Gentlemen,

Is there a switcher topology where the incoming mains supply is *not*
rectified before being fed into the chopper transistor? IOW, AC mains
is fed to some kind of PNPN triac-esque device which performs the
chopping?

CD

Cursitor Doom <cd@notformail.com> wrote:
> Gentlemen,
>
> Is there a switcher topology where the incoming mains supply is *not*
> rectified before being fed into the chopper transistor? IOW, AC mains
> is fed to some kind of PNPN triac-esque device which performs the
> chopping?
>
> CD
>

BITD hybrid bridge rectifiers were used occasionally. SCR pre-regulators on
the primary side of a bog-standard transformer-rectifier supply also work,
sorta.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC /
Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics

onsdag den 27. december 2023 kl. 16.03.18 UTC+1 skrev Phil Hobbs:
> Cursitor Doom <c...@notformail.com> wrote:
> > Gentlemen,
> >
> > Is there a switcher topology where the incoming mains supply is *not*
> > rectified before being fed into the chopper transistor? IOW, AC mains
> > is fed to some kind of PNPN triac-esque device which performs the
> > chopping?
> >
> > CD
> >
> BITD hybrid bridge rectifiers were used occasionally. SCR pre-regulators on
> the primary side of a bog-standard transformer-rectifier supply also work,
> sorta.

like the Carver amplifiers that used a triac "dimmer" on what looks like a 50VA
transformer to deliver kWs

On Wed, 27 Dec 2023 15:03:11 -0000 (UTC), Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>Cursitor Doom <cd@notformail.com> wrote:
>> Gentlemen,
>>
>> Is there a switcher topology where the incoming mains supply is *not*
>> rectified before being fed into the chopper transistor? IOW, AC mains
>> is fed to some kind of PNPN triac-esque device which performs the
>> chopping?
>>
>> CD
>>
>
>BITD hybrid bridge rectifiers were used occasionally. SCR pre-regulators on
>the primary side of a bog-standard transformer-rectifier supply also work,
>sorta.
>
>Cheers
>
>Phil Hobbs

I did a bunch of power amps that had a giant resistor with a shunt
triac on the transformer primary as a 1-step pre-regulator. It tended
to kill wirewould resistors from thermal fatigue, but the Welwyn
porcelain-steel resistors fixed that.

I don't recall seeing a real switcher that ran off line AC. It doesn't
sound economical.

John Larkin <jl@997PotHill.com> wrote:
> On Wed, 27 Dec 2023 15:03:11 -0000 (UTC), Phil Hobbs
> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>
>> Cursitor Doom <cd@notformail.com> wrote:
>>> Gentlemen,
>>>
>>> Is there a switcher topology where the incoming mains supply is *not*
>>> rectified before being fed into the chopper transistor? IOW, AC mains
>>> is fed to some kind of PNPN triac-esque device which performs the
>>> chopping?
>>>
>>> CD
>>>
>>
>> BITD hybrid bridge rectifiers were used occasionally. SCR pre-regulators on
>> the primary side of a bog-standard transformer-rectifier supply also work,
>> sorta.
>>
>> Cheers
>>
>> Phil Hobbs
>
> I did a bunch of power amps that had a giant resistor with a shunt
> triac on the transformer primary as a 1-step pre-regulator. It tended
> to kill wirewould resistors from thermal fatigue, but the Welwyn
> porcelain-steel resistors fixed that.
>
> I don't recall seeing a real switcher that ran off line AC. It doesn't
> sound economical.
>
>

I’ve never designed one, but I recall seeing SCR bridges as the HV
preregulator in an argon laser supply.

For regulating medium amounts of AC, the regular way was a ferroresonant
transformer (Sola).

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC /
Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics

On Wed, 27 Dec 2023 10:32:40 -0800 (PST), whit3rd <whit3rd@gmail.com>
wrote:

>On Wednesday, December 27, 2023 at 8:26:29?AM UTC-8, John Larkin wrote:
>
>> >Cursitor Doom <c...@notformail.com> wrote:
>> >> Gentlemen,
>> >>
>> >> Is there a switcher topology where the incoming mains supply is *not*
>> >> rectified before being fed into the chopper transistor? IOW, AC mains
>> >> is fed to some kind of PNPN triac-esque device which performs the
>> >> chopping?
>
>> I don't recall seeing a real switcher that ran off line AC. It doesn't
>> sound economical.
>
>Well, yeah, I've seen one; it was OLD and that was thirty or forty years ago.
>The 100VDC for some lecture halls/labs (arc lamp power for projectors?) was
>generated by an AC motor driving a DC (switch mode because it
>was commutated) generator. 100VDC output, stabilized by
>feedback to the stator current.
>
>Hey, old auto radios with vibrators were switch mode power supplies, too!

