Resonant SMPS.

[artem_bond]

Hello all.

I'm working on SMPS. And I'm choice a series-loaded schematics. I mean
that I use a capacitor series with primary coil transformer. But I
can't understand. Is possibly to build resonant power supply with full
range of load. From 0 to max? I can't understand how to this SMPS can
work without load. When I try to switch on SMPS without load, on low
(start) frequency my transformer comes to saturation. I understand why
this coming but I can't understand what I can do for prevent this. Or
resonant SMPS can't work without load physically?

With load all ok.

 

[R.Legg]

artem_bond@mail.ru (artem_bond) wrote in message news:<4bea06d1.0404290018.35fdeae6@posting.google.com>...

Hello all.

I'm working on SMPS. And I'm choice a series-loaded schematics. I mean
that I use a capacitor series with primary coil transformer. But I
can't understand. Is possibly to build resonant power supply with full
range of load. From 0 to max? I can't understand how to this SMPS can
work without load. When I try to switch on SMPS without load, on low
(start) frequency my transformer comes to saturation. I understand why
this coming but I can't understand what I can do for prevent this. Or
resonant SMPS can't work without load physically?

With load all ok.

Series resonant SMPS are typically at maximum frequency at minimum
load, while parallel resonant types are typically at minimum frequency
at minimum load.

The old Philips app notes are some help here:

http://www.vlsi.com/acrobat/applicationnotes/APPCHP2.pdf

RL

 

[legg]

On 29 Apr 2004 01:18:31 -0700, artem_bond@mail.ru (artem_bond) wrote:

Hello all.

I'm working on SMPS. And I'm choice a series-loaded schematics. I mean
that I use a capacitor series with primary coil transformer. But I
can't understand. Is possibly to build resonant power supply with full
range of load. From 0 to max? I can't understand how to this SMPS can
work without load. When I try to switch on SMPS without load, on low
(start) frequency my transformer comes to saturation. I understand why
this coming but I can't understand what I can do for prevent this. Or
resonant SMPS can't work without load physically?

With load all ok.

Resonant section of the Philips app note is section 2.5.2 starting on
page 123 of the electronic format document. (Index is at end.)

Also easy to check out is:

"A Comparison of Resonant Topologies in High Voltage DC Applications"
Johnson, Witulski and Erickson
IEEe trans AES'88, APEC'87

http://ece-www.colorado.edu/~rwe/papers/APEC87.pdf

RL

 

[Jim Yanik]

artem_bond@mail.ru (artem_bond) wrote in
news:4bea06d1.0404290018.35fdeae6@posting.google.com:

Hello all.

I'm working on SMPS. And I'm choice a series-loaded schematics. I mean
that I use a capacitor series with primary coil transformer. But I
can't understand. Is possibly to build resonant power supply with full
range of load. From 0 to max? I can't understand how to this SMPS can
work without load. When I try to switch on SMPS without load, on low
(start) frequency my transformer comes to saturation. I understand why
this coming but I can't understand what I can do for prevent this. Or
resonant SMPS can't work without load physically?

With load all ok.

The Tektronix 7904 o'scope has a series-resonant switchiing supply,and an
excellent circuit description of how it works.It runs off-resonance,max
power is developed AT resonance.It does need a minimum load,or it goes into
"burst" mode,where the PS chirps on and off.
(the PS has several regulating and current-sense loops)


IMO,no switcher likes to run unloaded.
--
Jim Yanik
jyanik-at-kua.net

 

[legg]

On Thu, 29 Apr 2004 16:15:59 +0000 (UTC), Jim Yanik <jyanik@abuse.gov>
wrote:

artem_bond@mail.ru (artem_bond) wrote in
news:4bea06d1.0404290018.35fdeae6@posting.google.com:


The Tektronix 7904 o'scope has a series-resonant switchiing supply,and an
excellent circuit description of how it works.It runs off-resonance,max
power is developed AT resonance.It does need a minimum load,or it goes into
"burst" mode,where the PS chirps on and off.
(the PS has several regulating and current-sense loops)


IMO,no switcher likes to run unloaded.

