Microcontrollers

[archaeon]

I'm trying to rig up a few sensors (temp, dissolved oxygen, etc.) and pumps/aerators in a biodigester. Generating the right beneficial microbes is one thing, but I'd like to setup a microcontroller to see how I can evolve the process through data and automation. Potentially run it from my phone. Anyone have any experience with sensors and microcontrollers (like the Arduino) and how they could be used to develop certain populations based on various data parameters? 

[Nathan McCorkle]

I would look into the Parallax Propeller, it's a really cool chip that
is as easy to use as Arduino with a large community, has 8 cores, and
is $8.

> --
> 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.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/diybio/-/MFglM7aAIHQJ.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>



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

[Sebastian S. Cocioba]

Phidgets are stand alone usb devices that can control and sense a large variety of things quite accurately. The come complete with operational code in many programing languages like c# cpp java etc etc. I use them in my rov and around the house helper robot. Im a .net kind of guy so i was thrilled when it was written in c# too. the c# code is nicely written and event based so all the data streams in real time. They offer ph sensors, thermocouples, along with an analog device hookup board which can read signals from up to 8 different sources simultaneously. Phidgets.com . Give em a shot! :)

Sebastian S Cocioba
CEO & Founder
New York Botanics, LLC

Sent via Mobile E-Mail

[Simon Quellen Field]

Phidgets are quite nice if you don't mind paying $218.25 for one, and then adding

all the really nice bits they have for sensors, etc.

If you are more willing to trade C++ coding skills for money, the MSP430 Launchpad

is $4.35 each, and you get two processors for that price, so one can remain in the

USB development kit and talk to the other one using I2C or SPI or even old fashioned

serial.

I usually buy five at a time. Like all of the other microcontrollers, there is a large

code base of free software, and a vigorous user community, and support from one

of the larger semiconductor companies.

If you're rather have a Linux based computer, complete with Ethernet port and two

USB ports (for keyboard and mouse, or for an extra hard drive, webcam or sensors),

and an HDMI port for 1080p video, the Raspberry Pi is only $35. I bought a couple

of those, and one is connected to the big screen in my home theater for web surfing.

I'm in the middle of setting up the other as a web server. I'll let the list know when

I give it a dynamic DNS URL, and you can all bring up its web page and play with it.

(That may be a while, I have a couple of book deadlines that I'd like to meet first.)

So while I love the Phidget and all of its nice accessories and ease of programming,

I just can't see trading one for 6 Raspberries or 50 Launchpads, given the big price

differences.

Given that an Arduino with Ethernet is $20 more than the Raspberry and doesn't

support Linux or HDMI etc., and the cheap ones without Ethernet are less powerful

than the Launchpad but cost 5 times more, I have stopped recommending them as well.

The Propeller Demo Board is also quite nice for $40, but it is VGA only, and doesn't

have enough memory to run Linux, and it's $5 more than the Raspberry anyway.

All of the devices I am comparing to the Launchpad and Raspberry Pi require

less learning and soldering than my less plug-and-play alternatives, but for the

DIY crowd, trading skill and learning time for money is a common tradeoff.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[John Griessen]

On 09/23/2012 05:11 PM, Simon Quellen Field wrote:
> quite nice if you don't mind paying $218.25 for one, and then adding
> all the really nice bits

+1 for MSP430.

[ByoWired]

On Sunday, September 23, 2012 3:13:03 AM UTC-4, archaeon wrote:... Anyone have any experience with sensors and microcontrollers...

I, too, suggest you look at the Propeller chip made by Parallax.com.  Its 8 parallel processors allow you to do all sorts of automation at the same time.  It uses a language called SPIN which isn't too hard to learn.   In my first real project with the chip, I built a coincident photon counter that scanned various energy levels (via digital pots), measured temperature via DS18B20s, operated a real time clock, displayed the real time data on a VGA screen, accepted some commands via mouse clicks, and recorded all the data on an SD card.  Not bad for a single chip at $8.  It has a built-in capability for providing output to a TV or VGA screen.

If you want to stick with a BASIC computer language, then Parallax also sells the BASIC Stamp 2, which costs more and does less, but it has great tutorials and is easy to use and understand.  It's been around a long time, too.  But it can't do parallel processing.  

Also using a BASIC type of language is the PicAxe, which ranges from about $3 to $10.  It's pretty good for projects that aren't too sophisticated.  It can't do actual parallel processing (although it can kinda fake it sometimes) but these chips have some nifty built-in functions like reading the DS18B20 temp sensors, analog-to-digital, etc.  

I don't know anything about the Arduino.  I hate the name, and we all know names make or break computational capability, so I've avoided it at all costs.  

[ByoWired]

On Sunday, September 23, 2012 6:12:02 PM UTC-4, Simon Field wrote:...

The Propeller Demo Board is also quite nice for $40, but it is VGA only....

Actually, it has an RCA jack that will allow you to plug in a TV, too.

And because it has 7 or 8 unused pins, I think you could easily use that board to run a serial LCD or something like that, too.

But Parallax also sells Propeller Proto boards for about $20, which give you access to all the pins, so practically anything is possible with that.

[Simon Quellen Field]

When I said VGA only, I meant it can't display more than 640x480.

On a TV you'll get even less.

