Electro-magnetic (IH) power injectors and a need to get rid of ghosts in the machine!

[Egzoset]

Salutations,

Ever heard of Vladimiro Mazzilli (Milano/Italia) and more precisely some popular modules which inherited his name?

Here's the author, captured in photo years ago when he reached 55 i think:

Now this is the reason why those affordable semi-DiY kits gained popularity:

Eventually enhanced by creative authors from around the globe:

Etc.

But all along it appears this particular concept had "a ghost in the machine" (e.g. by design), so i figure some potentially increasing number of enthousiast users will want to know if there still are any problems with it today - and the are indeed IMHO!!



For example when dealing with powerful Radio-Frequency emissivity one has got to wonder about such details as evoked below, not to mention "shielding" (...):

Then after digesting some intense reading one's own creativity may begin to bloom, possibly with something similar to this short suite of variations on a theme - contributed by myself:

And each single time plugging a hole of such leaky boat somehow invites another to pop up!

Actually any reader who can read & interpret such hieroglyphs has fair chances to see it's still awaiting for more revisions - which is on my menu effetively:

Why the PLL agravation? M'well, here's a hint:

It's already been demonstrated, 10 years ago (...), that a proper combination of IH-driving + Curied alloys does promise valuable benefits in terms of novel features.

Euh... No, wait! The year was:

[ https:// www.jai.co.jp/english/company/history.html]www.jai.co.jp/english/company/history.html ]
Nov. 1969 -- Released the first Curie point pyrolyzer JHP-1.

Which is as far as i can perceive using my foggy story-teller crystal ball at this present moment. Except perhaps i already know i'll want to separate the opto-couplers in order to simultaneously support both the interlock safety function and quadrature demodulation to help "close" (e.g. complete) a PLL control loop, to gain access to Curie-based experiments (when it's time) - not in an immediate future i can foresee anyway.

Yet i keep looking forward in that direction with hope and confidence.

Good day, have fun!! ☮

 

[LittleDabbie]

Are you as drunk as i am?

 

[Egzoset]

You tell me...

 

[Galvatron]

Is this a project being built from scratch?

 

[Egzoset]

Salutations Galvatron,

Is this a project being built from scratch?

Sorry for the delay, i vaguely recall disabling my e-Mail flags after concluding to an impossibility to get real dialogues on the matter, perhaps that was premature.

This dedicated thread is a continuation of this other:

[ https:// www.thcfarmer.com/threads/weapon-of-mass-seduction.94780/ ]​

Weapon Of Mass Seduction (2018-Mar-15)​

Actually IH drivers have been on the menu since i created the Egzoset alias in mid-2010... The "Plan-B" (butane-based) made such progress it's about time i start thinking of my "Plan-A". Most unfortunately, if anyone pays attention when i write it the fact is that i build on the ideas of others like inventor and owner of VaporGenie mister Dan Steinberg, although i can't be this precise on others i also remember having a crucial inspiration from the VapMan "PinHole". More recently i had consider dealing with the Mazzilli ZVS topology only to observed that's unfinished work as far as i'm concerned, and i ain't alone: just last night i was reading interesting stuff by the Toshiba Electronic Devices & Storage Corporation, about some new generation of Power iGBT/MOSFET Gate Drivers it seems...

In any case if i'm ever going to build it rest assured i won't settle with the present frenzy imposed as a result of popular Mazzilli "modules" found on eBay! To me too many years have been lost already, i can't afford more time-wasting even if it's just a hobby. Oh, well. Of course i can blow electronics with the same virtuosity as anyone else, but lets simply mention even major companies as Toshiba still won't recommend such top-notch technology for military applications and similar, for a reason i figure: because single-chip solutions imply all functions can be lost should there be damage cause to it somehow - and it turns out i'm satisfied already to finally "tumble" on relevant technical literature concatenating the efforts of a whole colony of technicians. All i can do now is browse through catalogs and search for more inspiration, here's a morning scoop since you may find that entertaining:

​

Egzoset garantees this wouldn't work. It only helps me to focus my mind on details like the need for low-impedance gate-driving, but above all i want some form of fail-proof "interlock" protection and i won't pretend inventing in a couple of work hours what technicians in top-notch laboratories still try to fix today...

Quite frankly the literature is overwhelming and there's got to be a natural process of assimilation, then more rationalization trying to account for a most basic aspect of the Mazzilli modules: their parts count. Not to mention over-confidence relative to their true applications range as sold.

