phil Hobbs >>> <pcdhSpamMeSenseless@electrooptical.net> wrote: >>> >>> >Jim Thompson wrote: >>> >> >>> >> In the past I have designed boost converters in burst mode, with peak >>> >> current control, but never with a PWM-controlled loop. >>> >> >>> >> Outputs: +5 @ 25mA, -5V @ 20mA >>> >> >>> >> Can someone point me to a tutorial? >>> >> >>> >> Thanks! >>> >> >>> >> [Don't point me to an off-the-shelf part. This has to go into a >>> >> _custom_ chip... just a little thing... 1mm x 25mm :-] >>> > >>> >Hmm, an inch long and a millimeter wide--it's for a robotic cockroach, >>> >right? >>> > >>> >Cheers >>> > >>> >Phil Hobbs >>> >>> Nope. It goes in something that virtually every person owns :-) >>> >>> ...Jim Thompson >> >>So you decided to take someone's advice about what to do with your >>design? >> >>Cheers >> >>Phil Hobbs > > This is but a small portion of the whole chip. As I said, I have no > experience (*) with PWM'd boost converters, though lots of those with > low output current, running burst mode (they have little or no > stability issues). > > As Larkin would snark, If you've got something technical to say, then > say it, otherwise STFU :-{) > > (*) Unlike some posters lurking here, I readily admit areas of which I > have little or no expertise. Then I listen, learn and ultimately > become the expert ;-)
Yeahbut you tried to cram a burst mode at one of my questions and I hadn't even asked for a design. It wasn't because burst mode was best for my app, it was because you didn't know PWM - not exactly honest was it.
On Sat, 11 Feb 2012 11:22:38 -0700, Jim Thompson
<To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:

>In the past I have designed boost converters in burst mode, with peak >current control, but never with a PWM-controlled loop. > >Outputs: +5 @ 25mA, -5V @ 20mA > >Can someone point me to a tutorial? > >Thanks! > >[Don't point me to an off-the-shelf part. This has to go into a >_custom_ chip... just a little thing... 1mm x 25mm :-] > > ...Jim Thompson
Input? RL
Jim Thompson wrote:
> > On Sat, 11 Feb 2012 13:34:08 -0500, Phil Hobbs > <pcdhSpamMeSenseless@electrooptical.net> wrote: > > >Jim Thompson wrote: > >> > >> On Sat, 11 Feb 2012 13:25:00 -0500, Phil Hobbs > >> <pcdhSpamMeSenseless@electrooptical.net> wrote: > >> > >> >Jim Thompson wrote: > >> >> > >> >> In the past I have designed boost converters in burst mode, with peak > >> >> current control, but never with a PWM-controlled loop. > >> >> > >> >> Outputs: +5 @ 25mA, -5V @ 20mA > >> >> > >> >> Can someone point me to a tutorial? > >> >> > >> >> Thanks! > >> >> > >> >> [Don't point me to an off-the-shelf part. This has to go into a > >> >> _custom_ chip... just a little thing... 1mm x 25mm :-] > >> > > >> >Hmm, an inch long and a millimeter wide--it's for a robotic cockroach, > >> >right? > >> > > >> >Cheers > >> > > >> >Phil Hobbs > >> > >> Nope. It goes in something that virtually every person owns :-) > >> > >> ...Jim Thompson > > > >So you decided to take someone's advice about what to do with your > >design? > > > >Cheers > > > >Phil Hobbs > > This is but a small portion of the whole chip. As I said, I have no > experience (*) with PWM'd boost converters, though lots of those with > low output current, running burst mode (they have little or no > stability issues). > > As Larkin would snark, If you've got something technical to say, then > say it, otherwise STFU :-{) >
Like you, right? Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 845-480-2058 hobbs at electrooptical dot net http://electrooptical.net
> > The inductor _is_ a problem. They're already at their height limit > and their present implementation (at lower current) is just pushing > into saturation :-( > >> >>Are you planning on paying for the 0.3V drop of a Schottkey on each >>supply, or the time & trouble to drive the extra two switches & make it >>synchronous? > > That's up for debate. > > But this client is no onesy-twosy outfit... they just shipped their > BILLIONETH chip!
