I was reading a topic from Forest_Child( https://www.dmt-nexus.me...t=102115&find=unread ) about a possible new vape tech. I think that it could work, but there are some options in the market, so maybe it would be a bit redundant to make the wheel again.

That idea gave me another idea:
Maybe an automatic DMT extractor would be even more interesting. It could be safer and healthier. Imagine this: put MHRB,vinegar, lye, NAPHTA, and water in the machine, and the next day we have stardust ready to be used. Awesome. I can envision a DIY project for that, which could be replicated, possibly using Raspberry-Pi. An open-source project for both software and hardware, designed with replicability in mind (maybe utilizing lab glass?)...

In an ideal scenario, it might be possible to create an open-source "autolab" that can perform certain purification, extraction (or synthesis) processes for other psychoactives. This would involve following a shareable file containing instructions for the machine to manipulate specific chemical products (pre-fed to the machine), automating processes such as measuring, mixing, heating, freezing, and drying. It would be groundbreaking, for sure. A legend would undoubtedly be born.

If there is a community other than universities that could undertake this project, it is DMT-Nexus, for sure.

The beer world already made this for brewing beer with brew pi:
https://www.brewpi.com/
So... it is technically possible

If anyone would like to try a project like this, I would sketch a general project and post.

Some quick notes, off the top of my head:

Peltier elements can be used to cool the precipitation vessel, needs a heat extraction fan and heatsink.

Any gaskets on rotating fluid tight elements of the design will need to be tolerant to NPS

The electromechanical aspects will likely be the most challenging.

Leave the emptying out of left over soup and clean up to humans! Make sure mixing vessel is easily dismountable.

Likewise scraping the precipitation vessel best left to humans.

To pull NPS something like a peristaltic pump connected to a pipette with a linear motion drive to raise / lower it into the NPS layer could work, optical sensors could detect the surface of the NPS (reflective sensor) when soup is being pulled up (transmissive sensor).

Water wash of the NPS would be mandatory though, as there will always be a little soup being sucked up.

There is no way any commercial organisation would go near the liabilities involved in this!

I totally agree but what do you think of a pre-cleaning routine with lye in the system?



Yes! Thats why we have to make it.

I'm thinking in implement a cleaning routine.
Maybe it would be nice to use Stainless Steel and Glass.
I was thinking that it would be good to use just one stainless steel kettle and just a pyrex.
The pipe of Naphta could be made of stainless steel or water cooling glass tube.

Let's start by thinking in a DMT build and expand from there?


Here is the basic algorithm for DMT:

1. Check if the user wants to proceed with A/B DMT extraction
- Input: User response ("yes" or "no" ) .

2. If the user wants to proceed:
1. Set the desired quantity of water in the pot.
- Input: Water quantity.
2. Set the desired quantity of vinegar in the pot.
- Input: Vinegar quantity.
3. Set the desired quantity of caustic soda.
- Input: Caustic soda quantity.
4. Set the desired quantity of naphta.
- Input: Naphta quantity.
5. Set the number of mixing cycles.
- Input: Number of cycles.

3. Pump and monitor the water quantity until the desired amount is reached.
- Input: Water quantity.

4. Pump and monitor the vinegar quantity until the desired amount is reached.
- Input: Vinegar quantity.

5. Mix and maintain the temperature at 60°C for 3 hours.

6. Add caustic soda while controlling the temperature.
- Input: User-defined temperature constraints.

7. Mix, heat, and maintain the temperature for the specified quantity of naphta.
- Input: Naphta quantity.

8. Intermittently mix for the user-defined number of cycles.
- Input: Number of cycles.

9. Pump the naphta to a refractory and freeze for 24 hours.
- Input: Freeze duration.

10. Send additional naphta to the pot.
- Input: Additional naphta quantity.

11. Mechanically drain and air-dry the refractory.

12. Activate coolers in the chamber and release gases into the atmosphere.

13. Notify the user after 4 hours that DMT is ready.
- Output: Notification to scrape the Pyrex and a question about whether to send new naphta to the Pyrex or abort.

14. If the user chooses to continue:
1. Repeat step 10 (sending additional naphta to the pot).
- Input: Additional naphta quantity.
2. Repeat steps 10-13 in a loop until the user chooses to abort.

