On Wednesday, January 10, 2024 at 4:02:01 AM UTC+11, Fred Bloggs wrote:
> On Monday, January 8, 2024 at 3:48:49 PM UTC-5, john larkin wrote: 
> > On Sat, 6 Jan 2024 07:42:22 -0800 (PST), Fred Bloggs 
> > <bloggs.fred...@gmail.com> wrote: 
> > >On Friday, January 5, 2024 at 5:13:11?PM UTC-5, john larkin wrote: 
> > >> On Fri, 5 Jan 2024 13:04:59 -0800 (PST), Fred Bloggs 
> > >> <bloggs.fred...@gmail.com> wrote: 
> > >> >On Wednesday, January 3, 2024 at 2:29:19?PM UTC-5, john larkin wrote: 
> > >> >> There is a new-to-me power supply architecture, an H-bridge driving a 
> > >> >> load, but with the phases on the two sides slid around to control 
> > >> >> delivered power. TI does that in some chips, like UCC2895. I may have 
> > >> >> seen the architecture first in this ng. 
> > >> >> 
> > >> >> Anyhow, I was thinking about a high-voltage power supply with the 
> > >> >> phase-shifted bridge driving a series-resonant transformer. 
> > >> >> 
> > >> >> https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1 
> > >> >> 
> > >> >> I wonder if one of the cores of an RP2040 could do this without an 
> > >> >> FPGA. It would of course need a voltage feedback loop in software too. 
> > >> > 
> > >> > 
> > >> >May I ask how much power is being consumed in the 500 V circuit? Since your step-up is ~ 20:1, that would be a reduction of load resistance by a factor of 400 in the primary circuit, in parallel with the resonant inductance. It doesn't look like a hopeful candidate for high Q, which is usually expected when people say resonance. 
> > >> I'm thinking maybe 1.5 or 2 KV at a couple of watts, per isolated 
> > >> channel. 
> > >> > 
> > >> >It does have the advantage of a single ended power supply delivering the same power as a bipolar, which could be a big simplification. And if it had any Q, it does eliminate a lot of broadband spikey stuff that creates radiated emi. So the circuit is not a total dud. 
> > >> High praise! 
> > > 
> > >I'm pretty sure you're going to have insert inductance in the secondary. My first cut/ guess/ wish/ ( no idea iow ) would be to make the L/Requivalent be about 10/w_resonant to achieve a Q of about 10. Doing that should allow your series C + transformer L to actually be resonant and not pulled too much if not totally extinguished by the secondary load. And that spikey capacitor voltage doubler isn't helping the situation. Those hv capacitors won't be cheap, might as well go with the transformer all the way. 
>
> > I'd expect low loaded Q, 5ish or even less maybe. 
> > 
> > A C-W multiplier uses low voltage diodes and caps. That's nice. Maybe 
> > one HV cap at the output. 
> > 
> > We have a 22nf 500v 1206 cap in stock, 13 cents. 22nF 1500V is $1.06, 
> > not awful. 
> > > 
> > >As you're probably aware, the transformer primary voltage will be Q x Vinput at resonance. 

Hardly ever happens. Q tends to be temperature dependent, so nobody wants that.

That could be a good thing if it's reliable, as less step-up is required, or it can be a bad thing if it blows something. 

> > >Does your switching controller sense the resonance condition and modify its frequency to track it? 
> >
> > With low Q, I would use a fixed frequency and use the phase shift 
> > trick to regulate the HV output, with ADCs to provide voltage and 
> > current feedback of course.
>
> After reading up the resonants and their beginnings there were a few standout reasons for going that route: 
> 
> 1. reduction of switching losses by using ZVS/ ZCS techniques made feasible by working with a single resonant frequency sine signal and the ease of synchronizing the H-bridge switch with it, then from this: 
> 
> 1a) reduced switching losses means higher switching frequencies are possible, 
> 
> 1b) elimination of discontinuous switching eliminates a big source of EMI 

The voltage excursions may be smaller, but there are still discontinuiities in current flow which radiate EMI.

> 2) the original instances didn't need a whole lot of dynamic range in their amplitude control, so they would adjust the switching frequency off resonance frequency to get the controlled attenuation needed, a simplified technique at the time, and widely used previously in applications such as FM demodulation, 

It was known as pulse width modulation and goes back a long way.

 E A Faulkner and D W Harding 1966 J. Sci. Instrum. 43 97 is three transistor phase detector from that period.
 
> 3) keeping the highest amplitude signal a more or less spectrally pure sinusoid again bodes well for reduced EMI.

You get lots of odd harmonics at higher frequencies - it's not all that much of a reduction.

