Electronic Cigarette Smoking Machine Controller Needed

[Chris Havel]

I am an Analytical Chemist working at the University of California San Francisco. We are currently funded by the FDA to study electronic cigarettes.

I have built an e-cigarette smoking machine that consists of a custom built controller that sends 12 V dc to a 3-way valve that directs a vacuum to the e-cigarette vaporizer and at the same time triggers a relay to direct the current from a constant voltage power supply to the coils in the vaporizer. The controller sends the 12 volt 4 second long pulse every 30 seconds to simulate e-cigarette smoking. The vapor is then collected in a series of gas washing bottles for chemical analysis.

I am looking for a better and cheaper controller. Does anybody know of an off the shelf controller that will deliver variable length pulses at variable timed intervals? Ideally, I would like to have 2, 3, 4, and 6 second long 12 volt pulses at 15, 30, 60 and 90 second intervals. It would be really great if there was a digital read out of the timing. I have attached a pictures of the smoking machine in action, the relay and the valve with the e-cigarette clearomizer attached.
I am using a 30 watt power supply to power the coil in the e-cigarette.

 

[KeepItSimpleStupid]

I'm not sure better and cheaper go together and I'm not sure you specified your system properly yet either.

I see a missing parameter or one that isn't specified. How do you plan to "stop" the process? Elapsed time? Smoking time? or something else or doesn't it matter?
e.g. elapsed time 20 minutes or say 15 min worth of draws or maybe some other way to detect done? (Pressure?)

How many systems? or outputs?

I'll; give you a few terms: www.picaxe.com. flowcode PLC, smart relay, I2C and Arduino

A smart relay (e.g. IDEC) might be able to handle 6 outputs or so and might be able to simultaneously do one experiment. So you have 4 rates and say one delay. Do them all at once. Stop each independently when so many seconds have been sampled or after an elapsed time. The IDEC IDE is essentially free, so you can try it out without buying it. So, you could have outputs and a done flag and one start button. the keypad can set the other parameters.

 

[BobW]

Maybe a "smart relay" such as a Schneider Zelio:
https://www.schneider-electric.com/...-plc-for-commercial-machines/531-zelio-logic/

 

[Flyback]

sounds like a job for a small microchip pic.
You can learn them best via the babani book on pic microcontrollers.
Microchip have a good forum that can help you with this.

All you need is simple delay code, to get your intervals, and an output to go via a tranasistor to activate what you want.

Having said this is as easy as falling out of bed, and it is, the act of learning to do this may be quite long due to the lack of good beginners tutorials...but as I said, the babani book is best and is cheap and short time to read.

 

[Dr_Doggy]

very nice , i have taken recent interest and just started smoking one of these , Im sure you know but my pen only goes to about 3.5 v, some go up to 5v, and my coil is around 1 ohm i think, are you doing the formaldehyde tests?

I would advise getting separate dc supplies a 12v for fans. and a variable voltage power supply up to 18v for the ecig supply,
Also will you want your ecig to shut off at the same time as the air pump? or a sec later to clear chamber?
I have heard of cases or "raunchy smoke" which is a) traces of bad compounds, b) decayed coils. Are you doing tests of power surges, dead batteries , exploding coils(end of life)? which are common in the field?

Also I wonder if you thought of testing different flow rates as well, for example when i really pull on my cig i get small cloud of smoke on full lungs and no noise, but if i dont draw it fast enough i hear a crackling, and i think i taste liquid that splashes up., but anyway these crackles are "hotspots" generating, which may be concern. so maybe variable fan/pump supply is good idea too.. and a flow-o-meter .. to track it

it will be interesting if you are able to post links to results!

 

[KeepItSimpleStupid]

I did see the use of a 3-way valve, but I'd like to see the final setup in a hood or somewhere where the effluent (vent and after the traps) goes some place else other than the room.

The OP probably doesn't work weekends .

 

[Chris Havel]

I'm not sure better and cheaper go together and I'm not sure you specified your system properly yet either.

I see a missing parameter or one that isn't specified. How do you plan to "stop" the process? Elapsed time? Smoking time? or something else or doesn't it matter?
e.g. elapsed time 20 minutes or say 15 min worth of draws or maybe some other way to detect done? (Pressure?)