The NoFlame 3rd-party ready-made commercially-available option felt compelling a while ago but this is less than an optimal design either i'm afraid. The LAVACapsule will require time-compressed IH energy injection in an effort to exploit its potential for transitional performance. Thermostatic thinking only collapses here, i'm attempting to achieve "Hot-Spot" Induction Heating which target the susceptor inside a non-magnetic metal mass that sucks parasitic heat losses before these can even threaten to cause skin burns. My concept is meant to be inclusive enough for 1-handed persons to use it i believe, and since i'm aiming for the ferro-magnetic type of IH drivers anyway it seems i'm alone. Maybe this other graphic may help to see why:

​

The temptation to rely too much on the "eddy" effect for Induction Heating might happen to explain the present situation gravitating around Mazzilli modules: it's "working" with DynaVap's VapCap, although i'm convinced that's an overweight susceptor to say the least. They rrely on a bimetal clicker, i foresee more durable satisfaction aiming for the proper valorization of Curie aloys, probably similar to JAi's PyroFoils available in 21 target ("ceiling") temperatures. Such technology may be 50 years old it's clearly awaiting for our collective attention. There again i only depend on the works of others to decide where to go next anyway.

Maybe the path remaining to be walked is suitable for no others than myself exclusively, but i was gambling on the opposite premice so far.

YMMV.

Good day, have fun!! ☮

 

[Galvatron]

Perhaps look into modification of existing high power RF amplifiers to your purpose?

 

[Egzoset]

Hi again Galvatron,

Perhaps look into modification of existing high power RF amplifiers to your purpose?

There's quite some good logic in such idea, which is why i actually tried my luck on some "High Voltage" forums board already - which ended up with my cancellation of the account... Yet it's radio hobbyists i should ask about ready-made modules compatible with my purpose, but not only; the others went as far as to invite me to put the SiC puck in a micro-wave and check what happens next, then the subject of "perpetual machines" prompted me to terminate that brief parenthesis and start looking elsewhere.

Yeas ago i was wondering about hot hair rolls heated by induction, but i'm tempted to think it's similar to cheap IH cookers, the NoFlame/SJK (dental wax) tools: those can be expected to be 1 power transistor applications only, even auto body work in garages perhaps, etc. I vaguely considered semiconductor enterprises offer prototyping kits but i yet got to see one able to convince me it's justifying its own price-tag. So i'll keep reading & reflecting on "Smart gate Drivers" and what not, at least it forces me to exercise under "the influence" and i can manage with the pressure of that...



My bet is that it's already been done before, ultimately i may find an easy-go solution acceptable by the masses, but just not yet!!



In any case the initial conceptualization phase feels as important if not a lot more than simply switching to action. It's a slow process for an isolated individual, that's exactly why i rely on previous testimonies, some being real eye-openers. Like this:

To me the benefits of "Pulse Stealing" never seemed so evident, i so keep hoping i can experience it to the 1st person, someday...

Good day, have fun!! ☮

 

[Egzoset]

So...

What's to be put in my mystery box??

​

Lets admit the revision presiously attributed to Jakub Tejiščák and Jan Martis has left me a durable impression, compared to so-called "ZVS" Mazzilli/Royer circuit as currently sold to unsuspecting/unprepared enthousiasts...



If i choose opto-couplers to implement a galvanic barrier then i may as well need to justify using closely-matched twin-units in order to extend service life, by comparing/adjusting light transmission/response within the twin mirror-half (against a stable TL431 reference possibly). Or maybe not, because those devices can already promise light emitting/receiving degradation won't take place in decades, after all... But there are competiting technologies claiming to tripple such life-span (e.g. beyond 1 century overall!), though in exchange for significant features of that same isolation barrier which i'm attempting to include as my priority design criterias.

Implying more browse hours obviously remain necessary!

The matter of "interlock" protection divides as 2 main groups i'd say: either it's built-in and the driver chip is actually a dual-channel gate-driver, either it's a single-channel version and then interlock logic has to be implemented externally. As previously illustrated, opto-couplers (or equiv.) can serve to reflect state(s) in my power stage and hence also help to detect a forbiden "catastrophic failure" situation, e.g. when echoes no longer reflect the commands...

Alternately, the more "intelligent" a chip gets, the less reliable its whole system shall tend to prove in case of component failure(s), and while i'm at it perhaps i better think of today's most suitable semiconductor technology to be adopted, since now Gallium Nitride power devices have started to emerge, while the MOSFET/iGBT/SiC ones all shine best in high-voltage applications, leaving the portable low-voltage "niche" applications for GaN to reclaim i think:

​

Consequently whatever the selected chip, my bet is that it's got to be GaN-compatible or my path to future convenient/easy upgrades might get blocked... Better now than too late!



