update on the liquid handler

[Tom Eberhard]

Howdy,

here's an update on the liquid handler I've been working on. I got X and Y working, still need to do Z and the pump. Yay progress.

The video of the prototype shows going around two 96 well plates, and then going from each well in the bottom row to a small petri dish:

https://www.youtube.com/watch?v=c6CyS06c8tI

Looking for a co-founder with a background in the life-sciences. Email me directly. You'd have to move to the San Francisco Bay area eventually.

Tom.

[Yuriy Fazylov]

I understand open tron is looking for collaborators. Their project was open source. Did you check with them? https://groups.google.com/forum/#!topic/diybio/NhY12FaSMgA.

http://www.opentrons.com/ (check out the background video on their website) 

[Otto Heringer]

Nice pippeting bot!

Will Canine from OpenTrons appears to be building a consistent supply chain on China, I gess. This is determinant to put a low cost DIY equipment on the market. You really should consider talk to him.
There is also some biotech accelerators on SF, right!?

I'm building with some friends a liquid handler prototype for 96 well plates too. Hope to finish it next months and share the design online also. :)

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.
To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/bf6e9cb4-7cf1-499c-bcbe-33fa5d8a2089%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Tom Eberhard]

Got the Z axis to work, see video here:

https://www.youtube.com/watch?v=ICd0xJmoino

I'm planning on going to the SF hardware meetup at Counsyl on Wednesday (11/12/2014) for a demo of what I have so far.

Tom.

[Tom Eberhard]

Hi Otto,

can you post a pic of your pipetting bot? Would be fun to compare designs a little.

Tom.

[Otto Heringer]

Sure, Tom!

Don't mind our design, is a little not conventional, haha

The idea is verticalize the working area on "floors". So our pippeting bot ended looking like a guillotine, I know.

https://www.youtube.com/watch?v=AgH6X_9N7D4

We still need to implement the x axis as you may see. And we have a problem with the displacement of the z axis connectors wen the pipette moves on y axis - I think that this is because of the belt tightening.
It was used a core H (something like this) for the mechanical control using two step motors.

Tha plan is build another prototype, more traditional, functional and smaller, with a working area for just one 96 wells plate.

OBS: in the video we're controlling using the directional buttons of a keyboard, only for testing.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/47c68a0a-9bc3-4e8e-9fa9-d3b8bcdff594%40googlegroups.com.

[Yuriy Fazylov]

Otto, That's interesting.

Just a thought. Some labs don't have much horizontal space due to clutter of clunky secondhand equipment.
Vertical space is something they have to find ways to exploit. Your bot could be better if it had to navigate its way around something like a leaning bookshelf. Could you lean your robot diagonally and put an extending swivel arm? Multi tier design would allow you to see what is going on on every "floor" as you put it.

[Tom Eberhard]

Hi Otto,

thanks for the video! Intersting design... I had thought of putting the motors off-axis as well, but figured the belt tension might be tricky. Hence going with a more conventional design, but at least it works. 

It looks like you're in a lab in the video, is that part of a University project?

Tom.

[Yuriy Fazylov]

Why use a belt when you can use a threaded pole?

[Tom Eberhard]

at least two reasons:

- price difference: a good leadscrew will set you back $50 to $100+

- leadscrew is slow but very accurate. belt is fast and accurate enough for this application.

[Yuriy Fazylov]

Ok. Good to know.

As you can tell I am no expert in these matters.

[Yuriy Fazylov]

How about a telescoping lift design?

[Otto Heringer]

Yeap, it is on a academic fablab at our university (university of são paulo).

This started as a project for a class last semester but the fun led the project beyond that. We were supposed to improve the OpenTrons design in some way, but again the fun led us more "far" - "far" from the original idea, haha.

This is exactly what we were thinking on space aptimization, Yuriy.

Researchers would also organize their experiment routines whitin the levels - imagining that the space of a single level is enough for each different routine.

And: wow! You have so many nice ideas, Yuriy! A leaning bookshelf! It would "kill" the appearance of a guillotine that our bot have right now (ha!).

The ergo centric is beautiful, it kinda remembered me of this awesome commercial bot (the "andrew" robot"); but I don't now if it will be light and slim (to easily pass through the levels), we want a pipette car really light and tight. This is something to research about.

We choosed the belt instedad a endless screw (this is how we call a threaded pole here on the south :p) because we thought it would be a faster design, as Tom guessed.

