Breaktru Forum

eCigarette Forum => Modding => Topic started by: CraigHB on June 10, 2012, 04:33:16 AM

Title: It's Alive
Post by: CraigHB on June 10, 2012, 04:33:16 AM
I've been working on this mod here and there for quite some time and it's finally done, yay.

I did a video instead of posting pics because my camera finally gave up the ghost and I haven't replaced it yet.  I call it the Powerblok 2200 because it's a box and it puts out a lot of power, up to 30 Watts max or 5A at 6V.  That's not a particularly big number compared to some of the series cell buck mods, but it's pretty darn good for a booster.  The number is for the battery size, a 2200 mAh LiPo flat cell.  The cell size is tacked on there to differentiate it from a couple others I'm also working on, one with a 1000 mAh LiPo and the other with a couple NCR18650As.  Those are also going to be boosters.  As such, this one has a range of 4 to 6 Volts.

The enclosure is a fiberglass nylon composite produced using a process called selective laser sintering or SLS for short.  The stuff is very strong and light.  Really nice to work with.  It comes out of the machine a grayish color so I threw some black spray paint on it.  I suck at painting and I was anxious to get this thing together so the finish is pretty rough.  Though, it's good enough for me.  SLS is not the cheapest way to go, but it's an affordable way to get a factory looking custom enclosure.

That's a tri-color LED next to the touch switch.  It's green for when the charge is okay, orange when it's under 10% and red for low battery.  With my last mod, I found that I don't often check the battery level indicator and sometimes get caught with a dead battery by surprise.  Even though there's a digital readout, the light changes color to let me know when the battery is about to run out.  The light flashes red if there's a fault like a discharged battery, an atomizer short,  or an over-current condition.  The display also illuminates describing the fault.

The charger uses a 1A rate so it's not something you can plug into a computer data port.  A wall wart is required.  There's two charging LEDs.  One is green for when the USB power supply is good.  The other is red for charging then turns green when charging is complete.  Charging goes into standy when the device is active so that's why you see the red light go out when the display is lit or when the trigger is pressed.

The volt down and volt up buttons also turn the device on and off.  Push and hold the top button for on, push and hold the bottom button for off.  I forgot to show that in the video.  The device doesn't actually turn off, but rather "off" mode disables voltage setting and the trigger.  It goes into an ultra-low power mode when inactive so it never actually needs to be switched off from the battery.

Well there it is finally.  Nice to see it finished.  Nice to use it too.  It feels really good in my hand and works beautifully.

I posted a photo of the PCB in the forum here before, It's here (http://breaktru.com/smf/index.php/topic,533.msg3078.html#msg3078).

http://www.youtube.com/watch?v=L33PHLYP0q4
Title: Re: It's Alive
Post by: Topper on June 10, 2012, 05:17:16 AM
Amazing work, again!
Title: Re: It's Alive
Post by: Breaktru on June 10, 2012, 08:08:23 AM
So that's what you've been up too. I was wonder what happened to Craig.
 :thumbsup: Wow, Wow, Wow. Hi-Tech for sure. Fantastic work  ;bow; Luv everything in it. Display, charger, dig pot, size, functionality, enclosure, power, ...  ;hubba;

Title: Re: It's Alive
Post by: Pantera on June 10, 2012, 08:51:41 AM
Pure genius. you sir are one of the elite. I am totally impressed.
 :rockin smiley:
Title: Re: It's Alive
Post by: Haileah on June 10, 2012, 09:04:07 AM
Good narrow size. I thought it would be a lot bigger when I saw it on the PCB in another thread.
Really cool. Hat off to Craig  ;cheers;
Title: Re: It's Alive
Post by: Breaktru on June 10, 2012, 09:31:10 AM
You don't have to answer this but I was curious.....
Is that a bezel over the display or did you just bevel the case. Looks really neat and cool.
Is that a CS controlled D.P.
I like the fact that the display retains the last settings. I assume that the calculations are not in a loop and are refreshed on triggering.
Title: Re: It's Alive
Post by: CraigHB on June 10, 2012, 04:29:38 PM
Thanks guys, I'm pretty proud of myself with this one.  It was a long time in the making, a real labor of love.

The enclosure simply has a chamfer in the rectangular hole over the display.  Just a matter of putting that in the engineering drawing used to produce it.

The digital pot is a Microchip MCP4152-103E/MS.  It has 256 positions and it's 10k Ohms.  It interfaces with a Microchip PIC24FJ16GA002-I/SS micro-controller via SPI.  Voltage control is open loop.  In other words, values are hard coded and do not self adjust.  I have to manually calibrate each one I make.

That is correct.  Values are retained from the last triggered state.  So, changing the voltage clears stored values.  Using the trigger collects new data which is retained until something changes.

I had actually debated on the point in my mind for some time.  I finally decided that I did not want to allow the display to be active and the trigger engaged at the same time.  It solves a couple problems.  For one, I don't have to filter data because it's not dynamically displayed.  Also, I don't have to provide a way to shut off the display.  I would find that annoying if I went to use the device and the display was lit when I did not want it to be.  That thing puts out a surprising amount of light if it's dark at all.  The way it is, I can just tap the trigger to turn off the display if I want.
Title: Re: It's Alive
Post by: Topper on June 10, 2012, 04:40:44 PM
Simple, genius and clever - the way it must to be!
Thanks Craig to share this peace of art with us.
Title: Re: It's Alive
Post by: FlameOut on June 10, 2012, 06:30:33 PM
Dam that is beautiful. I am in Awe  ;hubba; of your super work.
Don't understand the technical jargon but I understand the functionality.
Title: Re: It's Alive
Post by: wwwest on June 10, 2012, 06:39:31 PM
Move over Darwin, Craigwin is here. This is something to be proud. Congrats Craig.  :applaude:
Title: Re: It's Alive
Post by: Barryg41 on June 10, 2012, 06:57:23 PM
That is brilliant! I have been reading about your project for what seems like a long time. I like the size and all that little techie info that it displays. Especially the ohms.