15. If the user chooses to abort, end the process.

This modification adds a loop to repeat the process of sending additional naphta to the pot until the user chooses to abort, enhancing the user interaction and flexibility in the process.

8. Intermittently mix for user-defined cycles.
- Input: User-defined cycle parameters.

9. Pump the naphta to a refractory and freeze for 24 hours.
- Input: Freeze duration.

10. Mechanically drain and air-dry the refractory.

11. Activate coolers in the chamber and release gases into the atmosphere.

12. Notify the user after 4 hours that DMT is ready.


I think it would be interesting to try to use hardware store supplies and Aliexpress gadgets at most. It must be accessible

If you guys have any improvement to this simple DMT algorithm, just do it.

Things to think/implement:
- How we would monitor quantity of liquids? By its Weight on reservoir? By quantity of liquid pumped?
- What other routines could we implement to make a more generic machine in the future? (Capable of extract BHO or synthesize MDMA, for example)
- How we make it the most efficient, safe, cheapest and modular way possible?
- We need to think in a VERY GOOD UI (User Interface). The user must be able to easily edit the process and send to other people his routine.
- We need to implement a terminal where the user can type simple commands and do a lot of stuff if he wants to
- It must accept settings from phone, computer etc: be multi-platform, practical, smart.

+ We need to implement generic attributes like:
- Heating
- Ventilation
- Freeze
- Transfer
- Agitate
- UV light NOT A GOOD IDEA
- Internal/external cameras (security in general)
- presence sensor (automatic shutdown?)
- Autodestruct raw materials (incinerate in chamber) if a certain condition is reached? (too much time without presence sensor activated or no login for X days) NOT A GOOD IDEA

+ I'm thinking in a basic materials list to think this project:
- Stainless steel kettle --> easy to find, easy to adapt and clean, can stand lye, naphta and other harsh products. Can do reactions inside of it.
- Temperature sensors --> DS18B20 ?
- PH sensors
- Cameras
- Pump --> must pump air (can't be a submersive pump). Must be a pump that does not touch or get in contact with chemicals. Let's make a system that uses 1 pump and it changes what it is pumping with a system. It will make the machine cheaper
- Lab glass is expensive but modular, easy to find and generally safe, we need to use it wisely. Can't stand much pressure.
- Hose coupling--> stainless steel camlocks? (https://en.wikipedia.org/wiki/Cam_and_groove) Find cheap and easy on aliexpress
- Stainless Steel/Glass pipes

+ Maybe we can use a Horizontal/vertical freezer housing to quick and democratically protect the machine. (Cheap, easy to find, easy to adapt)

More quick notes:

rather than measure qty of vinegar and lye, use a ph sensor.

The auto destruct thing is far too complicated, solves a non problem (are you really going to have *all* your suspect raw materials in the machine at once?) and will burn your house down when you get called away for a few days during an extraction because a family member is in hospital or something.

Why UV light?

You'll not find such a thing as a non contact pump for anything other than molten metals AFAIK, peristaltic is the best I think, it meters automatically and can be made with PTFE tube like this:
https://www.adhesivedispensers.co.uk/TT-24G.htm for minimum possible effect on the fluids it is pumping.

I like the PH sensor idea, but if it will only measure PH, we will not include the user that just want to make a "recipe" following quantities.
Autolab should be reasonably adaptive. I believe that the amount of solvent, vinegar or any product involved should be electronically measurable. Remember, it must have an adaptive capacity to be generic enough to be adaptable to other syntheses and extractions and must be independent enough to do it with not much human intervention.
Ideally, the user can add all the naphta he has and the system will manage it (including used naphta, etc.)

I believe we need to implement recrystallization with heptane as well

I like the pump idea too

+ For the kettle:
- Stainless Steel kettle
- Relay Module
- Temperature Sensor (DS18B20 ?)
- Resistance

+ For the Scale
- Load Sensor (Strain Gauge HX711?)
- Instrumentation Amplifier (if required, because weight can be very small: INA125P ?)

recrystallisation needs a human scraping step first, so not so automatic.