How many systems? or outputs?

I'll; give you a few terms: www.picaxe.com. flowcode PLC, smart relay, I2C and Arduino

A smart relay (e.g. IDEC) might be able to handle 6 outputs or so and might be able to simultaneously do one experiment. So you have 4 rates and say one delay. Do them all at once. Stop each independently when so many seconds have been sampled or after an elapsed time. The IDEC IDE is essentially free, so you can try it out without buying it. So, you could have outputs and a done flag and one start button. the keypad can set the other parameters.

Good idea on adding a prescribed puff number control. Right now I sit there with a timer and turn off power supply after our standardized 15 puffs. The great thing about studying e-cigarettes as opposed to t-cigarettes is that you can weigh the device before and after smoking so you know exactly how much e-liquid was vaporized. If you know the nicotine concentration, which we do, you know the dose. Something that is not possible with t-cigs because of side stream smoke.

Thanks I'll take a look at the link.

 

[Chris Havel]

I did see the use of a 3-way valve, but I'd like to see the final setup in a hood or somewhere where the effluent (vent and after the traps) goes some place else other than the room.

The OP probably doesn't work weekends .

90+% of the vapor is in the first trap, 10% in the second and 0.1% in the third. Nothing gets out. No side stream smoke with e-cigs. When I run tobacco cigarettes I'm definitely in a hood. The side stream smoke is 10 times more dangerous than main stream smoke.

Thanks for the concern.

 

[Chris Havel]

Maybe a "smart relay" such as a Schneider Zelio:
https://www.schneider-electric.com/...-plc-for-commercial-machines/531-zelio-logic/

I'll read about it. Thanks.

 

[KeepItSimpleStupid]

I thought there could be residual smoke from your vent port. e.g. your drawing while heating or your venting the e-cig.

Some valves aren't rated for vacuum, but your probably OK.

IDEC, Schneider are Smart relay manufacturers. Some one makes a Flow Code version too. I did play with the IDEC simulator. You need a development system though.

Other options:

https://www.tri-plc.com/t100md.htm

Here https://www.kg4jjh.com/arc.html is a project using the CuBlock series of controllers.

 

[Chris Havel]

Why can't I buy a digital I/O setup like this **broken link removed** or https://www.ni.com/digitalio/? I think all I'm really looking for is a USB or PCI PC controller that will supply variable length contact closures at defined times. Being able to specify the total of cycles before shutting off the run would be a nice to have. Control through a PC especially USB so I can use as laptop seems the way to go?

 

[MCU88]

You don't need an computer. Do something standalone. 5-minute job using an 16F series PIC and the mikroC language.

 

[BobW]

Why can't I buy a digital I/O setup like this **broken link removed** or https://www.ni.com/digitalio/? I think all I'm really looking for is a USB or PCI PC controller that will supply variable length contact closures at defined times. Being able to specify the total of cycles before shutting off the run would be a nice to have. Control through a PC especially USB so I can use as laptop seems the way to go?

These should work fine, but of course you'll still need software to control it. If you already have NI Labview, and familiar with it, then this may be your best choice. If you need to buy the software then there may be cheaper options. I've used the NI products for several projects and it's good stuff.

I've also done projects using PICs and projects using Programmable Logic Controllers (PLCs) and smart relays. So I am familiar with all of these options. All of the suggestions given in this thread will work, but what works best for you will depend on your familiarity with the required hardware and software.

I suggested the smart relay because it's a self contained unit with terminal blocks for all of the wiring, so there's no soldering required, and it should have more than enough I/O for your application. The programming software is a free download from the manufacturer's site. So the hardware is the only thing you have to buy. Once it has been programmed, then it can be standalone, and you can disconnect your computer from it. However I haven't checked their prices recently. So they may be more expensive than the NI or Contec products.

I would only suggest a PIC or other microprocessor if you already have experience with them. With all due respect to MCU88, if you haven't previously used a PIC, it will take a lot longer than five minutes getting up to speed—more like a week. Plus you still have hardware interfacing issues and printed circuit boards to build.

 

[Chris Havel]

You don't need an computer. Do something standalone. 5-minute job using an 16F series PIC and the mikroC language.