You can adjust the drive frequency to track the resonant frequency on the fly - I've never done it but  it seems to be perfectly practical.

http://sophia-electronica.com/Baxandall_parallel-resonant_Class-D_oscillator3.htm

-- 
Bill Sloman, Sydney

On Wednesday, January 10, 2024 at 4:11:47 AM UTC+11, Fred Bloggs wrote:
> On Monday, January 8, 2024 at 9:55:01 PM UTC-5, Anthony William Sloman wrote: 
> > On Sunday, January 7, 2024 at 2:42:31 AM UTC+11, Fred Bloggs wrote: 
> > > On Friday, January 5, 2024 at 5:13:11 PM UTC-5, john larkin wrote: 
> > > > On Fri, 5 Jan 2024 13:04:59 -0800 (PST), Fred Bloggs 
> > > > <bloggs.fred...@gmail.com> wrote: 
> > > > >On Wednesday, January 3, 2024 at 2:29:19?PM UTC-5, john larkin wrote: 
> > > > >> There is a new-to-me power supply architecture, an H-bridge driving a 
> > > > >> load, but with the phases on the two sides slid around to control 
> > > > >> delivered power. TI does that in some chips, like UCC2895. I may have 
> > > > >> seen the architecture first in this ng. 
> > > > >> 
> > > > >> Anyhow, I was thinking about a high-voltage power supply with the 
> > > > >> phase-shifted bridge driving a series-resonant transformer. 
> > > > >> 
> > > > >> https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1 
> > > > >> 
> > > > >> I wonder if one of the cores of an RP2040 could do this without an 
> > > > >> FPGA. It would of course need a voltage feedback loop in software too. 
> > > > > 
> > > > > 
> > > > >May I ask how much power is being consumed in the 500 V circuit? Since your step-up is ~ 20:1, that would be a reduction of load resistance by a factor of 400 in the primary circuit, in parallel with the resonant inductance. It doesn't look like a hopeful candidate for high Q, which is usually expected when people say resonance. 
> > > > I'm thinking maybe 1.5 or 2 KV at a couple of watts, per isolated 
> > > > channel. 
> > > > > 
> > > > >It does have the advantage of a single ended power supply delivering the same power as a bipolar, which could be a big simplification. And if it had any Q, it does eliminate a lot of broadband spikey stuff that creates radiated emi. So the circuit is not a total dud. 
> > > > High praise! 
> > > I'm pretty sure you're going to have insert inductance in the secondary. My first cut/ guess/ wish/ ( no idea iow ) would be to make the L/Requivalent be about 10/w_resonant to achieve a Q of about 10. Doing that should allow your series C + transformer L to actually be resonant and not pulled too much if not totally extinguished by the secondary load. And that spikey capacitor voltage doubler isn't helping the situation. Those hv capacitors won't be cheap, might as well go with the transformer all the way. 
> > This is ill-informed nonsense. With big step-up ratios its hard to get the secondary inductance low enough to keep the resonant frequency high enough to keep filter capacitors small enough to be affordable. Coroft-Walton multipliers let you get away a smaller steep-up ratio in the transformer and less secondary inductance 
> > > As you're probably aware, the transformer primary voltage will be Q x Vinput at resonance. That could be a good thing if it's reliable, as less step-up is required, or it can be a bad thing if it blows something. 
>
> > It won't. In almost all the ones I've worked on it has been pi/2 times the primary supply voltage, with the exception of my current mirror variation on the Baxandall circuit which had controlled voltage feedback to make sure the peak voltage across the primary was more than didoe drop beflow the supply rail 
> > > Does your switching controller sense the resonance condition and modify its frequency to track it? 
>
> > I've never heard of one that does. I once simulated a circuit which used controllable quadrature feedback to keep the oscillation frequency constant despite small shifts in the capacitance and inductance in the resonator, but that didn't get anywhere.
>
> Q of pi/2, really? You're becoming a Baxandall cultist. 

Nothing to do with Q. It's the height of a half-sine wave averaged over the half-cycle. It does come up in the Baxandall circuit, but anybody who knew any math would be aware of it - it's scarcely the hall mark of a cultist (not that Peter Baxandall was the kind of person who had a cult following). The nearest thing to a Baxandall cult I know was the objectivist group in Hi Fi News and Record Review  which also included Peter J Walker who founded Quad Electroacoustic in 1936.

http://www.audiopolitan.com/blog/quad-electroacoustics-the-closest-approach-to-the-original-sound/peter-j-walker-founded-quad-in-1936-audiopolitan/

They were strictly rational, which is about as non-cultist as you can get.

