Seeking advice about my own squonk-box, I'm designing the chip and all.

[µDavid]

Hello.

I have a project for building a squonk-box, but I can't resist from designing my own complete system. Built with linear regulators... No I'm just kidding.


The core for the system will be build around a AVR XMEGA MCU that will be controlling the display and make the calculations, I also have planned to use some external circuits to enhanch the accuracy and resolution of the ohm measurement and such.

I'm prone to want to go for DC-DC converter solution, I had a thought that maybe I could use a boost converter to create a higher voltage and then use a 16-bit timer to generate a PWM waveform that is filtered, you might ask then "why employ both a DC-DC converter and PWM when I can just adjust the DC-DC converter?" But the idea was to gain the efficiency of a PWM switched MOSFET(s), I plan to use several MOSFETs in order to get very low RDSon and to spread the heat around. But these are only first thoughts.

I am quite capable of designing circuits if I find some kind of resource, does any one know of any that would be suitable for me to read for this project?
Or can someone suggest some search terms?

I also have a few questions that I was hoping someone could answe, the problem of PWM regulation making noises would be solved as far as I know by increasing the switching frequency to 20kHz or above, and if this was an ordinary SMPS situation then 20kHz would be a very low frequency, why does MOD makers use frequencies such as 33Hz or 800Hz?

Regards

 

[Diescum88]

www.breaktrue.com there's tons on modding and circuits

Sent from my SAMSUNG-SM-G930A using Tapatalk

 

[HondaDavidson]

My $.02.... i dont know shit about the tech... but, if I were to design an regulated Squonker it would be 2x18650 or 26650 powered. And simple buck/boost voltage regulated. 10ml bottle. NOT rectangular. Sliding door. Squonk hole on body... TC and VW would n
Be nice but not required.

Sent from my SCH-I605 using Tapatalk

 

[µDavid]

I have done some work on this,quite a lot actually and Ihave come to an cross-road where I need to make a choice that could change the design very much.

Thus far I have chosen to go with two parallel 18650 batteries since I want at least 40A capability so that I can make it work with coil builds down to 0,05Ω(assuming a 1V minimum voltage) and on that path I have in theory designed a non-inverting buck-boost converter to be controlled by a XMEGA MCU switching at 100kHz. But the hole circuit are beginning to become more complex than I am happy with since I need a separate integrated buck-boost converter to generate the 3,3V for all the support circuits(oh by the way in order to switch the converter at 100kHz I have planned to overclock the microcontroller,the datasheet says that it's internal 32Mhz is the highest it can manage but on AVRfreaks.net I have talked to people whom have run XMEGAs at 72MHz and I believe that I will go for something like 50MHz or 42MHz. But due to that fact I want the full 3,3V power rail).

But then I also need a boost converter to generate the voltage rail for the gate drivers, the best circuit I have found for that thus far by using TI's WEBENCH tool is a buck-boost circuit that dissipate about 2W while not rising above 70degC, and that was under the assumption that I need 1,5A peak for both MOSFETs combined(750mA each). But I have not yet learned how to properly calculate the current required to turn on/off specific MOSFETs but I am well into the process of learning that.

The circuits aren't even complete yet but still I start to feel it is growing out of hand, considering that I have other options.

The resulting devices dimensions is always a consideration and for me the most important aspect is that it is comfortable to grab with one hand, although I wouldn't mind if it was longer than any commercial MOD on the market. So I have been planning to build a device which is noticeably longer than normal mods while quite small around it's diameter(though it won't be round but you know the way around which you put your hand, I can't think of how to say it in English right now).

I also have plans to 3D-print the squonk-bottle in order to use the space inside the mod most efficiently.

But onwards to the crossroad, given the increasing size and complexity if the design I have begun wondering if I shouldn't opt for ether 4*18650 in order to get 8,4V - 6V while still be able to output 40A or use 2*26650 but I am not sure those can output 40A safely, do you know?

Because that would make the circuitry much simpler since:

• I don't need a buck-boost converter any more and can go for a buck converter
• I can choose to use an integrated synchronous buck MOSFET driver which can be driven directly from the battery voltage, I wouldn't complain about getting a synchronous converter at the same time as I don't need to think about the specifics of the gate driver circuit. All the while being able to forget about that boost converter for driving the FETs that the other design needs.
  
• I'm not sure but I have gotten it into my head that most LCD or OLED displays require a higher than 3,3V to function, if that is true then I would this way also make that easier.
• I feel as though I am forgetting something...

If I would go that way I would design the box such that the 4 18650 batteries are all next to each other as if you had put all 4 in parallel, and those four batteries next to each other as a square would dictate the circumference of the box. With the electronics and bottle fitted above the batteries.
This is when a custom designed approx. 8ml bottle would be superb, you can probably imagine now that such a box would be quite small in the hand but at the same time longer than any other mod you have ever seen(probably, not counting those ridicules "super sub-ohm" monsters. I don't mean to say that they are ridicules as such, but I would never want to carry around such a monster honk of thing).

Oh and another little feature I have decided to incorporate in my design, a few times I have fallen asleep in the sofa while holding my vaporizer and then awaken to discover that the liquid as run out onto and into the sofa.
So I will design in a feature that through the menu can be activated and while that is active a buzzer(or some tiny sound transducer) will sound if the device falls over

But what do you think about changing into using more batteries?
The battery time would certainly increase.

 

[DevAuto]

This is definitely an ambitious project. It certainly looks like a lot of thought has gone into the design. With regards to the batteries however, obviously the larger the mod, and the more batteries you use, the heavier the mod will be. At what point does the design become "too big" to be usable? .............

 

[~Don~]

Not sure if you have come across the use of 20700A cells yet...Mooch has done some preliminary testing and they seem promising.

https://www.e-cigarette-forum.com/f...-great-performing-30a-3100mah-battery.784543/

 

[µDavid]

Not sure if you have come across the use of 20700A cells yet...Mooch has done some preliminary testing and they seem promising.

https://www.e-cigarette-forum.com/f...-great-performing-30a-3100mah-battery.784543/

Wow that is awesome!

I have actually decided upon the series batteries solution since it makes the design much simpler, I have picked out a synchronous buck MOSFET driver IC that accepts one PWM signal as the input and then it manages the synchronous switching, "synchronous" means that it uses another MOSFET instead of the free-wheeling schottky diode. Such a diode in my design could be required to dissipate as much as 28W as heat while a MOSFET would only need to dissipate a little under 3W as heat.

And since Mooch have proved that those 20700A can be vaped with at up to 45A I can simply use two of those in series.
Which is great since as DevAuto points out the mod could get too heavy.

But if I would print the box in plastic, 4*18650 would probably be ok but 2*20700A is so much better and more practical.

Super useful suggestion, thanks.
This will mean that I can build a device the same size as dual 18650 mods(mine will be a little bigger, or rather longer while more slim) but one that can use as low as 0,025Ω. Which is cool, I don't know if I will ever be using such coil builds but I definitely want to be able to and see how they vape.

I am very close to building the first prototype which will be great, although the design is made for 40A and 1V-5V output voltage it is a long way until I would push any serious current through the circuit. I will of course not put it near my face until I have tested it and made 100% sure that it is made safe.