I was comparing it to the cheaper, more capable, Raspberry Pi,

which has a 32 bit ARM chip running Linux (lots and lots of simultaneous

processes, not just 8), and has 1920x1080 resolution using HDMI.

The Propeller is quite nice, and has a lot going for it.

It's just that bang-for-buck isn't one of them.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

--

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.

To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/YDoasUdOmnAJ.

[Nathan McCorkle]

Well Linux doesn't actually give you parallel processing power, it just looks that's way to most people (there's a process scheduler handing the processor over to different jobs, quite fast of course). Now whether people need true parallel processing or not is a good question.  I don't know how raspberrypi has progressed but last I checked it still wasn't as easy to start using as the arduino or propeller.

There's also the new TI launchpad ARM board for about $5

I like the approach of learning about all the options available, then deciding which is most appropriate.

[Jonathan Cline]

On Sunday, September 23, 2012 12:13:03 AM UTC-7, archaeon wrote:

I'm trying to rig up a few sensors (temp, dissolved oxygen, etc.) and pumps/aerators in a biodigester.

I would strongly recommend using Microchip PICs if you would like your design to be industrial-strength.  As for MSP, I was talking to a couple colleagues who had TI MSP chips in different designs and both had real engineering complaints, in fact they would like to dump these a.s.a.p. and move to Microchip PICs like the rest of the designs which are very successful.   Arduino is not used frequently in industry for prototyping and across the board most dislike Atmel (due to many problems).  BTW You can run python byte-code on a PIC, search for python-on-a-chip.  I have not yet tried this myself though it seems solid.
 
I have recently switched most of my engineering dept to using Microchip PIC UBW boards for rapid prototyping and test.   Results are very positive.  One comment was "It seems expensive for what it is" yet the engineer was up & running within 2 days with his custom etched board and customized UBW firmware vs. putting up with other lackluster environments and chip limitations and a learning curve, etc.  You can use the UBW with Processing.org (simplified Java-style language) for getting things up quickly, then modify the UBW source code to run stand-alone.     An example is here: http://88proof.com/synthetic_biology/blog/archives/270    At my lab we are using UBW with python, controlling UBW via serial port for prototyping, then later custom microcontroller code runs stand-alone without the PC.  To repeat the often FAQ, Microchip does have a free C compiler and the software tools are very good (the new tools run on windows/linux/osx and are based on Netbeans IDE).


I recently got a Raspberry Pi board to play with, yet calling it $35 is a misnomer since it still needs a flash card, power supply, and other development hardware to program (separate monitor/keyboard for non-embedded work, flash card reader/writer to make the file system, etc).

Propeller is nice for toy projects though I can't see it being used for something real.  "It can run multiple things at a time"  ummm well, just run a well-timed scheduler on a single core of a good microcontroller instead.  Remember the purpose of the Propeller is to replace the Basic STAMP, which was a typical 6th/7th/8th grade level hardware kit (with associated limitations), so as long as that fits the bill, then OK; otherwise, it's easy to outgrow the capabilities (either hardware or environment) relatively quickly -- similar to what happens with most Arduino/Atmel projects -- hardware or software doesn't fit well for industrial-strength code. 


Anyway, whichever you chose, the best thing to do is to prototype your system using a PC-attached-to-a-board, then when that is mostly working, re-implement for the non-PC-attached version.  Make sure the sampling rate of your sensors is compatible with the hardware you choose, and if you're doing remote data-logging, that there's enough flash memory to store the data you want to log.


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

[Gavin Scott]

The Propeller chip is indeed very cool, as it works as a synchronous interrupt-free kind of deal which makes interfacing with other devices comparatively simple in software (you basically dedicate one of the 8 "processors" to an I/O device and it just sits and watches the I/O lines).

But you can do pretty much all the same stuff with an Arduino and there's a huge amount (i.e. much more than the Propeller) of both hardware and software out there to help.

For the type of sensors and control you're talking about, I don't think you need something terribly fast, and you may be more interested in the number of I/O lines and how many you can drive simultaneously and with how much current in order to do simplify interfacing to your other hardware. In this respect an Arduino may be more capable than a Propeller-based solution I believe.

Sparkfun is a great place to look for both microcontrollers and sensors (in module or "breakout" form) that are really easy to interface to one another, and you can often get away with thinking of these things as Lego bricks and not have to worry too much about the electrical stuff (at least relatively speaking; you can still smoke things by hooking them up wrong, expecting a microcontroller to drive or sink an amp of current through one pin, etc.)

http://www.sparkfun.com/

Another interesting option is the Netduino which is an Arduino hardware compatible controller that runs the .Net micro edition so you get to use the (free) Microsoft C# Express development tools with it (if you like that sort of thing).

https://www.sparkfun.com/products/10186

For the 'duinos they have a standardized physical layout that allows one to plug daughterboard "shields" on top of them for simple expansion. Though when you start talking about stacking multiple shields you can quickly run into conflicts between them, and there's actually a web site dedicated to documenting shield resource usage:

http://shieldlist.org/

G.

[Cathal Garvey]

The Raspberry Pi itself isn't really well suited to a role as a
Microprocessor on its own. Not just because it vastly overpowers the
tasks you generally use microcontrollers for, but because it's not
designed for hardware hacking like an Arduino (et al) is.