And some synchronous rectified, too.

There were cool regulators and amps before power semis came along.
MG-sets, amplidynes, metadynes, carbon piles, magamps, thyratrons,
ignitrons, variacs.

On Wed, 27 Dec 2023 14:49:55 +0000, Cursitor Doom <cd@notformail.com>
wrote:

>Gentlemen,
>
>Is there a switcher topology where the incoming mains supply is *not*
>rectified before being fed into the chopper transistor? IOW, AC mains
>is fed to some kind of PNPN triac-esque device which performs the
>chopping?
>
>CD

There are some situations where the 'chopper' and 'bridge' are
placed in series with the AC line and load. Rectification still
present in most practical circuits, unless simple phase angle
controls or saturable reactor distortion (and weight) is
permissible.

There's not much point in high frequency conversion, unless
it avoids the size and weight of lower frequency components
at the regulator, isolation barrier or the load itself.

Topologies, as such, don't necessarily require real components or
practical circuits - they are conceptual.

RL

On 27/12/2023 14:49, Cursitor Doom wrote:
> Gentlemen,
>
> Is there a switcher topology where the incoming mains supply is *not*
> rectified before being fed into the chopper transistor? IOW, AC mains
> is fed to some kind of PNPN triac-esque device which performs the
> chopping?
>
> CD

Yes. Some so called true sine stage dimmers and smart solar PV immersion
heater controls build a kind of all silicon variac. Think a buck
switcher but where the series switch transistor and catch diode are
replaced by inverse series mosfets. Buck inductor and smoothing
capacitor are sized for kHz pwm switch rate but output is 50/60Hz
reduced amplitude version of the input supply.

piglet

Lasse Langwadt Christensen wrote:
=============================> Phil Hobbs:
>
> > BITD hybrid bridge rectifiers were used occasionally. SCR pre-regulators on
> > the primary side of a bog-standard transformer-rectifier supply also work,
> > sorta.
> like the Carver amplifiers that used a triac "dimmer" on what looks like a 50VA
> transformer to deliver kWs
>

** The PSU used in most Carver amplifierts was *very dodgy*.
The triac pre-regulator acted to stabilise the DC rail voltages derived from the way undersized AC transformer allowing it to deliver useful power levels. But not for long, rated audio power output could only be sustained for a few minutes before serious overheating began.
The scheme relied on the peak to average ratio of music or speech being at least 10 dB or so.
The use of multi level DC supplies aided by greatly improving the efficiency of the class B output stage.

...... Phil

On a sunny day (Wed, 27 Dec 2023 11:49:29 -0800) it happened John Larkin
<jl@997PotHill.com> wrote in <ifvooidah3tlmkl8mc1bh8mgmoms61q1f6@4ax.com>:

>On Wed, 27 Dec 2023 10:32:40 -0800 (PST), whit3rd <whit3rd@gmail.com>
>wrote:
>
>>On Wednesday, December 27, 2023 at 8:26:29?AM UTC-8, John Larkin wrote:
>>
>>> >Cursitor Doom <c...@notformail.com> wrote:
>>> >> Gentlemen,
>>> >>
>>> >> Is there a switcher topology where the incoming mains supply is *not*
>>> >> rectified before being fed into the chopper transistor? IOW, AC mains
>>> >> is fed to some kind of PNPN triac-esque device which performs the
>>> >> chopping?
>>
>>> I don't recall seeing a real switcher that ran off line AC. It doesn't
>>> sound economical.
>>
>>Well, yeah, I've seen one; it was OLD and that was thirty or forty years ago.
>>The 100VDC for some lecture halls/labs (arc lamp power for projectors?) was
>>generated by an AC motor driving a DC (switch mode because it
>>was commutated) generator. 100VDC output, stabilized by
>>feedback to the stator current.
>>
>>Hey, old auto radios with vibrators were switch mode power supplies, too!
>
>And some synchronous rectified, too.
>
>There were cool regulators and amps before power semis came along.
>MG-sets, amplidynes, metadynes, carbon piles, magamps, thyratrons,
>ignitrons, variacs.

