What equipment can we develop?

757 views
Skip to first unread message

J. S. John

unread,
Dec 18, 2010, 12:11:48 PM12/18/10
to diy...@googlegroups.com
I just glanced over the article "GLOWING Fluorescent E. coli samples
at GenSpace in Brooklyn." and it describes a nanodrop for $10000. I
worked with those before and can't figure out why they're worth so
much. It's a box that fits on my hand not some mass spec. Is there is
list of OS alternatives that need to be created out there? I'll have
to check but I can't remember anything right now.

Bryan Bishop

unread,
Dec 18, 2010, 12:32:21 PM12/18/10
to diy...@googlegroups.com, Jeswin John, Bryan Bishop, t...@postbiota.org
On Sat, Dec 18, 2010 at 11:11 AM, J. S. John wrote:
much. It's a box that fits on my hand not some mass spec. Is there is
list of OS alternatives that need to be created out there? I'll have
to check but I can't remember anything right now.

This should really be on the wiki. I have added it here:
http://openwetware.org/wiki/DIYbio/FAQ/Equipment#Equipment_list
.. so feel free to edit.

Here's a basic list of instruments:

accelerometer
ammeter
caliper
calorimeter
dna sequencer
dna synthesizer
dynamometer
electroscope
gravimeter
inclinometer
interferometer
magnetograph
mass spectrometer
micrometer
microscope
nmr spectrometer
ohmmeter
oscilloscope
seismometer
spectogram
spectrometer
telescope
time-of-flight mass spectrometer
theodolite
thermocouple
voltmeter
spin coater
transilluminator
gel box

Also, here's a list of material analysis methods:

AED - Auger electron diffraction
AES - Auger electron spectroscopy
AFM - Atomic force microscope
APS - Appearance potential spectroscopy
CAICISS - Coaxial impact collision ion scattering spectroscopy
CL - Cathodoluminescence
DVS - Dynamic vapour sorption
EBSD - Electron backscatter diffraction
EDX - Energy dispersive X-ray spectroscopy
EID - Electron induced desorption
EPMA- Electron Probe Microanalysis
ESCA - Electron spectroscopy for chemical analysis; see XPS
ESD - Electron stimulated desorption
EXAFS - Extended x-ray absorption fine structure
FEM - Field emission microscopy
FIM-AP - field ion microscopy-Atom probe
FTIR - Fourier transform infrared absorption spectroscopy
    ATR (Attenuated Total Reflection)
    GI (Grazing Incidence)
    DRIFTS (Diffuse Reflectance)
GDMS - Glow discharge mass spectrometry
GDOS - Glow discharge optical spectroscopy
GISAXS - Grazing Incidence Small Angle X-ray Scattering
GIXD - Grazing Incidence X-ray Diffraction
GIXR - Grazing Incidence X-ray Reflectivity
HAS - Helium atom scattering
HREELS - High resolution electron energy loss spectroscopy
HRTEM - High-resolution transmission electron microscopy
IAES - Ion induced Auger electron spectroscopy
IGA - Intelligent gravimetric analysis
IIX - Ion induced X-ray analysis
INS - Ion neutralization spectroscopy
IRS - Infra Red spectroscopy
ISS - Ion scattering spectroscopy
LEED - Low energy electron diffraction
LEEM - Low-energy electron microscopy
LEIS - Low energy ion scattering
LIBS - Laser induced breakdown spectroscopy
LIPS - Laser induced plasma spectroscopy
LOES - Laser optical emission spectroscopy
LS - Light (Raman) scattering
MEIS - Medium energy ion scattering
NDP - Neutron depth profiling
NEXAFS - Near edge X-ray absorption fine structure
PD - Photodesorption
PDEIS - Potentiodynamic electrochemical impedance spectroscopy
PED - Photoelectron diffraction (also called XPD, PhD, ARPEFS)
PIXE - Particle (or proton) induced X-ray spectroscopy
RBS - Rutherford backscattering spectroscopy
Atomic absorption spectroscopy (AAS)
Atomic fluorescence spectroscopy (AFS)
Alpha particle X-ray spectrometer (APXS)
Capillary electrophoresis (CE)
Chromatography
Colorimetry
Cyclic Voltammetry (CV)
Differential scanning calorimetry (DSC)
Electron paramagnetic resonance (EPR)
Electron spin resonance (ESR)
Field flow fractionation (FFF)
Fourier transform spectroscopy (FTIR)
Gas chromatography (GC)
Gas chromatography-mass spectrometry (GC-MS)
High Performance Liquid Chromatography (HPLC)
Ion Microprobe (IM)
Inductively coupled plasma (ICP)
Instrumental mass fractionation (IMF)
Ion selective electrode (ISE) eg. determination of pH
Laser Induced Breakdown Spectroscopy (LIBS)
Mass spectrometry (MS)
Mossbauer spectroscopy
Nuclear magnetic resonance (NMR)
Particle induced X-ray emission spectroscopy (PIXE)
Pyrolysis - Gas chromatography - Mass spectrometry (PY-GC-MS)
Raman spectroscopy
Refractive index
Resonance enhanced multi-photon ionization (REMPI)
Scanning transmission X-ray microscopy (STXM)
Transmission electron microscopy (TEM)
X-ray fluorescence spectroscopy (XRF)
X-ray microscopy (XRM)
Scanning Probe Microscopy (SPM)
Scanning Tunneling Microscopy (STM)
Transmission Electron Microscopy (TEM)