One other matter captured my attention and that's about I/O issues (impedance, isolation), always with "A.C. Decoupling" in mind - contrary to eBay's popular/cheap/over-simplified Mazzilli boards. Which ain't been a neglected detail in this colour/enhanced publication:

​

[ http:// www.piers.org/piersonline/pdf/Vol4No4Page417to420.pdf ]​

Piers - Anti-Interference design of Quasi-Resonant tank for magnetic Induction Heating system (2008-Jan)​

Naturally there's even more precautions to account for but this aspect alone should suffice to provide hints of my present concerns: the basics are in option with Mazzilli modules, imagine subtilities as "snubber" networks!... To the very least this means it's the battery packs which take punishement normally reserved to capacitors, or even perhaphs "super-capacitors" i figure. That's why i'm not going to waste further resources on a toy broken "By-Design".

On the positive side i liked the "Hybrid Hysteretic Control" concept by Texas Instruments (patent US20180048236A1); if memory serves me correctly it's removing delays & phase variations caused by PLL loop-filtering. The next dedicated page felt like some vast mine of inspiration in any case:

​

[ https:// www.ti.com/tool/TIDA-01495 ]​

480W, <17 mm, Thin Profile, 94% Efficiency, Fast Transient Response AC/DC SMPS Reference Design (2018-Feb-1)​

So, i begin to foreseee some mix of brute-force/fail-safe interlock protection juxtaposed to clever tricks intergrated as a single chip, whith the usual bells & whistles of classic power circuitry. It's a blurred perspective requiring me to renew perspectives then leave it aside to mature some more, until ready for another round.

With a bit of luck i may find a ready-made or semi-DiY "Eval" board/kit only requiring minor adaptations to fulfil such ambitious goal. The item below demonstrates performant "Transient Immunity" (I/O isolation) better than 125 kilo-Volts per microsecond in presence of a symetric (sink/source) 4 A-peak gate drive:

​

[ https:// www.silabs.com/development-tools/isolation/isolated-gate-driver/si823h9-evaluation-kit ]​

SiLabs - Si823H9-KIT, Si823H9 Isolated Gate Driver Evaluation Kit, 49 $​

By the way i believe this V/μs parameter hardly even appeared as a criteria in 2008 papers dating back to the Mazzilli days, to be exact.



Take note that eval board only has empty pads where power devices must be installed, which could be a blessing because of the multiple options it renders available. Of course i won't pretend being able to come up with a perfect solution years after Mazzilli himself left his work unfinished, as far as i'm concerned... The legacies left by a multitude of other prolific contributors is precious too, which reminds me of this VERY nice tutorial covering a vast range of Gate-Driving issues exactly:

[ https:// toshiba.semicon-storage.com/info/docget.jsp?did=59460 ]​

Application Note - MOSFET Gate Drive Circuit (2017 - 2018)​

​

4.4.3. Gate malfunction prevention (p18/22)​

With complementary perspectives below:

​

[ https:// e2e.ti.com/blogs_/b/powerhouse/archive/2018/04/17/how-to-achieve-higher-system-robustness-in-dc-drives-part-1-negative-voltage ]​

How to achieve higher system robustness in DC drives, part 1: negative voltage (2018-Apr-17)​

​

[ https:// e2e.ti.com/blogs_/b/powerhouse/archive/2018/05/30/how-to-achieve-higher-system-robustness-in-dc-drives-part-2-interlock-and-deadtime ]​

How to achieve higher system robustness in DC drives, part 2: interlock and deadtime (2018-May-30)​

​

[ https:// e2e.ti.com/blogs_/b/powerhouse/archive/2018/09/19/how-to-achieve-higher-system-robustness-in-dc-drives-part-3-minimum-input-pulse ]​

How to achieve higher system robustness in DC drives, part 3: minimum input pulse (2018-Sep-19)​

Etc., etc.

Honestly that's a huge amount of information to read and assimilate, but it also gives plenty of evidence revealing the eBay module shorcomings IMHO. Which is enough motivation for me to invest spare-time elsewhere, because i see no value in repeating such historical mistake...

Nonetheless please don't get me wrong. It's not a blame for Mazzilli who performed a great job, the problem relies with those who capitalize on his idea with minimal concern for the end-user's own needs...

M'well, i sure wouldn't want to risk having to field-service a defective Mazzilli-based unit just when i'm about to enjoy an IH-driven toke during some once-per-century eclipse event or similar, for example. Few of us wander around with spare parts and their butane-operated soldering iron in a pocket!...

Good day, have fun!! ☮

 

[Galvatron]

If the circuit you are trying to build is just the low voltage side if a high voltage power supply, have you looked at Tesla coil designs?
Depending on frequency and power required, look at a simple push-pull A/B class RF power amplifier. They are pretty stable already. The older CB amps used a pair of bipolar transistors but newer designs use FETs. It depends on frequency and power required. The design is generally for a low impedance input/output (50 ohm) but for you, it would depends on the output transformer. (Primary of your HV transformer.)
The exciter can be a signal generator for precise frequency control.
So, what is the frequency range and power required?