Great job!  :thumbsup:
Title: Re: It's Alive
Post by: rrtwister on June 10, 2012, 08:04:56 PM
Amazing what you mcu guys are doing.
well done craig  :rockin smiley:
Title: Re: It's Alive
Post by: cccsteel on June 10, 2012, 09:41:55 PM
good going...  :rockin smiley:
Bet you can sell a lot of these if you ever considered it.    :thumbsup:
Title: Re: It's Alive
Post by: jester on June 11, 2012, 07:16:49 AM
i would love to be able to do somthing that good but for now i will keep reading and learing for you guys cheers jester
Title: Re: It's Alive
Post by: CraigHB on June 11, 2012, 05:57:33 PM
Bet you can sell a lot of these if you ever considered it.

Yes, that would be very cool to poduce them en mass, but that's a gigantic step going form a prototype to a production unit.  It's not out of the question, but it's definitely a big one.  I think people take for granted what it takes to actually mass produce something.  Even on a small scale, there's big startup costs.  If I could spare the 100 grand I needed to do an initial production run, I probably would, but I don't have that kind of money laying around.  People borrow money or seek out investors to do that kind of thing all the time, but it's always a big risk.

In any case, my primary motivation is to build something for myself that is built to my own specifications.  Maybe I can spread the love a little and sell a few off here and there.  But yeah, very complimentary of you wwwest to use the term Craigwin, implying its on the same level.
Title: Re: It's Alive
Post by: Breaktru on June 11, 2012, 06:25:16 PM
Yeah, 100 grand is a lot not to mention the headaches.

That would be cool maybe making a half a dozen at a time at your leisure and recoup parts money with a bit left over as fun money or other hobby projects, tools, equipment.

wwwest got that right, craigwin  :yes" That VW stuff is no better than what you have right here. If I had a choice of Darwin or Craigwin, Craigwin would be my preferred choice.
Title: Re: It's Alive
Post by: CraigHB on June 11, 2012, 07:03:50 PM
What's Interesting is with the electronics my device comprises, I could quite easily switch to a wattage controlled mode, just a matter of programming.  The hardware is already there.  I prefer voltage control.  To me it's like the difference between a manual and automatic transmission in a car.  The highest performance cars always have a manual transmission.  Though, there are plenty of people that would argue that.  Everyone is entitled to their preference.  For me, it's voltage control.  I'm with Dave on that.
Title: Re: It's Alive
Post by: jomurp on June 12, 2012, 09:02:44 AM
Voltage control has been working for me thus far. i'm happy with it.  Ahhhh who needs wattage control.
i'm with you guys  :beer-toast:
Title: Re: It's Alive
Post by: Tameron on June 18, 2012, 06:22:40 PM
This is truly breath taking. Fantastic work.
I have read a couple of your posts on ECF about you using the TPS43000 and it is a PITA to get stable or something like that.
I know nothing about type III compensation network stuff. Wish there was an easy peezy way.  :laughing2:
I looked into piggybacking two ptn04050c for a possible 24watt output on the TI.com forum but the tech there said it is NOT do able. Damm that would have been an easy way to achieve higher wattage  :crying:
Title: Re: It's Alive
Post by: CraigHB on June 19, 2012, 04:34:41 PM
Thanks, mod is working very well so far, using it every day all the time, love it.  Getting busy on a smaller version, trying to stuff 10 lbs of crap in a 5 lb bag.

Boosters are just a pain all the way around.  It's hard to get them to put out a lot of power and that's why you don't see many high power plug-in modules.   They're also not in much demand.   What we do with them is a pretty specialized application.

I don't see any particular reason why you wouldn't be able to parallel a couple boosters.  You'd have to use a really low DCR battery like a 20C LiPo.  Though, it would be like paralleling diodes.  Whichever one is putting out the higher output will carry the brunt of the load.  I'd try it and see if it works.  Sometimes you think something will work and it doesn't, other times you think something won't work and it does.
Title: Re: It's Alive
Post by: Tameron on June 19, 2012, 06:40:01 PM
Great, you are truly a master to figure out the TPS4300 circuity and pull off 30watts.  :thumbsup:

This is the TI forum about the parallel PTN04050 http://e2e.ti.com/support/power_management/powermodules/f/501/t/85903.aspx (http://e2e.ti.com/support/power_management/powermodules/f/501/t/85903.aspx)
He says better off building one w/ a TPS4300 but doesn't mention the complexity
Title: Re: It's Alive
Post by: CraigHB on June 19, 2012, 08:48:57 PM
I looked at that post.  The operative term is "likely" meaning not for sure.  Hiccup mode is only entered when there's an output over-current condition.  I still don't see how two in parallel could cause that to happen.  Of course it's better to use a single converter, but there's a world of difference between plugging in a module and designing one yourself.  Tom at TI took that for granted in his reply.

Oh, BTW, there is an easier way, but you give up efficiency for it.  The TPS43000 is the most efficient because it's a "synchronous voltage mode" controller.  This is also the least stable of all the controller types and requires the most tuning.

You can use an "asynchronous current mode" controller.  You'll find them relatively easy to build.  They require very little tuning.  They are also much more tolerant to variations in PCB layout and variations in component values.  I would suggest the Maxim MAX1709.

