02/12/2021 at 18:32 •
« it's the follow through that gives us trouble »
Oh yeah? Maybe the real "trouble" is pictures silently dropped into digital oblivion, hence demotivaing difficult reads in a major way IMHO. What You See Is NOT What You Get seems the worst enemy here and i'm not even started on who's timescale matters in a consumerist environment where lurkers remain lurkers and objectors are practically all that's left anyway.
10 is faster than 1, obviously.
12/29/2020 at 08:59 •
M'yep! This time the title should get a few curious lurkers simply hooked i guess.
Actually if i didn't know better this might sound most puzling to me too! Euh...
Well lets begin with obvious evidence, for example:
I've theorized that a flux-concentrating core shaped into a flattened "C" could for all practical purposes create a pair of South & North poles right within the limits of the core "slot" intended to coïncidentally serve as an "IH Aperture" too. With a duo of those flux-concentrating structures facing each other at a 90 degrees angle it becomes similar to a 4-Poles stator, except it's a pair of very loosely-coupled horizontal magnetic "circuits" made parallel to the PCBoard, while the resulting IH Aperture actually passes through it. Now, what if those IH WorkCoil Assemblies can be stacked to create a 3rd magnetic circuit, so to speak: e.g. one which is at a square angle relatively to both the "X" and "Y" axis, simultaneously. Hence i'd call that a "Z" axis, except it's of a discontinuous nature directly reflecting the number of stacked modular layers. 2 being the minimum necessary to add a Z axis, or sort of, which is why i'd rather qualify it as half a dimension instead.
Anyway anything closer to 3-D feels desirable to me because that should allow more complex susceptor strategies, where different alloys would behave differently. That's especially appropriate in a context as that of vaporism where the bowl may require a pre-heating phase before the main convective/radiative heat vector floods it all real brief, yet just long enough to carry away whatever noble molecules happened to be ready for their "Release", followed by convective "Transport". In addition to some 3rd "purge" phase one might wish for a 4th one about maintenance of the screen(s) if present (by heat-scorching it)... So in the end it would seem reasonable to expect that having a 2½-D IH Driver should provide some fairly-genuine fun once all the bells & whistles have been fine-tuned, i presume.
Flux-concentrating material is commercially available, no insurmountable difficulty there. Lets just keep in mind it better not be manpower-intensive! That's a nice opportunity to explore the full potential of flat magnetics, quite on the contrary; and considering it's now public the DV VC is overweight by about (540 - 100) = 440 Joules! Euh... i mean WOW!! The significance of this is that portability shouldn't be considered a major issue anymore, not after reading a bit about what multilevel Class-D switched-capacitor topologies accomplish in the industry today.
And it's damn small even under the form of an evaluation board, plus there's NO D.C. path to worry about during catastrophic failure events caused by an overweight WorkLoad being inserted/removed, for example...
Using AWG#12 (2 mm dia.) wire (preferably Litz-type, with silver-coating option...) + thin magnetic core elements measuring between 2.5 and 3 mm thick, euh... I figured it must be OKay to expect a finished module not to exceed ~1.6 mm in thickness, without its casing. To say it all i'm even ready to gamble that time will push significant progress in this direction, exploiting 3-axis IH driving WorkCoil Assemblies to gain increasingly-fine control over eventual phases more and more specific to the delivery tool, but not only... Ultimately i can envision IH drivers replacing the cigarette pack both in terms of portable functionality and volume occupancy in the pocket, combined with a consumption method that won't need much more than ~100 Joules, like my LAVACapsule in its basic configuration.
What's so enthousiasmic about such prospect is the idea that a vast field of creativity awaits after module pairs can be safely emulated by the masses, so people can try/develop talents by focussing more on code than hardware. Then i can only imagine the karakoe-style...Read more »
12/28/2020 at 09:10 •
Explanation: an installed troll on reddit keeps moving stuff around to make it look like zero useful / on-topic suggestions were ever made...
Replies here have been deleted before, but i find it most deplorable to think you could use your "talents" so much more productively, for example by investing in some collective project. An open-design perhaps, and preferably based on smart chips since those solve the delicate issues related to gate-driving, as that's now on-chip besides plenty of protective features. Doing the same using discrete parts is overwhelming in comparison, but take a good look at economic alternatives like Infineon's MA12070(P) Multilevel Class-D 2x30 PWM "Digital Audio", in this case specified for a switching rate of 96 KHz i think. And made "filterless" because of a topology which by the way has NO path for D.C. during a catastrophic event (e.g. when ZVS becomes a forbiden stalled state leading to brutal failure). In comparison a smart chip ideally signals the fault requiring maintenance or correctives, ultimately a reset... Never mind the series-connected precision sub-Ohmic power-sampling resistor(s) for closed-loop control feed-back purposes! Forget about a 4th "Kelvin" pin and its attached monitoring circuitry, whatever. Acually the people who come here just don't even need to care about electronics design as long as the manufacturer's provided PCBoard artwork (and eventual corrections) are compatible with our final intended application. The best part of it all being that we're also getting rid of classic silicon issues related to junction reverse-recovery charges, because it's gone with Gallium Nitride for all practical purposes and it turns out this factor plays a major role in power loss from power-switching transitions, by moving away from analog linear behaviour that invites over-heating in MOSFET stages (if not worse!), especially the suicidal Mazzilly topology when submitted to excessive oscillations-damping (think of a heavy workload's insertion/removal! Does that ring any bell??)... Etc., etc.
But no, modiocrity is the rule it seems. You've got the means to print so much better circuits on your own designs which a community could control and improve quickly. I think you people call it crowd funding, which is when the painful reminescences resurface of a pizza eater (read Loto Labs Evoke/Lux...) who started to make the news around 2014. Which i think killed the spirit indefinitely it seems, yet the principle might have worked if the Loto Labs project hadn't quickly fallen into a spiral of conceptual dead-ends, its delivery device in particular. Which to me only seems like déjà vu after the NoFlame/SJK episodes.
If at least your power stage had some level of (retro-)active protection, not to mention only synchronizing information actually needs to feed a self-control loop; Mazzilli's "KISS" trick came at a price and that's analog coupling, essentially, while a digital version known as the "Fast ZVS Mazzilli Driver" of Francisco Piernas Diaz already been published long long ago... And all your time & efforts are still focussed on:
Me! Me! Me! ... You! You! You! ... In plain typical Sniff 'n Liche fashion, simply to "amuse", somehow. So please go ahead, give it a last sprint! Actually i think i've waited enough and been far too tolerant of those mediocre head games, socio-toxic entertainment for immature/deranged/criminal-minded individuals, whatever, etc. Quite a non-topic if you dare to ask me!!
You! You! You!
Poor thing. SYSTEMATICALLY SABOTAGED since i initiated the idea on FuckCombustion, in 2012...
Egzoset:Read more »