Building a Digital Vaporizer (Brain Dump)

[skippymcware]

[EDIT: I took the table one step farther to necessary heater power]

OK. Here is the best thing I will ever give this forum.

I caved and bought a rotameter from mcmaster. This one, actually:

http://www.mcmaster.com/#41945k35/=39v1r8

With all the other testing, this has give me a pretty good idea of the flow rates existing while you are hitting a vape. I haven't had time to put this into anything formal, but the data is very reliable given that I am measuring flow rate directly with the rotameter. I have a video of the tests. I can post that later.
Heater Power [W]:
Hit Length : Average Flow Rate [CFM]: @375 F @450 F
5 .85 86 108
8 .5 51 63
15 .33 34 42
20* .17* 17* 22*

*Assumes outlet temp of 375 and 450
*The 20 seconds hit is extrapolated from the other ones. I consider this about as long a hit as you would want to take.

[EDIT]
Be aware that the heater power is calculated using the formula:

power = flow*(Tout-Tin)/3

as you can see, this formula doesn't say anything about your vapes thermal mass, the parasitic heat losses going everywhere else but into the air, inefficiencies in the control system, or any other details that might affect the final answer. In general, you can consider the power numbers this formula gives as a low estimate. Specifically, this formula addresses a steady state system. That is, you have been hitting your vape exactly the same for so long that nothing is changing and the heater is just giving the air the exact amount of heat it needs to exit at your desired exit temp (Tout). That, minus all the details, is the amount of power this equation is giving you.
[/Edit]

So... there you go. That should make heater choice a bunch easier now.

Hope that helps someone. That damn thing wasn't cheap.

-skippy

 

[Hippie Dickie]

i'm not following ... lung volume is about 6 liters, say, 2 cubic feet maybe. And some of my crew take 30 second hits and then hold it for another 30 seconds (which totally kills the natural rhythm of the session, by the way). So that makes 4 cfm for a hit, or 2 cfm for a 15 second hit.

or are you talking about the fan speed?

 

[skippymcware]

Hippie,

I think your volumes are very near the maximum, which is why we are seeing differences in our results. Check wiki:

http://en.wikipedia.org/wiki/Lung_volumes

Vital Capacity:
The amount of air that can be forced out of the lungs after a maximal inspiration. Emphasis on completeness of expiration. The maximum volume of air that can be voluntarily moved in and out of the respiratory system.[3]

Male: 4.6L = .167 ft^3
Female: 3.6L = .127 ft^3

Now, using the male vital capacity, we can calculate flow rates from this for different hit lengths. Remember, this is the maximum amount of air, so it will be a high estimate.

Hit Length: 5s

CFM = (.167 ft^3) / (5 s) * 60 (s/min) = 2 CFM

Hit Length: 15s

CFM = (.167 ft^3) / (15 s) * 60 (s/min) = .668 CFM

I went through this same process once before. I did all this testing, and it was coming out with flow rates that were well below what I was predicting. It was really just that I was using less lung volume over a longer period of time than I had predicted.

Let me know what you think, but I am pretty confident on this after buying the flow meter and seeing it with my own eyes.

-skippy

 

[Hippie Dickie]

okay, i was off by a factor of 10 in guessing 2 cubic feet = 6 liters.

 

[skippymcware]

I don't even attempt to do that stuff in my head. I have already screwed myself so many times doing that.

Here is a pic of the rotameter:

That's SCFH, not SCFM. Just to avoid confusion.

 

[vap999]

[EDIT: I took the table one step farther to necessary heater power]

OK. Here is the best thing I will ever give this forum.

I caved and bought a rotameter from mcmaster. This one, actually:

http://www.mcmaster.com/#41945k35/=39v1r8

With all the other testing, this has give me a pretty good idea of the flow rates existing while you are hitting a vape. This is just off the cuff, but is very accurate, nonetheless. I have a video of the tests. I can post that later.
Heater Power [W]:
Hit Length : Average Flow Rate [CFM]: @375 F @450 F
5 .85 86 108
8 .5 51 63
15 .33 34 42
20* .17* 17* 22*

*Assumes outlet temp of 375 and 450
*The 20 seconds hit is extrapolated from the other ones. I consider this about as long a hit as you would want to take.

So... there you go. That should make heater choice a bunch easier now.

Hope that helps someone. That damn thing wasn't cheap.

-skippy

Thank you very much for posting this data. It will be useful for anyone designing a vaporizer, and is a significant contribution to the community.