I bet they'll pay you an ARM and a leg for the right solution<snigger>
"Jim Thompson" <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote 
in message news:1mfdj7l31utddhcf8ppa27lv9eke0r11h2@4ax.com...
> On Sat, 11 Feb 2012 13:36:38 -0500, "Martin Riddle" > <martin_rid@verizon.net> wrote: > >> >>"Jim Thompson" <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> >>wrote >>in message news:m3cdj7hc1nqt27bn08n86pjmjsdpgl8gtm@4ax.com... >>> In the past I have designed boost converters in burst mode, with >>> peak >>> current control, but never with a PWM-controlled loop. >>> >>> Outputs: +5 @ 25mA, -5V @ 20mA >>> >>> Can someone point me to a tutorial? >>> >>> Thanks! >>> >>> [Don't point me to an off-the-shelf part. This has to go into a >>> _custom_ chip... just a little thing... 1mm x 25mm :-] >>> >>> ...Jim Thompson >>> -- >> >>TI app notes? >><http://www.ti.com/analog/docs/analogtechdoc_hh.tsp?viewType=mostuseful&rootFamilyId=64&familyId=661&docTitle=boost&docCategoryId=1> >> >>The 3525 has been around for a long time... >>Maybe DN-62 has what your looking for. >><http://www.ti.com/product/uc3525b> >> >>cheers >> >> > > Thanks, Martin! I owe you another sip or two or... > > Though probably at some other location than Long Island. I sent those > horse's-asses in Huntington (and only you will appreciate why I say > that) an E-mail, "As I sit here in Rochester, in my _office_ (with a > _door_), looking out my _window_ at the melting snow, with a > _door-pass_ badge hanging around my neck, and with my laptop sitting > there connected to the customer's _network_ and the _Internet_, I was > thinking of you fellows at @#$%... >:-} " > > I doubt they'll invite me back ;-) Though you never know... I was the > only one who could convey to Cadence what was needed with the "w-real" > modeling. > > ...Jim Thompson > --
Snicker... I also came across a voltage vs. current move paper by TI (maybe unitrode) not sure. That may have something in it too. It should be some where in those links above. Cheers
> I bet they'll pay you an ARM and a leg for the right solution<snigger> > >
the old unitrode seminars are the best http://www.smps.us/Unitrode.html Mark
On Sat, 11 Feb 2012 11:26:18 -0700, the renowned Jim Thompson
<To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:

>On Sat, 11 Feb 2012 13:25:00 -0500, Phil Hobbs ><pcdhSpamMeSenseless@electrooptical.net> wrote: > >>Jim Thompson wrote: >>> >>> In the past I have designed boost converters in burst mode, with peak >>> current control, but never with a PWM-controlled loop. >>> >>> Outputs: +5 @ 25mA, -5V @ 20mA >>> >>> Can someone point me to a tutorial? >>> >>> Thanks! >>> >>> [Don't point me to an off-the-shelf part. This has to go into a >>> _custom_ chip... just a little thing... 1mm x 25mm :-] >> >>Hmm, an inch long and a millimeter wide--it's for a robotic cockroach, >>right? >> >>Cheers >> >>Phil Hobbs > >Nope. It goes in something that virtually every person owns :-) > > ...Jim Thompson
Display driver. Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
On Sat, 11 Feb 2012 13:18:28 -0700, Jim Thompson wrote:

> On Sat, 11 Feb 2012 13:59:29 -0600, Tim Wescott > <tim@seemywebsite.please> wrote: > >>On Sat, 11 Feb 2012 11:22:38 -0700, Jim Thompson wrote: >> >>> In the past I have designed boost converters in burst mode, with peak >>> current control, but never with a PWM-controlled loop. >>> >>> Outputs: +5 @ 25mA, -5V @ 20mA >>> >>> Can someone point me to a tutorial? >>> >>> Thanks! >>> >>> [Don't point me to an off-the-shelf part. This has to go into a >>> _custom_ chip... just a little thing... 1mm x 25mm :-] >> >>Actually, I'd go looking at off-the-shelf parts to snag ideas from. >> >>Efficiency? > > As high as practical. > >>Regulation bounds? > > Loose. Lots of compensating innards for power supply variation. > >> % of total product cost that'll be taken >>up by the converter? > > They've already committed to external switch(es) > > >>Unless the above answers are "low, loose, and low", I think your biggest >>challenge won't be the silicon, it'll be finding small, economical & >>high performance inductors, designing the thing to take advantage of >>them, and dancing around the inevitable unstable zero in the boost >>converter. > > The inductor _is_ a problem. They're already at their height limit and > their present implementation (at lower current) is just pushing into > saturation :-( > > >>Are you planning on paying for the 0.3V drop of a Schottkey on each >>supply, or the time & trouble to drive the extra two switches & make it >>synchronous? > > That's up for debate. > > But this client is no onesy-twosy outfit... they just shipped their > BILLIONETH chip!