That is, Arduinos can take a bit of punishment, voltage-wise, without
taking much damage. If you accidentally dump a little too much current
on an input pin for a second, probably no big deal. Although I've never
tried, I get the impression that Raspberry Pi is not so robust. Doesn't
take much imagination to see why; most Arduinos are large-ish packages
designed for hand-soldering or at least handling, probably with
large-ish traces inside, whereas the RPi runs on a tight
system-on-a-chip, a marvel of minimalisation that was never designed for
live outside a commercial embedded system.

The RPi is for hacking software in an embedded environment; if you want
to hack hardware with it, the prevailing suggestion is to link it up
with an Arduino and let the Arduino do the hardware interfacing, and the
RPi handle the crunchy data.

That said, if you want to risk ~�38, the RPi has a good few GPIO pins
you can repurpose, and some header slots for small cameras and LCDs that
are apparently due to be released in the near future. Lots of fun to be
had. :)

On 24/09/12 03:54, Nathan McCorkle wrote:
> Well Linux doesn't actually give you parallel processing power, it just
> looks that's way to most people (there's a process scheduler handing the
> processor over to different jobs, quite fast of course). Now whether people
> need true parallel processing or not is a good question. I don't know how
> raspberrypi has progressed but last I checked it still wasn't as easy to
> start using as the arduino or propeller.
>
> There's also the new TI launchpad ARM board for about $5
>
> I like the approach of learning about all the options available, then
> deciding which is most appropriate.
> On Sep 23, 2012 9:37 PM, "Simon Quellen Field" <sfi...@scitoys.com> wrote:
>
>> When I said VGA only, I meant it can't display more than 640x480.
>> On a TV you'll get even less.
>>
>> I was comparing it to the cheaper, more capable, Raspberry Pi,
>> which has a 32 bit ARM chip running Linux (lots and lots of simultaneous
>> processes, not just 8), and has 1920x1080 resolution using HDMI.
>>
>> The Propeller is quite nice, and has a lot going for it.
>> It's just that bang-for-buck isn't one of them.
>>
>> -----
>> Get a free science project every week! "http://scitoys.com/newsletter.html
>> "
>>
>>
>>
>>
>> On Sun, Sep 23, 2012 at 6:26 PM, ByoWired <byow...@gmail.com> wrote:
>>
>>>
>>>
>>> On Sunday, September 23, 2012 6:12:02 PM UTC-4, Simon Field wrote:...
>>>>
>>>>

>>>> The Propeller Demo Board is also quite nice for $40<http://www.parallax.com/Store/Microcontrollers/PropellerDevelopmentBoards/tabid/514/CategoryID/73/List/0/SortField/0/Level/a/ProductID/340/Default.aspx>,

>>>> but it is VGA only....
>>>>
>>>> Actually, it has an RCA jack that will allow you to plug in a TV, too.
>>> And because it has 7 or 8 unused pins, I think you could easily use that
>>> board to run a serial LCD or something like that, too.
>>> But Parallax also sells Propeller Proto boards for about $20, which give
>>> you access to all the pins, so practically anything is possible with that.
>>>
>>> --
>>> 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.
>>> To view this discussion on the web visit
>>> https://groups.google.com/d/msg/diybio/-/YDoasUdOmnAJ.
>>>
>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>
>>>
>>>
>>
>> --
>> 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 https://groups.google.com/groups/opt_out.
>>
>>
>>
>

--
www.indiebiotech.com
twitter.com/onetruecathal
joindiaspora.com/u/cathalgarvey
PGP Public Key: http://bit.ly/CathalGKey

[Jonathan Street]

I'm not sure that argument is entirely sound.  It's only really the Uno that uses a through hole chip. The Leonardo, Mega, Ethernet and BT all use surface mount chips. The conclusion may still be sound though. On a (very) quick look I'm struggling to find detailed info on the raspberrypi.

On 24 September 2012 15:49, Cathal Garvey <cathal...@gmail.com> wrote:

The Raspberry Pi itself isn't really well suited to a role as a
Microprocessor on its own. Not just because it vastly overpowers the
tasks you generally use microcontrollers for, but because it's not
designed for hardware hacking like an Arduino (et al) is.

That is, Arduinos can take a bit of punishment, voltage-wise, without
taking much damage. If you accidentally dump a little too much current
on an input pin for a second, probably no big deal. Although I've never
tried, I get the impression that Raspberry Pi is not so robust. Doesn't
take much imagination to see why; most Arduinos are large-ish packages
designed for hand-soldering or at least handling, probably with
large-ish traces inside, whereas the RPi runs on a tight
system-on-a-chip, a marvel of minimalisation that was never designed for
live outside a commercial embedded system.

The RPi is for hacking software in an embedded environment; if you want
to hack hardware with it, the prevailing suggestion is to link it up
with an Arduino and let the Arduino do the hardware interfacing, and the
RPi handle the crunchy data.

That said, if you want to risk ~€38, the RPi has a good few GPIO pins

[Jelmer Cnossen]

The hacking-proofness that Cathal mentions could be the resettable fuse on the Arduino (which the SMD versions also have I think). I'm always amazed that the thing still works after shorting random pins and accidentally resetting the thing because I dropped a wire on it. Not sure it's really a good argument though, I would say the large amount of libraries is the most useful aspect of arduino, especially for a beginner. 

ARM is both cheaper and faster so personally I'm on the verge of switching.