The down side is you give up 10 to 15% in efficiency.  With synchronous voltage mode, you get an average around 95%.  With asynchronous current mode, that can be as low as 80%.  So, as a consequence in rough terms, you'll lose 10 to 15% of your run time on a charge.  It's a reasonable trade-off for the simplicity.
Title: Re: It's Alive
Post by: Breaktru on June 19, 2012, 09:49:28 PM
I think maybe I would like to try my hand at the MAX1709  :yes"

So Craig if I follow the datasheet schematic in Fig 1, page 8 and Fig 4, page 10, I won't have to tune the circuit?
http://www.google.com/url?sa=t&rct=j&q=max1709&source=web&cd=4&ved=0CHIQFjAD&url=http%3A%2F%2Fdatasheets.maxim-ic.com%2Fen%2Fds%2FMAX1709.pdf&ei=8SnhT6uaNIe88ASr3PmRDQ&usg=AFQjCNHBbeGl4fC_KvDVHOIYoBig24ST0A&cad=rja (http://www.google.com/url?sa=t&rct=j&q=max1709&source=web&cd=4&ved=0CHIQFjAD&url=http%3A%2F%2Fdatasheets.maxim-ic.com%2Fen%2Fds%2FMAX1709.pdf&ei=8SnhT6uaNIe88ASr3PmRDQ&usg=AFQjCNHBbeGl4fC_KvDVHOIYoBig24ST0A&cad=rja)

What are my chances to get 4amps?
Title: Re: It's Alive
Post by: CraigHB on June 19, 2012, 10:33:07 PM
It's super easy to build a booster with that controller.  Just follow the directions and use the standard application circuit on page 8 of the data sheet.

The one negative about that controller is it's limited to 5.5V output, but you can probably squeeze 5.8V out of it if you milk it.

I used that controller on the the first battery powered VV mod I built.  Worked beautifully.  Got about 85% efficiency average.  It could put out 4A at 5.5V continuously with no heating issues in the least.  Keep in mind, it is a booster so you'll need a good low DCR battery (either an IMR cell or a LiPo cell) especially if you want maximum output.

I used a single 330uF tantalum polymer (http://www.mouser.com/ProductDetail/Kemet/T520D337M006ATE009/?qs=F4vAmCFCeaslH59lTEZ3fnfXAqa2HDej2wbARLKX5sU%3d) on both output and input instead of the dual electrolytics shown in the typical application.  The tantalums are more expensive,  but they're much more compact.  You could probably just use a high value MLCC SMD cap on the input.  I used the tantalum anyway, but it probably was not necessary.  The output cap is the one that has to have high enough value and low enough ESR.

Also pick a good rectifier diode with low forward voltage.  Pick an inductor with low DCR and high saturation.  Here's what I used;

http://www.mouser.com/ProductDetail/Vishay-Dale/IHLM2525CZER1R0M01/?qs=bOnNDoERoR18uqA6UaJ%252bPgjMt2rhYMOEtdL3DIKYm04%3d

http://www.mouser.com/ProductDetail/Diodes-Inc/PDS1040L-13/?qs=rNZh4F5RthQSjKiPhlxJLw%3d%3d

Get the MAX1709 package with the thermal pad on the bottom of the chip.  When you design/make the PCB for it, put a couple big vias underneth the thermal pad so you can solder it to the ground plane from underneath.  The other option is to use reflow soldering with a bunch of small vias, but if you want to hand solder, that's how you can make the thermal connection.
Title: Re: It's Alive
Post by: Breaktru on June 21, 2012, 09:51:51 AM
It's super easy to build a booster with that controller.  Just follow the directions and use the standard application circuit on page 8 of the data sheet.

The one negative about that controller is it's limited to 5.5V output, but you can probably squeeze 5.8V out of it if you milk it.

I used that controller on the the first battery powered VV mod I built.  Worked beautifully.  Got about 85% efficiency average.  It could put out 4A at 5.5V continuously with no heating issues in the least.  Keep in mind, it is a booster so you'll need a good low DCR battery (either an IMR cell or a LiPo cell) especially if you want maximum output.

I used a single 330uF tantalum polymer (http://www.mouser.com/ProductDetail/Kemet/T520D337M006ATE009/?qs=F4vAmCFCeaslH59lTEZ3fnfXAqa2HDej2wbARLKX5sU%3d) on both output and input instead of the dual electrolytics shown in the typical application.  The tantalums are more expensive,  but they're much more compact.  You could probably just use a high value MLCC SMD cap on the input.  I used the tantalum anyway, but it probably was not necessary.  The output cap is the one that has to have high enough value and low enough ESR.

Also pick a good rectifier diode with low forward voltage.  Pick an inductor with low DCR and high saturation.  Here's what I used;

http://www.mouser.com/ProductDetail/Vishay-Dale/IHLM2525CZER1R0M01/?qs=bOnNDoERoR18uqA6UaJ%252bPgjMt2rhYMOEtdL3DIKYm04%3d

http://www.mouser.com/ProductDetail/Diodes-Inc/PDS1040L-13/?qs=rNZh4F5RthQSjKiPhlxJLw%3d%3d

Get the MAX1709 package with the thermal pad on the bottom of the chip.  When you design/make the PCB for it, put a couple big vias underneth the thermal pad so you can solder it to the ground plane from underneath.  The other option is to use reflow soldering with a bunch of small vias, but if you want to hand solder, that's how you can make the thermal connection.