If we solve the shiffting problem, things might get more functional. I was looking the structure yesterday and find that the conector of the z and y axis could be the guilty for the displacement. With higher mechanical tensions, the PLA 3D printed pieces allows a flexibility on the connection that we cannot have.

The telescopic lift is another wonderful idea to implement the x axis. It will be something probably harder to do (we were planning to hack a CD driver) but funnier. You gave me another thing to evaluate.

I found some fancy pictures of the gilloti... ops, pipetting bot, for better showing the structure.

2014-11-26 23:29 GMT-02:00 Yuriy Fazylov <yuriy...@gmail.com>:

How about a telescoping lift design?

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/a7d3e46d-17e6-46e0-976e-c3117ff0e790%40googlegroups.com.

[John Griessen]

On 11/25/2014 06:20 PM, Yuriy Fazylov wrote:
> could be better if it had to navigate its way around something like a leaning bookshelf

Yes, with a Y axis to get down to each shelf that has all the carriage to one side
plus a jut out, you could "cover" a large area on each shelf even if they overlap vertically.
By cover, I mean move liquid. Leaning suggests attaching to the building to gain rigidity,
yet that always involves unskilled, (at least in carpentry), installers getting the angle
a little different each time, so some kind of vision system to get accurately to the
centers of wells would be nice for this system. It could go blind most of the time,
but home in on some targets to calibrate position on each shelf.

Vertical could be good!

[Yuriy Fazylov]

John, for Otto's design you are offering a projection out of the relatively 2D frame to navigate about a series of cascading shelves. Is that correct?

Guillotine design could probably use the same belt with the added Z axis projection. It would need a suspension bridge support or a brace.

>I don't now if it will be light and slim (to easily pass through the levels), we want a pipette car really light and tight.

Make a smaller but broader pipetter w/ autoclavable parts. What are you losing if the pipette goes in at an angle and then swivels upright into action? Or dare I propose adding some distance between shelves? ;)

Level buildings: I doubt DIYbio community is going to have trouble with a level and a right angle or a string with a load on it. But in case they are, a circular multi tier (typical wedding cake type) workstation on a turnable platform (equidistantly spaced ball bearings needed here for support) could be used. This would make the handler X and Y axis dependent whereas Z for the major part would be left to the cake. Holding containers would have to be custom made. Shouldn't be issue with additive manufacturing.

>It could go blind most of the time, but home in on some targets to calibrate position on each shelf.

Sure, let the apparatus dry run on high speed with [food coloring] or color turning [pH] sensetive liquids. You'd probably even work out kinks in the system that way.

As far as accuracy goes, laser guided (Kinect maybe on a stationary platform) realignment could be implemented for auto alignment, once level.

---

I just looked at the YouTube video again. In case I didn't mention it before, that's an impressive piece of machinery, Tom. How will you go about actual pipetter handling? Will it be modeled after commercially available machines?

[John Griessen]

On 11/29/2014 02:20 PM, Yuriy Fazylov wrote:
> John, for Otto's design you are offering a projection out of the relatively 2D frame to navigate about a series of cascading shelves. Is that correct?

Yes. I was seeing how tilt can be an advantage. It allows lots of terraces of area to be accessed
without such a wide/heavy moving part on the gantry. A small projection away from vertical
would be enough to reach under the terrace above -- same as terrace overlap.
Terrace vertical spacing can be about human hands getting in there.

Being vertical can be seen as a disadvantage, but once you get past issues like "It's tippy!",
you might find people like its small footprint. Sell it with clamps to
fasten on the edge of lab benches, or even on the edges of upper lab bench shelves that are solid.

Here's an idea:

Q: What if it goes up vertically from a top lab bench shelf so high it's out of reach mostly
except for the bottom two terraces?

A: Make it a liquid handler with a latch thingy so it can also be a tray handler to
deliver trays down where people can reach them. Have designated lower shelves for
humans to use -- not needing very much vertical spacing, and each shelf has a readout
showing what the tray name and contents are. No confusion, and results of some programmed
sequences of mixology are ready to go on a lower shelf. Add a clean air box around all this
and voila -- automation! Maybe later even add humidity, incubation, CO2 to the liquid handling space...

And all that might take up 15 inches of lab bench top shelf space.
The shelves that are a little below eye level.