This is the first empirical data I have seen comparing actual hit length and average air flow rates. That you also include the amount of heat (watts) needed is a bonus. Is my assumption correct that "Heat Power" is the calculated power needed to maintain each temperature throughout the inhalation?

Your data also confirms my assumption that about 2 liters of air are taken in per vaporizer hit. In my crudely converting your data to minutes and liters/minute, which I find easier to comprehend than CFM (with a liter equaling about a quart), and multiplying this hit time x hit flow rate, I came up with a fairly consistent volume of 2 liters of air being inhaled, no matter how long the hit. This seems about right. So one way of thinking about inhalation rates with convection vaporizers is about 2 liters or less being inhaled in whatever number of seconds.

 

[skippymcware]

"This is the first empirical data I have seen comparing actual hit length and average air flow rates. That you also include the amount of heat (watts) needed is a bonus."

You don't see it posted because it will never be a perfect guide and it is a pain in the ass to collect. Although, I suppose that it would have been a lot faster if I had just bought the flow meter right off the bat.


"Is my assumption correct that "Heat Power" is the calculated power needed to maintain each temperature throughout the inhalation?"

I edited the post to clarify. I say again, though. This is not perfect data. Don't go and use this without doing any work on your own and then get pissed when it turns out you need more power. You seem like you have already done some testing, so you are probably well on your way even without this stuff. Anyway, I was told that I would need 600 watts by a heater manufacturer. Many told me I would need 200 or 300 watts. This should keep you from going through that same sort of crap.


"So one way of thinking about inhalation rates with convection vaporizers is about 2 liters or less being inhaled in whatever number of seconds."

Yeah, pretty much. It might not be perfect, but I think it is a good approximation to say that most people draw in the same amount of air regardless of the hit length.

 

[vap999]

While on the subject of power requirements to heat air, does anyone know the formulas for the amount of power stored and delivered by a heat sink/exchanger to passing air?

The Purple Days (PD) at <10 watts, my prototype vaporizer at <25 watts and other low-power convection vaporizers with a substantial, e.g. ?100 gram, heat sink/exchanger do a very good job. Obviously, from the data you developed, more power than this is being transfered to passing air for an effective hit. Heat/energy is being stored and delivered as needed by the metal heat sink/exchanger. What are the equations for this?

For example, how much power (watts) is stored and delivered to air passing over X sq. cm of metal heat sink/exchanger at X1 degrees weighing X2 grams with X3 heat conductivity?

 

[skippymcware]

vap,

that's a tough one. I'm not sure you will be able to calculate that without a significant amount of digging around in heat transfer books. I would suggest doing it through experiment. I will look around, though.

-skippy

 

[skippymcware]

ok... my mind was just blown. I just learned about something called arduino.

http://en.wikipedia.org/wiki/Arduino

Arduino is a system that has been developed to make working with micro controllers more accessible. Instead of programming in assembly, or even C, you work in a modified version of C called Wiring. It makes it so simple that I laughed when I saw it. Check the wiki link for some code examples.

For someone like hippie dickie, who is familiar with programming in assembly, this is not very useful. But, for someone like me it changes everything. The only price is that I have to use a bigger chip so it can support all the overhead interpretation. The chip is still .4" x 1.4". With a printed circuit board, this shouldn't be that big a deal. For those interested, here are the components involved. They are so cheap it is crazy.

The $29.95 development/programming board with usb, pinouts, and a reset button:
http://www.sparkfun.com/commerce/product_info.php?products_id=666

The $5.50 IC I will have to use:
http://www.sparkfun.com/commerce/product_info.php?products_id=9217

This almost seems too good to be true. I still have to make sure I can do this in a 1" x 3.25" x .63" space, but I think I will be OK.


PS

Check these CNC machines/robots building this circuit:

http://store.gravitech.us/arna30wiatp.html
*It's the video under "See your Arduino Nano build!"

I am pretty vaped, but that is awesome any day of the week.

 

[Hippie Dickie]

re: Arduino ... it looks like it would be simpler to just learn PIC assembler (LOL). But i think that's just me, after 43 years as a computer geek i'm tired of taking on a new operating system or language. Learning the PIC was like going back to the PDP-11 or the 6502.

The PIC chip in an 8-pin DIP socket is 0.4" x 0.4" x 0.25" high, chip cost is $1.35 in quantity one. The programmer board is $35.

Good luck! and vape on!!!

 

[skippymcware]

ok, update time.