[Simon Quellen Field]

I think you worry too much.

But also, you might want to actually play with a device before you go public

with your critiques. :-)

I have several Arduinos, a bunch of 430 Launchpads, and a couple Raspberry Pis.

The Launchpads are so cheap I don't worry at all about what I connect them to,

although I have yet to damage one. The entire USB development board with

two processor chips is cheaper than a plain Arduino chip all by itself.

The Arduino Uno boards generally go for $21 (but HobbyKing has them for $15.29).

For that you get an 8 bit computer with 0.032 megabytes of flash and 0.002

megabytes of RAM, no USB, No video, and it runs at  a very slow 16 MHz.

For the price of two of those, you can get the Raspberry Pi.

A 32 bit computer running at 700 MHz (with hardware floating point and

memory management), a GPU that can do BueRay quality video at 1920x1080

at 30 frames per second and 24 gigaflops of processing speed, stereo audio

capability, an Ethernet port, 256 MB of RAM, two USB ports, HDMI and RCS

video output, and an SD card slot. It runs Linux, whose software library vastly

exceeds that of any of the little 8 bit boards we've been talking about. You can

surf the web on this little board. You can host a web server on it, so your friends

can run their experiments in your lab from their homes.

If you wanted to add Ethernet to an Arduino, that board costs $65. And that board

has no USB ports, so you need another device to program it.

You can short the pins on the Raspberry Pi without damaging the device.

I just did, by wrapping some aluminum foil around all the header pins.

It rebooted, but came up just fine.

If you want to connect it to something that might deliver 240 volts, just add

a 20 cent optoisolator chip, like you would when you were protecting anything

else.

Let's look at some of the projects this group might want to put a microprocessor

in. Some folks are building spectrometers. Wouldn't it be nice to be able to just

connect a web cam to a USB port, and use Linux free software to operate it? And

then be able to send the results to the web automatically? Or control it from the

web? Or use floating point and lots of RAM to process the data right in the device,

instead of sending it to a desktop or laptop computer to process? And wouldn't it

be nice to be able to store the results on a USB flash drive, or a terabyte hard

drive, and still have all that data accessible on the web? Or be able to see the

processed results on your HD television?

$35 is a bit much for a processor to go into an incubator. Unless you'd like the

incubator to be something you could check on using your smart phone. Or record

it's operation on a web-connected storage device. Or let you sous vide cook your

dinner when you start it from the browser in your phone.

Or maybe your device needs to run 100 experiments at the same time. Or maybe

some people in your group like to program in Python, some in Ruby, some in C++,

and some in Java? They can all write software for this device, and test it on their

own computers, where they might have some enterprise-class software 

development tools.

The Raspberry Pi would give you headroom to expand your project, rather than

having to develop it on a 32 kb Arduino and worry about squeezing in the code,

or about whether it would be too slow to use that software floating point, or

whether your data buffer would fit in the 2kb of RAM.

The 430 Launchpad is in-between the 8 bit toys and the Linux box. It is a 16 bit

processor with hardware floating point. In the other respects it is as limited as

the 8 bit toys, but it is also a lot cheaper than any of them.

I love all the little 8 bit guys, they are lots of fun to play with.

But I just can't justify building anything on them these days when more powerful

boards are either cheaper or vastly more capable.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

On Mon, Sep 24, 2012 at 12:49 PM, Cathal Garvey <cathal...@gmail.com> wrote:

The Raspberry Pi itself isn't really well suited to a role as a
Microprocessor on its own. Not just because it vastly overpowers the
tasks you generally use microcontrollers for, but because it's not
designed for hardware hacking like an Arduino (et al) is.

That is, Arduinos can take a bit of punishment, voltage-wise, without
taking much damage. If you accidentally dump a little too much current
on an input pin for a second, probably no big deal. Although I've never
tried, I get the impression that Raspberry Pi is not so robust. Doesn't
take much imagination to see why; most Arduinos are large-ish packages
designed for hand-soldering or at least handling, probably with
large-ish traces inside, whereas the RPi runs on a tight
system-on-a-chip, a marvel of minimalisation that was never designed for
live outside a commercial embedded system.

The RPi is for hacking software in an embedded environment; if you want
to hack hardware with it, the prevailing suggestion is to link it up
with an Arduino and let the Arduino do the hardware interfacing, and the
RPi handle the crunchy data.

That said, if you want to risk ~€38, the RPi has a good few GPIO pins

[John Griessen]

On 09/24/2012 07:16 PM, Simon Quellen Field wrote:
> You can short the pins on the Raspberry Pi without damaging the device.
> I just did, by wrapping some aluminum foil around all the header pins.
> It rebooted, but came up just fine.

Thanks for this semi-scientific testing for us!

I want some R-Pi's! and a touch screen that connects easily... And a camera, and...

[Jonathan Cline]

On Monday, September 24, 2012 5:17:29 PM UTC-7, Simon Field wrote:

 Some folks are building spectrometers. Wouldn't it be nice to be able to just

connect a web cam to a USB port, and use Linux free software to operate it? And

then be able to send the results to the web automatically? Or control it from the

web? Or use floating point and lots of RAM to process the data right in the device,

instead of sending it to a desktop or laptop computer to process? And wouldn't it

be nice to be able to store the results on a USB flash drive, or a terabyte hard

drive, and still have all that data accessible on the web? Or be able to see the

processed results on your HD television?