Sorry to step on your Alive topic about the Max1709. I have a question about the package styles. The TSSOP has 28pins and rated at 4a while the SOIC is 16pins rated at 2.4a. Datasheet also says the SOIC can also achieve 4a in short bursts.
The schematic shows the 16pin version. If I was to use the 28pin, would I just have to tie the additional LX and PGND pins together and leave the NC pins open then wire it the same?

Not sure how to calculate R1 and C3 for 4amps. Do you remember what you use?
Also R2 is 2 ohms. Can I use any type? 1/8w okay?
Title: Re: It's Alive
Post by: CraigHB on June 21, 2012, 04:18:43 PM
No problem there, post what you like on this thread.

You could try the SOIC package I supose, but it may go into thermal limiting at higher outputs.  The TSSOP will not since it has the thermal pad.  Those are really effective at dissipating heat when soldered to a ground plane.  They work amazingly well.

Since the TSSOP is a significantly smaller package, the pins are much smaller and you need a bunch in parallel to carry the current.  So, to answer your question, the LX pins are all tied to together and the PG pins are all tied together.  I just laid down one big pad for each set.

The 2 Ohm resistor is there to act as a light filter for the feedback circuit.  Current draw is very low there.  You could use something as small as an 0402 package if you like.

The feedback capacitors I used are the ones as shown in figure 1.  R1 is only required for current limiting.  Since we don't want to limit current, R1 can be removed.  C3 governs the soft-start speed.  .1uF  (100nF) is a reasonable value for that one.

Oh, one other note, figure 1 shows a set output voltage.  Look at figure 4 to use a variable voltage.  Replace R3 with the variable resistor.  However, if using a digital pot, be careful not to power up the converter with the digital pot powered down.  That puts Vout on the FB pin since R3 is absent.  The FB pin is not tolerant to voltages greater than Vin.  I burned a couple chips doing that.
Title: Re: It's Alive
Post by: Breaktru on June 21, 2012, 05:49:05 PM
No problem there, post what you like on this thread.

You could try the SOIC package I supose, but it may go into thermal limiting at higher outputs.  The TSSOP will not since it has the thermal pad.  Those are really effective at dissipating heat when soldered to a ground plane.  They work amazingly well.

Since the TSSOP is a significantly smaller package, the pins are much smaller and you need a bunch in parallel to carry the current.  So, to answer your question, the LX pins are all tied to together and the PG pins are all tied together.  I just laid down one big pad for each set.

The 2 Ohm resistor is there to act as a light filter for the feedback circuit.  Current draw is very low there.  You could use something as small as an 0402 package if you like.

The feedback capacitors I used are the ones as shown in figure 1.  R1 is only required for current limiting.  Since we don't want to limit current, R1 can be removed.  C3 governs the soft-start speed.  .1uF  (100nF) is a reasonable value for that one.

Oh, one other note, figure 1 shows a set output voltage.  Look at figure 4 to use a variable voltage.  Replace R3 with the variable resistor.  However, if using a digital pot, be careful not to power up the converter with the digital pot powered down.  That puts Vout on the FB pin since R3 is absent.  The FB pin is not tolerant to voltages greater than Vin.  I burned a couple chips doing that.

Thanks Craig, placed my Mouser order, had a couple of Max1709 on hand from a previous sample order. Eagle cad program is such a pain to find components to match. Wish the makers of the library had better descriptions to aide the search feature. I previously and also today had to make my own components for the library. Another PITA.
Title: Re: It's Alive
Post by: CraigHB on June 21, 2012, 06:03:41 PM
I also use Eagle.

I don't use any of the canned components.  I build my own libraries for all my components.  It's mainly because all of the canned ones are designed for reflow soldering and I do everything by hand.  I need longer tails on the pads.

One thing I do in Eagle that is unusual is that I don't use the schematic editor.  I design my boards by manually placing signals in the board editor.  So, I only define the board symbols in my part libraries.  It's odd, but I find it easier, especially since I use a CAD program I like much better to build my schematics.

In any case, I'm not a big fan of Eagle, I find it clumsy and poorly engineered for the most part.  However, it's cheap and widely used so that's what I use.   Also, the fab I use accepts raw Eagle board files which is highly convenient.
Title: Re: It's Alive
Post by: dennis15 on June 24, 2012, 11:11:08 AM
 :rockin smiley: you rock Craig. Your MCU mod is terrific  :thumbsup:
Title: Re: It's Alive
Post by: CraigHB on June 24, 2012, 02:44:50 PM
Thanks, working on a smaller one right now, about a third shorter, found a 1300mAh cell for it that may work well.  Really pulling my hair out on the PCB.  It was already hard to stuff everything on the board size for the 2200.

Got some bitchen 510 connectors coming I had custom made.  They're made out of stanless steel with a brass post and delrin insulator.  The nut is nickel plated brass so it's solderable.  The fab sent me one to check before doing the production run on the rest of the order.  The video shows the base and nut without the post and insulator in there;

http://www.youtube.com/watch?v=jpsp6QAe3hE


Title: Re: It's Alive
Post by: Breaktru on June 24, 2012, 03:26:03 PM
Thanks, working on a smaller one right now, about a third shorter, found a 1300mAh cell for it that may work well.  Really pulling my hair out on the PCB.  It was already hard to stuff everything on the board size for the 2200.

Got some bitchen 510 connectors coming I had custom made.  They're made out of stanless steel with a brass post and delrin insulator.  The nut is nickel plated brass so it's solderable.  The fab sent me one to check before doing the production run on the rest of the order.  The video shows the base and nut without the post and insulator in there;

So your controller mod will be even more fantastic. Wow.