-- 
Bill Sloman, Sydney

On Monday, January 8, 2024 at 9:55:01 PM UTC-5, Anthony William Sloman wrote:
> On Sunday, January 7, 2024 at 2:42:31 AM UTC+11, Fred Bloggs wrote: 
> > On Friday, January 5, 2024 at 5:13:11 PM UTC-5, john larkin wrote: 
> > > On Fri, 5 Jan 2024 13:04:59 -0800 (PST), Fred Bloggs 
> > > <bloggs.fred...@gmail.com> wrote: 
> > > >On Wednesday, January 3, 2024 at 2:29:19?PM UTC-5, john larkin wrote: 
> > > >> There is a new-to-me power supply architecture, an H-bridge driving a 
> > > >> load, but with the phases on the two sides slid around to control 
> > > >> delivered power. TI does that in some chips, like UCC2895. I may have 
> > > >> seen the architecture first in this ng. 
> > > >> 
> > > >> Anyhow, I was thinking about a high-voltage power supply with the 
> > > >> phase-shifted bridge driving a series-resonant transformer. 
> > > >> 
> > > >> https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1 
> > > >> 
> > > >> I wonder if one of the cores of an RP2040 could do this without an 
> > > >> FPGA. It would of course need a voltage feedback loop in software too. 
> > > > 
> > > > 
> > > >May I ask how much power is being consumed in the 500 V circuit? Since your step-up is ~ 20:1, that would be a reduction of load resistance by a factor of 400 in the primary circuit, in parallel with the resonant inductance. It doesn't look like a hopeful candidate for high Q, which is usually expected when people say resonance. 
> > > I'm thinking maybe 1.5 or 2 KV at a couple of watts, per isolated 
> > > channel. 
> > > > 
> > > >It does have the advantage of a single ended power supply delivering the same power as a bipolar, which could be a big simplification. And if it had any Q, it does eliminate a lot of broadband spikey stuff that creates radiated emi. So the circuit is not a total dud. 
> > > High praise! 
> > I'm pretty sure you're going to have insert inductance in the secondary. My first cut/ guess/ wish/ ( no idea iow ) would be to make the L/Requivalent be about 10/w_resonant to achieve a Q of about 10. Doing that should allow your series C + transformer L to actually be resonant and not pulled too much if not totally extinguished by the secondary load. And that spikey capacitor voltage doubler isn't helping the situation. Those hv capacitors won't be cheap, might as well go with the transformer all the way.
> This is ill-informed nonsense. With big step-up ratios its hard to get the secondary inductance low enough to keep the resonant frequency high enough to keep filter capacitors small enough to be affordable. Coroft-Walton multipliers let you get away a smaller steep-up ratio in the transformer and less secondary inductance
> > As you're probably aware, the transformer primary voltage will be Q x Vinput at resonance. That could be a good thing if it's reliable, as less step-up is required, or it can be a bad thing if it blows something.
> It won't. In almost all the ones I've worked on it has been pi/2 times the primary supply voltage, with the exception of my current mirror variation on the Baxandall circuit which had controlled voltage feedback to make sure the peak voltage across the primary was more than didoe drop beflow the supply rail
> > Does your switching controller sense the resonance condition and modify its frequency to track it?
> I've never heard of one that does. I once simulated a circuit which used controllable quadrature feedback to keep the oscillation frequency constant despite small shifts in the capacitance and inductance in the resonator, but that didn't get anywhere.

Q of pi/2, really? You're becoming a Baxandall cultist.

 
> 
> -- 
> Bill Sloman, Sydney

On Monday, January 8, 2024 at 3:48:49 PM UTC-5, john larkin wrote:
> On Sat, 6 Jan 2024 07:42:22 -0800 (PST), Fred Bloggs 
> <bloggs.fred...@gmail.com> wrote:
> >On Friday, January 5, 2024 at 5:13:11?PM UTC-5, john larkin wrote: 
> >> On Fri, 5 Jan 2024 13:04:59 -0800 (PST), Fred Bloggs 
> >> <bloggs.fred...@gmail.com> wrote: 
> >> >On Wednesday, January 3, 2024 at 2:29:19?PM UTC-5, john larkin wrote: 
> >> >> There is a new-to-me power supply architecture, an H-bridge driving a 
> >> >> load, but with the phases on the two sides slid around to control 
> >> >> delivered power. TI does that in some chips, like UCC2895. I may have 
> >> >> seen the architecture first in this ng. 
> >> >> 
> >> >> Anyhow, I was thinking about a high-voltage power supply with the 
> >> >> phase-shifted bridge driving a series-resonant transformer. 
> >> >> 
> >> >> https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1 
> >> >> 
> >> >> I wonder if one of the cores of an RP2040 could do this without an 
> >> >> FPGA. It would of course need a voltage feedback loop in software too. 
> >> > 
> >> > 
> >> >May I ask how much power is being consumed in the 500 V circuit? Since your step-up is ~ 20:1, that would be a reduction of load resistance by a factor of 400 in the primary circuit, in parallel with the resonant inductance. It doesn't look like a hopeful candidate for high Q, which is usually expected when people say resonance. 
> >> I'm thinking maybe 1.5 or 2 KV at a couple of watts, per isolated 
> >> channel. 
> >> > 
> >> >It does have the advantage of a single ended power supply delivering the same power as a bipolar, which could be a big simplification. And if it had any Q, it does eliminate a lot of broadband spikey stuff that creates radiated emi. So the circuit is not a total dud. 
> >> High praise! 
> > 
> >I'm pretty sure you're going to have insert inductance in the secondary. My first cut/ guess/ wish/ ( no idea iow ) would be to make the L/Requivalent be about 10/w_resonant to achieve a Q of about 10. Doing that should allow your series C + transformer L to actually be resonant and not pulled too much if not totally extinguished by the secondary load. And that spikey capacitor voltage doubler isn't helping the situation. Those hv capacitors won't be cheap, might as well go with the transformer all the way.
> I'd expect low loaded Q, 5ish or even less maybe. 
> 
> A C-W multiplier uses low voltage diodes and caps. That's nice. Maybe 
> one HV cap at the output. 
> 
> We have a 22nf 500v 1206 cap in stock, 13 cents. 22nF 1500V is $1.06, 
> not awful.
> > 
> >As you're probably aware, the transformer primary voltage will be Q x Vinput at resonance. That could be a good thing if it's reliable, as less step-up is required, or it can be a bad thing if it blows something. 
> > 
> >Does your switching controller sense the resonance condition and modify its frequency to track it?
> With low Q, I would use a fixed frequency and use the phase shift 
> trick to regulate the HV output, with ADCs to provide voltage and 
> current feedback of course.