I have been busting my ass on this. Luckily, things are coming together pretty well. As far as the controller side goes, I got myself the arduino duemilanove usb board and worked my way through "Getting Started with Arduino" by Massimo Banzi. I then built the control circuit minus the power side of things. This includes the buttons and led's and temperature input. As far as prototyping is concerned, I will only have to add one PWM output to the MOSFET to make something get hot. I also wrote and debugged all the button and temp setpoint code. That is, I can now adjust the temperature setpoint and have a nice user interface setup. Here are some pics:

The big green LED is actually an RGB LED. This means that it can be red, green, or blue. I have it setup so that it is blue if the heater temp is too cold and red if it is too hot. At the moment, I am using a 10 degree window around the setpoint, but this will probably change. The other LED, which may be hard to see, illuminates when you press a button and stays on until you release it. I used three buttons on the vape. I think it is the same layout as Hippie's. One for -10 degrees, one for reset to default, and one for +10 degrees. I spent a bunch of time "debouncing" the buttons and making sure you could only increase the temp by 10 degrees at a time regardless of how you pressed the button.

I also managed to get a 75 watt 12 VDC 3/8" diameter x 2.5" long cartridge heater with an internal thermocouple. This is on the way. In the meantime, however, the guy sent me a bunch of similar 120V cartridge heaters for free. Here are some pics of those:

I should be able to make a usable prototype by this weekend. I won't have the 12V heater yet, so it will be a PID on/off system that maintains a constant heater temp. I'm not sure whether this will be with the coil heater or one of the new cartridge heaters. It would be nice to try out all those different wattage cartridge heaters, but i would also like to compare this system to the old one.

I had a bunch more to say, but my internet is struggling at the moment and making this painful. I will keep the updates coming as I move forward.

-skippy

 

[Hippie Dickie]

slight clarification on how i have my 3 buttons set up. Two of the buttons have dual function.

The -10F button is also the start button -- hold for 2 seconds to start the vape.

The reset button has two functions:
(1) click to take the current coil temperature as the new set point
(2) hold the button for 2 seconds to reset to factory default settings.

and like for car radio station select buttons, the PIC executes the button action when the button is released, not pressed. (also like in the Broderbund game "AE" for the Apple II)

 

[vap999]

I also managed to get a 75 watt 12 VDC 3/8" diameter x 2.5" long cartridge heater with an internal thermocouple.

-skippy

Can you relate any of your experience purchasing this heater? When I searched, I could not find any stock low-voltage (12-14 volt) cartridge heaters anywhere, with companies agreeing that this is what they would expect (no real market). Was this heater from stock or a custom order? About how much more is it to add a built-in thermocouple to a cartridge heater?

For the prototype vaporizers I'm slowly working on, I purchased 14 volt, 25 watt, 3/8" square, 2" long cartridge heaters from Marathon. I had to order 5 and let them get to it when they felt like it to purchase these made at a reasonable price (something over $20 each).

 

[skippymcware]

vap,

Good question. This heater was a total pain to find. In fact, I thought it didn't exist two weeks ago. The reality is quite the opposite, but let me explain. My first try at getting this heater was with high temp ind. Those guys sold me the coil heater I am using now. He said 3 and 6 volts would be really hard, but that 12 wouldn't be an issue. He did warn that anything over 9 amps would be tough. I later found out that it isn't quite that simple, but it would take me two pages to explain. I thought I was all set, but this guy was so busy that he just never got back to me.

So, at this point, I was relatively confident it could be done, but wasn't sure. I next contacted Hotwatt. Again, the initial reaction to 3 and 6 volts was very negative, but 12 volts got me a "this will be no problem." At this point, I thought the deal was done. Then, two weeks ago, I got an email from them saying that anything over 30 watts would be impossible. You see, the fact that this heater was going to be in open air was really scaring people. My best guess is that they were covering their asses by giving me such a low wattage heater. Regardless, I was devastated by this news.

Then I called Marathon (as you have done). I dealt with them before, and one of their people encouraged me to try the cartridge heater idea in the first place. Mind you, this was at the very beginning of this whole project (around when I started this thread). Anyway, I was close to giving up on this idea entirely when I got this same guy on the phone. He was by far the most helpful, and most interested, heater guy I have talked to. He said they make 12 volt cartridge heaters by the thousands, and even had a bunch of them lying around. Furthermore, he told me that watt densities in the 30's were no problem at all and would only be concerned if the heater diameter was going to be 1/4" or less. In fact, he was the first one to be able to relate to me that the heating elements in ovens are of a very similar construction and have watt densities in the 30's as well. So, without any more delay, I ordered a 12 volt cartridge heater.