That is really a cool 510 connector. Looks like a built in drip catch too. Is that a cut for a tool to hold the base while you tighten the bottom nut? Fantastic idea. And another plus is the length. Looks shorter so not to waste precious inside space.
Title: Re: It's Alive
Post by: Pantera on June 24, 2012, 03:32:19 PM
Double Plus + Plus for craig. Cool  :beer-toast:

How much will the 510's cost you? Looks expensive but you couldn't ask for a better connector.  :applaude:
Title: Re: It's Alive
Post by: CraigHB on June 24, 2012, 03:50:41 PM
Yes, that's a catch cup, but also acts as a flange to support the 3/8 panel nut, double duty there.  The connector makes the atomizer connection almost flush with the mod, short of about 2mm.  I have a slotted bit that fits the cutouts in the top of the connector, so yes, it's for holding the connector while tightening.  Most atomizer connectors have those slots to act as airways, but they're deep enough on the custom one to use a wide screwdriver.

The main feature is the base material, 304 stainless.  There are no stainless connectors out there.  All of them are brass or nickel plated brass.  I'm always replacing them because the threads wear out.  Brass is not a particularly hard metal.  Stainless is very hard so the threads will never wear out.  Plus, the insulator is delrin so it will never split and short out.

The one negative is that I couldn't get a high temp plastic like PEEK or polyimide for the insulator because of the cost.  It's surprising how expensive it is to have stuff made out of it.  I have to solder the post prior to the press fit installation because the heat of soldering can damage the insulator.

My cost on them is $15 a pop.  Gotta love those Chinese machine shops.  As much as I bag on the Chinese, I certainly don't have any problem taking advantage of them when I need something made cheap.  I may have to change my tune on those guys.
Title: Re: It's Alive
Post by: Breaktru on June 24, 2012, 04:17:33 PM
It's good to know that your 510 connector will out last your mods.

Is this the machine shop? ......

http://www.youtube.com/watch?v=yL1AgOqnYYE (http://www.youtube.com/watch?v=yL1AgOqnYYE)
Title: Re: It's Alive
Post by: CraigHB on June 24, 2012, 04:30:28 PM
Hehe, hopefully I won't get any severed fingers in my order.
Title: Re: It's Alive
Post by: tommygun on July 01, 2012, 07:45:03 AM
30 watts is amazing for a boost mod. I don't believe I saw a boost mod put out that much wattage.   :thumbsup:
Not even the Darwin which I think is spec'd at 15 watts max. Even stacking two DNA's are 24 watts max on paper.
Cheers mate  :beer-toast:
Title: Re: It's Alive
Post by: vaporhead on July 01, 2012, 08:24:48 AM
30 watts is amazing for a boost mod. I don't believe I saw a boost mod put out that much wattage.   :thumbsup:
Not even the Darwin which I think is spec'd at 15 watts max. Even stacking two DNA's are 24 watts max on paper.
Cheers mate  :beer-toast:

That is pretty amazing to see a 30 watt boost circuit. Hats off to you Craig  :rockin smiley:
Title: Re: It's Alive
Post by: CraigHB on July 01, 2012, 08:30:45 PM
Thanks guys.

I actually did a little more tuning on it the other day, I can get up to 40W out of it now.  Still, the MCU programmed limit will remain at 30W, but that's so it can for sure hit the programmed current limit before destabilizing with a duty cycle too high.  The last one I made could only get to 27W output at 6V with the battery close to fully discharged.  That problem is solved.  Now it can do 36W wiith the battery at 3V loaded which is a bit below the programmed low battery limit.

If the converter was tuned perfectly, it would never destabilize, but that's almost impossible to achieve for me.  Firstly, I would need some really expensive test equipment to measure frequency response on the bench and secondly, because I'm using a couple paralleled 100uF MLCC output caps, output ESR is very low and output capacitance is relatively low for a booster.  Both low capacitance and low ESR make the control loop less stable.

I tuned the control loop with a simulation then did some component swapping to get it close enough.  That's a really rough way to do it, like building a table with a chainsaw, but it works well enough and gets me a working circuit without spending a small fortune on test equipment.

Theoretically, I should be able to get up to 50W.  That's the controller's hard wired current limit and the MOSFETs I'm using can support those currents.  The 20C LiPo I'm using can easily support the drain.  However, I would need more output capacitance and more tuning for that.  I figure there's no reason to put in the effort to achieve an output that high.  I think 24W is about the most anyone would ever actually use and even then, it's a super hot vape no matter what you're using to do it.
Title: Re: It's Alive
Post by: columbusbk on July 08, 2012, 07:17:34 AM
Congrats on It's Alive  ;footloose;
Fantastic mate
Title: Re: It's Alive
Post by: poorboy on July 25, 2012, 07:47:23 PM
Craig i hope its not too much, you see i have a lot of questions for iam just a newbie with mcu. I understand now that the booster/buck regulator is just used for the mcu to get a stable voltage supply it needs if iam not mistaken. How do you attach the trimmer pot would it be the same as you wire them in the regulators? How would it signals to the mcu when adjusting voltage/wattage?
Title: Re: It's Alive
Post by: CraigHB on July 25, 2012, 08:19:12 PM
Well, you have a steep learning curve ahead of you, but Breaktru did it and he does not have a degree in EE, though I think he's a pretty sharp feller.

The MCU's job is not to regulate output.  The boost converter does that.  The MCU only sets the value of the digital pot and that determines the regulation point for the boost converter.

This mod is boost only so the range is 4 to 6V.  Boosters can not go below input voltage.  The reason is so that I can acheive the 30 Watts output I want.  Most of the buck-boost controllers out there (i.e. ones used by the Pro-Vari and LavaTube) have internal MOSFETs resulting in lower current limits.  Mine uses a TI TPS43000 controller that utilizes external MOSFETs providing much higher current limits.  However, I would not recommend this controller to a newbie since it requires very complicated tuning to stabilize. 

