On 03-10-2023 14:37, John Larkin wrote:> On Tue, 3 Oct 2023 09:46:44 +0200, Klaus Vestergaard Kragelund > <klauskvik@hotmail.com> wrote: > >> On 29-09-2023 04:09, John Larkin wrote: >>> On Thu, 28 Sep 2023 13:58:40 -0700 (PDT), Klaus Kragelund >>> <klaus.kragelund@gmail.com> wrote: >>> >>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: >>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> >>>>> wrote: >>>>> >>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus >>>>>> Kragelund <klaus.k...@gmail.com> wrote in >>>>>> <903144ac-6642-4a3f...@googlegroups.com>: >>>>>> >>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: >>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin >>>>>>>> <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: >>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin >>>>>>>>>> <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: >>>>>>>>>> >>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> >>>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin >>>>>>>>>>>> <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: >>>>>>>>>>>> >>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> >>>>>>>>>>>>> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> This is pleasingly weird. >>>>>>>>>>>>>> >>>>>>>>>>>>>> https://www.dropbox.com/sh/sif3efs69dxe1mg/AACY0RJGXl4k8CVvauUbJtYFa?dl=0 >>>>>>>>>>>>>> >>>>>>>>>>>>>> Sort of a baseline-boosted multi-auto-transformer voltage-doubler >>>>>>>>>>>>>> flyback. >>>>>>>>>>>>>> >>>>>>>>>>>>>> What's strange is that adding the two snubbers increases the LT spice >>>>>>>>>>>>>> sim speed radically, about 10:1. >>>>>>>>>>>>> >>>>>>>>>>>>> Since I couldn't find a suitable flyback transformer, and I only need >>>>>>>>>>>>> 500 volts at 10 mA, a flyback/C-W multiplier makes sense. >>>>>>>>>>>>> >>>>>>>>>>>>> https://www.dropbox.com/scl/fi/aquzpawfn8mfmfsnx70ra/T875_HV_6.jpg?rlkey=vfs9pexa0rgyxpo7pg1grqlb7&dl=0 >>>>>>>>>>>> >>>>>>>>>>>> Do you still need C11 R8 in that circuit? >>>>>>>>>>> >>>>>>>>>>> It doesn't need them to work, but there is a big high-Q ring when the >>>>>>>>>>> fet turns off, and the RC damps it. It would reduce radiated EMI. And >>>>>>>>>>> it speeds up the simulation some. >>>>>>>>>> >>>>>>>>>> Did you scope that? >>>>>>>>> >>>>>>>>> All Spice now. I'll build it when the customer gets serious. >>>>>>>>> >>>>>>>>>> There is a load of capacitance hanging from the MOSFET drain in that flyback multiplier ciruit >>>>>>>>>> I would expect it to damp RF oscillations but could be wrong... >>>>>>>>>> HV diodes have capacitance too.. >>>>>>>>> >>>>>>>>> The real circuit will have a lot of inductances too, what with those >>>>>>>>> long strings of parts. It needs to be built and tested. Dremeled, to >>>>>>>>> start. >>>>>>>>> >>>>>>>>> Here's an old auto-flyback C-W miltiplier prototype. I think that was >>>>>>>>> 1400 volts. Worked fine. >>>>>>>>> >>>>>>>>> https://www.dropbox.com/s/yd19osiwz1z74s4/HV_Proto_2.JPG?rsw=1 >>>>>>>> Yea, >>>>>>>> I did a PMT supply this way: >>>>>>>> http://panteltje.nl/pub/PMT_HV_supply_with_regulator_img_3175.jpg >>>>>>>> http://panteltje.nl/pub/PMT_HV_supply_componet_side_img_3180.jpg >>>>>>>> http://panteltje.nl/pub/PMT_regulated_power_supply_diagram_img_3182.jpg >>>>>>>> >>>>>>>> It has been in use now for many years, sits with PMT in the cardboard tube on the right, runs on batteries: >>>>>>>> http://panteltje.nl/pub/gamma_soectrometer_IMG_4505.JPG >>>>>>>> >>>>>>>> Of course that is very low power, but for a bit more power, and using a real TV muliplier: >>>>>>>> https://panteltje.nl/panteltje/pic/sc_pic/ >>>>>>>> >>>>>>>> No dremeling, boards are OK when they work.. >>>>>>>> >>>>>>>> And old TV circuits.. if you want a few kV there was / is? plenty stuff to be found on ebay. >>>>>>> For my case, I am working on a 5W 24V to 4.5kV converter. Right now looking to use a flyback topology, with possibly usage of >>>>>>> CCFL transformer bobbins. >>>>>>> >>>>>>> I am thinking about staggering the windings on 2 transformers, so the voltage across the windings will be lower, but still the >>>>>>> individual transformer needs to handle 4.5kV insulation (functional, not safety) >>>>>>> >>>>>>> Due to the limiting power, it could also be tempting to use the capacitive doublers as listed elsewhere in the thread >>>>>>> >>>>>>> My worry goes beyond the design, also about corona effects, which comes into play at +kV voltages. Potting is surely needed. >>>>>> >>>>>> For a HV transformer, old TV transformers (from the CRT times) may perhaps be an option: >>>>>> https://www.google.com/search?client=firefox-b-e&q=ebay+horizonatl+TV+output+transformer >>>>>> 4.5 kV diodes were used for focus in color TV sets... >>>>>> >>>>>> >>>>>> For a flyback, my rule of thumb (I do not use El Tea Spice), is 1 V per turn for DC input. >>>>>> So 24 turns primary for 24 V DC supply. >>>>>> The flyback voltage is (70/12) * 24 = 140V positive pulse if positive DC supply. >>>>>> To get a 4.5 kV pulse you would then wind 4500 / 140 is about 32, multiplied by 24 makes 768 turns. >>>>>> This for an EI core running at about 15 kHz (math comes from building TV output stages). >>>>>> 768 turns is not a big deal.. >>>>>> Spreading the turns over a segmented coil former is a good idea. >>>>>> If you want to use a voltage multiplier then less turns are needed, use a TV multiplier, those are 'potted' >>>>>> The 70/12 factor depends on the PWM on/off ratio of the drive you feed into the switch transistor. >>>>>> I use Microchip PICs for that, it has a PWM unit., hardware comparators that can directly stop the PWM drive >>>>>> for cycle by cycle current limiting, and ADCs to measure voltage and currents. >>>>>> Maybe an old TV horizontal output transformer with build in rectifier / multiplier would