After reading up the resonants and their beginnings there were a few standout reasons for going that route:

1. reduction of switching losses by using ZVS/ ZCS techniques made feasible by working with a single resonant frequency sine signal and the ease of synchronizing the H-bridge switch with it, then from this:

1a) reduced switching losses means higher switching frequencies are possible,

1b) elimination of discontinuous switching eliminates a big source of EMI

2) the original instances didn't need a whole lot of dynamic range in their amplitude control, so they would adjust the switching frequency off resonance frequency to get the controlled attenuation needed, a simplified technique at the time, and widely used previously in applications such as FM demodulation,

3) keeping the highest amplitude signal a more or less spectrally pure sinusoid again bodes well for reduced EMI.

On a sunny day (Mon, 08 Jan 2024 12:38:13 -0800) it happened john larkin
<jl@650pot.com> wrote in <ermopip67fn4u4hpl1jv3rgq7uam7jq2fm@4ax.com>:

>On Sat, 06 Jan 2024 06:02:38 GMT, Jan Panteltje <alien@comet.invalid>
>wrote:
>
>>On a sunny day (Fri, 05 Jan 2024 14:09:56 -0800) it happened john larkin
>><jl@650pot.com> wrote in <t5ugpilo3r4tg912oif2av49872u0ej8ob@4ax.com>:
>>
>>>On Fri, 05 Jan 2024 08:47:03 GMT, Jan Panteltje <alien@comet.invalid>
>>>wrote:
>>>
>>>>On a sunny day (Thu, 04 Jan 2024 20:59:16 -0800) it happened John Larkin
>>>><jl@997PotHill.com> wrote in <in2fpitju1b2163tpb5qhmhs97qeufvu4m@4ax.com>:
>>>>
>>>>>On Fri, 05 Jan 2024 04:18:40 GMT, Jan Panteltje
>>>>><pNaonStpealmtje@yahoo.com> wrote:
>>>>>
>>>>>>On a sunny day (Wed, 03 Jan 2024 11:29:02 -0800) it happened john larkin
>>>>>><jl@650pot.com> wrote in <olcbpihia58uugpv9g3ptumif14qru2007@4ax.com>:
>>>>>>
>>>>>>>
>>>>>>>There is a new-to-me power supply architecture, an H-bridge driving a
>>>>>>>load, but with the phases on the two sides slid around to control
>>>>>>>delivered power. TI does that in some chips, like UCC2895. I may have
>>>>>>>seen the architecture first in this ng.
>>>>>>>
>>>>>>>Anyhow, I was thinking about a high-voltage power supply with the
>>>>>>>phase-shifted bridge driving a series-resonant transformer.
>>>>>>>
>>>>>>>https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1
>>>>>>
>>>>>>Is it not much simpler to drive a capacitor with L to ground from a complementary output stage?
>>>>>>
>>>>>> +
>>>>>> |
>>>>>> T1
>>>>>> |      C1      D1
>>>>>> |------||-------|>|----------
>>>>>> |          |          |
>>>>>> T2         L1        ===
>>>>>> |          |         --- C2
>>>>>> :          |          | 
>>>>>>///        ///        ///
>>>>>>
>>>>>>Now Q of C1 L1 and load sets the voltage, no transformer needed.
>>>>>>Simple PIC micro can drive that.
>>>>>
>>>>>The transformer provides step-up and isolation. 
>>>>
>>>>That circuit has both sides connected to ground?
>>>>Or is that not a ground on V1 negative supply?
>>>
>>>My 24 volt supply will be grounded, but I want the HV output channels
>>>to float. 
>>>
>>>Are you commenting on the quality of my hand-drawn schematic?
>>>
>>>
>>>>
>>>>>I kind of like the
>>>>>full-bridge drive, to put 48 volts p-p into the transformer primary
>>>>>circuit.
>>>>
>>>>Yes, all depends on the amount of power needed...
>>>
>>>Not a lot, so I may be able to use that cute little IXYS driver.
>>>
>>>>
>>>>
>>>>>>
>>>>>>>I wonder if one of the cores of an RP2040 could do this without an
>>>>>>>FPGA. It would of course need a voltage feedback loop in software too.
>>>>>>
>>>>>>PIC micro.
>>>>>>Has hardware PWM generator and comparators that can directly switch the PWM off.