[Jonathan Cline]

On Monday, November 10, 2014 at 10:41:23 PM UTC-8, Tom Eberhard wrote:

Got the Z axis to work, see video here:

https://www.youtube.com/watch?v=ICd0xJmoino - tracked

I'm planning on going to the SF hardware meetup at Counsyl on Wednesday (11/12/2014) for a demo of what I have so far.

 

Hope you made the communication protocol command set Cavro compatible.   Also, acceleration needs to be programmable on all axis, especially Z.


## Jonathan Cline
## jcl...@ieee.org
## Mobile: +1-805-617-0223
########################

[Enrico]

guys this is really great! programming the machine to work with different protocols will be hard and cool at the same time

[Otto Heringer]

Sure, it will. I still dont looked if the online community is engaged to do this with the opentrons pipetting machine, but hope to work out for them. Anyone here already using it!?

By the way, I've put aside the guillotine design for a moment and now I'm working on an "almost a 3D printer" design for the pipetting machine. My focus now is on the pipet carriage and operation controlling system and the plan is to finish a presentable (for you guys) 3D model next week.

@Tom, are you working on something like this right now!? Critics and suggestions of experienced ppl (or not so experienced as well) would be welcome. :D

Em 01/02/2015 06:32, "Enrico" <enr.t...@gmail.com> escreveu:

guys this is really great! programming the machine to work with different protocols will be hard and cool at the same time

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/d2c3f6b7-0368-4ce8-8ac6-c51cad351acb%40googlegroups.com.

[John Griessen]

On 02/01/2015 06:08 PM, Otto Heringer wrote:
> plan is to finish a presentable (for you guys) 3D model next week.

What SW tool are you using?

[Otto Heringer]

Sketchup

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en

Learn more at www.diybio.org
--- You received this message because you are subscribed to the Google Groups "DIYbio" group.

To unsubscribe from this group and stop receiving emails from it, send an email to diybio+unsubscribe@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/54CEEBD1.2070409%40industromatic.com.

[Tom Eberhard]

Hey Otto and whoever's interested in the liquid handler and automation:

I'm working on gettig an off-axis pump working. That way you don't have to buy a pipette and you don't need 2 additional motors moving around. 

Right now I'm having a small vacuum leak, but I think I'll get that fixed soon. It works well when I clamp things down hard, but I want to make the whole design of the liquid handling part a bit tighter. And by "works well", I mean it aspires and dispenses liquid, but I have not measured the accuracy or repeatability.

At small volumes liquid handling becomes trickier because of surface tension and microfluidic effects. (and temperature, viscosity, etc)

Tom.

[John Griessen]

On 02/06/2015 01:42 AM, Tom Eberhard wrote:
> I'm working on gettig an off-axis pump working.

I suppose you are meaning some kind of positive displacement pump?

1. peristaltic tubing pump?
2. valved pump?
3. vane pump?
4. adjustable flow by variable-angle-of-drive ceramic piston-as-valve pumps?

I like the design of 4. but they are still patent protected and expensive.
They do volumes like .5 cc/second at pressures like 30 PSI easily, which is
so far beyond liquid handler needs that less expense makes more sense.

2. with pinch valves makes sense for tubing -- up to an accuracy point of diminishing returns
depending on tubing qualities.

2. with valves being made by driving a diaphragm all the way to seal against the chamber it
pumps in is a thought I had. Some kind of flat films assembly of low cost/throw-away
plastic layers fabbed to make a pump with some dynamics that
are not so easy to deal with, but it can be shut off to be drip free
is what I see as a next "good thing" for liquid bots. Some diligent motor code
experimenting and the "not so easy dynamics" come under fine control.

It won't look like a syringe, but rather a strip of tape That clamps between two blocks
that have squeeze actuators in them. It would have a row of 3 or 4 or 5
pockets in the tape with flow restricted paths between them and pump by careful timing of
the squeeze actions. Some really careful timing and maybe you cold stop it without
an extra unplanned droplet and so use it as a pipettor that would not dribble between
positionings.

[Tom Eberhard]

It's positive displacement, with a piston in a cylinder. 

I'm tyring to make it as accurate as a pipette, so a threaded rod is used to move the piston, just like in a pipette:

- When the piston moves towards the motor, it creates a vacuum which draws in fluid. 

- When the piston moves away from the motor, it creates pressure, which moves the liquid out of the pipette tip. (not pictured)

I'm not looking to move fluid continuously (for now), but rather do 5ul to 500ul reliably and accurately. So a peristatic pump will be for later, maybe for a washing station. 

Tom.

Alva Bio