do. >>>>>> https://www.ebay.com/b/flyback-transformer/bn_7024744628 >>>>>> to adapt use 15 kHz and add 24 turns to the core primary? >>>>>> ... >>>>>> :-) >>>>>> >>>>>> Maybe like this? >>>>>> https://www.ebay.com/p/1692562663?iid=374326473542 >>>>>> plenty space for 24 turns thick wire, its a 30 kV DC out, so use less PWM ... or more primary turns >>>>>> or whatever you can find... on ebay. >>>>>> >>>>> High turns ratios can have two problems: voltage breakdown in the >>>>> insulation, and winding capacitance limiting the flyback ratio. That's >>>>> where a C-W diode multiplier helps. >>>> >>>> I am driving a UV lamp, so I need to deliver a pulse. Possible with a multiplier, but doesn't make it easier >>> >>> No, the C-W multiplier isn't good for pulsing. Ixys has a 4700 volt >>> mosfet. A C-W multiplier could make the HV. >>> >>> Maybe use a Marx generator? With medium-high voltage mos or SiC fets? >>> >>> Some of the Zetex app notes, for their avalanche transistors, use a >>> Marx-type circuit. Charge the caps in parallel and discharge them in >>> series. You only have to trigger one in the string. >>> >>> What sort of speed and current do you need? >>> >>> >> I will generate a single pulse, about 5us long, shaped like a half wave >> sinusoid, at 100kHz repetition frequency. That's why I was looking at >> flyback, since I can control the voltage delivered simply by adjusting >> the primary peak current. Power seems to be less than 5W, at 4.5kV peak, >> duty of let's say 25%. probably below 10mA > > One can stuff a bunch of current into an inductor and then dump it > into a mostly-capacitive load. That can make a high-voltage half-sine. > > This boosts 48 volts to 1400v at MHz pulse rates. > > http://www.highlandtechnology.com/DSS/T850DS.shtml > > by switching an inductor around. The energy is recovered and returned > to the 48v supply after each pulse, which keeps the power dissipation > under control. > > https://www.dropbox.com/scl/fi/lkjg19hvcus9xz8aeecxy/T850_Rev_C.JPG?rlkey=0pitv0zdmb0x1tz0fkw0gj3ob&raw=1 > > The really tricky stuff is on the bottom of the board, dumping heat > into the water-cooled baseplate. > > At 5 us, you could make a 4KV DC supply and switch that into your load > with a high-voltage mosfet. Brute force. >Yes, that could also be an idea. Will have some losses though, since it's switching into a capacitive load
500 volt power supply
Started by ●September 24, 2023
Reply by ●October 3, 20232023-10-03
Reply by ●October 3, 20232023-10-03
On Wednesday, October 4, 2023 at 1:41:37 AM UTC+11, Klaus Vestergaard Kragelund wrote:> On 03-10-2023 14:37, John Larkin wrote: > > On Tue, 3 Oct 2023 09:46:44 +0200, Klaus Vestergaard Kragelund <klau...@hotmail.com> wrote: > >> On 29-09-2023 04:09, John Larkin wrote: > >>> On Thu, 28 Sep 2023 13:58:40 -0700 (PDT), Klaus Kragelund <klaus.k...@gmail.com> wrote: > >>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: > >>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> wrote: > >>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus > >>>>>> Kragelund <klaus.k...@gmail.com> wrote in <903144ac-6642-4a3f...@googlegroups.com>: > >>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: > >>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin > >>>>>>>> <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: > >>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> wrote: > >>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: > >>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> wrote: > >>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin > >>>>>>>>>>>> <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: > >>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> wrote:<snip>> > One can stuff a bunch of current into an inductor and then dump it into a mostly-capacitive load. That can make a high-voltage half-sine.A transformer would work better - perhaps several of them in series. Dump a bunch of current into the primaries in parallel - perhaps time shifted to allow the high voltage pulse to propagate through the series connected secondaries - and rely on the parallel capacitances of the secondaries to shape the pulse into a half-sine. An inductor to store a load of energy has to be a lot bulkier than a transformer which just couples it into a high impedance load. -- Bill Sloman, Sydney
Reply by ●October 3, 20232023-10-03
On Tue, 3 Oct 2023 16:41:27 +0200, Klaus Vestergaard Kragelund <klauskvik@hotmail.com> wrote:>On 03-10-2023 14:37, John Larkin wrote: >> On Tue, 3 Oct 2023 09:46:44 +0200, Klaus Vestergaard Kragelund >> <klauskvik@hotmail.com> wrote: >> >>> On 29-09-2023 04:09, John Larkin wrote: >>>> On Thu, 28 Sep 2023 13:58:40 -0700 (PDT), Klaus Kragelund >>>> <klaus.kragelund@gmail.com> wrote: >>>> >>>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: >>>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> >>>>>> wrote: >>>>>> >>>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus >>>>>>> Kragelund <klaus.k...@gmail.com> wrote in >>>>>>> <903144ac-6642-4a3f...@googlegroups.com>: >>>>>>> >>>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: >>>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin >>>>>>>>> <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: >>>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> >>>>>>>>>> wrote: >>>>>>>>>> >>>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin >>>>>>>>>>> <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: >>>>>>>>>>> >>>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> >>>>>>>>>>>> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin >>>>>>>>>>>>> <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: >>>>>>>>>>>>> >>>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> >>>>>>>>>>>>>> wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> This is pleasingly weird. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> https://www.dropbox.com/sh/sif3efs69dxe1mg/AACY0RJGXl4k8CVvauUbJtYFa?dl=0 >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Sort of a baseline-boosted multi-auto-transformer voltage-doubler >>>>>>>>>>>>>>> flyback. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> What's strange is that adding the two snubbers increases the LT spice >>>>>>>>>>>>>>> sim speed radically, about 10:1. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Since I couldn't find a suitable flyback transformer, and I only need >>>>>>>>>>>>>> 500 volts at 10 mA, a flyback/C-W multiplier makes sense. >>>>>>>>>>>>>> >>>>>>>>>>>>>> https://www.dropbox.com/scl/fi/aquzpawfn8mfmfsnx70ra/T875_HV_6.jpg?rlkey=vfs9pexa0rgyxpo7pg1grqlb7&dl=0 >>>>>>>>>>>>> >>>>>>>>>>>>> Do you still need C11 R8 in that circuit? >>>>>>>>>>>> >>>>>>>>>>>> It doesn't need them to work, but there is a big high-Q ring when the >>>>>>>>>>>> fet turns off, and the RC damps it. It would reduce radiated EMI. And >>>>>>>>>>>> it speeds up the simulation some. >>>>>>>>>>> >>>>>>>>>>> Did you scope that? >>>>>>>>>> >>>>>>>>>> All Spice now. I'll build it when the customer gets serious. >>>>>>>>>> >>>>>>>>>>> There is a load of capacitance hanging from the MOSFET drain in that flyback multiplier ciruit >>>>>>>>>>> I would expect it to damp RF oscillations but could be wrong... >>>>>>>>>>> HV diodes have capacitance too.. >>>>>>>>>> >>>>>>>>>> The real circuit will have a lot of inductances too, what with those >>>>>>>>>> long strings of parts. It needs to be built and tested. Dremeled, to >>>>>>>>>> start. >>>>>>>>>> >>>>>>>>>> Here's an old auto-flyback C-W miltiplier prototype. I think that was >>>>>>>>>> 1400 volts. Worked fine. >>>>>>>>>> >>>>>>>>>> https://www.dropbox.com/s/yd19osiwz1z74s4/HV_Proto_2.JPG?rsw=1 >>>>>>>>> Yea, >>>>>>>>> I did a PMT supply this way: >>>>>>>>> http://panteltje.nl/pub/PMT_HV_supply_with_regulator_img_3175.jpg >>>>>>>>> http://panteltje.nl/pub/PMT_HV_supply_componet_side_img_3180.jpg >>>>>>>>> http://panteltje.nl/pub/PMT_regulated_power_supply_diagram_img_3182.jpg >>>>>>>>> >>>>>>>>> It has been in use now for many years, sits with PMT in the cardboard tube on the right, runs on batteries: >>>>>>>>> http://panteltje.nl/pub/gamma_soectrometer_IMG_4505.JPG >>>>>>>>> >>>>>>>>> Of course that is very low power, but for a bit more power, and using a real TV muliplier: >>>>>>>>> https://panteltje.nl/panteltje/pic/sc_pic/ >>>>>>>>> >>>>>>>>> No dremeling, boards are OK when they work.. >>>>>>>>> >>>>>>>>> And old TV circuits.. if you want a few kV there was / is? plenty stuff to be found on ebay. >>>>>>>> For my case, I am working on a 5W 24V to 4.5kV converter. Right now looking to use a flyback topology, with possibly usage of >>>>>>>> CCFL transformer bobbins. >>>>>>>> >>>>>>>> I am thinking about staggering the windings on 2 transformers, so the voltage across the windings will be lower, but still the >>>>>>>> individual transformer needs to handle 4.5kV insulation (functional, not safety) >>>>>>>> >>>>>>>> Due to the limiting power, it could also be tempting to use the capacitive doublers as listed elsewhere in the thread >>>>>>>> >>>>>>>> My worry goes beyond the design, also about corona effects, which comes into play at +kV voltages. Potting is surely needed. >>>>>>> >>>>>>> For a HV transformer, old TV transformers (from the CRT times) may perhaps be an option: >>>>>>> https://www.google.com/search?client=firefox-b-e&q=ebay+horizonatl+TV+output+transformer >>>>>>> 4.5 kV diodes were used for focus in color TV sets... >>>>>>> >>>>>>> >>>>>>> For a flyback, my rule of thumb (I do not use El Tea Spice), is 1 V per turn for DC input. >>>>>>> So 24 turns primary for 24 V DC supply. >>>>>>> The flyback voltage is (70/12) * 24 = 140V positive pulse if positive DC supply. >>>>>>> To get a 4.5 kV pulse you would then wind 4500 / 140 is about 32, multiplied by 24 makes 768 turns. >>>>>>> This for an EI core running at about 15 kHz (math comes from building TV output stages). >>>>>>> 768 turns is not a big deal.. >>>>>>> Spreading the turns over a segmented coil former is a good idea. >>>>>>> If you want to use a voltage multiplier then less turns are needed, use a TV multiplier, those are 'potted' >>>>>>> The 70/12 factor depends on the PWM on/off ratio of the drive you feed into the switch transistor. >>>>>>> I use Microchip PICs for that, it has a PWM unit., hardware comparators that can directly stop the PWM drive >>>>>>> for cycle by cycle current limiting, and ADCs to measure voltage and currents. >>>>>>> Maybe an old TV horizontal output transformer with build in rectifier / multiplier would do. >>>>>>> https://www.ebay.com/b/flyback-transformer/bn_7024744628 >>>>>>> to adapt use 15 kHz and add 24 turns to the core primary? >>>>>>> ... >>>>>>> :-) >>>>>>> >>>>>>> Maybe like this? >>>>>>> https://www.ebay.com/p/1692562663?iid=374326473542 >>>>>>> plenty space for 24 turns thick wire, its a 30 kV DC out, so use less PWM ... or more primary turns >>>>>>> or whatever you can find... on ebay. >>>>>>> >>>>>> High turns ratios can have two problems: voltage breakdown in the >>>>>> insulation, and winding capacitance limiting the flyback ratio. That's >>>>>> where a C-W diode multiplier helps. >>>>> >>>>> I am driving a UV lamp, so I need to deliver a pulse. Possible with a multiplier, but doesn't make it easier >>>> >>>> No, the C-W multiplier isn't good for pulsing. Ixys has a 4700 volt >>>> mosfet. A C-W multiplier could make the HV. >>>> >>>> Maybe use a Marx generator? With medium-high voltage mos or SiC fets? >>>> >>>> Some of the Zetex app notes, for their avalanche transistors, use a >>>> Marx-type circuit. Charge the caps in parallel and discharge them in >>>> series. You only have to trigger one in the string. >>>> >>>> What sort of speed and current do you need? >>>> >>>> >>> I will generate a single pulse, about 5us long, shaped like a half wave >>> sinusoid, at 100kHz repetition frequency. That's why I was looking at >>> flyback, since I can control the voltage delivered simply by adjusting >>> the primary peak current. Power seems to be less than 5W, at 4.5kV peak, >>> duty of let's say 25%. probably below 10mA >> >> One can stuff a bunch of current into an inductor and then dump it >> into a mostly-capacitive load. That can make a high-voltage half-sine. >> >> This boosts 48 volts to 1400v at MHz pulse rates. >> >> http://www.highlandtechnology.com/DSS/T850DS.shtml >> >> by switching an inductor around. The energy is recovered and returned >> to the 48v supply after each pulse, which keeps the power dissipation >> under control. >> >> https://www.dropbox.com/scl/fi/lkjg19hvcus9xz8aeecxy/T850_Rev_C.JPG?rlkey=0pitv0zdmb0x1tz0fkw0gj3ob&raw=1 >> >> The really tricky stuff is on the bottom of the board, dumping heat >> into the water-cooled baseplate. >> >> At 5 us, you could make a 4KV DC supply and switch that into your load >> with a high-voltage mosfet. Brute force. >> >Yes, that could also be an idea. Will have some losses though, since >it's switching into a capacitive loadSwitching a DC supply into a capacitive load is inherently lossy, and people usually discharge the cap and waste that energy, too. So power dissipation is C*F*V^2. The inductor trick has, ideally, zero power dissipation... it just sloshes energy around. That's how I got 4 MHz out of a tiny Pockels Cell driver. If F is low, it wouldn't matter much.
Reply by ●October 4, 20232023-10-04
On Wednesday, October 4, 2023 at 3:00:46 AM UTC+11, John Larkin wrote:> On Tue, 3 Oct 2023 16:41:27 +0200, Klaus Vestergaard Kragelund<klau...@hotmail.com> wrote: > >On 03-10-2023 14:37, John Larkin wrote: > >> On Tue, 3 Oct 2023 09:46:44 +0200, Klaus Vestergaard Kragelund <klau...@hotmail.com> wrote: > >>> On 29-09-2023 04:09, John Larkin wrote: > >>>> On Thu, 28 Sep 2023 13:58:40 -0700 (PDT), Klaus Kragelund <klaus.k...@gmail.com> wrote: > >>>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: > >>>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> wrote: > >>>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus Kragelund <klaus.k...@gmail.com> wrote in <903144ac-6642-4a3f...@googlegroups.com>: > >>>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: > >>>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: > >>>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> wrote: > >>>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: > >>>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> wrote: > >>>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: > >>>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> wrote:<snip>> >>> I will generate a single pulse, about 5us long, shaped like a half wave sinusoid, at 100kHz repetition frequency. That's why I was looking at flyback, since I can control the voltage delivered simply by adjusting the primary peak current. Power seems to be less than 5W, at 4.5kV peak, duty of let's say 25%. probably below 10mA > >> > >> One can stuff a bunch of current into an inductor and then dump it into a mostly-capacitive load. That can make a high-voltage half-sine. > >> > >> This boosts 48 volts to 1400v at MHz pulse rates. > >> > >> http://www.highlandtechnology.com/DSS/T850DS.shtml > >> > >> by switching an inductor around. The energy is recovered and returned to the 48v supply after each pulse, which keeps the power dissipation under control. > >> > >> https://www.dropbox.com/scl/fi/lkjg19hvcus9xz8aeecxy/T850_Rev_C.JPG?rlkey=0pitv0zdmb0x1tz0fkw0gj3ob&raw=1 > >> > >> The really tricky stuff is on the bottom of the board, dumping heat into the water-cooled baseplate. > >> > >> At 5 us, you could make a 4KV DC supply and switch that into your load with a high-voltage mosfet. Brute force. > >> > >Yes, that could also be an idea. Will have some losses though, since it's switching into a capacitive load. > > Switching a DC supply into a capacitive load is inherently lossy, and people usually discharge the cap and waste that energy, too.It isn't, if the load hasn't got any resistance. Discharging the capacitor into a separate supply isn't clever.> So power dissipation is C*F*V^2. The inductor trick has, ideally, zero power dissipation... it just sloshes energy around. That's how I got 4 MHz out of a tiny Pockels Cell driver.What on earth is "F"? Klaus seems to want to charge a 10pF load to 2.5kV peak, as a half cycle of 100kHz, getting it from 0V up to 2.5kV and back to 0V over 5usec. If he sets up a 0.25H transformer secondary and charges it up with 10mA for about 2usec waits a usec and then discharges it at 10mA for another 2usec, he'd get close. Figuring on a 16:1 step=up from about 100V, and 10V per turn, that would be a 160 turn secondary, which implies a core with an Al of 10uH per root turn, which is bit high. Going to two transformers in series lets you drop the step-up to 8:1. Each secondary now needs to be O.125H which is an Al of 20uH. Three times more turns fixes that, but you are stuck with an 240 T secondary on each transformer and you don't want the parallel capacitance of the secondaries to get close to the 10pF in your load. If you allow 20pF across each secondary - 10pF in parallel with the load your 0,25H tuning inductance has to drop to 0.18H which makes life easier. -- Bill Sloman, Sydney
Reply by ●October 4, 20232023-10-04
