• Members of the previous forum can retrieve their temporary password here, (login and check your PM).

Pulse-Heating DiY « Cube-of-Borg » concept vs ghosts in the machine.

Egzoset

Vaporist of Borg
Salutations,

This dedicated thread ain't about "semi-DiY" modding, hacking nor cloning.

Years ago i started to focus on workcoil geometries applied to Induction Heating, unsatisfied with the over-simplistic 1-D single-phase 2-poles air-core selenoïd. Instead, stepper motor parts prompted my imagination, wondering what the benefits of flux concentration might be like after making such transition as evoked below:


Please take note my next illustration was only meant to communicate this alternate perspective in a visual manner, e.g. it never been actually tested through real-life experimentation:

full

If for example we dare consider Camouflet's Inductor selenoïd then the presence of a flux concentrating core will become obvious at a glance alone, no need to dissect it and search inside; what's less evident though is the fact that their design strategy implies only ONE pole can face its target IH susceptor directly... As a consequence the field emanating from its complementary pole gets dispersed unproductively, only to vanish into infinite space.

In comparison the 2-D tri-phase 9-poles layout would always insert our IH heater element between poles of opposing polarity, which completes a magnetic "circuit" that is said to be "closed" rather than "opened". In other words the fields of a stepper motor form rotating closed-loop magnetic paths to maximize their electro-motive efficiency; as a bonus little energy leaks outside its containment shield and that's quite a suitable feature in a context where such stealthy asset can help to improve privacy...

But back in the days when i first developped my 'Next Generation Vaporizer Fantasy' it still wasn't clear what shape the susceptor would have, using a VaporGenie as "Prototyping Platform" i needed a very long while before my findings finally converged with those of the original patent owned by Dan Steinberg, which was about his forgotten fig. #11/#12 item exactly:
vaporgenie-us-7434584-2008-oct-14-egzosets-fig-11-12-supplement-5b480x640-5d-png.87283
full

My humble contribution was to define it more clearly, setting its diameter close to ~12.6 mm after including some heat-decoupling pads. Thickness needed to be 1.0 mm and there were 17 (#53) Holes distributed around 2 concentric orbits, 'C1' (orange) located 2.66 mm away from its center (red) and 'C2' (yellow) at 4.81 mm, while each tiny hole measured 1.5113 mm (dia.) which corresponds to the #53 drill bit standard. Lucky me those 17 holes were already present on some "UFO" shaped smoker filters found on Ste-Catherine street in Montréal during its jazz festival, so it only had to be flattened until i got a disc - the trade-off here being unable to select a proper susceptor material.

Nonetheless my basic Hybrid Core configuration worked great in "Plan-B" (butane torch) mode, not to mention its structure couldn't be any simpler with a pair of refurbished silicon carbide foam slices and my 17-holes Metal Disc in between:

full

A terminating screen was added for convenience as this proved having a tendency to stay clean because of the intense heat.

My LAVACapsule concept calls for quasi-symetric Hybrid Cores, one 'basic' with 3.11 & 2.43 mm thick SiC "pucks" and a 'deluxe' complementary option having a thinner
"SiC Front Puck" but also a thicker "Top-of-Bowl SiC Puck", set at 2.95 & 2.51 mm thick respectively. My definitive Basic & Deluxe configurations were published just in time for the summer solstice of 2018 and i generally settle for 3.1 & 2.4 mm thick. The satisfaction is such it pleases me to tag that solution as my « Weapon Of Mass Seduction ».

Yet there's a catch... Initially i wanted to dream of a portable IH Driver with some Smart-Slave options fitting a pocket format using 2 power cells and a bifilar or center-tap workcoil, some vague idea i once happened to express like this:
full

Many years later i refined that vision by exploiting Cardan's legacy that i had seen in action so often on the farm of my father as a teen, adapted to flat-magnetics through another creative fiction of mine:
full

It represents a pair of identical modules facing each other at a 90 degrees angle, both inter-penetrate to horizontally align their fields in a 2-D dual-phase 4-poles model (instead of 2-D tri-phase 9-poles as with the more complex stepper motor). It's genuinely "creative" because i never seen nor heard of "R.F. Chokes" getting implemented that way, but this would save precious space if it did work for sure!

My problem with it was the very long magnetic bridge linking each pair of selenoïds, on top of its central twin arms... Well that just felt "too much", so i had to fold the fields differently and eventually came up with this (during a particularily enlightening vaporist session i might add):
full

My reflexions were leading somewhere at last! Besides the optimized dipoles proximity such weaving of the windings results in a mechanically rigid layout when compared to side-by-side selenoïds and that's a good thing because i had microphonics in minds...

Finally a workcoil geometry conveniently fitting the flat magnetics format with a flux-concentrating core inside:
full

Starting from there we have the pin positions & assignments necessary to proceed even further with Printed Circuit Board design. Better yet, at 22.5 x 18.5 x 9.5 mm it wasn't too extravagant to think of putting it inside the volume of a King Size cigarette pack, just like those i once used to carry around all day long when a young man. For example:
full

Important aspects start falling in place like pieces of a fancy puzzle. My 17-holes Metal Disc susceptor can get IH "excited" from all sides within its horizontal plane while the number of poles is raised to 18 in presence of 2 trios of "U" shaped workcoils, e.g. vertically supperposed with a PC Board in between - which adds a 3rd dimension and makes it a 3-D dual-phase 18 poles scenario because the workcoils also form dipoles with those on other face! So, 6 + 6 +6 = 18...