- Bryan
http://heybryan.org/
1 512 203 0507

John Griessen

unread,
Dec 18, 2010, 12:34:28 PM12/18/10
to diy...@googlegroups.com
On 12/18/2010 11:11 AM, J. S. John wrote:
> Is there is
> list of OS alternatives that need to be created out there?

That's a good idea. I should put mine on the web...
The nanodrop idea is one that would have you paying patent royalties.

There would not be any way around that since the nanodrop concept is so pure.
Put a drop that's inaccurately sized down on a solvent-wiped-clean inverted microscope
objective glass surface, then cause the sample shape to be precise thickness
by lowering a detector with glass lens in a super repeatable way to contain it
and then do a measurement through the same old calibrated optical path
with one thing changed, the sample.

The fact that the $10K product is "It's a box that fits on my hand not some mass spec."
is a value and costs more than if it were in a file cabinet. Think of
making a prism spectrophotometer... to get fine wavelength resolution
you need fine angle measurement, so to get that in a small box requires
a microfabricated glass angle scale, (also called optical encoder), that can be expensive.

Cheaper would be a 15 inch radius semicircular encoder with marks made by an ordinary
laser printer, then toner transferred to a metal plate. but then "box that fits on my hand"
has gone out the window.

John

J. S. John

unread,
Dec 18, 2010, 11:21:25 PM12/18/10
to diy...@googlegroups.com
On Sat, Dec 18, 2010 at 12:34 PM, John Griessen <jo...@industromatic.com> wrote:

> That's a good idea.  I should put mine on the web...
> The nanodrop idea is one that would have you paying patent royalties.
>
> There would not be any way around that since the nanodrop concept is so
> pure.
> Put a drop that's inaccurately sized down on a solvent-wiped-clean inverted
> microscope
> objective glass surface, then cause the sample shape to be precise thickness
> by lowering a detector with glass lens in a super repeatable way to contain
> it
> and then do a measurement through the same old calibrated optical path
> with one thing changed, the sample.
>

Hmm, well, what if we came up with a novel method to do what they do,
i.e., we came up with a different method to quantify the amount of
DNA.

I recall (very vaguely, the details might be wrong so tell me the
right story) a story where a company needed to come up with a BIOS(?)
alternative so they put some engineers in a room and told them what to
do. The result was the same but the way the thing worked was
significantly different so no patent problems.