On Tue, 3 Oct 2023 14:45:43 +0200, Klaus Vestergaard Kragelund <klauskvik@hotmail.com> wrote:>On 30-09-2023 21:10, John Larkin wrote: >> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <jl@997arbor.com> >> wrote: >> >>> >>> This is pleasingly weird. >>> >>> https://www.dropbox.com/sh/sif3efs69dxe1mg/AACY0RJGXl4k8CVvauUbJtYFa?dl=0 >>> >>> Sort of a baseline-boosted multi-auto-transformer voltage-doubler >>> flyback. >>> >>> What's strange is that adding the two snubbers increases the LT spice >>> sim speed radically, about 10:1. >> >> OK, this has become sort of a background hobby, waiting on the >> customer to get serious. >> >> https://www.dropbox.com/scl/fi/60nb1pcrnl17c07tk8xeu/T875_HV_9.jpg?rlkey=9uv466edcuvddmjldgepcyi0u&raw=1 >> >> All the caps are a part that we have, 100nF 630 volts 1812. Of course >> the capacitance plummets as voltage goes up, so the Spice values have >> to be adjusted appropriately. C is only 20nF at 500 volts. >> >> The RCD snubber is nice. It softens the rise at the fet drain and >> dissipates less power than a simple RC that damps ringing about the >> same. >> >Isn't the RCD clamp done wrong? > >Normally you do it like this: > >https://www.ridleyengineering.com/images/SPM/11/article11_02.jpg >It's actually a current snubber, not a voltage clamp. Conventionally, for uncomplicated function, it's applied across the switch that is turning off. In Larkin's application, it will only function as well as layout strays and supply line decoupling allow. RL
Reply by ●October 4, 20232023-10-04
On Tuesday, October 3, 2023 at 10:30:48 PM UTC-5, Anthony William Sloman wrote:> On Wednesday, October 4, 2023 at 3:00:46 AM UTC+11, John Larkin wrote: > > On Tue, 3 Oct 2023 16:41:27 +0200, Klaus Vestergaard Kragelund<klau...@hotmail.com> wrote: > > >On 03-10-2023 14:37, John Larkin wrote: > > >> On Tue, 3 Oct 2023 09:46:44 +0200, Klaus Vestergaard Kragelund <klau...@hotmail.com> wrote: > > >>> On 29-09-2023 04:09, John Larkin wrote: > > >>>> On Thu, 28 Sep 2023 13:58:40 -0700 (PDT), Klaus Kragelund <klaus.k...@gmail.com> wrote: > > >>>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: > > >>>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> wrote: > > >>>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus Kragelund <klaus.k...@gmail.com> wrote in <903144ac-6642-4a3f...@googlegroups.com>: > > >>>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: > > >>>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: > > >>>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> wrote: > > >>>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: > > >>>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> wrote: > > >>>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: > > >>>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> wrote: > <snip> > > >>> I will generate a single pulse, about 5us long, shaped like a half wave sinusoid, at 100kHz repetition frequency. That's why I was looking at flyback, since I can control the voltage delivered simply by adjusting the primary peak current. Power seems to be less than 5W, at 4.5kV peak, duty of let's say 25%. probably below 10mA > > >> > > >> One can stuff a bunch of current into an inductor and then dump it into a mostly-capacitive load. That can make a high-voltage half-sine. > > >> > > >> This boosts 48 volts to 1400v at MHz pulse rates. > > >> > > >> http://www.highlandtechnology.com/DSS/T850DS.shtml > > >> > > >> by switching an inductor around. The energy is recovered and returned to the 48v supply after each pulse, which keeps the power dissipation under control. > > >> > > >> https://www.dropbox.com/scl/fi/lkjg19hvcus9xz8aeecxy/T850_Rev_C.JPG?rlkey=0pitv0zdmb0x1tz0fkw0gj3ob&raw=1 > > >> > > >> The really tricky stuff is on the bottom of the board, dumping heat into the water-cooled baseplate. > > >> > > >> At 5 us, you could make a 4KV DC supply and switch that into your load with a high-voltage mosfet. Brute force. > > >> > > >Yes, that could also be an idea. Will have some losses though, since it's switching into a capacitive load. > > > > Switching a DC supply into a capacitive load is inherently lossy, and people usually discharge the cap and waste that energy, too. > It isn't, if the load hasn't got any resistance. Discharging the capacitor into a separate supply isn't clever. > > So power dissipation is C*F*V^2. The inductor trick has, ideally, zero power dissipation... it just sloshes energy around. That's how I got 4 MHz out of a tiny Pockels Cell driver. > What on earth is "F"?To an engineer it means, in this context, frequency. So to you it might mean something else in your fancy vocabulary.> Klaus seems to want to charge a 10pF load to 2.5kV peak, as a half cycle of 100kHz, getting it from 0V up to 2.5kV and back to 0V over 5usec. > > If he sets up a 0.25H transformer secondary and charges it up with 10mA for about 2usec waits a usec and then discharges it at 10mA for another 2usec, he'd get close. > > Figuring on a 16:1 step=up from about 100V, and 10V per turn, that would be a 160 turn secondary, which implies a core with an Al of 10uH per root turn, which is bit high. > > Going to two transformers in series lets you drop the step-up to 8:1. Each secondary now needs to be O.125H which is an Al of 20uH. Three times more turns fixes that, but you are stuck with an 240 T secondary on each transformer and you don't want the parallel capacitance of the secondaries to get close to the 10pF in your load. > > If you allow 20pF across each secondary - 10pF in parallel with the load your 0,25H tuning inductance has to drop to 0.18H which makes life easier. > > -- > Bill Sloman, Sydney
Reply by ●October 5, 20232023-10-05