Peristaltic pumps will deliver a specific volume per revolution with pretty good accuracy, the ph meter is needed to compensate for different vinegar strengths, I'm not sure I follow why anyone would like to follow a recipe when proper measurements are available? No reason why both options shouldn't be available, but designing a high tech system is to avoid having to blindly trust recipes

There will need to be electronic scales under the reaction vessel I guess, so they can double up on fluid measuring duty to keep an eye on the metering via pump (use geared down stepper motors on the pumps to have decent precision fluid delivery, plus it's very difficult to make a stepper fail "on" so it's safer than a normal motor drive )


... later: develop a whole ecosystem of reagent cartridges, pre filled with the right amounts of whatever.. add DRM, offer subscriptions that are a nightmare to cancel, profit!

Haha Yeah... but it's an open source project. Just trying to help humanity. Apple should do it btw.
The recristalization idea was just to user to put what he wants to recristalize in the dish and the system will deal with the heptane and drain it when it is done. The idea is that the user comes into contact with chemicals and their vapors as little as possible. But yeah... can be a little difficult so let's think on the simplest scenario now.


Yes! this will work. We just have to pour 1 liquid at a time.

Now to freeze.... maybe just use a old freezer and monitor it with the same raspberry?

Look up peltier effect devices like I mentioned before, they are what powers electric camping freezers.

bright idea: for pulling: have pipette with two (platinum or stainless) wires sticking 2mm or so below the mouth, measure resistance between the wires when lowing it in, NPS will be highly insulating, soup will be very conductive, stop when resistance drops, use pump to ensure steady gentle suckage... == easy foolproof pipette positioning (actually I might just try making that bit on its own )

Interesting fractals! I didn't know what to do in this part.
I'll call this a Fractal pull for the default user.

I think we have almost all the basics now. Now I will make a project, search for all the small things and start to code.
My code game is not very good, but I think I can do it.
Wish The Traveler could help.
open positions for programmers!
We need: code for the machine (raspberry pi) AND a GUI guy.
Until we find those guys I'll do what I can.
See u guys later

Note for later: use gravity to reduce the need for pumps or many pump circuits.

I could implement that as i developed besides of everything else an automatic cocktailachine using a msp430f2274 chip and controlled the cocktail machine via bluetooth using a self written xamarin forms mobile application.
But i sadly do not have the time for it although this sounds interesting.

You could either measure the weights with a scale using adc conversion.
Or you could use throughput and time calculation of the substances.
First one is more accurate but harder to implement.
If you dont need a lot accuracy then it wouldnt probably be that relevant.

This is something I've pondered to a greater or lesser extent over the years, largely just for fun, but perhaps I should dig out some of my sketches to share. You're definitely coming from more of an automation perspective than I ever did.

Liquid/liquid extraction equipment already exists, so it's worth looking at that for some inspiration. It does tend to be rather expensive though! They are sometimes known as a Clevenger apparatus.

Issues like lye carry-over are essentially taken care of with a Clevenger design if it is correctly filled and operated. There would be ways of incorporating a trap, maybe with a detector electrode assembly, if it proved to be less than 100% effective with clean separation.

The Peltier could be used for warming the base soup as well as chilling the naphtha, maybe via a heat-transfer fluid for greater ease of control.

I do wonder what the use-case for such a device might be. Perhaps it could quickly churn out dose-sized amounts as and when needed, especially considering the overall spirit of the Nexus attitude rules.

Such a shame you don't have time. I completely understand, but the door is open... maybe if you see the evolution of the project some night time could appear
Just kidding. I'm full of work too, sometimes I have more time to surf on Nexus or Extract DMT than actually using DMT.

I Think it would be possible to implement a calculation on the kettle or in a intermediary measure cup... I would like to keep things a bit small, for obvious reasons.
I really appreciate your feedback, and yeah... accuracy won't be relevant for this project but if this thing ideally evolves, it will be a general purpose Autolab. Let's keep it small for now, so we don't drown in our own pretense. But it would be something beautiful. That's why I ask for help from people more experienced than me (we would cut costs and time, as well as unnecessary hassles).

My knowledge of chemistry is certainly limited, so I'm automating the process I know most intimately, although perhaps that's not the best process to automate. In any case, as I already had the opportunity to explain, the project will initially deal with DMT but the tendency is that, as it will be open source, it will evolve into a general purpose autolab, which can even be used for other applications besides these (Biohacking? ).