There is one high output buck-boost controller I've found from Linear Technology, but it also requires complicated tuning and has a high part count since it uses 4 external MOSFETs.  It's this one (http://www.linear.com/product/LTC3785).

Communication with the digital pot is performed via SPI which stands for serial peripheral interface.  It's a widely used chip interface topology that uses 4 pins for chip to chip communications.  Most MCUs have a built-in SPI module and many peripheral chips support it.  There are other interface topologies as well.  I2C is another chip to chip communications topology that is also commonly used.
Title: Re: It's Alive
Post by: silverslayer on July 26, 2012, 12:56:55 AM
Nice Mod Craig. I think those 510 connectors are pretty cool too.
Title: Re: It's Alive
Post by: CraigHB on July 26, 2012, 01:47:48 AM
Thanks. 

Yes, those connectors were a key component in my mind.  Being forced into using those crappy brass connectors with the soft insulators that flatten and split would have resulted in endless frustration.

On my previous mods, I was using a shorty 510 adapter to keep the mod's connector from wearing.  It was lame.  Those shorties would last about 3 months before the insulator was flattened and the threads were worn out.  Otherwise, it would have been the connector in the mod that wore out. 

These new connectors are pretty bulletproof with a Delrin insulator and a stainless body.  So far, I have not detected any wear.

Here's how the connector looks installed;

http://www.youtube.com/watch?v=9QIB2jazLAA
Title: Re: It's Alive
Post by: poorboy on July 26, 2012, 06:12:13 AM
Its getting more exciting craig. Im learning a lot. So if i were to build a mod with a DNa module, what micro controller would you recommend that is easily to work on? I would ask breaktru aswell. I cant wait to play with this new stuff!
Title: Re: It's Alive
Post by: silverslayer on July 26, 2012, 01:38:45 PM
Yes I hear ya on those cheap connectors. I am really surprised that someone has not came out with a decent connector. All this modding has been going on for a long time, and I would think some one could make a small fortune just on that part. How much are they setting you back if you don't mind me asking? I've often thought about modding a v3 top cap just for this purpose.
Title: Re: It's Alive
Post by: CraigHB on July 26, 2012, 01:44:22 PM
They cost me $15 a pop from a machine shop in China.  The quality is really good  on them even without considering where they are made.  It would cost 10 times more to have a machine shop make them in the US.
Title: Re: It's Alive
Post by: sterling101 on July 29, 2012, 05:26:33 PM
The question is Craig - do you have any connectors for sale?
Sure would beat the brass press fit ones I've got here that's for sure!

;)
Title: Re: It's Alive
Post by: CraigHB on July 30, 2012, 02:56:30 AM
I can probably sell a few here and there.  I have more than I'll need.  Say $15 plus shipping a pop.  That's my cost on them.  PM me.
Title: Re: It's Alive
Post by: synchro on August 21, 2012, 10:56:21 PM
Are you using I2C for the data bus between the pot, display and MCU? 

That is an amazing mod Craig, very impressive.  Bet you're loving vaping away on a mod you built yourself that bests anything available on the market! :rockin smiley:
Title: Re: It's Alive
Post by: CraigHB on August 22, 2012, 04:30:25 PM
Thanks synchro,

Yes, I am enjoying a mod that performs better than anything else you can buy off the shelf.  I've been making coils with rebuildable atomizers that run nicely around 15W.  Not many devices you can run at outputs like that without some kind of issue.  I'm working on a smaller 1400mAh version of this mod presently, same thing, but 2/3 the length.

The data bus is SPI for both the LCD and digital pot, similar to I2C.  SPI is more common though I2C has its advantages.
Title: Re: It's Alive
Post by: synchro on August 22, 2012, 10:57:30 PM
Very cool, SPI is interesting in that its not formalized like I2C so the Arduino has its own implementation, PIC has its own etc. Definite speed advantage tho.

Are you having the boards built out or are you soldering components yourself?
Title: Re: It's Alive
Post by: CraigHB on August 23, 2012, 01:00:36 PM
The boards are sent out for fabrication.  I can make simple boards myself and I often do for prototyping SMD on breadboards.  However, for actual finished projects, I always have my boards fabbed.  It's worth the expense in my opinion.  You can find shops that will do them cheap enough.

I2C is a more advanced bus architecture, but the main reason I use SPI is actually because of the MCU I'm using.  The I2C module is not included in the MCU's peripheral pin selection feature.  That's a really nice feature because it allows you to assign functions to pins.  That makes circuit board design easier.  It's a hassle to code the pin functions, but it's a lot more hassle to arrange circuit traces in sometimes odd configurations.  Also, SPI tends to be the default bus architecture for many components.  I can get the same digital pot with an  I2C interface, but the LCD I'm using is only available with SPI.
Title: Re: It's Alive
Post by: asdaq on September 04, 2012, 04:25:00 PM
Craig, the circuit, the case and the connector, wow. You make me feel like a caveman whose eyes have been opened to a number of possibilities I really hadn't thought about.  Excellent work, and long time no see.  :)

Best of luck on the smaller versions too!
Title: Re: It's Alive
Post by: CraigHB on September 04, 2012, 04:42:27 PM
Thanks asdaq, nice to see you as well.

I'm excited about the smaller one, should have PCBs today, enclosures are in the mail, and LiPo cells should be here end of the month.  I'll post some pics here when I finish it.  Finally ordered a camera so I'll actually be able to take photos.  My camera died like a year ago and it took me that long to order another one.  That's why I posted a video using my PC camera.