Wouldn't it be nice to have it be *inexpensively* ?
The best way to create a cost-reduced design is to strip out the cost at the very beginning, not add the kitchen sink from the start.  OpenPCR costs >$600 when it was supposed to be "the affordable thermocycler" for example; by chosing Arduino, the resultant circuit board was 3x to a comparable design, with the mistaken rationale: "we wanted to get going quickly"  when in retrospect the choice of Arduino probably cost as much time in working around it's limitations as any supposed development savings.   Development time & effort are one-shot events, whereas production costs are recurring.

Minimize the design at the start and don't pick the "because it's cool" product.  Pick the one that fits best.  In some cases the resultant design can easily end up as a commercial product if it's designed as industrial-strength from the beginning.

Linux is a perfect example, it will add at least 30% to the cost of anything just from that choice alone, and "send to the web automatically or control it from the web" can be done in other, simpler ways than embedding a bloated embedded Linux kernel.  Another example, most video recorders don't use a file system per se (as Linux would view a file system, and as Linux would carry around the baggage for), so recording video doesn't need a flashfs, even if using an external hard drive.

Minimize minimize minimize...   get the design down to reasonable cost and simplicity at the same time!

[Michael Turner]

Jonathan Cline wrote: "As for MSP, I was talking to a couple

colleagues who had TI MSP chips in different designs and both had real
engineering complaints, in fact they would like to dump these a.s.a.p.
and move to Microchip PICs like the rest of the designs which are very
successful."

What were their specific complaints?

Regards,
Michael Turner
Project Persephone
1-25-33 Takadanobaba
Shinjuku-ku Tokyo 169-0075
(+81) 90-5203-8682
tur...@projectpersephone.org
http://www.projectpersephone.org/

"Love does not consist in gazing at each other, but in looking outward
together in the same direction." -- Antoine de Saint-Exupéry

> --
> 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.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/diybio/-/q1sEbaKHfxsJ.

[Simon Quellen Field]

In part, you're making my point.

I think that might be because we agree more than disagree.

Why use a $15 Arduino when you have a $4.35 Launchpad?

And if the problem was working around the limitations of the Arduino,

starting with a cheaper and more capable system might have helped.

The comment about 'resigned as industrial-strength' speaks to that as well.

For DIY folks, I don't think Linux is adding 30% to anything.

I'd love to hear where you have found a web capable microcontroller system

for less than $35. Can it run PHP and MySQL and Apache too?

For DIY one-offs, or products where the expected market is in the hundreds

or at most thousands, such as OpenPCR, I'm also not convinced that the

development cost is dwarfed by the recurring production costs. Certainly

not at the rates that I charge for software development. A man-month of

coding in my neighborhood is probably more than the profit OpenPCR

is ever going to make.

I tend to go in the opposite direction.

I would build something that connected to an expensive PC or laptop, and

use the capabilities of that machine, in much the same way that I would

use a car to transport things instead of putting an engine and wheels on

anything too big to carry. People have cars and computers.

So when something comes along for $35 that essentially has all the power of

the computers we used 5 years ago (but with much better graphics and a GPU),

doesn't require me to learn a new limited toy language, and has tons of software,

multitasking, disk space, RAM, Ethernet, video, USB, and all the rest, it makes

sense to me to build around that. It is extensible, I know I won't be cramped by it,

it will have the speed I need, and I can use it on many different pieces of

equipment at once, since it can run them all in separate processes, each with its

own web server. So I amortize the CPU across a whole lab full of gadgets.

Now my incubator, spectroscope, PCR device, microscope controller, and all the

rest can split the $35 between them.

I think blaming the $600 price of OpenPCR on a $15 Arduino might be going a bit

too far as well. :-)

There is a lot to be said for giving even simple devices a decent user interface.

How do you want to set the temperature on your incubator?

A simple knob works fine, but then you have to calibrate it.

A web interface I can access on my phone allows me to get a digital readout,

enter the temperature on a keyboard or use sliders, and the temperature is

set without having to calibrate a potentiometer value to the temperature

sensor. And because my device doesn't need an LCD screen, a potentiometer

and dial, beepers, LEDs, and such, I don't have to buy them and solder them

in -- I have my phone. And I can check on the incubator from anywhere.

Devices with enormous computing power are becoming the bricks of modern

design. Thirty years ago someone would have complained that a design was

using a computer chip when it could have been done with transistors.

These days, the Linux box is becoming the computer chip of thirty years ago.

It is a building block. It is capable of a whole lot more than any one design will

ever use it for, and that's OK. Because it allows you to solve problems you didn't

expect, and add features you didn't know you needed. And all for $35 (and there

is a $25 version of the Raspberry Pi too).

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

--
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.

To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/6a22WgRR268J.

[Tristan Eversole]

I've followed this thread with great interest, because I'm working on a project with requirements similar those mentioned in the original post: I'm trying to keep brachiopods in closed aquaria. Nobody knows how to do this, although apparently someone was able to do it in the 1970's; whenever I show experienced aquarists the paper which outlines how she did it, they... don't believe it. So I'm expecting this to be a very difficult, long-term project-- just getting the brachiopods is going to be an adventure in itself.

