On Wednesday, August 5, 2020 at 7:38:00 PM UTC-4, John Larkin wrote:> On Wed, 5 Aug 2020 23:32:02 +0200, Piotr Wyderski > <peter.pan@neverland.mil> wrote: > > >jlarkin@highlandsniptechnology.com wrote: > > > >> I think people design transformers for > >> equal copper loss in the primary and secondary > > > >BTW, what is the source of this and similar rules of thumb (e.g. equal > >copper and core losses)? Logic says that one should always design for > >minimal total losses, given the economic constraints. > > > > Best regards, Piotr > > I don't think transformers are usually designed for equal core and > copper losses. This one runs cold at full AC voltage but no load. > > They are designed for equal primary and secondary copper losses at > full load, I think. > > Cooling depends on surface area. You get more surface area by adding > more copper, and that relationship is not linear (the volume-surface > thing, like mice and elephants) so core loss might require a lot of > expensive copper.I'm intrigued by the matrix transformer concept: two transformers with primaries in series and secondaries in parallel have both 1/4th the i^2*r losses each, and more dissipation surface per watt compared to a single giant lump of copper buried in steel. Cheers, James
transformer thermals
Started by ●August 4, 2020
Reply by ●August 7, 20202020-08-07
Reply by ●August 7, 20202020-08-07
On 2020-08-07 09:59, dagmargoodboat@yahoo.com wrote:> On Wednesday, August 5, 2020 at 7:38:00 PM UTC-4, John Larkin wrote: >> On Wed, 5 Aug 2020 23:32:02 +0200, Piotr Wyderski >> <peter.pan@neverland.mil> wrote: >> >>> jlarkin@highlandsniptechnology.com wrote: >>> >>>> I think people design transformers for >>>> equal copper loss in the primary and secondary >>> >>> BTW, what is the source of this and similar rules of thumb (e.g. equal >>> copper and core losses)? Logic says that one should always design for >>> minimal total losses, given the economic constraints. >>> >>> Best regards, Piotr >> >> I don't think transformers are usually designed for equal core and >> copper losses. This one runs cold at full AC voltage but no load. >> >> They are designed for equal primary and secondary copper losses at >> full load, I think. >> >> Cooling depends on surface area. You get more surface area by adding >> more copper, and that relationship is not linear (the volume-surface >> thing, like mice and elephants) so core loss might require a lot of >> expensive copper. > > I'm intrigued by the matrix transformer concept: two transformers > with primaries in series and secondaries in parallel have both 1/4th > the i^2*r losses each, and more dissipation surface per watt compared > to a single giant lump of copper buried in steel. > > Cheers, > James >You'd probably gain more than that on the primary at least (especially at 240V) because less of the volume would be used up by the insulation--you'd have more than twice the copper cross-section and the volts per turn would be the same. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Reply by ●August 7, 20202020-08-07
On Fri, 7 Aug 2020 06:59:00 -0700 (PDT), dagmargoodboat@yahoo.com wrote:>On Wednesday, August 5, 2020 at 7:38:00 PM UTC-4, John Larkin wrote: >> On Wed, 5 Aug 2020 23:32:02 +0200, Piotr Wyderski >> <peter.pan@neverland.mil> wrote: >> >> >jlarkin@highlandsniptechnology.com wrote: >> > >> >> I think people design transformers for >> >> equal copper loss in the primary and secondary >> > >> >BTW, what is the source of this and similar rules of thumb (e.g. equal >> >copper and core losses)? Logic says that one should always design for >> >minimal total losses, given the economic constraints. >> > >> > Best regards, Piotr >> >> I don't think transformers are usually designed for equal core and >> copper losses. This one runs cold at full AC voltage but no load. >> >> They are designed for equal primary and secondary copper losses at >> full load, I think. >> >> Cooling depends on surface area. You get more surface area by adding >> more copper, and that relationship is not linear (the volume-surface >> thing, like mice and elephants) so core loss might require a lot of >> expensive copper. > >I'm intrigued by the matrix transformer concept: two transformers >with primaries in series and secondaries in parallel have both 1/4th >the i^2*r losses each, and more dissipation surface per watt compared >to a single giant lump of copper buried in steel. > >Cheers, >JamesUtility transformers are very efficient, over 99%. They eliminate surface area from the efficiency tradeoff by oil cooling, sometimes with radiators and fans. Absent concern about cooling, the matrix thing sort of collapses with the observation that all the windings can share one core. Flex PCB winding transformers are neat. Power densities are outrageous. https://www.coilcraft.com/en-us/products/transformers/planar-transformers/ Hey, are these new? https://www.coilcraft.com/en-us/products/power/coupled-inductors/1-1-shielded-coupled/lpd_v/lpd8035v/ 1500 volt dual inductors! -- John Larkin Highland Technology, Inc Science teaches us to doubt. Claude Bernard
Reply by ●August 7, 20202020-08-07