In any case, the purpose of this device would not be to produce a lot of dmt (although this is in fact more easily scalable), but simply to produce dmt in a safe and easy way, without much contact with chemicals.

I'm beginning to think, however, that it can democratize too much, and bring about those fears that Plato had about democracy (being in the hands of the many and foolish). But I'm too small and irrelevant to make a decision like that. my concern is just harm reduction and making my life easier (I'm not a particularly fast DMT cooker).

I thought about adding the lye to a kind of mat, without diluting it, taking advantage of the exothermic reaction of its hydration to save energy. Do you think it would be more beneficial to keep it in an aqueous environment?

Liquid environment means things mix and react optimally, I'm not sure what you are envisaging with this mat idea TBH, the heat of hydration can be conserved with simple insulation of the vessels, surely?

Anyway, I came here to post this interesting link to methods of liquid / liquid extraction using microfluidics ... which would be the way to go for single dose extractions, if things swing that way in the design.

https://www.elveflow.com...liquid-phase-extraction/

Happy reading!

Thanks Fractals! I will sure read it after work

Hi guys... I was designing a simple GUI.
I had some problems with widgets sizes (buttons) but I think I can handle it.. I'm a bit tired because I spended a part of the night coding...

So as I was saying, the proposal is to make an AutoLAb that is capable of carrying out a myriad of actions, (initially extracting DMT, but it can also be used to make chocolate milk for children )

Anyway, it has a HUD that displays the products registered in the system and their respective quantity.

It also has a procedure import, procedure export (we need to create your own ".al" extension)

The procedure tab is where the magic will happen: it has a menu in cascade with the possibility of creating a procedure, or to open a procedure or even execute a procedure previously registered in the system.
The creation of the procedure will be, I think, a kind of timeline, with the actions that the machine can carry out, where it will send each thing, etc. It will be editable graphically or using a very simple code (for the lazy ones)

There is the Auto-lab configuration button that will show you how to assemble the machine.

There is a basic cleaning cycle to clean connections and of course, a button to lock the screen.

Additionally, there is a button to register the chemicals you put into the machine. you register the slot where it is, the name and the system that returns how much of it is on the machine and the "recipes" in which it is registered.

That's it for now. My incompetence in Python made me take much longer than necessary.

Below are some images (and the code until now). I accept suggestions. I will soon register a git hub for anyone interested.

Maybe it would be very cool if in future we make a possibility to install it on a old phone and it can control a simpler AutoLAb board (a generic and cheaper microcontroller, like some models of arduino) by bluetooth. It would make the thing even easier, cheaper, funnier and kinda impossible to prohibit.

Please programmers, appear and save me.

Good start in figuring out what sort of things you need on a gui... but "here comes the but", all just my opinion :

Personally I'd go for a web based interface, that way the autolab doesn't need much of a screen on it, and it easily opens the way for remote checking on the process.
Also there are many more web devs than tk devs these days, probably.

You can host a web server over WiFi and control processes all on an esp32 based microcontroller that costs about $5 these days. Just add power control things like mosfets and relays for switching the heaters and pumps.

There are also handy little touch screens you can use with them that are real cheap.
Downside is c++ dev is a bit more difficult than python to get started on, upside is there are plenty of examples out there.

You could do worse than repurposing a 3d printer controller board for the hardware, IMHO, they have multiple channels of power control and stepper motor control, plus are open source and being mass manufactured. Very good value for money.

I'd further say leave the firmware alone in the printer controller, and have an ESP32 running your process control and gui talking to it over a serial interface to send G code commands to move motors or switch heaters and fans on and off or get temperature readings.

Don't think of phone based web operation as an added extra for later, design it in from the start and you can do all your gui design in nice html javascript and css. The events and messages and gui elements you have already done will be useful, just think about leaving TK behind as you've only invested one night in it so far

... all written by an old greybeard engineer who won't have time to seriously work on this I'm afraid. The day job is too demanding. There will be many possible ways to do this job, the above would be my recommendation as a quick way to get started on a prototype to test things out.

Cheers,
Frac4