>>>>>>Used many times.
>>>>>
>>>>>I'm plannng to use RP2040 in this product line.
>>>>
>>>>If more tasks are needed and that RP2040 has a hardware facility to make PWM yes
>>>>If not using a PIC is simple, interrupt increments a counter, sets a pin, again counter resets a pin.
>>>>Or use the build in PWM generator.
>>>>Main routine reads ADCs sets counter values. drives an LCD, etc...
>>>>The 18F14K22 has 4 ADCs, a hardware comparator (for cycle by cycle current liming for example), 64 MHz internal PLL clock..
>>>>This uses the PIC's PWM generator:
>>>> https://panteltje.online/panteltje/pic/pwr_pic/power_box_diagram_img_1817.jpg
>>>>use ctrl+ in browser to enlarge and read circuit diagram.
>>>>I use a current transformer to sense transistor current and trigger the PIC hardware comparator
>>>>making the cycle by cycle urrent protection.
>>>>The PIC ADCs read current and voltage and display it on an LCD, no interrupst anywhere needed.
>>>> http://panteltje.online/panteltje/pic/pwr_pic/
>>>>
>>>>You could reduce diode voltage drop, think I got the idea here in this group?
>>>> https://panteltje.online/pub/power_pic/power_pic_synchronous_rectifier_diagram_img_0968.jpg
>>>> https://panteltje.online/pub/power_pic/power_pic_synchronous_rectifier_working_img_0965.jpg
>>>
>>>It will probably need a C-W voltage multiplier after the transformer,
>>>depending on the transformer I can find. I'm thinking 1500 volts out
>>>per channel, maybe 2KV. Diode drops won't matter at HV, low current.
>>
>>Dunno what he power requirement is, this worked for me for my PMT:
>> https://panteltje.online/pub/PMT_regulated_power_supply_diagram_img_3182.jpg
>> https://panteltje.online/pub/PMT_HV_supply_with_regulator_img_3175.jpg
>> https://panteltje.online/pub/PMT_HV_generator_solder_side_img_3172.jpg
>
>Yeah, something like that but switchmode with voltage and current
>sense ADCs and a digital control loop.  It's still just an idea. I
>have a couple of nice boxes and need stuff to go in them.
>
>
>>
>>Or just get an old TV voltage multiplier:
>> https://panteltje.nl/panteltje/pic/sc_pic/sc_pic_pcb_test_in_box_img_2452.jpg
>>
>>
>>>A C-W multiplier lets me use cheap SOT-23 diodes and reasonable caps,
>>>to.
>>>
>>>I'm planning a new product line and was just toying with possible
>>>boxes. A dual HV supply might be fun.
>>
>>There are likely many possibilities, I usually just use what I find laying about..
>>Or get cheap stuff from ebay...
>>Old CRT monitors had nice transformers, rectifiers, 15625 kHz 25 kV
>>Must be millions still available...
>>I keep am old CRT color monitor in the attic as my own personal particle accelerator :-)
>>Tried all sorts of things:
>> https://panteltje.online/pub/multi_transformer_2_kV_PMT_supply_img_3126.jpg
>>And there is the very high voltage ebay thing:
>> https://www.ebay.com/b/High-Voltage-Generator/117000/bn_7023272732
>>cannot beat the price, I have one :-)
>
>Does it really make 400 KV?

I tried to measure it, but no way could I see that much voltage.
Still fun, its in a box somewhere:-)


>I design stuff to be manufractured and sold, preferably surface mount,
>so I only use ebay or amazon parts for breadboards.

That is what China does too, hundreds of these HV generators are manufactured and sold.

If you scrolled down on that ebay link or search a bit more: there is plenty of HV generating stuff.
You can always learn from their designs
Look at that 1 Ge transistor circuit HV generator for the radiation monitoring modules.
That was MIL standard, and still works up to this day.
Cannot go into what I did for Plactre.