If I'm on the roof messing with antennas, I probably do have spare parts for a butane-operated soldering iron, and a few other things, in my pockets!

 

[Galvatron]

Wait...Is this what you are trying to accomplish?
https://www.amazon.com/Lighter-Elec...s=Triple+Plasma+Lighter&qid=1597050633&sr=8-2

 

[Egzoset]

Salutations Galvatron,

Wait... Is this what you are trying to accomplish?

Thank you very much for the tip!! That's some innovation i'm real glad to discover, yet the truth is i'm just no combustion guy: ex-smoker since mid-2007 and vaporist (after a couple years of total abstainance) initiated late in life on February 15, 2011...

In any case i find this plasma-driven lighter shares common desirable features with portable IH-driven dry-herb vaporizers. But i happen to prioritize safety above portability since i'm only vaporizing cannabis at home, away from privacy invasions which already resulted in "legal" evictions in my province, and i think others too.

Consequently, among design criterias which don't necessarily involve the production of a PC Board (because it's meant to remain as replicable as possible), right on top of my criterias list there's Electro-Magnetic Noise and more precisely INTERFERENCE with my neighbours devices...

The IH driver must deliver ~100 joules into a 0.75 g susceptor within a period of 2 ~ 3 seconds max. That susceptor is installed inside a faintly-magnetic Stainless Steel tube itself tightly inserted into an aluminium mass large enough to disperse heat losses before they cause any concern for skin burns. The trick being that it's essentially a Pulse-Mode process only existing during a brief transition between the stand-by state taking place between 2 inhalations, vs the dynamic "Release/Transport" state during which thermalization is expected to be near-instantaneous because plenty of energy is made available, until the heat charge gets depleted to prevent undesired "baking" effects, that is. I often refer to this excerpt relative to Pulse Heating:

​

If the circuit you are trying to build is just the low voltage side if a high voltage power supply, have you looked at Tesla coil designs?

Absolutely, i did find a lot of inspiration in Tesla Coil forums indeed. Of course i don't really require a High-Voltage transformer, the workcoil and susceptor will do fine, eventually. But in reason of serious environmental concerns as explained above i'm afraid the matter of RF "SHIELDING" simply can't be avoided. That's why i reasoned that i must experiment with power Stepper Motors and most especially 3-Phase ones: after all their materials already been optimized as a function of conversion efficiency, from electrical to mechanical. So, what if i dare remove its rotor and drill a hole large enough to insert some glass delivery wand with 0.75 g of SS430 at its core? For example:

​

...what is the frequency range and power required?

That would fall around 38 KHz, because i'm hoping to experience with Curie alloys if lucky, and also in hope to find narrow-band "reed" (or Surface Acoustic Wave?) filters perhaps, to simplify filtering in some future PLL experiment(s). One important criteria being not to close doors at priori. 38 KHz is a commonly-used frequency low enough in the radio band for my purpose, we're talking about spitting out RF radiation in the 250 ~ 350 Watts range i believe, which ain't exagerated knowing i must allow for some margin if it's to be made "ajustable" (by stealing pulses, hence my previous referal to Tesla music).

...look at a simple push-pull... ... ...newer designs use FETs.

Do you mean the type of bridge with complementary N/P transistor pairs, MOSFET or even iGBT/SiC?

Still environmentally speaking it appears quasi-resonant scenarios with proper shielding (and maybe low-voltage/high-current GaN soon it seems...) may be my best bet. Battery power for portability doesn't appeal to me but i like the idea of something run on 12 Volts for automotive compatibility, with a wall-plug and some black box PSU attached.

Anyway i can only gamble the RF emissions can be conveniently kept inside some NEMA-17 (or 23?) stepper casing, combining power in 3-Phase style. If all 3 coil sets "resonate" with a rarallel-connected power capacitor installed directly on that casing to minimize losses then at least it should be possible to A.C. decouple a significant part of the energy near their High-Q operating range, euh... Which is starting to align the "IFs".

The exciter can be a signal generator for precise frequency control.

PLL 4046-based Tesla Coils have been demonstrated, while at Texas instruments i found their "hysteretic" Gate Driver documention most compatible with my own objectives, except i also want galvanic isolation so i'm still sorting through datasheets and their "eval" kits, with PCB artwork preferably... Their implementation of a "gating" closed-loop regulation process stealing as many pulses as necessary to restore stable/efficient neo-ZVS energy conversion implies the insertion/removal of my delivery tool would get automagically detected and safely managed by reducing output power until loading conditions are optimal again. So consumption would get reduced in stand-by "scanning" mode, essentially...