Russell Durrett

unread,
Dec 19, 2010, 12:57:42 AM12/19/10
to DIYbio
It would definitely be worthwhile to investigate building a
spectrophotometer! It doesn't have to fit in your hand - just in a
lab.

The nanodrop is an elegant design that only requires ~1 uL of your
sample to get a very accurate measurement, but DIY labs just doing
cloning you would really only need to know approximate amounts of DNA.
So, you could use older (and dare I say larger) design that requires a
larger amount of sample, but just dilute your sample until you had
that amount. I used to use one that required 50uL of a sample, so I
would just dilute 5uL of my prep tenfold and take my concentration up
a level of magnitude. Voila.

If you end up producing something worthwhile be sure to hit up the
good folks at GenSpace to see it they still need one ;)

On Dec 18, 10:21 pm, "J. S. John" <phillyj...@gmail.com> wrote:

Joseph Jackson

unread,
Dec 19, 2010, 1:27:50 AM12/19/10
to DIYbio
I recently attended a point of care diagnostics meeting where one
argument was that the lab of the future would not consist of cheaper
and miniaturized versions of traditional equipment, but entirely new
multi-purpose platform devices...eg a universal reader with bio-nano
chips/cartridges that have customizable chemistry and the whole
process...sample/optics is packaged and handled by this bench top
device.

That was the argument for diagnostics at least, but does it hold more
generally...ie...we won't see the mass spread of DIY instruments but
instead replace old school lab infrastructure entirely with a
different way of doing participatory biology either with new tech that
obsoleces the traditional instruments by combining several of their
functions or with a different model?

Guess I am asking if maybe I and others were barking up the wrong tree
thinking that biology would follow the home brew computer club model
then to the equivalent of the microcomputer and finally the PC. What
if it skips directly to the era of Cloud Computing/Cloud Biology. Eg,
what is the real killer app for a distributed Bio Device vs out
sourcing things like sequencing or synthesis to on demand wet lab
facilities and getting your results back.

Certain things like biosensors are going to be inherently cheap and
distributed, but I'm beginning to question my own earlier conviction
that the instruments listed in Bryan's post above will end up being
that crucial to distributed DIY bio.

Desktop/personal synthesis machines still seems disruptive to me.

Jay Woods

unread,
Dec 19, 2010, 7:55:50 AM12/19/10
to diy...@googlegroups.com
There is a mathematical relationship between the drop height and
diameter that is also dependent on the surface tension. This can be
used to get the optical path close without "a super repeatable way"
being needed.

Alternatively by using an additional wavelength and an additional
component that is optically dense at that wavelength (such as a pH
indicator dye), an exact thickness can be determined.

John Griessen

unread,
Dec 19, 2010, 10:47:33 AM12/19/10
to diy...@googlegroups.com
On 12/18/2010 10:21 PM, J. S. John wrote:
> so they put some engineers in a room and told them what to
> do. The result was the same but the way the thing worked was
> significantly different so no patent problems.

They gambled a large capital amount. Won't happen like that in this setting.
Will end up paying royalties if they allow or being shut down by huge $$
lawsuits more likely. The nanodrop patent is likely a very fundamental one,
a process patent with wide ranging claims.

JG

John Griessen

unread,
Dec 19, 2010, 10:57:30 AM12/19/10
to diy...@googlegroups.com
On 12/19/2010 12:27 AM, Joseph Jackson wrote:
> I am asking if maybe I and others were barking up the wrong tree
> thinking that biology would follow the home brew computer club model
> then to the equivalent of the microcomputer and finally the PC.

I think the rising costs of colleges, (where the researchers are), will drive many
to offer online courses and many more to shut op for good. In that atmosphere,
some low cost gear could be bought by university purchasing agents if good enough,
and taking up a small enough desktop footprint. There will be many paths in biology --
it's big field. Some will still go for status appeal and have the shiny latest,
while many will make do with far less status and just a little less
functionality for the money.