>

On Mon, 8 Jan 2024 17:01:05 -0800 (PST), whit3rd <whit3rd@gmail.com>
wrote:

>On Monday, January 8, 2024 at 12:48:49?PM UTC-8, john larkin wrote:
>
>> A C-W multiplier uses low voltage diodes and caps. That's nice. Maybe 
>> one HV cap at the output. 
>> 
>> We have a 22nf 500v 1206 cap in stock, 13 cents. 22nF 1500V is $1.06, 
>> not awful.
>
>Oh, the story doesn't end there.   The capacitor for 1500V isn't
>awful, but have you priced a 1500V connector?   
>

Yes, the connector problem is worse than finding a transformer. SHV
ain't too bad, but I'd prefer one with two isolated contacts. Needs
research.

I've used BNCs and RG59 at 10 KV, but that wouldn't be popular. SMAs
are actually fine at 2KV, ditto.

The Phoenix wire-screw connectors are a possibility. Skip some pins
maybe.    V- skip GND skip V+    on a couple of 5-pins maybe.



>It's not an OUTPUT filter capacitor until you have a connector.
>
>Me, I've got a ceramic disk capacitor rated for 45kV, and 
>attaching endcaps to it requires a machine shop to carve
>up big hunks of aluminum into spherical-like blobs.

Not surface mount?

On Sunday, January 7, 2024 at 2:42:31 AM UTC+11, Fred Bloggs wrote:
> On Friday, January 5, 2024 at 5:13:11 PM UTC-5, john larkin wrote: 
> > On Fri, 5 Jan 2024 13:04:59 -0800 (PST), Fred Bloggs 
> > <bloggs.fred...@gmail.com> wrote: 
> > >On Wednesday, January 3, 2024 at 2:29:19?PM UTC-5, john larkin wrote: 
> > >> There is a new-to-me power supply architecture, an H-bridge driving a 
> > >> load, but with the phases on the two sides slid around to control 
> > >> delivered power. TI does that in some chips, like UCC2895. I may have 
> > >> seen the architecture first in this ng. 
> > >> 
> > >> Anyhow, I was thinking about a high-voltage power supply with the 
> > >> phase-shifted bridge driving a series-resonant transformer. 
> > >> 
> > >> https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1 
> > >> 
> > >> I wonder if one of the cores of an RP2040 could do this without an 
> > >> FPGA. It would of course need a voltage feedback loop in software too. 
> > > 
> > > 
> > >May I ask how much power is being consumed in the 500 V circuit? Since your step-up is ~ 20:1, that would be a reduction of load resistance by a factor of 400 in the primary circuit, in parallel with the resonant inductance. It doesn't look like a hopeful candidate for high Q, which is usually expected when people say resonance. 
> > I'm thinking maybe 1.5 or 2 KV at a couple of watts, per isolated 
> > channel. 
> > > 
> > >It does have the advantage of a single ended power supply delivering the same power as a bipolar, which could be a big simplification. And if it had any Q, it does eliminate a lot of broadband spikey stuff that creates radiated emi. So the circuit is not a total dud. 
> > High praise!
> I'm pretty sure you're going to have insert inductance in the secondary. My first cut/ guess/ wish/ ( no idea iow ) would be to make the L/Requivalent be about 10/w_resonant to achieve a Q of about 10. Doing that should allow your series C + transformer L to actually be resonant and not pulled too much if not totally extinguished by the secondary load. And that spikey capacitor voltage doubler isn't helping the situation. Those hv capacitors won't be cheap, might as well go with the transformer all the way. 

This is ill-informed nonsense. With big step-up ratios its hard to get the secondary inductance low enough to keep the resonant frequency high enough to keep filter capacitors small enough to be affordable. Coroft-Walton multipliers let you get away a smaller steep-up ratio in the transformer and less secondary inductance
 
> As you're probably aware, the transformer primary voltage will be Q x Vinput at resonance. That could be a good thing if it's reliable, as less step-up is required, or it can be a bad thing if it blows something. 

It won't. In almost all the ones I've worked on it has been pi/2 times the primary supply voltage, with the exception of my current mirror variation on the Baxandall circuit which had controlled voltage feedback to make sure the peak voltage across the primary was more than didoe drop beflow the supply rail

> Does your switching controller sense the resonance condition and modify its frequency to track it?

I've never heard of one that does. I once simulated a circuit which used controllable quadrature feedback to keep the oscillation frequency constant despite small shifts in the capacitance and inductance in the resonator, but that didn't get anywhere.

-- 
Bill Sloman, Sydney

On Monday, January 8, 2024 at 12:48:49 PM UTC-8, john larkin wrote:

> A C-W multiplier uses low voltage diodes and caps. That's nice. Maybe 
> one HV cap at the output. 
> 
> We have a 22nf 500v 1206 cap in stock, 13 cents. 22nF 1500V is $1.06, 
> not awful.

Oh, the story doesn't end there.   The capacitor for 1500V isn't
awful, but have you priced a 1500V connector?   

It's not an OUTPUT filter capacitor until you have a connector.