Transductors driven by thyristors regulator stuff I worked on in the sixties
hundreds of amps at 48 V output for example.
Huge transductors and transformers 3 phase..

On Wed, 27 Dec 2023 14:49:55 +0000, Cursitor Doom <cd@notformail.com>
wrote:

>Gentlemen,
>
>Is there a switcher topology where the incoming mains supply is *not*
>rectified before being fed into the chopper transistor? IOW, AC mains
>is fed to some kind of PNPN triac-esque device which performs the
>chopping?
>
>CD

Grid tie chargers connected to the line synchronize to the grid and
turn into a boost converter to PFC charge batteries in the opposite
direction using full wave bridge switchers and inductors tied to the
AC line itself.

There is more circuitry of course.

Full wave diode bridges work pretty well too. Just not PFC corrected.

boB

On 2023-12-27, Cursitor Doom <cd@notformail.com> wrote:
> Is there a switcher topology where the incoming mains supply is *not*
> rectified before being fed into the chopper transistor? IOW, AC mains
> is fed to some kind of PNPN triac-esque device which performs the
> chopping?

Does "bridgeless PFC" meet the requirements?

https://toshiba.semicon-storage.com/info/docget.jsp?did=68570
chapter 3.3

https://www.ti.com/lit/pdf/slua517

This replaces two or all four of the diodes in a rectifier bridge with
transistors, and places an inductor between line input and the bridge, thus
integrating bridge rectifier and step-up switcher in one block.

cu
Michael
--
Some people have no respect of age unless it is bottled.

On Wednesday, January 17, 2024 at 12:47:47 AM UTC+11, Michael Schwingen wrote:
> On 2023-12-27, Cursitor Doom <c...@notformail.com> wrote:
> > Is there a switcher topology where the incoming mains supply is *not*
> > rectified before being fed into the chopper transistor? IOW, AC mains
> > is fed to some kind of PNPN triac-esque device which performs the
> > chopping?
> Does "bridgeless PFC" meet the requirements?
>
> https://toshiba.semicon-storage.com/info/docget.jsp?did=68570
> chapter 3.3
>
> https://www.ti.com/lit/pdf/slua517
>
>
> This replaces two or all four of the diodes in a rectifier bridge with
> transistors, and places an inductor between line input and the bridge, thus
> integrating bridge rectifier and step-up switcher in one block.

It might, but Cursitor Doom doesn't know much and understands less.

He was asking for " AC mains fed to some kind of PNPN triac-esque device which performs the
chopping?" and the bridgeless circuits rely on MOSFETs to do the chopping
and the power factor correction - which he doesn't seem to have thought about.

--
Bill Sloman, Sydney

On 16 Jan 2024 13:47:39 GMT, Michael Schwingen
<news-1513678000@discworld.dascon.de> wrote:

>On 2023-12-27, Cursitor Doom <cd@notformail.com> wrote:
>> Is there a switcher topology where the incoming mains supply is *not*
>> rectified before being fed into the chopper transistor? IOW, AC mains
>> is fed to some kind of PNPN triac-esque device which performs the
>> chopping?
>
>Does "bridgeless PFC" meet the requirements?
>
>https://toshiba.semicon-storage.com/info/docget.jsp?did=68570
>chapter 3.3
>
>https://www.ti.com/lit/pdf/slua517

Fig 8 says "bidgeless" when there is a diode bridge in plain sight.

Looks like a ton of complexity at little advantage.

On 2024-01-16, John Larkin <jl@997PotHill.com> wrote:
>>
>>https://www.ti.com/lit/pdf/slua517
>
>
> Fig 8 says "bidgeless" when there is a diode bridge in plain sight.

Replacing the remaining 2 diodes with FETs is an options. I believe the term
"bridgeless" is supposed to mean "no separate bridge in front of the PFC
stage".

> Looks like a ton of complexity at little advantage.

Figure 3 in the TI Appnote has a breakdown of losses for both topologies -
the bridgeless version seems to improve efficiency. Figure 9 shows the
resulting efficiency. For compact designs where cooling is the limit (think
laptop chargers etc.), this may be significant.

cu
Michael
--
Some people have no respect of age unless it is bottled.