MEMS bio "chip" diagnostics won't be the only way things are assessed,
but it will be available, and the DIY/DICheapo gear won't use that
method with its built in high capital chip fabbing required.

I do see a coming of semi-custom microfluidics plus instrument heads that
clamp on modularly for building your own light and IR and UV and microwave
measurements of stuff going by in a 70 um channel. Some of the less expensive
MEMS fabbing will be DIY available, just not the custom integrated electronics and MEMS
assay instruments.

JG

John Griessen

unread,
Dec 19, 2010, 11:00:55 AM12/19/10
to diy...@googlegroups.com
On 12/19/2010 06:55 AM, Jay Woods wrote:
> Alternatively by using an additional wavelength and an additional
> component that is optically dense at that wavelength (such as a pH
> indicator dye), an exact thickness can be determined.

That's a good concept that would sidestep nanodrop. Good thinking.
Is there such a thing as a cover slip just 3 mm square? It
could be dropped on top of a droplet of fluid on a slide...then
proceed as you say above.

JG

Russell Durrett

unread,
Dec 20, 2010, 3:06:13 AM12/20/10
to DIYbio
Joe -

The trend in biology will follow the homebrew computer club, we just
need to make the equipment to test hypotheses cheap enough for
everyone to afford.

If people can get their hands on all the equipment to test a PCR
reaction, then there are literally thousands of ways for them to
advance biological research. Cheap, reliable equipment just has to be
available before we can observe the 'citizen science revolution'

Phil

unread,
Dec 20, 2010, 11:33:35 AM12/20/10
to DIYbio

On Dec 19, 1:27 am, Joseph Jackson <joseph.jack...@gmail.com> wrote:
> I recently attended a point of care diagnostics meeting where one
> argument was that the lab of the future would not consist of cheaper
> and miniaturized versions of traditional equipment, but entirely new
> multi-purpose platform devices...eg a universal reader with bio-nano
> chips/cartridges that have customizable chemistry and the whole
> process...sample/optics is packaged and handled by this bench top
> device.

Interesting. I hope robotics is the way of the future; big question
is whether we will see common large lab robotics setups before that
gets replaced by microfluidics setups.

Either way, though, DIYbio is probably going to be constrained to
using 10-year old equipment. Even if we go to microfluidics labs on a
chip, the companies that will develop that hardware will charge so
much for it that we won't be able to afford it. This will be $2000/
sample stuff when it is introduced 10 years from now; the price over
time is likely to mirror the price of Affymetrix chips over time after
their introduction. So these things won't enter into DIYbio for 20
years.

(I am making some assumptions based on my conclusions from Moore's
Law. The rate of increase in semiconductor density is too constant to
reflect the underlying difficulty of technology. I believe it instead
reflects the more-constant market pressures. Companies will only
invest in technological improvement to the extent that they are
profitable; a tech is profitable if it motivates people to buy new
chips; a new chip that is 100x as dense as the previous generation is
no more profitable than a new chip that is 2x as dense if the 2x-as-
dense chip motivates just as many people to upgrade. What I'm saying
is that people have a threshold of improvement above which they want
to upgrade; and this threshold, not physics, dictates the speed at
which technology progresses and the rate at which prices drop. A
notable contrast between semiconductors and biotech is that the price
of old Affy chips may have dropped, but it doesn't matter, because the
price of processing an old chip is about as great as the price of
processing a new chip.)

What I'm saying is, this issue has no impact on us unless we can
address the underlying problem that there is no perceived DIYbio mass
market.

Lee Nelson

unread,
Dec 20, 2010, 12:36:23 PM12/20/10
to diy...@googlegroups.com

I heard these chips were available in England from vending machines for a dollar.

http://www.telecareaware.com/index.php/mobile-phone-chip-for-std-self-testing-uk.html

google.com/q=std+chip+phone

John Griessen

unread,
Dec 20, 2010, 3:09:29 PM12/20/10
to diy...@googlegroups.com
On 12/20/2010 10:33 AM, Phil wrote:
> What I'm saying is, this issue has no impact on us unless we can
> address the underlying problem that there is no perceived DIYbio mass
> market.