As far as prices go, if you don't want any frills on a 3/8" cartridge heater it will run you around $30-$40. That is one at a time. If you buy them in large batches, even a few at a time, the price goes down significantly. Some companies offer exotic sheath materials, but this guy said that a large portion of their heaters are stainless steel. An internal thermocouple seems to run around $20-$30 extra. He mentioned that they often cost more than the heater themselves. After that you need to decide what kind of leads/wires you want on it. What angle they will come out at and what sheath material. You also need to specify if you want a fitting on the end or not. This can be pipe thread or a flange. Lastly, you need to specify what kind of potting around the wires you would like. You don't need this, but the heater draws in air, and moisture, as it cools off it you don't have something there. Most people say that this isn't necessary, but I figure I would get it anyway.

When all is said and done, my cartridge heater, with almost every bell and whistle possible, will be somewhere around $80-$90. I haven't gotten the quote yet, though. My goal in the long run is to modify the design so that I can eliminate many of those expensive parts. I just want something to get going with, though. Another good goal will be to lower the heater wattage as much as possible. The smaller I get this thing, while still getting enough heat out of it, the better this vape will be. I can make everything smaller, lighter, last longer, blah, blah, blah.


Another quick update:

Beyond my best expectations, I got my controller to work on my old heater setup. Honestly, this blew me away and I am really pumped about it. I still have to tune the control parameters, but it should work just fine after that. As it is, it takes forever to settle to the setpoint after a huge overshoot, but then it will stay within 10 degrees. I can't really expect any better, since I just used the tuning parameters that came with the code. Anyway, here is a pic:

-skippy

 

[Hippie Dickie]

congratulations on getting the controller working with your old heater. It is a rush when the prototype starts to work.

But, man, i just don't get the heater cartridge thingie. that's just a nichrome heater in a stainless steel shell, no? and the thermocouple is not providing control is it? it's just for measurement, no? i do admit positioning the thermocouple probe on my heater coil is still a work in progress. and positioning the digital thermometer probe next to it is also too hard -- too temporary and prone to displacement. however, i think i've nailed the "insulating the probe tip" problem with some teflon tubing on part of two coil loops.

And, how are you isolating the vapor/air flow from the cartridge heater?

also, it seems to me, if you reduce the wattage, you give up responsiveness. i don't get any temperature drop with a toke, regardless of how much air is moving through the bud.

yeah, the startup overshoot is a bitch. i have a COILTEMPtrip parameter -- when the temperature exceeds this value, the PIC shuts off the MOSFET until the peak of the overshoot occurs. The algorithm adjusts the "trip" value so the peak happens at the desired set point temperature. This isn't quite right yet, since the overshoot is also affected by the charge of the batteries to some degree (pun?).

blah, blah, blah -- i'm just speaking from my perspective and we're going different design paths i know, but there's some knowledge overlap, no?

 

[jackalope]

I ran across this forum just now and registered to post this entry. I apologize for not reading any other threads before posting.

An arduino seems like a perfect device for making a vaporizer. I've been toying with Arduinos on and off for about a year now so I'm no arduino expert but I've looked at enough code and prototypes to see how useful the arduino platform can be for something like this. With a bit of tuning the Arduino PID libraries should be able to regulate the temperature to within 2 or 3 degrees Fahrenheit. No overshoot, faster heat-up time, etc.

The Thermocouple mounted to the heating element is worthwhile but I would think you'd want to be monitoring the vapor temp as well. In my experience, the thermal mass of the herbs, the amount of heated air contacting the herbs and the flow rate of the hit can vary quite a bit and should be accounted for dynamically.

As for the cost of using an arduino, it will be cheap. Once a prototype is built you no longer need the full Arduino board. Each vaporizer would only need the Atmega chip (~$2) and a couple small parts like a capacitor and a resonator (another $2). Of course, you'll still need the heating element, sensors and a MOSFET which will be the expensive parts but the brain will be a miniscule cost.

Can I suggest incorporating a light that comes on when the heating element is at optimal temp? I have a cheap-o vaporizer that takes a while to heat up. It would be nice to know when it's ready.

Good luck on the arduino project. It looks like fun. I've been trying to get off my ass and start a project like this myself. Of course that means I'd probably have to spend some time not smoking so much.