Parallel resonance generally offers better independance from loading
effects. Series resonance tends to offer better dynamic response. The
characteristics of both are often combined to give partial benefits of
both.

I shouldn't think that a tube scope would have much trouble presenting
a hefty minimum load. When you try to migrate this to semiconductor
versions, without suitable adaptation, you'd probably get this
chirping.

Intermitent operation is only really benign in damped circuits; in a
resonant converter it invites chaotic behavior. This is not
automatically destructive, but it is difficult to characterize for
stress level compliance under all conditions.

RL

 

[Jim Yanik]

legg <legg@nospam.magma.ca> wrote in
news:73n2901fpi46toklepj53kjnikpk925st2@4ax.com:

On Thu, 29 Apr 2004 16:15:59 +0000 (UTC), Jim Yanik <jyanik@abuse.gov
wrote:

artem_bond@mail.ru (artem_bond) wrote in
news:4bea06d1.0404290018.35fdeae6@posting.google.com:


The Tektronix 7904 o'scope has a series-resonant switchiing supply,and
an excellent circuit description of how it works.It runs
off-resonance,max power is developed AT resonance.It does need a
minimum load,or it goes into "burst" mode,where the PS chirps on and
off. (the PS has several regulating and current-sense loops)


IMO,no switcher likes to run unloaded.

Parallel resonance generally offers better independance from loading
effects. Series resonance tends to offer better dynamic response. The
characteristics of both are often combined to give partial benefits of
both.

I think series resonant PS have better efficiency.

I shouldn't think that a tube scope would have much trouble presenting
a hefty minimum load. When you try to migrate this to semiconductor
versions, without suitable adaptation, you'd probably get this
chirping.

While at TEK,I built dummy loads for most of the 7K series that used
switcher PS,as the supplies unplugged from the mainframes,and it was easier
to TS them with a dummy load.
My DL's weren't perfect subs,but they worked.(pat on back)

Intermitent operation is only really benign in damped circuits; in a
resonant converter it invites chaotic behavior. This is not
automatically destructive, but it is difficult to characterize for
stress level compliance under all conditions.

RL

Curent and voltage limiting keeps the PS from frying the more valuable
scope circuits.

The TEK 7000 series scopes were the best lab scopes of their time,too.
Many of them are still in use.

--
Jim Yanik
jyanik-at-kua.net

 

[artem_bond]

Thank for all.

I'm fixed my problem. And I'm get a resonant at 30kHz frequency. But I
can't get a necessary load range at frequency range from 8 to 30kHz.
Maybe it's can be solved by DSP for changing mode resonant/PWD at
small loads, or something like this but I have no time for this.

Parallel resonance generally offers better independance from loading
effects. Series resonance tends to offer better dynamic response. The
characteristics of both are often combined to give partial benefits of
both.

I shouldn't think that a tube scope would have much trouble presenting
a hefty minimum load. When you try to migrate this to semiconductor
versions, without suitable adaptation, you'd probably get this
chirping.

Intermitent operation is only really benign in damped circuits; in a
resonant converter it invites chaotic behavior. This is not
automatically destructive, but it is difficult to characterize for
stress level compliance under all conditions.

RL

 

[Winfield Hill]

Jim Yanik wrote...

The TEK 7000 series scopes were the best lab scopes of their
time, too. Many of them are still in use.

Still in use, perhaps. I paid big bucks for mine, way back when,
and more recently supplemented it with every conceivable plug-in,
thanks to the Flea at MIT and eBay. Yes it's plugged in, but I'm
sorry, it sits waiting in all its glory off to one side, while I
use the digital scope on my bench. That's life in the big city.


Thanks,
- Win

whill_at_picovolt-dot-com (use hill_at_rowland-dot-org for now)