On Friday, August 7, 2020 at 11:03:04 AM UTC-4, jla...@highlandsniptechnology.com wrote:> On Fri, 7 Aug 2020 06:59:00 -0700 (PDT), dagmargoodboat@yahoo.com > wrote: > > >On Wednesday, August 5, 2020 at 7:38:00 PM UTC-4, John Larkin wrote:> >> Cooling depends on surface area. You get more surface area by adding > >> more copper, and that relationship is not linear (the volume-surface > >> thing, like mice and elephants) so core loss might require a lot of > >> expensive copper. > > > >I'm intrigued by the matrix transformer concept: two transformers > >with primaries in series and secondaries in parallel have both 1/4th > >the i^2*r losses each, and more dissipation surface per watt compared > >to a single giant lump of copper buried in steel. > > > >Cheers, > >James > > Utility transformers are very efficient, over 99%. They eliminate > surface area from the efficiency tradeoff by oil cooling, sometimes > with radiators and fans. > > Absent concern about cooling, the matrix thing sort of collapses with > the observation that all the windings can share one core.Well sure they can use one core, that's where we started. The matrix transformer gets its advantages from /distributing/ the core and the windings. - Heat is spread out over more area, and current-sharing reduces i^2*r loss in each transformer-element. - Leakage inductance is proportional to turns^2, which is reduced by 1/n when you split, e.g., the secondary into n sections on separate cores. - And of course switching loss is lowered, since leakage inductance energy is lost to the snubber each cycle in most switchers. You could make GaN switchers run at super speeds. A further advantage of low leakage inductance claimed is that the primary and secondary currents are in such tight phase, that the primary winding and synchronous rectifier on the output can share the exact same timing.> Flex PCB winding transformers are neat. Power densities are > outrageous. > > https://www.coilcraft.com/en-us/products/transformers/planar-transformers/I was surprised to read recently that transformers' power-handling capabilities aren't core-size limited, but by one's ability to get out the heat (core-loss, and i^2*r).> Hey, are these new? > > https://www.coilcraft.com/en-us/products/power/coupled-inductors/1-1-shielded-coupled/lpd_v/lpd8035v/ > > 1500 volt dual inductors!That's a nice upgrade. Best I've seen 'til now were 500V. Cheers, James
Reply by ●August 7, 20202020-08-07
On Fri, 7 Aug 2020 13:51:22 -0700 (PDT), dagmargoodboat@yahoo.com wrote:>On Friday, August 7, 2020 at 11:03:04 AM UTC-4, jla...@highlandsniptechnology.com wrote: >> On Fri, 7 Aug 2020 06:59:00 -0700 (PDT), dagmargoodboat@yahoo.com >> wrote: >> >> >On Wednesday, August 5, 2020 at 7:38:00 PM UTC-4, John Larkin wrote: > >> >> Cooling depends on surface area. You get more surface area by adding >> >> more copper, and that relationship is not linear (the volume-surface >> >> thing, like mice and elephants) so core loss might require a lot of >> >> expensive copper. >> > >> >I'm intrigued by the matrix transformer concept: two transformers >> >with primaries in series and secondaries in parallel have both 1/4th >> >the i^2*r losses each, and more dissipation surface per watt compared >> >to a single giant lump of copper buried in steel. >> > >> >Cheers, >> >James >> >> Utility transformers are very efficient, over 99%. They eliminate >> surface area from the efficiency tradeoff by oil cooling, sometimes >> with radiators and fans. >> >> Absent concern about cooling, the matrix thing sort of collapses with >> the observation that all the windings can share one core. > >Well sure they can use one core, that's where we started. The matrix >transformer gets its advantages from /distributing/ the core and the >windings. > >- Heat is spread out over more area, and current-sharing reduces i^2*r >loss in each transformer-element. > >- Leakage inductance is proportional to turns^2, which is reduced >by 1/n when you split, e.g., the secondary into n sections on >separate cores. > >- And of course switching loss is lowered, since leakage inductance >energy is lost to the snubber each cycle in most switchers. > >You could make GaN switchers run at super speeds. > >A further advantage of low leakage inductance claimed is that the >primary and secondary currents are in such tight phase, that the >primary winding and synchronous rectifier on the output can share >the exact same timing. > >> Flex PCB winding transformers are neat. Power densities are >> outrageous. >> >> https://www.coilcraft.com/en-us/products/transformers/planar-transformers/ > >I was surprised to read recently that transformers' power-handling >capabilities aren't core-size limited, but by one's ability to get >out the heat (core-loss, and i^2*r). > > >> Hey, are these new? >> >> https://www.coilcraft.com/en-us/products/power/coupled-inductors/1-1-shielded-coupled/lpd_v/lpd8035v/ >> >> 1500 volt dual inductors! > >That's a nice upgrade. Best I've seen 'til now were 500V. > >Cheers, >JamesThey don't spec max winding voltage. They are sending me 10 samples on the condition that I blow up a couple.