As for my late Opto-Coupler interlock ideas i like that the HV workcoil supply must be superior in voltage to the main 12 Volts line that feeds low-power control circuitry. Some systems even offer elaborate power up/down functions, etc. Yet in any case the power devices wouldn't receive any positive bias until the HV line is ready; despite general failure of a blown "smart" gate driver chip my fail-safe mechanism would/must still prevail, at least in principle. While the effect of an interlock forbiden condition whould result in a same shut-down scenario.

Among the options i'm considering twin PLLs, and even Software-Defined Radio if this can help keep the parts-count down...

So now the year is 2095 and my brain has been transplanted so i could complete my mission in this time-frame: It's already been done before, just not yet!



...

Well, rest assured it's a long path to contemplate (and walk!) so i feel modest in my personal situation. Yet i find necessary to perform some planning early in this whole process, away from the Mazzilli mirage.

Perhaps there are power 3-Phase ready-made "eval" boards capable of 38 KHz IH-driving as describe above. Even if such preliminary think-tank never translates into a real prototype this may still suffice to identify 3rd-party "module" candidates close enough to spare me the need to actually produce one.

My bet is that it's easier to find something once knowing what it looks like. Go figure what other commercially-available product i should investigate, starting with automotive A.C. power inverters i presume. But would it detect workcoil loading? Handle interlock catastrophic failure?? Support provisions for bidirectional haptic/audible user-interfacing?... Etc., etc. That's why i rely on the work of others including the last decade of Tesla Coil enthousiast work. Which reminds me of this intreaguing little gem of creative inspiration IMO:

​

[ https:// www.analog.com/en/design-center/reference-designs/circuits-from-the-lab/cn0196.html ]​

​

[ https:// www.analog.com/-/media/analog/en/reference-circuits/images/cn0196-00-1024.gif ]​

Analog Devices ADuM7234​

Ain't that an analog-digital implementation of the logical interlock protective function?...

Honestly i'm still unsure as i just can't focus enough! Hence i must trust my intuition and seek for more hints, but i certainly appreciate those interludes anyway!

If I'm on the roof messing with antennas, I probably do have spare parts for a butane-operated soldering iron, and a few other things, in my pockets!

Fair enough, obviously this is what makes humanity's vast variety of lifestyles most unique if not precious, including its exceptions: it's a wonderful world.



Good day, have fun!! ☮

 

[Galvatron]

I like the idea of plasma but I understand the desire to eliminate actual combustion. I have always favored a HUGE cold bong rip, but with time, the lungs tend to protest.
If I get this right, you are trying to accomplish thermal vaporization via rapid induction?
In the end, isn't it still just heating up the chamber just very quickly?
It's 2095...
Theory: Perhaps consider a higher portion of the RF spectrum.
Direct material vaporization in a resonate microwave chamber? X-Rays? Gamma Rays?

 

[Egzoset]

Hi again Galvatron,

...with time, the lungs tend to protest.

Perhaps when that happens you might like this alternative for water toys:

​

I've tagged it "FogBong!" and what it does is to add just a tiny bit of moisturized air to the cannabic path, on its way to the lungs. The trick is to create a mixing aperture controlled by negative pressure vs surface tension & gravity... The "PinHole" at the base of an elbow tube is rapidly plugged by accumulated water droplets condensing inside that tube, which preserves the mixing ratio between 2 inhalations. Very little suffices, that's about when i realized the potential of VaporGenie pipes i think, the year was 2012.

It was "divine" but after only ~1 week it took its toll, because that had promoted abuse...

If I get this right, you are trying to accomplish thermal vaporization via rapid induction?

It's the vaporization which gets done quickly, my customized butane-operated VG pipe already provides some fair perspective on what i call "Micro-Bursting", which is about thermal bursts targeting contact-surface features as trichome glands (on the basis of mass/size differenciation).

It may help to play to "Heat-Shave the Spider", a cruel treatment reserved to large summer arachnoïds for invading balconies... More exactly it's about turning their legs hairs incandescent while also avoiding to kill the rest under an exo-skeleton. Long, slow-steady blows are the most deadly because this allows ample time for all matter located at the flame's tip to thermalize...

So in short i'm exploiting a physical impossibility for heat to move instantly. Using an amount of energy which appears to be "infinite" within the range of operation that implies a heat pulse of well-define duration can be expected to "Release" whatever molecules happening to be ready for the next airborne phase, then depletion occurs before serious "BAKING" damage is given a chance to accumulate. So at 1st there's all sorts of compounds being released and i believe it confirms in aroma/taste appreciation of the 1st few puffs from a fresh bowl, then i'd say the easy-gone fragrances have left after the 3rd toke or so. At least i can tell it's not like thermalizing a 300 mg bowl for 3 minutes.