I need something more turn key. the 16F PIC is a chip and I don't know how to integrate it into an IO box let alone program it. Here is a picture of the box I want to replace. It cost$700. It does the job of doing 4 second puffs every 30 second but it is clumsy and doesn't give me much control over the puff duration and timing as I would like. There are other labs that want to build this machine and I would like to refine it a bit more as well as save them some money. Although the guy who built and programmed the controller might charge a little less now that it has been developed.

 

[KeepItSimpleStupid]

It looks a bit like chewing gum and paper clips. The case, power supply and real-estate (PCB space) are some of the things that don't seem to be addressed.

I did ask earlier, how many setups do "you" plan to control? We know now that there is to be more than one or instructions to make them. The smart relay does look like a winner.

You might have a site license for LabView and the license terms may or may not apply. Labview has a STEEP learning curve. The smart relay doesn't. You can test out the program without buying the relay.

 

[Chris Havel]

It looks a bit like chewing gum and paper clips. The case, power supply and real-estate (PCB space) are some of the things that don't seem to be addressed.

I did ask earlier, how many setups do "you" plan to control? We know now that there is to be more than one or instructions to make them. The smart relay does look like a winner.

You might have a site license for LabView and the license terms may or may not apply. Labview has a STEEP learning curve. The smart relay doesn't. You can test out the program without buying the relay.

I'll only be controlling set up at a time. I talked with National Inst. They're going to send me a quote. He said UC has a site license for Labview. I don't have the time or patience for steep learning curves though.

 

[KeepItSimpleStupid]

If VI's (Virtual Instruments) exist for your hardware, then it becomes easier. LV is very version specific and once it's converted to a higher version,you generally can't go back. Sometimes, you have to incrementally upgrade. They get students hooked so when they go out into the world, they are already familiar. VI's (Instrument drivers) are only available in specific versions or only the latest version. You can create executables that are freely distributable with certain versions of LV. As usual NI hardware with NI software works well. Measurement computing make some interesting hardware.

Data-Flow programming is hard to get used to, but it makes it super simple to code processes that run in parallel. It has a side-effect of having the possibility of "race conditions". You can think of the VI as "When all the data is present, execute"

Error handling takes a bit getting used to because it's more like a "bypass". If error exit; If not error execute this. Then the last one at the end of the "wire" handles the Error.

In the major system I designed, I built in simulations where the hardware did not have to be present. This meant development could occur without the hardware and the 4 minute acquisition would translate to seconds.

 

[Chris Havel]

Finding a simple a off the shelf turn key controller isn't looking all that simple. I think I'll send the guy who built my controller a link to this forum and see if his electronics guy can come up with a better box. I want one where it is easy to set a larger number of puff duration and frequency times than the one I have now. I would also like to be able to set the number of puffs and then stop. Also, a nice digital display showing the settings and with timer displays would be great.

If you want to see what commercial t-cigarette machine look like check out borgwaldt: https://borgwaldt.hauni.com/en/instruments/smoking-machines/linear.html. I think they start at $30K.

Thanks for all your comments and help.

 

[KeepItSimpleStupid]

Actually, what I really think your after is a machine with a better HMI or Human Machine Interface. A couple of buttons and knobs won't do, but

Input x
Input y
Input z

would.

The Arduino development environment combined with a PC might work.

Let the Arduino do the timing and interfacing and let the PC do the inputing from the IDE (Integrated Development Environment)

As long as you don't need logging of other stuff or need something more complex, it might work too.

I think you can ask for input similar to a old CRT terminal and I know for sure you can output to a window text.
You MAY be able to use VT100 emulation and output to specific screen positions. My Arduino experience is a "Hello World" , blink the LED under Linux.

 

[KeepItSimpleStupid]

Here's https://www.arduino.cc/en/Tutorial/ConsoleRead a Arduino Console example/tutorial.

While not as trivial as Input "Duration", X. I think it does offer the ability for you to say monitor the lab from your office.

With an IP to USB serial device server, you could use the PC in your office to start/stop/monitor the lab.
Maybe not both at the same time, though. Add an LED for "Working" or add a local "Start/Stop" button.

==
LabVIEW offers a webpage view of a process too. I forget how write was handled, but I THINK, the webpage was read-only.