Yes, that's a problem, but starting small and with contributions to
easily updated open hardware designs can get a perception of a small market
at first. Then some of the makers who only respond to a larger one will jump in.

I have some designs planned to build with 3D printing, inexpensive circuit assembly,
and free open source mechanical and electronics design tools we have now
and would like help from you if can get your attitude more towards possibility
and away from the seeming defeatist position you are on.

John Griessen

Ruediger Trojok

unread,
Dec 20, 2010, 4:08:44 PM12/20/10
to DIYbio
I worked with liquid handling robots and can tell you they are crap.
It really only pays out if you have a totally standardized protocol
that you want to get done a million times.
Nothing useful for tinkering or basic research. Only high throughput
systems biology or commercial applications are worth the trouble and
costs
you have with this robots.

Microfluidics will definitively replace these robots very soon.
Optical tweezers are also pointing in the same direction.
Really useful tools if you manage to handle them. The technology is
not much more complex than lets say more
advanced microscopes. So I predict these items to be widely used soon,
combined with (for universities) affordable prices.

The paper that was posted recently about pcr using microfluidic
knowledge is one more example to me that this is the right path.

BUT, in the longer run, we will probably replace almost all robots,
instruments and tools by developing artifical cells
able to do all the desired steps in molecular biology on its own.
There were already several igem project pointing in that way, like
self lysing cells, self pelleting cells, light inducable promotors,
nano/ bio assembly lines for chemical compounds and so on. Not to
forget craig venters artifical life.
My dream is to do field research in future only equipped with a pipet
and a few eppendorf tubes containing multifunctional cells.
I guess this will be the case within the next 3 decades.

John Griessen

unread,
Dec 20, 2010, 8:04:10 PM12/20/10
to diy...@googlegroups.com
On 12/20/2010 03:08 PM, Ruediger Trojok wrote:
> BUT, in the longer run, we will probably replace almost all robots,
> instruments and tools by developing artifical cells

If you could direct your view back to the nearer term,
I'd be interested to hear your take on optical tweezer
and microfluidic valves and surface moving of drops by electrowetting
variability via Voltage changes.

JG

Lee Nelson

unread,
Dec 20, 2010, 9:59:59 PM12/20/10
to diy...@googlegroups.com

Patrik

unread,
Dec 21, 2010, 1:51:46 AM12/21/10
to DIYbio
On Dec 20, 1:08 pm, Ruediger Trojok <ruedige...@googlemail.com> wrote:
> I worked with liquid handling robots and can tell you they are crap.
> It really only pays out if you have a totally standardized protocol
> that you want to get done a million times.
> Nothing useful for tinkering or basic research. Only high throughput
> systems biology or commercial applications are worth the trouble and
> costs
> you have with this robots.

Would that be your assessment, even if we could bring the price down
to somewhere around $1000, and have a completely open design that you
can fix and tinker with at will?

I've been pointing out to a number of people that open source 3D
printers like the Makerbot have a lot more accuracy than you'd need
for liquid handling, and that it should be relatively easy to design
an open source liquid handling robot along similar lines. But as you
point out, I've really only heard them being used for a fairly small
range of high-throughput applications.

If we could make them into cheaper, more general purpose platforms, do
you think we might be able to come up with a wider range of
applications?