Oh joy -- so, do you have the frequency headroom to use a lower inductance coil with the same core, to get the saturation current up? -- My liberal friends think I'm a conservative kook. My conservative friends think I'm a liberal kook. Why am I not happy that they have found common ground? Tim Wescott, Communications, Control, Circuits & Software http://www.wescottdesign.com
On 2/11/2012 3:18 PM, Jim Thompson wrote:
> On Sat, 11 Feb 2012 13:59:29 -0600, Tim Wescott > <tim@seemywebsite.please> wrote: > >> On Sat, 11 Feb 2012 11:22:38 -0700, Jim Thompson wrote: >> >>> In the past I have designed boost converters in burst mode, with peak >>> current control, but never with a PWM-controlled loop. >>> >>> Outputs: +5 @ 25mA, -5V @ 20mA >>> >>> Can someone point me to a tutorial? >>> >>> Thanks! >>> >>> [Don't point me to an off-the-shelf part. This has to go into a >>> _custom_ chip... just a little thing... 1mm x 25mm :-] >> >> Actually, I'd go looking at off-the-shelf parts to snag ideas from. >> >> Efficiency? > > As high as practical. > >> Regulation bounds? > > Loose. Lots of compensating innards for power supply variation. > >> % of total product cost that'll be taken >> up by the converter? > > They've already committed to external switch(es) > >> >> Unless the above answers are "low, loose, and low", I think your biggest >> challenge won't be the silicon, it'll be finding small, economical& high >> performance inductors, designing the thing to take advantage of them, and >> dancing around the inevitable unstable zero in the boost converter. > > The inductor _is_ a problem. They're already at their height limit > and their present implementation (at lower current) is just pushing > into saturation :-( >
Has the value of inductor been selected properly? The rule of thumb for a boost converter in continuous conduction mode is L = (V * D)/(R*I_l*f), where the parameters are evaluated at the _lowest_ specified (worst case for a boost) input voltage. V is the input voltage, I_l is the DC inductor current (I_o/(1-D)), D is the duty cycle, f is the switching frequency. R is the "ripple current ratio" and has an optimum value of 0.4 for continuous conduction mode in most cases. At low voltages the inductor can be sized solely based on the DC load current requirements - if the current limiting is fast enough it doesn't matter if the inductor saturates under abnormal conditions. If the inductor is still too big maybe use discontinuous mode instead? It'll be harder to stabilize, unfortunately.
On Sat, 11 Feb 2012 11:26:18 -0700, Jim Thompson
<To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:

>On Sat, 11 Feb 2012 13:25:00 -0500, Phil Hobbs ><pcdhSpamMeSenseless@electrooptical.net> wrote: > >>Jim Thompson wrote: >>> >>> In the past I have designed boost converters in burst mode, with peak >>> current control, but never with a PWM-controlled loop. >>> >>> Outputs: +5 @ 25mA, -5V @ 20mA >>> >>> Can someone point me to a tutorial? >>> >>> Thanks! >>> >>> [Don't point me to an off-the-shelf part. This has to go into a >>> _custom_ chip... just a little thing... 1mm x 25mm :-] >> >>Hmm, an inch long and a millimeter wide--it's for a robotic cockroach, >>right? >> >>Cheers >> >>Phil Hobbs > >Nope. It goes in something that virtually every person owns :-)
Car ignition key?