On Thursday, October 5, 2023 at 2:24:13 AM UTC+11, John Smiht wrote:> On Tuesday, October 3, 2023 at 10:30:48 PM UTC-5, Anthony William Sloman wrote: > > On Wednesday, October 4, 2023 at 3:00:46 AM UTC+11, John Larkin wrote: > > > On Tue, 3 Oct 2023 16:41:27 +0200, Klaus Vestergaard Kragelund<klau...@hotmail.com> wrote: > > > >On 03-10-2023 14:37, John Larkin wrote: > > > >> On Tue, 3 Oct 2023 09:46:44 +0200, Klaus Vestergaard Kragelund <klau...@hotmail.com> wrote: > > > >>> On 29-09-2023 04:09, John Larkin wrote: > > > >>>> On Thu, 28 Sep 2023 13:58:40 -0700 (PDT), Klaus Kragelund <klaus.k...@gmail.com> wrote: > > > >>>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: > > > >>>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> wrote: > > > >>>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus Kragelund <klaus.k...@gmail.com> wrote in <903144ac-6642-4a3f...@googlegroups.com>: > > > >>>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: > > > >>>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: > > > >>>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> wrote: > > > >>>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: > > > >>>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> wrote: > > > >>>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: > > > >>>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> wrote: > > <snip> > > > >>> I will generate a single pulse, about 5us long, shaped like a half wave sinusoid, at 100kHz repetition frequency. That's why I was looking at flyback, since I can control the voltage delivered simply by adjusting the primary peak current. Power seems to be less than 5W, at 4.5kV peak, duty of let's say 25%. probably below 10mA > > > >> > > > >> One can stuff a bunch of current into an inductor and then dump it into a mostly-capacitive load. That can make a high-voltage half-sine. > > > >> > > > >> This boosts 48 volts to 1400v at MHz pulse rates. > > > >> > > > >> http://www.highlandtechnology.com/DSS/T850DS.shtml > > > >> > > > >> by switching an inductor around. The energy is recovered and returned to the 48v supply after each pulse, which keeps the power dissipation under control. > > > >> > > > >> https://www.dropbox.com/scl/fi/lkjg19hvcus9xz8aeecxy/T850_Rev_C.JPG?rlkey=0pitv0zdmb0x1tz0fkw0gj3ob&raw=1 > > > >> > > > >> The really tricky stuff is on the bottom of the board, dumping heat into the water-cooled baseplate. > > > >> > > > >> At 5 us, you could make a 4KV DC supply and switch that into your load with a high-voltage mosfet. Brute force. > > > >> > > > >Yes, that could also be an idea. Will have some losses though, since it's switching into a capacitive load. > > > > > > Switching a DC supply into a capacitive load is inherently lossy, and people usually discharge the cap and waste that energy, too. > > It isn't, if the load hasn't got any resistance. Discharging the capacitor into a separate supply isn't clever. > > > So power dissipation is C*F*V^2. The inductor trick has, ideally, zero power dissipation... it just sloshes energy around. That's how I got 4 MHz out of a tiny Pockels Cell driver. > > > > What on earth is "F"? > > To an engineer it means, in this context, frequency. So to you it might mean something else in your fancy vocabulary.An engineer wouldn't capitalise it. And they'd use radians per second - 2.pi.f - rather than frequency to get the numbers to come out right. https://en.wikipedia.org/wiki/Radian_per_second <snipped the bit an engineer might have found interesting> -- Bill Sloman, Sydney
Reply by ●October 5, 20232023-10-05
On 03-10-2023 16:59, Anthony William Sloman wrote:> On Wednesday, October 4, 2023 at 1:41:37 AM UTC+11, Klaus Vestergaard Kragelund wrote: >> On 03-10-2023 14:37, John Larkin wrote: >>> On Tue, 3 Oct 2023 09:46:44 +0200, Klaus Vestergaard Kragelund <klau...@hotmail.com> wrote: >>>> On 29-09-2023 04:09, John Larkin wrote: >>>>> On Thu, 28 Sep 2023 13:58:40 -0700 (PDT), Klaus Kragelund <klaus.k...@gmail.com> wrote: >>>>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: >>>>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> wrote: >>>>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus >>>>>>>> Kragelund <klaus.k...@gmail.com> wrote in <903144ac-6642-4a3f...@googlegroups.com>: >>>>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: >>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin >>>>>>>>>> <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: >>>>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> wrote: >>>>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: >>>>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> wrote: >>>>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin >>>>>>>>>>>>>> <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: >>>>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> wrote: > > <snip> > >>> One can stuff a bunch of current into an inductor and then dump it into a mostly-capacitive load. That can make a high-voltage half-sine. > > A transformer would work better - perhaps several of them in series. > > Dump a bunch of current into the primaries in parallel - perhaps time shifted to allow the high voltage pulse to propagate through the series connected secondaries - and rely on the parallel capacitances of the secondaries to shape the pulse into a half-sine. >That's what I was considering doing, but still the individual transformer needs to be able to handle the 4.5kV isolation from winding to winding.
Reply by ●October 5, 20232023-10-05