Me, I've got a ceramic disk capacitor rated for 45kV, and 
attaching endcaps to it requires a machine shop to carve
up big hunks of aluminum into spherical-like blobs.

On Sat, 6 Jan 2024 07:42:22 -0800 (PST), Fred Bloggs
<bloggs.fredbloggs.fred@gmail.com> wrote:

>On Friday, January 5, 2024 at 5:13:11?PM UTC-5, john larkin wrote:
>> On Fri, 5 Jan 2024 13:04:59 -0800 (PST), Fred Bloggs 
>> <bloggs.fred...@gmail.com> wrote:
>> >On Wednesday, January 3, 2024 at 2:29:19?PM UTC-5, john larkin wrote: 
>> >> There is a new-to-me power supply architecture, an H-bridge driving a 
>> >> load, but with the phases on the two sides slid around to control 
>> >> delivered power. TI does that in some chips, like UCC2895. I may have 
>> >> seen the architecture first in this ng. 
>> >> 
>> >> Anyhow, I was thinking about a high-voltage power supply with the 
>> >> phase-shifted bridge driving a series-resonant transformer. 
>> >> 
>> >> https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1 
>> >> 
>> >> I wonder if one of the cores of an RP2040 could do this without an 
>> >> FPGA. It would of course need a voltage feedback loop in software too. 
>> > 
>> > 
>> >May I ask how much power is being consumed in the 500 V circuit? Since your step-up is ~ 20:1, that would be a reduction of load resistance by a factor of 400 in the primary circuit, in parallel with the resonant inductance. It doesn't look like a hopeful candidate for high Q, which is usually expected when people say resonance.
>> I'm thinking maybe 1.5 or 2 KV at a couple of watts, per isolated 
>> channel.
>> > 
>> >It does have the advantage of a single ended power supply delivering the same power as a bipolar, which could be a big simplification. And if it had any Q, it does eliminate a lot of broadband spikey stuff that creates radiated emi. So the circuit is not a total dud.
>> High praise!
>
>I'm pretty sure you're going to have insert inductance in the secondary. My first cut/ guess/ wish/  ( no idea iow ) would be to make the L/Requivalent be about 10/w_resonant to achieve a Q of about 10. Doing that should allow your series C + transformer L to actually be resonant and not pulled too much if not totally extinguished by the secondary load. And that spikey capacitor voltage doubler isn't helping the situation. Those hv capacitors won't be cheap, might as well go with the transformer all the way.

I'd expect low loaded Q, 5ish or even less  maybe. 

A C-W multiplier uses low voltage diodes and caps. That's nice. Maybe
one HV cap at the output.

We have a 22nf 500v 1206 cap in stock, 13 cents. 22nF 1500V is $1.06,
not awful.

>
>As you're probably aware, the transformer primary voltage will be Q x Vinput at resonance. That could be a good thing if it's reliable, as less step-up is required, or it can be a bad thing if it blows something.
>
>Does your switching controller sense the resonance condition and modify its frequency to track it?

With low Q, I would use a fixed frequency and use the phase shift
trick to regulate the HV output, with ADCs to provide voltage and
current feedback of course.

On Sat, 06 Jan 2024 06:02:38 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