Reply by ●August 7, 20202020-08-07
On Friday, 7 August 2020 22:09:56 UTC+1, John Larkin wrote:> On Fri, 7 Aug 2020 13:51:22 -0700 (PDT), dagmargoodboat@yahoo.com > wrote: > >On Friday, August 7, 2020 at 11:03:04 AM UTC-4, jla...@highlandsniptechnology.com wrote:> >> Hey, are these new? > >> > >> https://www.coilcraft.com/en-us/products/power/coupled-inductors/1-1-shielded-coupled/lpd_v/lpd8035v/ > >> > >> 1500 volt dual inductors! > > > >That's a nice upgrade. Best I've seen 'til now were 500V. > > > >Cheers, > >James > > They don't spec max winding voltage. They are sending me 10 samples on > the condition that I blow up a couple.Years ago I discovered from hipot testing transformers that it was too easy for them to suddenly not withstand as much voltage any more. The 1 person manufacturing line swore nothing had been changed. NT
Reply by ●August 7, 20202020-08-07
legg wrote more absurd crap: ============================> > >> For windings, it's a 'fill factor'. > >> > >> If you check the ratio of copper vs air in the inner > >> diameter of a toroid, as the fill approaches a certain > >> percentage, you'll see a diminishing return for agravated > >> difficulty in fabrication. > > > >** Cost is everything in manufacture. > > > >But as here like to consider what physics allows, it is clearly a fact that filling a toroidal with as much copper as possible maximises the VA for a given core. > > > >This IS the practice for all other types of core shape. > > > >Most toroidals are low and flat - again not optimum but liked by many customers. Same goes for R-cores and U cores which result in low height products. > > > Those last few turns also have the longest length > per turn, further diminishing their 'return'. >** What utterly stupid garbage. One fills the available space by using the heaviest gauge wire possible - not by increasing the turns - that is already set by the core cross section, voltage and frequency. Yo have no point to make and just post bullshit for the sake of having the last word and big noting your pathetic self. FOAD you ridiculous ass. ..... Phil
Reply by ●August 7, 20202020-08-07
On Fri, 7 Aug 2020 17:15:53 -0700 (PDT), Phil Allison <pallison49@gmail.com> wrote:>legg wrote more absurd crap: > >============================ > >> >> >> For windings, it's a 'fill factor'. >> >> >> >> If you check the ratio of copper vs air in the inner >> >> diameter of a toroid, as the fill approaches a certain >> >> percentage, you'll see a diminishing return for agravated >> >> difficulty in fabrication. >> > >> >** Cost is everything in manufacture. >> > >> >But as here like to consider what physics allows, it is clearly a fact that filling a toroidal with as much copper as possible maximises the VA for a given core. >> > >> >This IS the practice for all other types of core shape. >> > >> >Most toroidals are low and flat - again not optimum but liked by many customers. Same goes for R-cores and U cores which result in low height products. >> >> >> Those last few turns also have the longest length >> per turn, further diminishing their 'return'. >> > >** What utterly stupid garbage. > >One fills the available space by using the heaviest gauge wire possible - not by increasing the turns - that is already set by the core cross section, voltage and frequency. > >Yo have no point to make and just post bullshit for the sake of having the last word and big noting your pathetic self. > >FOAD you ridiculous ass. > > >..... PhilIf the MLT drops 20%, You get the same loss with 20% reduction in copper x-section. Not so? RL
Reply by ●August 8, 20202020-08-08
On 2020-08-07, dagmargoodboat@yahoo.com <dagmargoodboat@yahoo.com> wrote:> On Wednesday, August 5, 2020 at 7:38:00 PM UTC-4, John Larkin wrote: >> On Wed, 5 Aug 2020 23:32:02 +0200, Piotr Wyderski >> <peter.pan@neverland.mil> wrote: >> >> >jlarkin@highlandsniptechnology.com wrote: >> > >> >> I think people design transformers for >> >> equal copper loss in the primary and secondary >> > >> >BTW, what is the source of this and similar rules of thumb (e.g. equal >> >copper and core losses)? Logic says that one should always design for >> >minimal total losses, given the economic constraints. >> > >> > Best regards, Piotr >> >> I don't think transformers are usually designed for equal core and >> copper losses. This one runs cold at full AC voltage but no load. >> >> They are designed for equal primary and secondary copper losses at >> full load, I think. >> >> Cooling depends on surface area. You get more surface area by adding >> more copper, and that relationship is not linear (the volume-surface >> thing, like mice and elephants) so core loss might require a lot of >> expensive copper. > > I'm intrigued by the matrix transformer concept: two transformers > with primaries in series and secondaries in parallel have both 1/4th > the i^2*r losses each, and more dissipation surface per watt compared > to a single giant lump of copper buried in steel.It seems to me that you'd get most of those benefits buy just using a single transforer of twice the mass. I think you may be counting some of the gains twice. -- Jasen.