Brachiopods are filter feeders, and marine filter feeders turn out to be hard to maintain in closed aquaria. The most important parameters for maintaining filter feeders, apparently, are 1) laminar flow of water through the tank, 2) cleanliness of the water, and 3) evenness of dissolved oxygen throughout the tank. One must also know what the filter feeders eat, and in the great majority of cases nobody does. I plan on trying the seawater/beef heart extract approach described in McCammon's 1973 paper-- animals frequently thrive on weird things which aren't found in their natural habitats-- but I believe that I'll go through an awful lot of brachiopods before I get it right.

Proprietary aquarium controllers which could do the job exist. Proprietary controllers for biodigesters probably exist as well, and if they don't, you might be able to adapt something from fermentation tanks; perhaps it'd be worth looking at amateur brewing operations? In my case, the aquarium controllers are quite expensive. (The whole project will be expensive: it'll take a large tank with a good filter, a laminar flow setup, and a refrigeration unit.) I know that this kind of controller project is likely to come up in the amateur biology world quite often, and I don't have much money (another reason why this is a very long-term project!), so if I can build an open monitoring and control system which comes out to being cheaper than the proprietary stuff, I think it'd be useful for lots of people.

However, time spent doing programming and electrical engineering is time not spent doing paleontology, evolutionary developmental biology, or population genetics, and I'm an evolutionary biologist first and foremost; although loads of things are interesting, I'll never be effective if I try to do everything. Knowing very little about microcontrollers, I thought I'd go for the Arduino: lots of people who know things about Arduinos are around, sensors have already been made for them, there are nice software libraries, and so on. As far as I can tell, the Pi would be nice, but it's frankly a lot more power than I could handle, because I don't know Linux, Apache, PHP, etc., and I would have to spend a lot of time not learning about R, morphometrics, stratigraphy, or coalescence theory if I wanted to use the Pi's full potential. I looked at Phidgets, but the expense killed that line of inquiry right quick-- I'd be better off with a proprietary aquarium controller. I've never heard of a Microchip PIC and have no idea how to use one-- what shields are available? Sensors? Software libraries? I have a Launchpad on order, though, because the price was good and I might ultimately need more power than the Arduino offers.

Casting around the internet, I found a few things. I haven't put very much time into this yet, because I have other projects which are more feasible and/or closer to completion(*), and I need to actually get a Master's and so on. Nailing the monitoring aspect comes before figuring out the control aspect-- imagine trying to deal with bugs on the control side without knowing whether the sensors are working-- so sensors are the place to start.

Hope this helps-- I expect these approaches could be adapted for a biodigester:

http://practicalmaker.com/documentation/arduarium-controller-ultimate-documentation

> The Arduarium Controller Ultimate has been designed to take care of everything that you may be interested in with an aquarium, aquaponics setup, hydroponics setup etc. Expandability, function and cost have been the three largest factors that were considered when designing this board.
>
> What's come out of the design process is a shield that costs under $120 (barely). If you include a relay board, Port Expander Shield, Arduino, and ethernet shield you have an amazing controller for ~$250 that does more than controllers that are multiple times more expensive.

http://reefprojects.com/wiki/Main_Page

> This site will revolve around the open source hardware and software of the Arduino Microcontroller. The goal will be to automate as much of the home aquarium as possible to both maintain a healthy aquarium and increase safety.

The list of planned sensors includes temperature, pH, water level, humidity, moisture, and voltage, which isn't directly applicable to a biodigester, for the most part, but still might be helpful. The site looks old, but the forums aren't dead.

> http://web.cecs.pdx.edu/~gerry/class/EAS199B/howto/

These are class project thingies from Portland State University which concern aquarium monitoring. Although it's not that expansive, there's some valuable information about building a salinity sensor here (http://web.cecs.pdx.edu/~gerry/class/EAS199B/howto/fishtank/wiring/salinity_sensor_wiring.html ). I guess that's not useful to you, but salinity sensors aren't that easy to come by (I haven't found too many so far, anyway), and it might be useful to other people.

I bought two Digisparks in a fit of enthusiasm and assume that they'll eventually find some use even if they're much too limited for my brachiopod project. (Maybe I can use them to grow plants.) I'm interested in them partly because they work with the Grove sensors, which look pretty dang impressive-- they've already got some water flow sensors (remember, that's super important to filter feeders) and temperature sensors (http://www.seeedstudio.com/depot/sensors-c-144.html ). The Seeed stuff in general seems pretty cool.

--T.


*One of my other projects, which is relatively far along, is to sequence developmentally important genes from the Triops genus. I don't recall whether I've mentioned this on the DIYBio list before, so I'll explain it here:

Triops is a genus of "living fossils", which allegedly haven't changed much in millions of years. A recent paper in PLoS One (Vanschoenwinkel et al. Toward a global phylogeny of the "living fossil" crustacean order of the Notostraca. PLoS ONE (2012) vol. 7 (4) pp. e34998 Available here: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0034998 ) presented a very nice phylogeny of the Triops species based on mitochondrial DNA, and also made the claim that they're not living fossils, based on a molecular clock analysis performed with a standard assumption of mutation rates in Crustacea. I'm not entirely sure it's reasonable to use standard rates of mutation in what might be a very weird group, and I also think that it would be highly interesting to know whether the actual morphology of the Trops has changed much in millions of years, a question which can't be addressed with mitochondrial DNA. One step down that path would be to compare gene sequences of critical developmental genes between Triops species and between Triops and the Crustacea generally. We have a few very short (approx. 100bp) sequences of these genes from T. longicaudatus, and we finally have a full crustacean genome sequence in the form of the Daphnia pulex genome, so I thought I'd try to pull sequences from T. cancriformis, the Triops species which is commonly alleged to be the Triops species mostly closely resembling its fossilized forebears.