>On a sunny day (Fri, 05 Jan 2024 14:09:56 -0800) it happened john larkin
><jl@650pot.com> wrote in <t5ugpilo3r4tg912oif2av49872u0ej8ob@4ax.com>:
>
>>On Fri, 05 Jan 2024 08:47:03 GMT, Jan Panteltje <alien@comet.invalid>
>>wrote:
>>
>>>On a sunny day (Thu, 04 Jan 2024 20:59:16 -0800) it happened John Larkin
>>><jl@997PotHill.com> wrote in <in2fpitju1b2163tpb5qhmhs97qeufvu4m@4ax.com>:
>>>
>>>>On Fri, 05 Jan 2024 04:18:40 GMT, Jan Panteltje
>>>><pNaonStpealmtje@yahoo.com> wrote:
>>>>
>>>>>On a sunny day (Wed, 03 Jan 2024 11:29:02 -0800) it happened john larkin
>>>>><jl@650pot.com> wrote in <olcbpihia58uugpv9g3ptumif14qru2007@4ax.com>:
>>>>>
>>>>>>
>>>>>>There is a new-to-me power supply architecture, an H-bridge driving a
>>>>>>load, but with the phases on the two sides slid around to control
>>>>>>delivered power. TI does that in some chips, like UCC2895. I may have
>>>>>>seen the architecture first in this ng.
>>>>>>
>>>>>>Anyhow, I was thinking about a high-voltage power supply with the
>>>>>>phase-shifted bridge driving a series-resonant transformer.
>>>>>>
>>>>>>https://www.dropbox.com/scl/fi/tbiioti0gt2emknjc61p7/Res_HV_Supply.jpg?rlkey=f1gryr8vr4jdu2y46wsz6ch2k&raw=1
>>>>>
>>>>>Is it not much simpler to drive a capacitor with L to ground from a complementary output stage?
>>>>>
>>>>> +
>>>>> |
>>>>> T1
>>>>> |      C1      D1
>>>>> |------||-------|>|----------
>>>>> |          |          |
>>>>> T2         L1        ===
>>>>> |          |         --- C2
>>>>> :          |          | 
>>>>>///        ///        ///
>>>>>
>>>>>Now Q of C1 L1 and load sets the voltage, no transformer needed.
>>>>>Simple PIC micro can drive that.
>>>>
>>>>The transformer provides step-up and isolation. 
>>>
>>>That circuit has both sides connected to ground?
>>>Or is that not a ground on V1 negative supply?
>>
>>My 24 volt supply will be grounded, but I want the HV output channels
>>to float. 
>>
>>Are you commenting on the quality of my hand-drawn schematic?
>>
>>
>>>
>>>>I kind of like the
>>>>full-bridge drive, to put 48 volts p-p into the transformer primary
>>>>circuit.
>>>
>>>Yes, all depends on the amount of power needed...
>>
>>Not a lot, so I may be able to use that cute little IXYS driver.
>>
>>>
>>>
>>>>>
>>>>>>I wonder if one of the cores of an RP2040 could do this without an
>>>>>>FPGA. It would of course need a voltage feedback loop in software too.
>>>>>
>>>>>PIC micro.
>>>>>Has hardware PWM generator and comparators that can directly switch the PWM off.
>>>>>Used many times.
>>>>
>>>>I'm plannng to use RP2040 in this product line.
>>>
>>>If more tasks are needed and that RP2040 has a hardware facility to make PWM yes
>>>If not using a PIC is simple, interrupt increments a counter, sets a pin, again counter resets a pin.
>>>Or use the build in PWM generator.
>>>Main routine reads ADCs sets counter values. drives an LCD, etc...
>>>The 18F14K22 has 4 ADCs, a hardware comparator (for cycle by cycle current liming for example), 64 MHz internal PLL clock..
>>>This uses the PIC's PWM generator:
>>> https://panteltje.online/panteltje/pic/pwr_pic/power_box_diagram_img_1817.jpg
>>>use ctrl+ in browser to enlarge and read circuit diagram.
>>>I use a current transformer to sense transistor current and trigger the PIC hardware comparator
>>>making the cycle by cycle urrent protection.
>>>The PIC ADCs read current and voltage and display it on an LCD, no interrupst anywhere needed.
>>> http://panteltje.online/panteltje/pic/pwr_pic/
>>>
>>>You could reduce diode voltage drop, think I got the idea here in this group?
>>> https://panteltje.online/pub/power_pic/power_pic_synchronous_rectifier_diagram_img_0968.jpg
>>> https://panteltje.online/pub/power_pic/power_pic_synchronous_rectifier_working_img_0965.jpg
>>
>>It will probably need a C-W voltage multiplier after the transformer,
>>depending on the transformer I can find. I'm thinking 1500 volts out
>>per channel, maybe 2KV. Diode drops won't matter at HV, low current.
>
>Dunno what he power requirement is, this worked for me for my PMT:
> https://panteltje.online/pub/PMT_regulated_power_supply_diagram_img_3182.jpg
> https://panteltje.online/pub/PMT_HV_supply_with_regulator_img_3175.jpg
> https://panteltje.online/pub/PMT_HV_generator_solder_side_img_3172.jpg

Yeah, something like that but switchmode with voltage and current
sense ADCs and a digital control loop.  It's still just an idea. I
have a couple of nice boxes and need stuff to go in them.


>
>Or just get an old TV voltage multiplier:
> https://panteltje.nl/panteltje/pic/sc_pic/sc_pic_pcb_test_in_box_img_2452.jpg
>
>
>>A C-W multiplier lets me use cheap SOT-23 diodes and reasonable caps,
>>to.
>>
>>I'm planning a new product line and was just toying with possible
>>boxes. A dual HV supply might be fun.
>
>There are likely many possibilities, I usually just use what I find laying about..
>Or get cheap stuff from ebay...
>Old CRT monitors had nice transformers, rectifiers, 15625 kHz 25 kV
>Must be millions still available...
>I keep am old CRT color monitor in the attic as my own personal particle accelerator :-)
>Tried all sorts of things:
> https://panteltje.online/pub/multi_transformer_2_kV_PMT_supply_img_3126.jpg
>And there is the very high voltage ebay thing:
> https://www.ebay.com/b/High-Voltage-Generator/117000/bn_7023272732
>cannot beat the price, I have one :-)

Does it really make 400 KV?

I design stuff to be manufractured and sold, preferably surface mount,
so I only use ebay or amazon parts for breadboards.