Ruediger Trojok

unread,
Dec 21, 2010, 6:29:35 AM12/21/10
to DIYbio
For fundamental research it simply doesnt pay out to standardize
protocols to make them applicable to a robot system.
Its too time consuming and error prone. I tried that with a protocol
in my uni and a commercial perkin elmer robot.
The robot costs (at least did cost some years ago) around 100000euros.
I simply wanted a protocol to create protoplasts from arabidopsis
leaves
to be done by that robot, because there were a lot of time wasting
gaps in that (waiting for digestion of cell wall and so on),
so I thought its a good job for a robot.
But then it took me 3 weeks to get the robot exactly doing what i
wanted it too. Then it was not as precise as it needed to be.
And in the end the robot had technical problems and a repairman from
the company had to come. He charged 6000€ for 2h of work :D
I thought I'm in the wrong business.... I'll never get this payment as
a biologist.

Well, also ich you manage to have these systems at NO cost, it is not
useful for basic research or tinkering.
It will probably be very useful in things like open source drug
discovery, where you have to screen thousands of probes.


Concerning optical tweezers: I worked with them in a physics
department here. Its an amazing technology.
You had video from a camera sitting on a microscope in which you could
very easily set the optical trap be a mouseclick.
You could adjust strength, shape, speed, depth, xy-position of the
scanned area just with the software.
Even applying two traps were possible, giving you the possibility to
grab things and pull it apart, rotate it and so on.
And all this on a scale where you could easily move single bacteria
around!
The department was also working with microfluidic systems in
combination with the traps.
E.g. having a stream of water passing by some basins on the side. You
could flip your bacteria
or whatsoever from one basin into the other by passing through the
stream. Each basin could be filled up with a different chemical.
It wasnt well established yet, but if the researchers gain more
experience, I am sure that this kind of applications will be
a standard in all bio labs in the next 10 years.
I heard that there is the first commercial optical tweezer from zeiss
on the market, that can be plugged on a fluorescent microscope...
http://www.vision.zeiss.com/__c12567be0045acf1.nsf/Contents-Frame/923f48116e9897ab41256a7200352cfc?opendocument&click=

John Griessen

unread,
Dec 21, 2010, 4:25:12 PM12/21/10
to diy...@googlegroups.com
On 12/21/2010 12:51 AM, Patrik wrote:
> If we could make them into cheaper, more general purpose platforms, do
> you think we might be able to come up with a wider range of
> applications?


Back when discussing incubators, shakers, centrifuges, and photometric "done incubating" sensors,
I noticed that the same basic 3DP part could be the basis of all of those
and allow moving vials between sensors. http://ecosensory.com/diybio/carousel_greycoded-1.jpg

Here's a sketch of how to minimize heat sinking of the vials to the holder:
http://ecosensory.com/diybio/carousel_tube_wells-1.jpg
(That's good for incubator or PCR ramp up/down speed and accuracy when air-heater air-cooler driven)

It's also a good simple handler station for a dropper bot that only moves up and down, (very cheap),
so the price could get very low by sharing part making volume with other assemblies used
in a diy or official lab with low budget and perfectly good performance.

JG

Patrik

unread,
Dec 21, 2010, 6:00:39 PM12/21/10
to DIYbio
On Dec 21, 1:25 pm, John Griessen <j...@industromatic.com> wrote:
> It's also a good simple handler station for a dropper bot that only moves up and down, (very cheap),
> so the price could get very low by sharing part making volume with other assemblies used
> in a diy or official lab with low budget and perfectly good performance.

Heh - I remember the Church lab at one point had a liquid handling
"robot" that they had rigged up on the circumference of a horizontally
mounted bicycle rim, very much along the same principle. I don't
remember what they were using it for, but it struck me as a great
example of a cheap DIY approach to something that might otherwise
require a big expensive commmercial robot (or equivalently, a horde of
students doing mind-numbing pipetting steps).

Nathan McCorkle

unread,
Dec 29, 2010, 12:14:49 AM12/29/10
to diy...@googlegroups.com

The reason it costs so much is because it has an Ocean Optics USB2000
(nice linear CCD, and the volume of sales are much lower than say a
pocket camera (which use CMOS generally, which are not as good for
spectroscopy purposes due to larger pixels))... there's also a
perkin-elmer light source inside them, a xenon flasher if I recall
correctly, so again volume-of-sale is low there.