Getting these sequences won't be as easy as I thought, but I believe it's still doable. Fortunately, Triops species are easy to get and easy to keep. I got T. cancriformis eggs off of eBay for $13, successfully hatched them, and currently have six specimens living in a donated 3 gallon tank. Triops are hardy and easy to keep, as well, so it only took me one try before I got the hang of it-- and the ones that didn't survive my first attempt are in the freezer awaiting DNA extraction. Alas, the PCR side of it may be more complex or laborious than I'd hoped, because the Daphnia genome is only available as a contigs (I think?) and I'll have to use RACE or primer walking to get the full gene sequences.

...The relevance of all this to an email about microcontrollers is that I have a tank set up and waiting for an Arduino-based sensor system. :)

[Simon Quellen Field]

You might want to reconsider the Raspberry Pi.

First of all, it is silly to worry that you won't use all of its capabilities.

You won't be using all of the capabilities of any of the microcontrollers you

might evaluate. But given that the processor is such a small part of the

cost of your system, why put all of your effort into a $15 Arduino when

you could use a $35 Pi and not worry about whether your computer is going

to limit you?

Because the Pi is basically a laptop computer without the screen and keyboard,

what you know about the computers you use for home, school, and work will

apply to it. The Arduino is nothing like what you are used to, and will take

longer to learn, even though it is very simple.

There is free open source software for controlling things like your project that

have already been written for Linux-based computers:

"http://code.google.com/p/marelab-solar/"

"http://www.youtube.com/watch?v=nayBCN4foL8"

"http://code.google.com/p/aquacontroller/"

"http://ronnybull.com/linux-reef-controller/"

"http://www.homebrewtalk.com/f84/brewtarget-1-2-4-free-brewing-software-277810/"

"http://www.youtube.com/watch?v=QheVIVao0us"

"http://brewpi.com/survey/index.php?sid=72293&lang=en"

"http://www.homebrewtalk.com/f51/monitoring-controlling-linux-cheap-240955/"

"http://www.linuxjournal.com/article/4627"

...and many, many more. Embedded Linux controllers have been controlling

industrial systems for many years.

The advantage of a multitasking web connected controller is that you can monitor

it from anywhere on your phone or laptop, and it can send you alerts if something

goes wrong with pumps, temperature, power failure (use a UPS for your router),

oxygen, pH, or anything you'd like to monitor.

It can keep a record of all the sensors on the hard drive, so when you kill your

first batch, you can find out what went wrong and fix it before trying again.

You can watch graphs of all the things you are measuring, and adjust things in

response over the web, without having to be home watching it all the time.

It can control off-the-shelf USB gadgets such as USB controlled power strips

so you can control mains power without needing to solder anything (although

I would probably build my own such device for a couple bucks -- I did that while

I was working at Google by putting a 3 solid state relays into a power strip so

that I could control a traffic light using three parallel port pins).

Rather than build around a single threaded limited microcontroller and wish you

had more power, features, or flexibility, consider spending another $20 and get

a full computer with a real operating system and a huge software base and tons

of people who know the operating system and have a habit of helping out for free.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[Cathal Garvey]

> I think you worry too much.
> But also, you might want to actually play with a device before you go
public
> with your critiques. :-)

I'm probably more critical of fragile tech because Ireland has long
shipping lead-times for replacements; there are virtually no reasonably
priced stores for electronics in Cork, possibly the whole Island, and
the only ones to stock Arduinos are overpriced by far.

So, robustness is critical to me because I can't afford to wait a week
or two for a replacement when something goes wrong. :P

On 25/09/12 01:16, Simon Quellen Field wrote:
> I think you worry too much.
> But also, you might want to actually play with a device before you go public
> with your critiques. :-)
>
> I have several Arduinos, a bunch of 430 Launchpads, and a couple Raspberry
> Pis.
>
> The Launchpads are so cheap I don't worry at all about what I connect them
> to,
> although I have yet to damage one. The entire USB development board with
> two processor chips is cheaper than a plain Arduino chip all by itself.
>
> The Arduino Uno boards generally go for $21 (but HobbyKing has them for

> $15.29<http://www.hobbyking.com/hobbyking/store/__26838__Arduino_Uno_R3_Microcontroller_Atmel_ATmega328.html>
> ).

> it's operation on a web-connected storage device. Or let you *sous
> vide*cook your

>> That said, if you want to risk ~�38, the RPi has a good few GPIO pins

[Cathal Garvey]

But yea, great to know the RPi is tougher than I'd previously read. :)

[Simon Quellen Field]

I sympathise. I live on a mountain, and often there are weeks when I only

go into town on Wednesdays for a meeting. So I usually buy multiples of

whatever I need. That said, my mountain overlooks Silicon Valley, so if I

have incentive to get in the car, there are dozens of hobbyist friendly

electronics vendors half an hour away. But I do most of my shopping via

the Internet. Given that the Launchpad development board is cheaper

than the Arduino chip all by itself, you can afford more Launchpads than

Arduinos. And the Raspberry Pi comes from England, so the shipping

time shouldn't be that bad. :-)

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

>> That said, if you want to risk ~€38, the RPi has a good few GPIO pins

[Tristan Eversole]

On Sep 25, 2012, at 3:42 PM, Simon Quellen Field wrote:

> Because the Pi is basically a laptop computer without the screen and keyboard,
> what you know about the computers you use for home, school, and work will
> apply to it. The Arduino is nothing like what you are used to, and will take
> longer to learn, even though it is very simple.