On Wednesday, January 17, 2024 at 3:21:20 AM UTC+11, John Larkin wrote:
> On 16 Jan 2024 13:47:39 GMT, Michael Schwingen
> <news-15...@discworld.dascon.de> wrote:
>
> >On 2023-12-27, Cursitor Doom <c...@notformail.com> wrote:
> >> Is there a switcher topology where the incoming mains supply is *not*
> >> rectified before being fed into the chopper transistor? IOW, AC mains
> >> is fed to some kind of PNPN triac-esque device which performs the
> >> chopping?
> >
> >Does "bridgeless PFC" meet the requirements?
> >
> >https://toshiba.semicon-storage.com/info/docget.jsp?did=68570
> >chapter 3.3
> >
> >https://www.ti.com/lit/pdf/slua517
> Fig 8 says "bidgeless" when there is a diode bridge in plain sight.
>
> Looks like a ton of complexity at little advantage.

Power factor correction is obligatory for all but the smallest of mains driven power supplies.

Not getting prosecuted for selling a non-compliant power supply may strike you as a minor advantage, but some people take their legal obligations more seriously.

--
Bill Sloman, Sydney

On Wednesday 27 December 2023 at 14:50:03 UTC, Cursitor Doom wrote:
> Gentlemen,
>
> Is there a switcher topology where the incoming mains supply is *not*
> rectified before being fed into the chopper transistor? IOW, AC mains
> is fed to some kind of PNPN triac-esque device which performs the
> chopping?
>
> CD

It's a standard approach to improving smpsu efficiency. I don't know of any using PNPN devices, modern hf smpsus want fast switching, not latching on. These psus require 2 switching devices to drive the transformer, so no cost saving.

On Wed, 17 Jan 2024 22:54:38 -0800 (PST), Tabby <tabbypurr@gmail.com>
wrote:

>On Wednesday 27 December 2023 at 14:50:03 UTC, Cursitor Doom wrote:
>> Gentlemen,
>>
>> Is there a switcher topology where the incoming mains supply is *not*
>> rectified before being fed into the chopper transistor? IOW, AC mains
>> is fed to some kind of PNPN triac-esque device which performs the
>> chopping?
>>
>> CD
>
>It's a standard approach to improving smpsu efficiency. I don't know of any using PNPN devices, modern hf smpsus want fast switching, not latching on. These psus require 2 switching devices to drive the transformer, so no cost saving.

Yes, isolated gate drives reduce the appeal of many switcher
topologies. We recently designed a programmable ac/dc dummy load
module and tried to use an "ideal bridge" followed by a big mosfet as
the active load device. The gate divers shot that idea down. There are
a few active-bridge driver ICs, but they are too slow.

The trend lately is a plain bridge rectifier off the AC line, a filter
cap, and some sort of GaN switcher running at high frquency.

On 2024-01-18, John Larkin <jl@997PotHill.com> wrote:
> The trend lately is a plain bridge rectifier off the AC line, a filter
> cap, and some sort of GaN switcher running at high frquency.

Only up to 75W (?) - above that, PFC is mandatory, at least in the EU.

cu
Michael
--
Some people have no respect of age unless it is bottled.

Michael Schwingen wrote:
------------------------------------------
>
> Only up to 75W (?) - above that, PFC is mandatory, at least in the EU.
>>

** No way is using PFC mandatory.
The EU standard covering harmonic currents is chock full of exceptions and limitations.
Except for catagories like PCs, electronic lighting and TV sets, near everything else is allowed to continue to be sold with PSUs made the same way they have long been.

See this article:

https://www.emcstandards.co.uk/files/61000-3-2_mains_harmonics.pdf

........ Phil

On Friday, January 19, 2024 at 3:20:04 PM UTC+11, Phil Allison wrote:
> Michael Schwingen wrote:
> ------------------------------------------
> >
> > Only up to 75W (?) - above that, PFC is mandatory, at least in the EU.
> >>
> ** No way is using PFC mandatory.
> The EU standard covering harmonic currents is chock full of exceptions and limitations.
> Except for catagories like PCs, electronic lighting and TV sets, near everything else is allowed to continue to be sold with PSUs made the same way they have long been.
>
> See this article:
>
> https://www.emcstandards.co.uk/files/61000-3-2_mains_harmonics.pdf

It points up the exceptions - every standards committee is infested with industry representatives who write in exceptions for their own industrial products. They are one of the reasons why sitting on standards committee is a chore to be avoided.

Anything that squirts out lots of higher harmonics peeves the people who have been silly enough to buy it, so it makes sense to conform, even if there is a handy loophole. You will sell more product in the end.