Actually, the workhorses of the NanoDrop aren't even made by the
company, they basically only make the metal shell and the armature
assembly, and the software... which all make usability very
user-friendly.

I've been thinking about designing a spectrometer based on the same
sensor for a while now, but I don't know how to layout boards, and as
I mentioned about volume-of-sales... who would buy it to help recover
my costs.... now an instrument for just analysing DNA concentrations,
that is something that is less flexible, but yes, probably cheaper to
design, make and manufacture.

> --
> 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 http://groups.google.com/group/diybio?hl=en.
>
>

--
Nathan McCorkle
Rochester Institute of Technology
College of Science, Biotechnology/Bioinformatics

Giovanni Lostumbo

unread,
Dec 30, 2010, 1:00:05 PM12/30/10
to DIYbio
I'd like to add:

a device that can de-ionize water.

also, one of these: http://www.scientificamerican.com/media/inline/blog/Image/NASA-urine-recycler.jpg
http://www.scientificamerican.com/blog/post.cfm?id=nasa-all-systems-go-for-space-urine-2008-11-25
http://www.csmonitor.com/Environment/Bright-Green/2008/1126/should-we-recycle-urine-on-earth-too
http://www.watseco.com/images/WSC4-Photo.jpg

then using one of those devices to further de-ionize the water with an
added modification outlet.

Adding CAD/STL(?) files to a repository like SKDB for the thousands of
parts of one of those sounds interesting.


Bryan Bishop

unread,
Dec 30, 2010, 1:08:48 PM12/30/10
to diy...@googlegroups.com, Giovanni Lostumbo, Bryan Bishop
On Thu, Dec 30, 2010 at 12:00 PM, Giovanni Lostumbo wrote:
Adding CAD/STL(?) files to a repository like SKDB for the thousands of
parts of one of those sounds interesting.

Actually the basic way of doing that is:
* store your content in a git repository
* add a basic metadata file in yaml that lists out dependencies, a BOM either in this file or another
* put this git repo somewhere it'll be identified for inclusion in a package mirror of open source hardware, or i.e. ask me to remember it

Cathal Garvey

unread,
Dec 30, 2010, 1:34:06 PM12/30/10
to diy...@googlegroups.com

Quick suggestion for deionised water: I'm unsure of the purity, but deionised water is available in the hardware section of Tesco for car batteries etc, so ought to be available from auto or camping supply stores and such as well.

On 30 Dec 2010 18:08, "Bryan Bishop" <kan...@gmail.com> wrote:

On Thu, Dec 30, 2010 at 12:00 PM, Giovanni Lostumbo wrote:
>

> Adding CAD/STL(?) files to a reposito...

Actually the basic way of doing that is:
* store your content in a git repository
* add a basic metadata file in yaml that lists out dependencies, a BOM either in this file or another
* put this git repo somewhere it'll be identified for inclusion in a package mirror of open source hardware, or i.e. ask me to remember it

--

You received this message because you are subscribed to the Google Groups "DIYbio" group.

To post to...

Brian Degger

unread,
Dec 30, 2010, 3:15:10 PM12/30/10
to diy...@googlegroups.com
Also try aquarium shops selling marine fish, they sometimes sell DI
water, as its used for making up artificial salt water.
Cheers
Brian

Giovanni Lostumbo

unread,
Dec 30, 2010, 3:56:06 PM12/30/10
to diy...@googlegroups.com
In the short term, a hardware/camping/aquarium store might be convenient, while a RepRap-made device that can convert to d-i using tap water/dehumidifer and electrodeionization/filtration/carbon filter/reverse osmosis technologies would be a long-term goal and in the spirit of DIY bio and openmanufacturing.

 "Water for microbiology experiments needs to be completely sterile, which is usually accomplished by autoclaving. Water used to analyze trace metals may require elimination of trace metals to a standard beyond that of the Type I water norm."