In the end, isn't it still just heating up the chamber just very quickly?

That's still not accounting for a crucial anti-baking aspect of my LAVA Hybrid Core: the Heat Charge is actually injected outside the bowl with a protective SiC Foam layer behaving like an airflow-actuated thermal valve inbetween... It's not well-defined that much in my analog "Plan-B" equivalent but i can predict a nice surprize when "Micro-Bursting" becomes IH-driven, nonetheless.

It's 2095...

M'yeah, these future-past senarios tend to feel confusing, after a while there's no telling what came 1st... But i know for sure that JAi's original Curie Pyrolyzer dates back to 1969, which is why it's already been done before, though somehow just not yet in our parallel universe of "stoner$"/"droÿé$". Imagine, even its PyroFoils selling in 21 Curie temperatures have been on the real market since i got to wonder when!

Perhaps consider a higher portion of the RF spectrum.

No, the DynaVap IH enthousiasts using Mazzilli modules are doing exactly that and i was reading recently that it resulted in heating some metal casings too. They must be operating in the 80 ~ 120 KHz region i guess, while that's moving away from such precious characteristic(s) as illustrated below:

​

It's strangely similar to the curve attributed to Irshad Khan in my introduction post, isn't it?

Even in '95 i bet the Curie effect shall still prove to offer an "edge" over non-magnetic metals heated by eddy currents and that's thermostatic behaviour locked into the alloy's structure itself. It means ZERO distance to an hypothetic temperature detector and hence zero delay, zero components to fail, etc. Briefly put one would have to destroy the susceptor's Curie alloy itself to anihilate such precious Self-Regulating property, as revealed (and demonstrated in 1969...) by IH-driving.

So i've ideally settled for Curie alloys as JAi's PyroFoils long ago. Although i'm also considering "Magnetic" alternatives to explore, for example:

​

​

Go figure if SS430 ain't featured with a magnetic signature reminescent of the Curie effect, which itself happens to conveniently fall right into a temperature range most compatible with our common purpose if/when the metallurgy is correctly adapted:

​

Simple intuition tells me i can find an alloy with some measurable/reproductible signature.



In a pulse "gating" senario as previously evoked i'd try to energize my LC resonator tank 1st and then only after the pulse is finished launch a scanning window meant to detect "echoes" of this energy being reflected back to a pick-up coil. Having the local oscillator signal at hand for direct reference comparisons i suppose quadrature demodulation may suffice to extract some I/Q signal useful enough to determine when the Curie point is reached, or even better: extrapolate relative temperature on the way to it, with the Curie temperature only serving as a "ceiling" meant to prevent combustion in the most ultimate fail-safe manner imaginable. Because there's absolutely nothing to break to begin with...

Direct material vaporization in a resonate microwave chamber?

Such heat originates from inside, necessarily implying that *ALL* of the matter gets thermalized as a bloc. There's also the messy problem of dosing its energy bubbles as these could just as well result in bursts from the sudden internal pressure build-up... In comparison my LAVACapsule has provisions for post-vaporization re-heating meant to cause further shrinking of the microscopic sticky droplets, contrary to invisible yet brutal/uncontrolled explosive fractures.

X-Rays? Gamma Rays?

Egzoset follows the sacred/illuminated "Shortest Path of Lesser Transformation", consequently disqualifying denaturation by radiation burns. We want to vaporize it, certainly not promote noble molecule mutations!...

Good day, have fun!! ☮

 

[Egzoset]

Variations on a theme (from a previous post):

​

Notice the similarities when compared to so-called "ZVS" Mazzilli modules with a pair of toroïd "chokes", except this time only 1 toroïd core would be required i think, while also performing some additional A.C. related (RF) "decoupling" function. Further adaptation might include complementary "shunt" (by-pass) capacitors for improved separation between the DC and AC paths, so to speak... But that's little more than wishfull thinking rooted in good old "analog" technology dating back to a remote era, as when man set foot on the moon... Go figure: it's already been done before, just not yet!

Good day, have fun!! ☮

 

[Galvatron]

Now that looks like a nice, basic RF amp! The upper portion, in red, is for filtering the DC input. That's the AC decoupling. T2 is wired to prevent run away. Sometimes a single loop through that torroid can be applied as negative feedback to the input for additional stability. The secondary of T1 provides the low impedance required for the input to the FETs. T3, in conjunction with C8 is the tuned output. What normally follows the output is usually a low pass filter. At 12VDC, this should provide at least 100 watts of RF!