On 29-09-2023 18:16, Lasse Langwadt Christensen wrote:> fredag den 29. september 2023 kl. 15.41.10 UTC+2 skrev Klaus Kragelund: >> On Thursday, 28 September 2023 at 23:09:44 UTC+2, Lasse Langwadt Christensen wrote: >>> torsdag den 28. september 2023 kl. 22.58.46 UTC+2 skrev Klaus Kragelund: >>>> On Thursday, 28 September 2023 at 17:26:46 UTC+2, John Larkin wrote: >>>>> On Thu, 28 Sep 2023 14:59:48 GMT, Jan Panteltje <al...@comet.invalid> >>>>> wrote: >>>>> >>>>>> On a sunny day (Wed, 27 Sep 2023 13:34:29 -0700 (PDT)) it happened Klaus >>>>>> Kragelund <klaus.k...@gmail.com> wrote in >>>>>> <903144ac-6642-4a3f...@googlegroups.com>: >>>>>> >>>>>>> On Wednesday, 27 September 2023 at 17:38:35 UTC+2, Jan Panteltje wrote: >>>>>>>> On a sunny day (Wed, 27 Sep 2023 08:10:02 -0700) it happened John Larkin >>>>>>>> <j...@997arbor.com> wrote in <d3h8hitm54oie4gcl...@4ax.com>: >>>>>>>>> On Wed, 27 Sep 2023 11:44:57 GMT, Jan Panteltje <al...@comet.invalid> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> On a sunny day (Wed, 27 Sep 2023 00:51:10 -0700) it happened John Larkin >>>>>>>>>> <j...@997arbor.com> wrote in <4bn7hit0n3rmcg855...@4ax.com>: >>>>>>>>>> >>>>>>>>>>> On Wed, 27 Sep 2023 05:14:49 GMT, Jan Panteltje <al...@comet.invalid> >>>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>>> On a sunny day (Tue, 26 Sep 2023 19:35:18 -0700) it happened John Larkin >>>>>>>>>>>> <j...@997arbor.com> wrote in <mf47hi15db7eq7mp2...@4ax.com>: >>>>>>>>>>>> >>>>>>>>>>>>> On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> >>>>>>>>>>>>> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> This is pleasingly weird. >>>>>>>>>>>>>> >>>>>>>>>>>>>> https://www.dropbox.com/sh/sif3efs69dxe1mg/AACY0RJGXl4k8CVvauUbJtYFa?dl=0 >>>>>>>>>>>>>> >>>>>>>>>>>>>> Sort of a baseline-boosted multi-auto-transformer voltage-doubler >>>>>>>>>>>>>> flyback. >>>>>>>>>>>>>> >>>>>>>>>>>>>> What's strange is that adding the two snubbers increases the LT spice >>>>>>>>>>>>>> sim speed radically, about 10:1. >>>>>>>>>>>>> >>>>>>>>>>>>> Since I couldn't find a suitable flyback transformer, and I only need >>>>>>>>>>>>> 500 volts at 10 mA, a flyback/C-W multiplier makes sense. >>>>>>>>>>>>> >>>>>>>>>>>>> https://www.dropbox.com/scl/fi/aquzpawfn8mfmfsnx70ra/T875_HV_6.jpg?rlkey=vfs9pexa0rgyxpo7pg1grqlb7&dl=0 >>>>>>>>>>>> >>>>>>>>>>>> Do you still need C11 R8 in that circuit? >>>>>>>>>>> >>>>>>>>>>> It doesn't need them to work, but there is a big high-Q ring when the >>>>>>>>>>> fet turns off, and the RC damps it. It would reduce radiated EMI. And >>>>>>>>>>> it speeds up the simulation some. >>>>>>>>>> >>>>>>>>>> Did you scope that? >>>>>>>>> >>>>>>>>> All Spice now. I'll build it when the customer gets serious. >>>>>>>>> >>>>>>>>>> There is a load of capacitance hanging from the MOSFET drain in that flyback multiplier ciruit >>>>>>>>>> I would expect it to damp RF oscillations but could be wrong... >>>>>>>>>> HV diodes have capacitance too.. >>>>>>>>> >>>>>>>>> The real circuit will have a lot of inductances too, what with those >>>>>>>>> long strings of parts. It needs to be built and tested. Dremeled, to >>>>>>>>> start. >>>>>>>>> >>>>>>>>> Here's an old auto-flyback C-W miltiplier prototype. I think that was >>>>>>>>> 1400 volts. Worked fine. >>>>>>>>> >>>>>>>>> https://www.dropbox.com/s/yd19osiwz1z74s4/HV_Proto_2.JPG?rsw=1 >>>>>>>> Yea, >>>>>>>> I did a PMT supply this way: >>>>>>>> http://panteltje.nl/pub/PMT_HV_supply_with_regulator_img_3175.jpg >>>>>>>> http://panteltje.nl/pub/PMT_HV_supply_componet_side_img_3180.jpg >>>>>>>> http://panteltje.nl/pub/PMT_regulated_power_supply_diagram_img_3182.jpg >>>>>>>> >>>>>>>> It has been in use now for many years, sits with PMT in the cardboard tube on the right, runs on batteries: >>>>>>>> http://panteltje.nl/pub/gamma_soectrometer_IMG_4505.JPG >>>>>>>> >>>>>>>> Of course that is very low power, but for a bit more power, and using a real TV muliplier: >>>>>>>> https://panteltje.nl/panteltje/pic/sc_pic/ >>>>>>>> >>>>>>>> No dremeling, boards are OK when they work.. >>>>>>>> >>>>>>>> And old TV circuits.. if you want a few kV there was / is? plenty stuff to be found on ebay. >>>>>>> For my case, I am working on a 5W 24V to 4.5kV converter. Right now looking to use a flyback topology, with possibly usage of >>>>>>> CCFL transformer bobbins. >>>>>>> >>>>>>> I am thinking about staggering the windings on 2 transformers, so the voltage across the windings will be lower, but still the >>>>>>> individual transformer needs to handle 4.5kV insulation (functional, not safety) >>>>>>> >>>>>>> Due to the limiting power, it could also be tempting to use the capacitive doublers as listed elsewhere in the thread >>>>>>> >>>>>>> My worry goes beyond the design, also about corona effects, which comes into play at +kV voltages. Potting is surely needed. >>>>>> >>>>>> For a HV transformer, old TV transformers (from the CRT times) may perhaps be an option: >>>>>> https://www.google.com/search?client=firefox-b-e&q=ebay+horizonatl+TV+output+transformer >>>>>> 4.5 kV diodes were used for focus in color TV sets... >>>>>> >>>>>> >>>>>> For a flyback, my rule of thumb (I do not use El Tea Spice), is 1 V per turn for DC input. >>>>>> So 24 turns primary for 24 V DC supply. >>>>>> The flyback voltage is (70/12) * 24 = 140V positive pulse if positive DC supply. >>>>>> To get a 4.5 kV pulse you would then wind 4500 / 140 is about 32, multiplied by 24 makes 768 turns. >>>>>> This for an EI core running at about 15 kHz (math comes from building TV output stages). >>>>>> 768 turns is not a big deal.. >>>>>> Spreading the turns over a segmented coil former is a good idea. >>>>>> If you want to use a voltage multiplier then less turns are needed, use a TV multiplier, those are 'potted' >>>>>> The 70/12 factor depends on the PWM on/off ratio of the drive you feed into the switch transistor. >>>>>> I use Microchip PICs for that, it has a PWM unit., hardware comparators that can directly stop the PWM drive >>>>>> for cycle by cycle current limiting, and ADCs to measure voltage and currents. >>>>>> Maybe an old TV horizontal output transformer with build in rectifier / multiplier would do. >>>>>> https://www.ebay.com/b/flyback-transformer/bn_7024744628 >>>>>> to adapt use 15 kHz and add 24 turns to the core primary? >>>>>> ... >>>>>> :-) >>>>>> >>>>>> Maybe like this? >>>>>> https://www.ebay.com/p/1692562663?iid=374326473542 >>>>>> plenty space for 24 turns thick wire, its a 30 kV DC out, so use less PWM ... or more primary turns >>>>>> or whatever you can find... on ebay. >>>>>> >>>>> High turns ratios can have two problems: voltage breakdown in the >>>>> insulation, and winding capacitance limiting the flyback ratio. That's >>>>> where a C-W diode multiplier helps. >>>> I am driving a UV lamp, so I need to deliver a pulse. Possible with a multiplier, but doesn't make it easier >>>>> >>>>> Old TV flybacks were pie wound to help with both problems. That makes >>>>> a hand-wound transformer yet nastier; Sloman would love to (have >>>>> someone else) wind that. >>>> Pie wound, so that's adding on layers instead of across the bobbin? >>>>> >>>>> There are some cool old parts from CRT TVs but we wouldn't use Ebay >>>>> parts for production. Even ISDN transformers are hard to get these >>>>> days. And most of our boards have a component height limit below about >>>>> 1". >>>> I agree, Ebay does not work for production items :-) >>>>> >>>>> I've gravitated to the classic C-W layout with all series strings of >>>>> caps. That's driven by the poorly-or-never defined C/V behavior of >>>>> available ceramic caps, which encourages their use at a small fraction >>>>> of rated voltage. >>>> In any case for the C-W multiplier to work properly, the charging of the caps needs to be kept low. >>>>> >>>>> There are lots of cute flyback controller chips that do peak current >>>>> limiting control. >>>>> >>>> I am using a microcontroller. Microcontroller and kV design does not sound good, right? >>>>> I don't know how much voltage we dare generate on a regular PCB >>>>> without coating or potting. I'm guessing that 1KV is safe and I know >>>>> that 7KV isn't. >>>> Potting or conformal coating is probably needed to avoid carbonization of PCB surface. >>> how about something crazy, a car ignition coil? ratio about 1:100, isolation good for +20kV >>> coil-on-plug can be pretty small, if you need access to both ends of the secondary a dual output coil for wasted spark has that >> These are used for old style flash: >> >> https://www.aliexpress.us/item/32521417603.html?gatewayAdapt=4itemAdapt > > coilcraft have some ccfl transformers, but the isolation rating is quite a bit lower than your 4.5kV > > https://www.coilcraft.com/en-us/products/transformers/power-transformers/ccfl-transformers/fl/Maybe the isolation can be achieved by potting. That's a reasonable cheap way to boost the isolation voltage.
Reply by ●October 5, 20232023-10-05
On Sunday, 1 October 2023 at 18:07:40 UTC+2, Jan Panteltje wrote:> On a sunny day (Sun, 01 Oct 2023 07:08:02 -0700) it happened John Larkin > <j...@997arbor.com> wrote in <4cuihitefaok29ca0...@4ax.com>: > >On Sun, 01 Oct 2023 05:31:31 GMT, Jan Panteltje <al...@comet.invalid> > >wrote: > > > >>On a sunny day (Sat, 30 Sep 2023 12:10:06 -0700) it happened John Larkin > >><j...@997arbor.com> wrote in <2grghi1vnnsmrkbcl...@4ax.com>: > >> > >>>On Sun, 24 Sep 2023 09:18:31 -0700, John Larkin <j...@997arbor.com> > >>>wrote: > >>> > >>>> > >>>>This is pleasingly weird. > >>>> > >>>>https://www.dropbox.com/sh/sif3efs69dxe1mg/AACY0RJGXl4k8CVvauUbJtYFa?dl=0 > >>>> > >>>>Sort of a baseline-boosted multi-auto-transformer voltage-doubler > >>>>flyback. > >>>> > >>>>What's strange is that adding the two snubbers increases the LT spice > >>>>sim speed radically, about 10:1. > >>> > >>>OK, this has become sort of a background hobby, waiting on the > >>>customer to get serious. > >>> > >>>https://www.dropbox.com/scl/fi/60nb1pcrnl17c07tk8xeu/T875_HV_9.jpg?rlkey=9uv466edcuvddmjldgepcyi0u&raw=1 > >>> > >>>All the caps are a part that we have, 100nF 630 volts 1812. Of course > >>>the capacitance plummets as voltage goes up, so the Spice values have > >>>to be adjusted appropriately. C is only 20nF at 500 volts. > >> > >>That much really? Th big caps I use do not change at all I think? > > > >I measured one. > > > >https://www.dropbox.com/scl/fi/zrrount2a5s7k5bmqh4bu/C-V.jpg?rlkey=3so16njq8md4ccnm610a8mf49&raw=1 > That is realy awfull! > Tt will likely create more ripple and less output. > >There are some big C0G caps around these days, but in this case I may > >as well use the X7R in stock and let it do what it does. > If the output stays in spec OK... > >It's fun to make an RC timing ramp from a Z5U or something. The > >exponential curves UP. > You could use that for frequency modulation, sort of a very big varicap... > Not very linear... > >> > >>>The RCD snubber is nice. It softens the rise at the fet drain and > >>>dissipates less power than a simple RC that damps ringing about the > >>>same. > >> > >>Nice, it delivers power back to the supply line :-) > > > >I didn't invent that, got that from a switcher textbook. > > > >There are actually "lossless" snubbers that store the snub energy in > >an inductor or a transformer and return it to the supply, but that's > >overkill in a 5 watt supply. > > > >The C-W multiplier actually clamps the mosfet drain to safe voltages; > >an inductor doesn't have leakage inductance! (Or it has 100% leakage > >inductance?) The snubber is there to damp the HF ringing and reduce > >radiated EMI. > CRT TV horizontal output stages are interesting, also have a 'tuning capacitor' > so the flyback is tuned to some harmonic (third harmonic Fred?) of the drive frequency, > I experimented with winding several transformers for TVs I build. > > This goes a bit into that, so no snubbers needed. > http://www.broadcaststore.com/pdf/model/793699/TT207%20-%205080.pdf > tv_horizontal_output_stage_operating__TT207-5080.pdf > see figure 6: > --------- -------- > | | | | > -- --- --- > . > / \ > ----------- ------------ flyback tunedThanks for the link, that is a very informative document, from the good old days :-)