Some microbes can use trace metals as an alternative energy source thus Type II-IV might be necessary in addition to autoclaving to have pure cultures. Cleaning/Scrubbing Pyrex/glass flasks thoroughly to assure trace metals/minerals aren't lodged in any scratch grooves does help.

Andrew Barney

unread,
Feb 12, 2011, 1:48:15 AM2/12/11
to diy...@googlegroups.com
Okay, so we want equipment right?

Alright, Lets see where we can start. (lets see where i can start)

I'm going to start small, and see where it takes me. I'm going to
start with a simple Arduino data-logging shield. Now, for some things
i think an arduino is going to be overkill and over expensive, so i
dont think the arduino is good for everything. But for now, it's an
abstraction layer that works well, and if your like me you already
have one unused sitting in the corner.

I found a Data-logging shield recently released by Ladyada. It looks
nice, and probably will work well for a lot of projects where data
logging is needed. But, this being a scientifically minded group, i
realized that the RTC clock included in that design is not super
accurate.

Okay, so i've only been working on this for a couple days. I could use
some help. I've decided to hijack the eaglecad files from ladyada, and
instead use a ChronoDot (macetech) and a OpenLog (sparkfun). I think
i'm almost done actually. I'm just running into issues wiring
everything up. The EagleCad files are here:

http://forums.adafruit.com/viewtopic.php?f=8&t=19675

-Andrew

John Griessen

unread,
Feb 12, 2011, 11:51:25 AM2/12/11
to diy...@googlegroups.com
On 02/12/2011 12:48 AM, Andrew Barney wrote:
> I found a Data-logging shield recently released by Ladyada. It looks
> nice, and probably will work well for a lot of projects where data
> logging is needed. But, this being a scientifically minded group, i
> realized that the RTC clock included in that design is not super
> accurate.
>
> Okay, so i've only been working on this for a couple days. I could use
> some help. I've decided to hijack the eaglecad files from ladyada, and
> instead use a ChronoDot (macetech) and a OpenLog (sparkfun). I think
> i'm almost done actually. I'm just running into issues wiring
> everything up. The EagleCad files are here:
>
> http://forums.adafruit.com/viewtopic.php?f=8&t=19675

I wish there was an easier starting open set of PCB and schematic tools
so people would stop using crippled free versions of commercial software
for open hardware projects. KiCAD might be the first to get the ease of use,
and there's also one called Fritzing that could be it. My favorites,
the gEDA tools, have a learning curve that puts off many users, especially
Windows users since even the install on Windows is difficult.

I'm not going to be able to chip in help on this project because I
have not seen many requests for components from biologists -- they mostly want
finished systems that are low cost, and so the modular Arduino approach
loads systems down with costs that disappear when you make your own
system boards. Another thing about modular Arduino is that it aims at
crude fabrication techniques instead of being miniature, so packaging
it into systems is harder. Instead of 0.1 inch pitch post connectors
I like to use flat flex connectors with an easy to solder pitch of 1.25mm
between lands. They're as inexpensive as 0.025" posts on 0.100" centers
and small and lay down flat to a board, (2mm high).

So now that I've blasted Arduino modules let me say that using the hardware
and Arduino code tools is great for standalone instruments, or instruments
connected to a PC and there are low cost modules to use that can be a
starting point for more compact physical designs:

http://wiblocks.luciani.org/
http://wiblocks.luciani.org/prices.html

They're nearly the same as your proposal -- see the specs for NB2AS module.

And they're made with open design tools so going from prototype to production is easy.

I think JCL's Altoids tin Sanguino/Arduino compatible modules would be even better with
smaller flat flex cables between rather than IDE cables, but the IDE cables are basically
free, and then, the flat flex cable stuff can be left for production.

John Griessen

plugging for openness in open hardware,
and plugging John Luciani's existing products too :-)

Reply all
Reply to author
Forward
0 new messages