 

[Egzoset]

Salutations Galvatron,

Now that looks like a nice, basic RF amp!

Indeed, sorry for being so negligent i somehow skipped including any references, as follows:

[ https://www.microsemi.com/document-portal/doc_download/14717-a-50-mhz-250w-amplifier-using-push-pull-arf448a-b ]​

Advanced Power Technology, Application Circuits #APT 9702a​

A 50 MHz, 250W Amplifier Using Push-Pull Arf448A/B (1997, 2003)​

The blue part was mostly accessorial but it evoked a common requirement of power devices, especially the analog ones.

And now i find myself stacking more decoupling-related ideas together, for example:

​

​

Although it's just about principles is it at least realistic to hope getting away with only 1 toroïd "choke", knowing the initial circuit had 2 actually?​

​

​

Good day, have fun!! ☮​

 

[Galvatron]

250 watts is really pushing it. Supply must be over 100VDC. The original design is for a 50MHz amp. (6 meters) You can get away with 100 watts at 12VDC if the frequency is lower, especially in the kHz range. Also, there are a variety of FETs that are more efficient at lower frequencies. T3 doesn't have to be made of torroids, it is more common to use a low to high impedance RF transformer similar to T1. Ultimately, your design must account for input power at J1, gain and frequency of Q1/Q2 and output impedance.

 

[Egzoset]

Hi again Galvatron,

...You can get away with 100 watts at 12VDC if the frequency is lower...

The DynaVap IH-driven clan seems to aim for 80 ~ 120 KHz, though i don't really pretend to follow their "progress" as i got something radically different in mind anyway. IMO if the Curie effect is to be evaluated i must tame that down to high ultra-sound frequencies, yet not rejecting the eventuality of some dual-bands system meant to extend compatibility both to IH by magnetic hysteresis losses and also IH by eddy surface-current ones, in a same setup.

Your previous interpretation of the 2003 RF amplifier schematic was brief and concise, somehow i hoped it would feel at home but maybe i should have been more careful with my selection. It's all involving inadequacies, not to mention in the context of my favourite vaporist application.

...your design must account for input power at J1, gain and frequency of Q1/Q2 and output impedance.

Truth is i'm not the right guy to perform an update of the Mazzilli topology. What i contemplate instead is isolated gate drivers as the Maxim Max22701E, or a more classic chip combined to Toshiba's TLP2348 opto-coupler, etc. Honestly i need to check the dates too for starters!... But i figure at least the booming GaN market makes room for even more versatile MOSFET/iGBT/GaN hybrid solutions, so time is working for us.

Good day, have fun!! ☮

 

[Egzoset]

Salutations,

Nearly a month has passed already, so it may now be appropriate to post a modest update...



Relatively to the Mazzilli-topology boards i've been defining/clarifying my perceptions even further recently:

​

By the way, although i'm probably not allowed to publish a direct link which might appear to look just like SPAM it's still a desirable opportunity IMO to inform the audience about RMCybernetic's "CRO-SM2" alternative board where all of those aspects seem suitably addressed, except my "sensor" item that corresponds to a large resistor supplying control-loop feedback in their circuitry.

Dealing with this level of power i personally find crucial to consider/include basic radio techniques as AC Decoupling/RF Shunting (& Shielding), etc. Actually, in the present context where portable battery operation should be incorporated into the global picture (...) it only makes sense to stop depending on Mazzilli's cross-connected feedback schottky diodes directly commanding its power device gates, because otherwise power supply variations from the output are bound to show up at the input of such powerful amplifiers normally characterized by only 2 safe states: On or Off. E.G. transitions are where the ghosts keep hiding precisely.

So, decoding Mazzilli's topology it becomes quite obvious it's no ideal solution for any portable IH driver once realizing that a voltage drop from weak batteries will interfere with the transfer function of gates already plagued with excess parasitic capacitance, on top of so-called "fast" diodes themselves causing more issues by virtue of their reverse recovery charge - another semiconductor junction parasitic, not to mention lead inductance and similar parasitic loops in the PC Board copper traces themselves... Briefly put the popular Amazon/e-Bay "bargains" are dirt-cheap as a result of even the most basic design criterias getting trivially sacrificed, pretending such modules are something they were never meant to be and proof for that is the absence of RC snubbers and other tweaks necessary for reliable dependability.



Actually i'm forced to observe the Mazzilli topology is now acting as an anchor blocking experimentation while it's supposed to fuel the imagination & creativity of "DiY" enthousiasts, truth being it's most often serving individuals after a "product" for thess than pizza delivery, ideally, no matter what's the trade-off. Yet there are serious/severe repercussions for failing to taking charge of (post-prototyping) debuging + corrections.

