Thank you for changing you're mind @RastaBuddhaTao. And free when i offered to pay, is very generous!
Thank you @virtualpurple , I did demonstrate early on in the thread my ability to work with wood with photos of electric guitars I have built from scratch using my own tools . I was quickly dismissed off the team on basis of the safety aspect of the heater construction . Which on that aspect too, you are right one could ask a friend if not having proper fused power source if not having . I did ask earlier in the thread if possible to have a concrete parts name list so that i could procure my own parts. Mesh size, what the copper plate is called , things as such ... but apparently that is unavailable or restricted to those chosen for the project
DoneThanks to everyone for the interest.
Here is the tentative list of builders and projects:
Base kit - heater material, 14/19 mm joints, wire, components, screens etc.
@Boden @oddjobold @Gator
Heater kit - heater, female glass screen assembly, wire and heater mounting hardware with mouthpiece and screen set.
@rz
Zion wood body and heater set.
@oddjobold @Lazy Lighting
Milaana parts kit - all parts minus the wood
@ZC @bellas
Mi1 kit - all parts for a Mi1
@Phenix has lead and has volenteered to make instructional videos. @slatihe and @P.A.M. @stickstones
As @RastaBuddhaTao has done in the past, these kits will be provided free of charge in order to encourage more builders to use the HSA technology here on FC.
If you are still interested please complete a contact form on the RBT www by 2/8 with your FC handle, name, address and a brief description of what you plan to build so that I can customize the kits to best suit your needs.
@RastaBuddhaTao
Will you be starting a project build thread for those of us making mods, or shal we just post here?
When there is a DIY design that is available for sale to all, then I think it would be appropriate to start a new DIY thread. In the meantime, I've moved this thread to the Upcoming and Unreleased section as it is more appropriate here given the direction of the thread.I will leave it up to the mods as to when a design warrants its own DIY thread.
Other than the fact there is no ground in this application (should be batt+) I see nothing inherently wrong.@RastaBuddhaTao Thanks! I'm looking forward to playing with those!
I plan to experiment with controlling power delivery to the resistive load. I also want to test how it behaves with the battery protection circuit I'm currently using.
Since this project is still open-ended, I'm not sure if it'll end up incorporate existing drivers like the DNA boards or end up with something else a little more customized, or a mix of this and that. Maybe a few different versions? If it does end up with some electronics of it's own, which may need some protection, I'll start some relevant discussion based on the limited experience I have.
After some research, I have incorporated over-voltage, under-voltage and over-current/short circuit protection into once of my recent boards, based on this S-8252.
The current detection on this chip works by monitoring the voltage difference between VSS and VM - the voltage built up on the control MOSFETS, so choosing the right transistors is critical. Two transistors are needed for all the features to work properly.
I (finally) found some that look great for the job, and they seem to be working well in my (real world usage) testing so far.
These little puppies (IRLHM620PBF) run at about 2.5mOhm.
This is for a 2S battery pack, with a nominal voltage of 7.4V.
Looking at 15 Amps as a starting point for maximum current, thats ~110Watts. Actual delivered power is a bit less due to losses here and there, so lets call that at least 80 Watts deliverable.
15Amps x (2x 2.5mOhm) = 0.075v = the voltage drop on the two MOSFETs at 15Amps.
I currently have the S-8252AAO variant with with over-current detection at 0.2V on the MOSFETs and short-circuit detection at 0.5v resulting in about 40A over-current detection (8ms window) and 100Amp short-circuit detection (280us window).
The thresholds can be changed by chip variant selection (ideally, when available), mosfet selection (RDS dependent), or lowered by adding series resistance to the MOSFETS. In my case, adding 5mOhm would halve those thresholds.
Ideally we'd like the threshold to be just above what we need - in the example above, 15Amps, however we need to leave some(??) headroom. For instance, If the load wants to draw a lot, and we use PWM to tame it, we have to watch the peaks, not the average. 8ms should be enough time to tame detection if PWMing above a few hundred Hz though. With my IH experiments, I only PWM at ~10s of Hz and a hungry IH easily tripped my circuit while PWMing with the 0.2V detection threshold, So it works! - but some headroom IS needed. I tamed my IH down a bit to suit the protection by removing some capacitance.
These are some of what I'd like to play with for this RBT project - driving a resistive load by either directly PWMing the heater element via some MOSFET at some frequency (maybe with some choke or capacitor - basically a noisy DC/DC converter, which may just be suitable enough for the application), or by the use of an integrated DC/DC converter that runs at a few hundred Khz with low ripple and probably a bunch of extra bells and whistles ( who doesn't like bells and whistles? ), and how all this behaves with the battery protection I've got built up already, making any necessary modifications.
Below is the current design I'm using. It has not been extensively tested in an anti-blow-up-your face lab cause I don't want to blow up my face and don't own a lab as extensive as I see fit, but I do what I can. I'm putting this out there cause well, I don't want anyone to blow up their face. If anyone sees anything wrong with this design, plz discuss. Otherwise, feel free to use I hope it saves your face
!!! DISCLAIMER --- THE FOLLOWING CIRCUIT MAY BLOW UP IN YOUR FACE - USE AT YOUR OWN RISK !!!