--
Bill Sloman, Sydney

On 2024-01-19, Phil Allison <pallison49@gmail.com> wrote:
> Michael Schwingen wrote:
> ------------------------------------------
>>
>> Only up to 75W (?) - above that, PFC is mandatory, at least in the EU.
>
> ** No way is using PFC mandatory.
> The EU standard covering harmonic currents is chock full of exceptions and limitations.
> Except for catagories like PCs, electronic lighting and TV sets, near everything else is allowed to continue to be sold with PSUs made the same way they have long been.

Sorry for the generalization. Yes, there are exceptions, but most devices
above a certain power limit nowadays *do* require PFC.

The exceptions are getting fewer with every new version of the standard.

https://www.lisungroup.com/news/technology-news/detailed-explanation-of-en61000-3-22019-limits-for-harmonic-current-emissions-standard.html
https://www.lisungroup.com/wp-content/uploads/2021/07/BSENIEC61000-3-2-2019.pdf

has a newer version than the document you linked.

There are different classes which have different limits, but few exceptions
remain (see chapter 7). Above 75W (or 5W for lighting), those are basically
phase control dimmers, heating elements and professional equipment >1kW.

There is a clause for "professional equipment" that does not comply to the
limits, but that includes "requirement to ask the supply utility for
permission to connect", so that is no easy way out.

I have not looked at the exact limits for each class, but I assume it is not
possible to meet them without some kind of PFC - correct me if I am wrong.

cu
Michael
--
Some people have no respect of age unless it is bottled.

Michael Schwingen wrote:
----------------------------------------
> Phil Allison
>>
> >> Only up to 75W (?) - above that, PFC is mandatory, at least in the EU.
> >
> > ** No way is using PFC mandatory.
> > The EU standard covering harmonic currents is chock full of exceptions and limitations.
> > Except for catagories like PCs, electronic lighting and TV sets, near everything else is allowed to continue to be sold with PSUs made the same way they have long been.

** Near everything else = class A items. It's a huge category.

> I have not looked at the exact limits for each class, but I assume it is not
> possible to meet them without some kind of PFC - correct me if I am wrong.

** You need to look carefully at the " permissible harmonic current " numbers cited in the table for class A.

Only a very small percentage of such items exhibit harmonic currents exceeding those values.
Note that the steady rms current drawn from the supply would have to be approach the 16A limit for there to be a risk.

...... Phil

On 2024-01-21, Phil Allison <pallison49@gmail.com> wrote:
> ** Near everything else = class A items. It's a huge category.
>
>> I have not looked at the exact limits for each class, but I assume it is not
>> possible to meet them without some kind of PFC - correct me if I am wrong.
>
> ** You need to look carefully at the " permissible harmonic current " numbers cited in the table for class A.
> Only a very small percentage of such items exhibit harmonic currents exceeding those values.
> Note that the steady rms current drawn from the supply would have to be approach the 16A limit for there to be a risk.

Ok, I see - *that* extends the range of the exceptions quite a bit, thanks.

cu
Michael
--
Some people have no respect of age unless it is bottled.

On Sat, 20 Jan 2024 23:13:51 -0800 (PST), Phil Allison
<pallison49@gmail.com> wrote:

>Michael Schwingen wrote:
>----------------------------------------
>> Phil Allison
>>>
>> >> Only up to 75W (?) - above that, PFC is mandatory, at least in the EU.
>> >
>> > ** No way is using PFC mandatory.
>> > The EU standard covering harmonic currents is chock full of exceptions and limitations.
>> > Except for catagories like PCs, electronic lighting and TV sets, near everything else is allowed to continue to be sold with PSUs made the same way they have long been.
>
>** Near everything else = class A items. It's a huge category.
>
>> I have not looked at the exact limits for each class, but I assume it is not
>> possible to meet them without some kind of PFC - correct me if I am wrong.
>
>** You need to look carefully at the " permissible harmonic current " numbers cited in the table for class A.
>
> Only a very small percentage of such items exhibit harmonic currents exceeding those values.
> Note that the steady rms current drawn from the supply would have to be approach the 16A limit for there to be a risk.
>
>
>..... Phil

Testing at lower current levels will pro-rate the 16A values as a
ratio.

What it avoids is phase displacement characteristic of motors.

That's why Power Factor isn't in the actual standards.

RL