In any case i wish to publish this fun experiment specifically intended to the attention of hobbyists after some inspiring distration and nothing too involving:

​

{ Scratch 'n Sniff! }​

Please be advised i'm not suggesting to attach anything near an IH workcoil, this is about my "Evaluate CHOKEs as SENSOR" item based on some hypothetical unification of a chokes pair to reduce weight while possibly gaining additional functionality. Also take note i've deliberately avoided the inductor pole dot notation as it's not certain in my mind there's no benefit in comparing both possible pole-orientation scenarios, while the chokes never get sollicitated together presumably: these work in sequence instead, one after the next anyway...

My reasoning is this. If there are microphonics powerful enough to send acoustic waves directly into dynamic earphones (with no power supply around!!) then i gather the same would apply relatively to any power stage depending on chokes as those. It implies galvanic isolation to begin with, which ain't a bad thing at all, quite on the contrary. So far the outcome(s) i foresee would translate as a simplified user-interface where a soft On/Off switch is implemented through the IH delivery tool itself, since it's already in the operator's hand precisely: tap-commands sent through an IH-compatible vaporist wand can be exploited to serve power-management functions as On/Off, for example tap twice at a short interval followed by one longer delay and a 3rd tap to exit stand-by/deep-sleep state and prompt attention for optional/additional adjustments, like output level.

Past this initialization phase then it would make great sense to "KICK" the IH "tank" resonator in order to "energize" it sufficiently for its self-sustaining (natural) oscillations to continue even after power-injection kick-pulsing has past or ceased, to "listen" at the true resonator main frequency long enough to sample it and regenerate it for the duration of a whole "blanking" sequence, which evoques reminescence of the colour-burst reference in the analog-days NTSC video standard. Only 9 ±1 cycles @ 3.58 MHz sufficed to "lock" a local reference signal for I/Q quadrature demodulation of an entire video line, which is about 4 % in comparison to the horiz. rate of 15738 KHz. Assuming an IH frequency of 100 KHz and 5 μsec pulses i figure if the reset phase is 10 μsec followed by a 10 μsec kick then another 10 μsec for PLL purposes then that leaves 85 % of a 20 pulses frame for output level adjustment, while i'm expecting a voiceband (5 KHz) side-product from IH level modulations (e.g. IMO it might sign too during the main power injection phase!)... Simple timeout logic glue can behave both as a "watch-dog" checking for forbiden states as loss of resonance (for self-protection), with adjustable blanking delay to proportionately control how many IH pulses are sent per frame, etc.

In absence of even a single LED the total Stand-By/Deep-Sleep consumption current should prove minimal, which is most desirable for battery operation. It's equally reasonable to envision yet more innovation by using the workcoil to empower circuitry located on the IH wand itself, possibly to support some BT sensor measuring opacity of the cannabic stream and why not its pressure/speed as well. As long as a tiny super-capacitor can sustain it - which is compatible with timing in my toking model i think.

Microphonic feedback from a pick-up coil wound on the same toroïd core as the 2 power chokes would return 2 easily-distinguishable signals: 1 in the voice band and another way up in the ultrasonic one, the 1st corresponding to mechanical vibrations created at the contact of hard magnetic materials in presence of a magnet, for example. Those tempted to duplicate my illustrated experiment must understand although transposition to a 3 windings choke will send lots of energy via its sensor coil that's still going to be 2 very distinct signals each with their own signatures, and there's more...

Considering the particular motivation of DV VC owners for IH driving it's obvious George's clever temperature-dependent "clicker" must be put to contribution and there again it pays not to block the path to evolution, because now that i've refered to tap-commands (maybe with insufficient focus on insertion/removal detection so far) it's time i argue that DynaVap clicker microphonics could as well complete the automation palette by reducing/stopping the heat after sending an acknowledgment user-feedback signal, most hopefully "haptic" in nature as the operator's hand would sense it and react intuitively by removing his IH wand once ready.

Etc., etc.

All things implicitely innaccessible to experimenters dealing with hard-wired limitations imposed to them "by design", to serve goals not necessarily in line with their own. Of course i wish nobody here is against prototyping, but the time for that is over when no new lessons can be learned - keep in mind the Mazzilli 2008 topology was never updated by its author while SiC diodes and GaN power switches arrived on the market years ago.

...

Good day, have fun!! ☮

 

[Galvatron]

If you are still battling ghosts, I still think your best bet is looking at RF amps. For the most part, parasitics have been dealt with.
Here's a thought... Instead of IH which is still heating, ever thought of ultrasonic vaporization at much lower temperature?