Board layout is a separate discussion, as is battery charging..
Thanks @Stu that sounds perfect!
Wow @rz !!! I am a mechanical engineer but ran hardware and firmware teams in India so I know enough to be dangerous lol. Everything you said seems sound to me but someone like @KeroZen or other electrically minded people should take a look.
Hopefully someone will take a 555 timer and make a simple adjustable PWM board as it might be small enough to fit into the Milaana electronics cavity. Dave at Mist and I have discussed it in the past. Or the poor man's version, a voltage regulator to set constant current. This is super inefficient at the regulator "burns off" the peak to get to constant voltage. Could work with a two battery set-up?
Other ideas... LED buttons or LED imbedded in the wood such that it lights up the female glass. Did a prototype on 2015 on the early Milaana days. CAUTION: ensure that the buttons are rated for the voltage, current, at required duty cycle / life.
https://m.ebay.com/itm/Mini-12mm-12...h-Button-Switch-For-Car-Boat-US-/162823591348
Have also thought a hole drilled for a stir stick would be cool in Milaana. You could have an o-ring fit so it didn't fall out.
I have a feeling some serious Highdeas are about to drop?
Other than the fact there is no ground in this application (should be batt+) I see nothing inherently wrong.
Current flowing through a switch is a bit dodgy. Adding a N channel mosfet with a 15k resistor bridging the gate and source and triggering with the switch will last a lot longer.
https://www.mouser.com/ProductDetail/Infineon/IRFZ44NPBF/?qs=9%2bKlkBgLFf24zghzPg2f9g==&gclid=Cj0KCQiA2NXTBRDoARIsAJRIvLzk0H5qLVFoxIeLDGib4F-nTt6vbmGP8qGJwgXP_HgL3JIHeQ8BPowaAl44EALw_wcB
Yes, thus the disclaimer to ensure the switch is robust enough for the application. The nice thing about using a MOSFET is that you can ise a small, light duty button. This opens up tons more options for buttons given that the MOSFET is taking the brunt of the load.
The button in #rbtmilaana has been endurance tested in an over voltage situation for the equivalent of 5 years of heavy use without failure. Only one button in the field has failed in the last two years which was most likely a manufacturing defect. Provided adaquate factor of safety is applied, a direct contact switch is a reliable solution. It's a matter of a cost benefit trade-off for a given design.
@Boden earth is at the bottom right. The transistors disconnect the Battery negative terminal from the board earth. There are also 'high-side' protection chips. I started working on a high side implementation to avoid ground difference issues that would appear if using the device while charging and may wonk the balancing circuit out. I opted to stick with the low side one for now since i know it works to some degree, and concentrate on the charger. if the issue is significant enough (I don't think it will be an issue.. but compromises.. everywhere), I'll move over to high side protection..
Gotcha. I generally design battery powered circuits with the “earth” +@Boden load is connected between batt+ and earth. Earth is essentially the batt-, unless the BMS disconnects it.
Hey @KeroZen and RBT Open Source vapor makers...can you share your plans for what project(s) you intend to work on? Ya'll gots to be havin' some highdeas, no?
This will help me when preparing the material kits.
Since a OKLT20 needs two cells in series I would not charge them in a mod.@oddjobold Battery and heater wiring looks good. The wire from just above B- (where a GND post is on the board) to the GND on the top left is not really necessary as these points are connected internally on the board.
These DNA boards look pretty cool. They have a load of useful features, including reverse battery protection, and seem pretty straight forward to wire up. I think it'd be a cool project that almost anyone could build with the right kind of kit and instructions with basic electronic tools.
A cheaper method with fewer features which seems to be used already are the integrated DC/DC modules.
Here's a kit to get an idea (this ones sold out, but I'm sure others exist..)
https://shop.findmyvapes.com/products/diy-okl-t20-box-mod-kit
and a build video (not by me):
If anyone wants to get started with builds like these and I'll do what I can to help with electronics.
There are some features lacking from that module, like a built in USB charger. It may be feasible to build a ~minimalistic motherboard for that module, with just a few extra features. This is something I could maybe get around to but no idea by when. I'll be getting myself an assortment of modules to play with in the meanwhile
Since a OKLT20 needs two cells in series I would not charge them in a mod.
Well yes you could add all that complexity and do it. Or... you could pop them in a dedicated charger, pop some freshly charged cells on the mod and be on your way.Not even with a designated 2S charger with relevant current limitations and an NTC for thermal throttling? built in balance charging? I'm not sure if there are other reasons not to.. ?
I've found that 2S batteries stay pretty close even after a bunch of charges. I'd say with a decent pair of same-batch cells you could get away with no balancing and limit your charge voltage to ~8.2V either with a simple on board charger or a suitable external one, while charging externally every few weeks to get back into max balance.
The advantage of adding a BMS circuit over, say, the built in over-current used in the OKL chips, is the per cell under/overvoltage protection. This may not even be a necessary addition with an on-board balancing charger.