Boxes are much more flexible than tubes as far as having room for electronics.  Boxes are all I've been doing.  This latest one will finally be small enough to compete in size with a tube mod.  It was quite a challenge to pack all the electronics onto such a small PCB.  It's still a box, but it will be pretty short, only 81mm with the same width and breadth.  It will have 1400mAh instead of 2200mAh, all other features the same.
Title: Re: It's Alive
Post by: asdaq on September 04, 2012, 05:36:51 PM
I just saw this thread today, so it was nice to read all at once. If you need help financing the camera, please count me in. ;)
Title: Re: It's Alive
Post by: Str8v8ping on September 10, 2012, 06:16:09 PM
Awesome work Craig . I dont think it can get much better then this.
Title: Re: It's Alive
Post by: CraigHB on September 11, 2012, 07:22:29 AM
Thanks.

Got circuit boards last week and I'm working on the smaller one now.  Running into some issues with the electronics.  The higher impedance of the smaller cell and the use of a smaller inductor with higher DCR is giving me trouble.  At maximal outputs, I'm getting a lot more heating and the booster's NMOS switch is getting too hot.  It drives up RDS(on) which causes the controller to hit the hard-wired "on" voltage limit across the NMOS drain-source.

After some long and tedious searching, I found a Vishay MOSFET with lower on resistance and low enough gate charge.  Hoping that will solve the problem, parts here later in the week.

Another issue I'm having is that the smaller board and lower heat dissipation is tripping the parallel PTC fuses I'm using.  You can trip any PTC fuse by getting it hot enough regardless of current flow.  They're on the same pad as the input PMOS and it generates a good amount of heat at maximal outputs.  I've decided to forgo the use of them in lieu of a temperature sensor on the booster's drain pad.

I've used all my pins on the MCU so I had to use one of the programming pins with a jumper for the temp sensor.  I don't like it since I'm now relying 100% on the MCU for safety features and I always like to avoid jumping programming pins, but I don't see any other way around the problem.  To keep input impedance down, I'd need bigger fuses and I just don't have room for them.  This board is tiny and I keep adding stuff as it is.

I was expecting this build to go quick, but it's turning into a whole new project now.  I feel like I keep digging myself deeper with this one.  I could just say screw it and limit output current to 4 Amps, but I want my 30 Watts dammit.  It's a bitch because at that output, the power consumption for any impedance is a factor of approximately 100 (I2R).  Though, it's a design challenge and I'm up for it.
Title: Re: It's Alive
Post by: sterling101 on September 11, 2012, 05:46:16 PM
If anyone can do it you can Craig!

I'm struggling with the new MCU now I've started setting up in the final config.  Can't seem to stop it resetting so might have to re-do the code from scratch bit by bit until it works.
Certainly a challenge but like you say - I'm up for it too :)
Title: Re: It's Alive
Post by: CraigHB on September 11, 2012, 07:58:30 PM
Hehe, thanks for the vote of confidence.

Unless you've programmed a reset into your code and it's getting called arbitrarily, spurious resets are usually indicative of an electrical noise problem.  It can be a real headache.  The higher the currents the greater the magnitude of noise.  I have a pretty far out ground plane on my PCB to avoid that sort of thing.  Under the booster it covers 3 layers and is isolated from the MCU plane.
Title: Re: It's Alive
Post by: sterling101 on September 11, 2012, 08:35:54 PM
The setup is the same as the old chip besides the internal oscillator but I've got an idea it could be one of the fuse settings doing it as it flickers the screen a lot then resets fully so maybe something to do with the WDT or brownout fuses.
Just need to build the code up bit by bit until it falls over to be 100% sure though as it could be something to do with the portb interrupt handler too using the new generation of MCU with enhanced CCP it seems to tie into the interrupt handler different to the older one.
Will post some info on my thread once I've narrowed it down a touch more so as not to hijack yours too much :)
Title: Re: It's Alive
Post by: CraigHB on September 12, 2012, 01:06:44 AM
Some code failures can cause spurious resets so that may be the issue.  For example, if a stack overflow occurs, the program counter may keep incrementing, loading each NOP in unprogrammed memory space until it rolls over to zero where the MCU's reset instruction lives.  Though it usually takes something pretty obvious for that to happen just from a programming error.  One of the many things that are nice about the 16 bit PICs compared to the 8 bit PICs is the stack is pretty bulletproof.  If there's a problem like that in my code, it gets trapped by an ISR that displays an error.


Title: Re: It's Alive
Post by: sterling101 on September 12, 2012, 03:42:49 AM
Found it, fixed it, vaped it :)  Floating pin messing up everything - so easy to miss when the code to hold it high has changed between MCU's!
Quick re-read of the data sheet followed by some face palming and it's now holding stable.

Next stop, tweaking the PCB a bit as I've had to move some pin assignments around to suit using the ECCP module correctly as it's pin re-assignments aren't quite what I was expecting so buttons and atty have had to be mixed up a touch....
Title: Re: It's Alive
Post by: t-minik on January 03, 2014, 06:52:45 PM
WOW !  freaked_out:

sorry it this one is my 1st post but :


awesome job Craig !

any chance to get some schematics pcb and firmware untill it's stable ?
I really would like to try to build one
Title: Re: It's Alive
Post by: CraigHB on January 03, 2014, 08:07:08 PM
Thanks,

I'm actually on to a new design that is  much nicer in terms of electronics, higher output, wider voltage range.  I'll post some preliminaries on that one once I get the first board up and running, actually getting pretty close to submitting the PCB to the fabricator just about now.