Reply by ●August 8, 20202020-08-08
jlarkin@highlandsniptechnology.com wrote:> On Thu, 6 Aug 2020 06:57:07 +0000 (UTC), Cydrome Leader > <presence@MUNGEpanix.com> wrote: > >>jlarkin@highlandsniptechnology.com wrote: >>> On Wed, 5 Aug 2020 01:21:10 -0700 (PDT), Phil Allison >>> <pallison49@gmail.com> wrote: >>> >>>>Cydrome Leader is Funny: >>>> >>>>======================= >>>>> >>>>> > So I ran it for a few hours with 10 amps DC in the primary. Temp rise >>>>> > was about 26C in free air. I think people design transformers for >>>>> > equal copper loss in the primary and secondary, so temp rise would >>>>> > double when loaded in the system. >>>>> >>>>> What about the heat generated by losses in the secondary? >>>> >>>> >>>>** Doubles the temp rise - as JL just claimed. >>>> >>>> >>>>> > https://www.dropbox.com/s/ylm8dc1e14dwv7y/P900_Xfmr_Thermal.jpg?raw=1 >>>>> > >>>>> > https://www.dropbox.com/s/5b55ybfoq2pkp9v/P900_Xfmr_Thermal_2.jpg?raw=1 >>>>> >>>>> You can even see on the label (looks sort of like a Noratel) the input is >>>>> 266VA and output is rated 240VA. So depending on power factor at full >>>>> load, it could be upto 26 watts of loss, although I doubt a toroid that >>>>> size would have such a poor efficiency. The iron losses will surely be >>>>> less than 50% on a toroid. Maybe the maker can tell you the ratio? >>>>> >>> >>> This transformer was made for us by Amgis. I specified it so I know >>> the ratios: 1 : 1.4 : 1.4 : 1.4 : 1.4. We have a relay board that >>> switches the secondaries to get four output voltage ranges. >> >>Are these ratios exact or is there a fudge factor for losses at rated >>load? > > Exact turns ratio. We know the other parameters: wire resistance, mag > inductance, leakage inductance, SRF, saturation. We plug all that into > the system Spice model.How did that work out? Did you give rought specs and the then the manufacture provided full expected specs?> Just curious. Retired transformer designer friend seemed to spend a >>bit of his time trying to get "the facts" from customers to make designs >>that would actually work under load. The stubborn customers would get full >>production runs of stuff that met specs, but didn't work for the >>application. >> >>>>** PF has no effect - VA is all that matters, effectively just the RMS current. Iron losses are a watt or so per kg or iron. I mag is tiny. >>>> >>>>So 20 watts copper loss, 6 watts for iron. >>>> >>>>Regulation about 8%. >>>> >>>>Toriods are very simple. >>>> >>>> >>>>.... Phil >>> >>> Unloaded, the AC operating primary current is essentially zero, so I >>> don't think core loss is significant. >>> >>> As an alternator simulator, voltage increases with frequency, which >>> keeps Imag low on the low end. >>> >>> It's a weird application. We specialize in weird. >> >>Was amgis friendly about making samples? > > Not free samples, but I didn't ask for that. We placed a PO that > included a few first-articles for verification, with the rest shipped > on approval.Sounds reasonable.