One benefit implied by this strategy is the accommodation of my LAVACapsule where 2 Hybrid Cores are required to heat up at once, with each of them aligned to one workcoil trio and the cannabic bowl residing right in the middle. Past that point the challenge is related to how much IH punch a pair of 14500 3.6 Volts cells can supply in burst mode @ 2600 mAh for non-rechargeable units. Assuming an energy conversion efficiency around 85 % i estimated the output would range close to ~15.9 Watts and hence the warm-up time @ 100 Joules becomes ~6.3 seconds, or ~12.6 seconds for my 200 Joules LAVACapsule.

Today's rechargeable cells still can't support such an ideal scenario but years will pass and energy density is bound to improve, so better prepare for that glorious future!

In the meantime lets continue to explore the possibilities awaiting ahead, for example dual-frequency IH driving intended to widen the range of admissible susceptor alloys:
full

My purpose with these graphic exercises ain't to seek on immediate perfection, expect to find errors overlooked here and there quite on the contrary. In this sample it's strongly suggesting the length of copper tracks calls for a global revision of the workcoil assembly nonetheless. My next drawing was an attempt to correct that and only confirmed the pin layout wasn't optimal:
full

At least it did provide more opportunities to improve the concept with a clarification on the micro-switch for mechanical detection on insertion/removal relative to the consumption tool.

Then eureka!
full

Why not make room for 2 micro-switches in order to add consumption tool identification leading to even more versatility of the design, possibly extending the capabalities in its Smart-Slave mode... In addition this other drawing session gave me an opportunity to initiate signals routing, attempting to clarify how that system could be organized around a pair of Hall-Effect sensors so to implement over-current protection and more - and then boom! Lets go back to the classic bifilar selenoïd with a twist (so to speak!):
full

Having rendered the twin bobins independant this meant earch "U" core arm could receive an assymetrical number of winding turns, consequently the terminals could be displaced to better suit the PC Board layout that followed, but 1st lets focus on those terminals exactly, especially those closer to the chips:
full

This way it should be easier to foresee how the pins are going to align when the "U" cores of each PC Board face are being supperposed. Here's the result:
full

Only the bifilar bobin pin positions could be shown, for more clarity. It's still a vertical stack of "U" core trios providing 18 poles in King Size cig-pak format.

Initially i didn't know if there would be sufficient space left where to store one of its associated consumption tools on top of a small amount of the vaporist goodies. Would there be 2 or 3 power cells? How many winding turns?? Etc.

What i do know with certainty is that 14 years later we yet got to see the venerable VapGong/Vaponic tool being finally adapted for Induction Heat:
full

Besides, my solution to avoiding energy waste (by recycling the stream of hot air) in a convection converter was illustrated more than 5 years ago:
full

No love for it neither. Que sera sera! 👴👵👶

Anyway my hope is that planting such seed will give fruits eventually. IMO the ideas are worthy enough to survive in a few minds, until a brave feels ready to take over.

Good day, have fun!!
full
 
Last edited:
Every Valentine's day reminds me of a crucial date, which was the next day in 2011, as i initiated myself to vaporism. That could have been too late in my life, i suppose i was lucky to experience the essentials while most individuals preferred to talk about cannabis only to juxtapose the topic with illness... And now that we got Trudeau's « légaleezation » we no longer exist in my province except when ma$$-media wants to plug more propaganda foccu$$ed on poli¢e and traffi¢, etc.

:rolleyes:

Lets forget it all today. To me it's time to celebrate, except this year i got no publication ready. The path has been walked while i left pieces of bread all along, yet some flat mediatic silence remains because times change: now consumerism is king and all potential audience is gone away, or never arrived... But sometimes i get a nice surprise and yesterday i got one of those quite worthy of sharing with my future readers, the guys who will want to take over when such hobby has gained popularity again. Well, History likes to repeat itself after all!

So, over 8 years ago, as i recall, there been a datasheet from Allegro explaining that their ACS733 chip performed galvanic isolation in a Hall-effect sensor rated 20, 40, 65 or 75 Amperes with a ZERO to 1 MHz bandwidth on its analog output. Better yet, the chip also included a level detector (a digital output labelled as "FAULT"), suggesting great potential in over-current protection and such:
full

This boost of imagination rendered my creativity most fertile, because i could finally conceive how some day a class of brave hobbyists might manage to experiment with synthetic "Royer ZVS" circuits no longer exposed to self-destruction as a result of dropping their FlyBack transformers (previously inherited from the forgotten Tesla-coil era, when its "clever" trick resided in magnetic saturation normally avoided in application types). In short a valuable but invisible asset linked to flux concentration material was sacrificed in cheap asian modules, thought to be convenient simply because « it works! »...

😉

My recent reading stated that Infineon's TLI4971 is rated at 25, 50, 75 or 120 Amperes! So far the better. Oh well, there's a bandwidth trade-off but it doesn't matter much at all, not to mention i learned about it through another paper related to kitchen (3 KW) IH cookers; no workload "scanning" s far but i keep my fingers crossed knowing now we even have radio implemented in software!

Christmass came a bit late this year, and yet the Providence provides when least expected. This is how progress happens, patiently:

Smart Induction Cooktop Reference Design (2024-Mar-26)​
See Fig. 36 page 32-of-115​
TLI4971 high precision coreless current sensor for industrial​
applications in 8x8mm SMD package (01-12-2021)​
See Fig. 2 page 4-of-20​
The beauty in it happens to be about the fact it's all been part of creation since its very begining, only waiting to get revealed 1 step at a time past the Big Bang, which eventually seems a fair bit like poetry.

Good day, have fun!! ☮️
 
Last edited:
Back
Top Bottom