I'm not really willing to to publish any of my designs in an open software/hardware sense, but if you're serious about building something similar, I'd be happy to offer any advice I can lend.
Title: Re: It's Alive
Post by: t-minik on January 04, 2014, 05:51:47 AM
Hi Craig.

Thanks for your reply.

Unfortunately I'll never be competent in software development, I triedmany times but it remains totally dark in my brain.
I'm not so bad to make a guitar fx but that's all, microcontrollers are another world to me.
Hope we get one day an open hard/soft ecigarette project.
Title: Re: It's Alive
Post by: CraigHB on January 04, 2014, 04:49:28 PM
I saw someone tried to start an MCU based mod in an open source manner on ECF.  I haven't visited that forum in quite some time so I don't know what ever became of it.

One of the big problems with microcontroller stuff and open source is it's highly dependant on hardware.  Unlike computer software that is limited to a handful of processor platforms using standard interface protocols, there's a truly vast array of microcontroller platforms each with its own set of software requirements.  Microcontroller platforms just don't cross over much.

So, anything open source would have to be both a hardware and software project.  Even small changes in hardware can require rather large changes in software.  That's the goal of Arduino, to eliminate that hardware dependance, but they really have not made much progress in that area.  The majority of Arduino stuff is still pretty hardware dependant.


Title: Re: It's Alive
Post by: Thurge on January 22, 2014, 01:51:40 PM
Hi Craig,

If someone were willing to try to learn how to build a similar device where should they start? I'm decent with a soldering iron and I can read a schematic (might need some reference at hand for symbols i don't remember) I'm not afraid to learn how to code (I've done some light coding in the past and understand how object oriented languages work etc) I just don't know enough about the which component and why part...
Title: Re: It's Alive
Post by: CraigHB on January 22, 2014, 03:07:52 PM
Hi Thurge,

There's a few of us here playing with mods that incorporate an MCU.  There's a number of platforms to choose from.  Most commonly, you can incorporate an Arduino based system, but you're not limited to that.  Arduino is nice because it provides an easy to use and free (that's the big one) way to program MCUs to your requirements.  Arduino does have some limitations so you may want to look at other options there as well.

You do need to program your MCUs typcially using the C programming language, but you can use assembly for platforms other than Arduino.  I like assembly a lot because it's down at the machine level, but it has the big disadvantage of being tedious and repetetive.  Depending on your experience with programming in general, it may or may not be a big task to become familiar with MCU programming.

In terms of circuit design, well, that's the hard part.  There's lots of issues you can run into in building your own circuits and it's where you'll find yourself spending the majority of development time second to programming.  That is, solving electrical problems.  There's just no shortcuts for that other than experience and study.  You just have to roll up your sleeves and do it.  I still learn things every time I do a new project and I've been doing stuff with electronics for decades.  Depending on your knowledge level, you may want to start with the tutorials at http://forum.allaboutcircuits.com

The soldering is not the hard part IMO.  The trick there is mainly the right tools and equipment.  Soldering itself is really not very hard.  There's some great tutorials on the net about soldering electronics that will get you going pretty quick there.  I would suggest the ones at EEVBlog; http://www.eevblog.com/episodes   See episodes #180, 183, 186, 434, 415.

Making your own PCBs is not terribly hard, but it takes some time to get things sorted out so you get good prints.  The best way I've found is the "toner transfer method".  Having them fabricated is a good option and I would suggest OshPark for that.  In some cases you may want to make them yourself and in some cases you may want to have them fabricated.  http://oshpark.com is a co-op that makes it quite inexpensive to have them fabbed.

Well, hopefully that covers it.
Title: Re: It's Alive
Post by: Thurge on January 23, 2014, 01:11:40 PM
Earlier in this thread (I think...) you mentioned the max1709 as being an easy chip to use. I found the max688 which supports higher output current and higher output wattage (in excess of 20 Watts.) It also has less pins which I would assume means it's more difficult to setup (less pins means you need additional external modules to control it) would that be the case?
Title: Re: It's Alive
Post by: CraigHB on January 23, 2014, 02:58:19 PM
Let me take a quick look at it, need to go to the Maxim site, I go there a lot.  I like their stuff since they always seem to have those osbcure things you might need, but their stuff is always expensive so I only use them when I can't find something cheaper.

Hmm, that part number is no good, comes up with an LDO regulator that has been discontinued.

In any case the MAX1709 is an easy one to use and it can do the job pretty well, but if you can find something with better specs, go for it.  The boost or buck-boost controllers are not nearly as common as the buck controllers so options are a lot more limited.
Title: Re: It's Alive
Post by: Thurge on January 24, 2014, 02:54:04 PM
What about a fixed voltage circuit? I mean something that always gives you 4 volts until your battery gets too low and can handle high enough currents to support sub ohm setups? Has this been done before?
Title: Re: It's Alive
Post by: CraigHB on January 24, 2014, 03:51:08 PM
You could do that with a step-down converter quite simply.  Things get a lot easier when you use series cells, but the issue that comes with that is charging if you want to use a non-removable battery.  Even then, only if you want to use a USB charger specifically. 

The way you do things is highly dependant on what you want.  Once you define your requirements in specific manner, then you can think about how to do things in a specific way. 

A mod that can put out a lot of power is easy to design.  It's when you start adding features that things get more complicated.  Define your feature set first, then think about design.  If you want a mod that only needs to put out 4V at a high power level, the design is almost trivial.  If you want to start adding gauges, controls, and other features then you get into something more complicated.
Title: Re: It's Alive
Post by: whoi on February 25, 2014, 09:09:42 AM
This is awesome. I want to step into microcontroller based mod, just not sure if I have the patience lol