 

[skippymcware]

HD,

Thanks. I'm just glad it came together that quickly. In reality, it doesn't matter that much, because the 12 volt heater is what I care about. It is definitely nice after all that work, though. Here is how I understand the heater choices. Either I do what you do and have a heater wrapped around an air tube, or put a heater in the middle of the airflow. Each has benefits and drawbacks. I don't know what the best solution is, but I will have tried both after this next version. Here is how I understand the tradeoffs:

Indirect heater: (like my coil heater and your vape)
no risk of contaminating the air
can use bare nichrome wire, but then must use glass tube or insulated nichrome wire (dubious)
less efficient due to heat loss to environment
more direct airflow
challenging to packing due to bare nichrome wire/heated elements

Direct heater:
air contamination risk (as long as we are careful it isn't a problem)
cannot use bare nichrome (this necessitates a stainless steel sheath)
more efficient because all the heat loss goes into air stream
heater obstructs air flow path
easy to package
airflow has the double benefit of insulating the heater from the rest of the device

That's my understanding, anyway. As far as what I am doing overlapping with what you have done and are doing, there seems to be a ton. It is really interesting to see how well some of your experiences have lined up with mine, and how others are distinctly different. For the most part, though, I think all the overlap occurs because this is a relatively simple process and the physics of the situation don't involve that many variables. Because of this, I think we are destined to converge on a very similar solution.

P.S. I just realized that the MAX6675 needs to be left alone for 220ms after doing a conversion. Man, I could not figure out why it wouldn't sample faster. At least I know now.


jackalope,

Glad you are interested. Definitely build one of your own, it is worth it. I learned the whole arduino side of this in about two weeks. Several of the concerns you have raised, if not all of them, have been discussed repeatedly in this thread and others. Take a look if you want to see what came of those discussions. As for the PID tuning, check this shit out:

Haha, victory is mine. I am going to play around with it some more, but I doubt I will be able to beat that. That's Temperature [F] vs. Time [seconds], btw. Almost forgot to mention that the red line is the heater output. 500 is 100%. You can see that it starts out at 100% for quite a while and then turns itself on and off in varying proportions throughout the rest of the process.

-skippy

 

[jackalope]

Skippy,
This is awesome@! I have a million questions but Im going to go read the rest of this thread thoroughly before I ask. That plot is pretty conclusive proof that you've nailed it. Does the graph cover any times when someone was drawing air across the the heating element?

 

[Hippie Dickie]

P.S. I just realized that the MAX6675 needs to be left alone for 220ms after doing a conversion. Man, I could not figure out why it wouldn't sample faster. At least I know now.

you could also say that it takes 170ms (typ) to 220ms (max) to do the conversion after you signal the MAX to start the conversion by setting ChipSelect high.

i read the MAX every 3rd clock tick, i.e. every 300 msec, just to be safe.

 

[jackalope]

I'm thinking of buying a cheap vaporizer so I can canabalize the heating element and the case. It seems like buying one of the cheap-o $30 vaporizers on ebay would be just as cheap as getting the element, case, and whip parts individually.

Is this a bad idea?

 

[skippymcware]

jackalope,

tough to say whether or not that is a good idea. It all depends on the unit and what is inside it. If nothing else, it sounds like a good way to try this out without dumping too much money into it.

 

[Beefmaster]

I'm thinking of buying a cheap vaporizer so I can canabalize the heating element and the case. It seems like buying one of the cheap-o $30 vaporizers on ebay would be just as cheap as getting the element, case, and whip parts individually.

Is this a bad idea?

First time post I usually have nothing to say but the heating element in those cheap ebay pieces are bare nichrome, even though the listing says ceramic. I bought one of the noble vapes and the first time i heated it up the glass male inside box cracked and a piece fell off, the metal must of expanded faster then the glass. The controllers are cheaply made and slowly decrease in temp over time. So really your buying a whip which you could get cheaper.

 

[jackalope]

Thanks for posting that info. Knowing that will save me a lot of time and money.

I'm surprised that heating elements are so expensive. Even the ceramic elements are simple components and I feel like the only reason they're so expensive is because I'm not trying to buy 1000 of them at once.

I wonder if there is some househod appliance I can pull a ceramic element out of.

 

[Hippie Dickie]

imho, heating elements are so expensive because 400F is very, very hot -- hot enough to cook a turkey. And then, it has to be safe and reliable. Good design is expensive.

And this is a very custom application at this point in time, so yes, buy 1000 and the price will drop.

When i first started my vape project i got some ceramic "washer": 1/2" dia, with a 1/8" hole, and 3/8" thick, for $1 each -- these were surplus from another guy's order. It is impossible to find such a thing for that price now. i think ALL ceramic is custom made for a particular project.

some soldering irons have a ceramic element - check ebay.