> The advantage of a multitasking web connected controller is that you can monitor
> it from anywhere on your phone or laptop, and it can send you alerts if something
> goes wrong with pumps, temperature, power failure (use a UPS for your router),
> oxygen, pH, or anything you'd like to monitor.
>
> It can keep a record of all the sensors on the hard drive, so when you kill your
> first batch, you can find out what went wrong and fix it before trying again.
>
> You can watch graphs of all the things you are measuring, and adjust things in
> response over the web, without having to be home watching it all the time.

You make a compelling argument for the Raspberry Pi. I hadn't spent much time looking at the Pi before, but it's clearly warranted now.

Nonetheless, I think you really hit the main issue on the head in that first paragraph above. The Pi is basically a laptop computer without the screen and keyboard, but that doesn't help me much:

1) I have no familiarity with Linux, Python, PHP, Apache, bash, etc. I currently have no idea how to set up THIS computer as a web server, or have it display continuously-monitored data as a graph. (Can I use R?) Maybe this is much, much easier than I think it is, or maybe I can get someone to help me, but I wouldn't count on it. I'm hardly averse to learning this stuff, either, but I have a very hard time not exploding into twenty different topics as it is, and I need to watch that.

2) I also have no idea how to connect this computer to a bunch of sensors designed to monitor environmental parameters. The Phidgets are easy, I think, but too expensive. In my particular case, the legacy of my long-standing and wholly unsuccessful attempts to get into robotics is a slightly greater familiarity with things like H-bridges than with things like sorting algorithms or data structures, so the low-level nature of the Arduino isn't as big a barrier to me as you might expect, although it's still substantial.


1) is unquestionably very nice. In the case of brachiopods (and possibly marine filter feeders in general), nobody actually knows what the relevant boundaries on these parameters are, so easy viewing of them would have serious value; I might have an easier time figuring out what kills brachiopods. But 2) is essential. The entire project is dead in the water (har) unless I can do 2).

The advantage of the Arduino, on this score, is that I know that I can actually go out and buy sensors for water flow, temperature, pH (https://www.sparkfun.com/products/10972 or http://atlas-scientific.com/product_pages/sensors/ph-sensor.html), dissolved oxygen (http://atlas-scientific.com/product_pages/sensors/do-sensor.html ), and oxidation-redox potential (http://atlas-scientific.com/product_pages/sensors/do-sensor.html ). Like, right now. If I want to go in the direction of a digester, a fermentation tank, a hydroponic system, an aeroponic system, or a terrarium, I can sense methane, carbon monoxide, alcohol, hydrogen (all at https://www.sparkfun.com/categories/146 ), humidity (http://www.seeedstudio.com/depot/grove-temperaturehumidity-sensor-pro-p-838.html ), and light. Sensing salinity is a problem, but at least I can find attempts at Arduino-compatible salinity sensors.

Similarly, I know that a bunch of people have tried to do these things before; the most recent find was the BioBoard project at Noisebridge (https://www.noisebridge.net/wiki/BioBoard ). It's clearly possible to make the Arduino do some data logging and PID control, as well, so I know I can continue beyond the sensing step using an Arduino.

Here, the sheer *seniority* of the Arduino is a substantial advantage-- unless it's trivial to connect all these things to the Raspberry Pi, in which case, hey, awesome. As this is a long term project, I could just sit around and wait for the Pi to "catch up", in terms of the availability or compatibility of sensors; how long do you think that would take?

Now, I obviously haven't looked into these issues in any remotely comprehensive way. It seems as if there's a significant overlap between many of these different environmental control problems, so it may be possible to design a board or something with general utility, one that could be used in everything from reef tanks to chemostats. (I guess this is the goal of the BioBoard project, really. Maybe I should talk to them, because they're roughly in the area and all.) DIYBio people: Would it be useful to you for me to go down this road? Do lots of you think you might need an environmental monitoring/control system? I'll probably do it anyway, because I'm pretty serious about the brachiopods, but I'd love to know.

--T.

[Jonathan Street]

I hadn't realised the raspberrypi is being made in the UK. Thanks for the heads up. More info at http://www.bbc.com/news/technology-19510040 for those interested.

Tristan, if you've learnt enough about R to feel comfortable with it I would say putting a few graphs on the web will be very simple. You probably could use R to run a web server but it isn't really the right tool for the job. Python is, as far as I understand it, the preferred language for the raspberrypi and as it is quite a nice fit for your project anyway I would go with that. It is a nice fit because it does very well at scientific computing tasks (check out numpy, scipy, matplotlib) AND websites/servers (you're spoilt for choice here but won't go far wrong with any of them).

That cuts the learning down to linux (most of the tasks you'll need to do here will be covered by tutorials), python and bash (the basics should get you a long way and be relatively easy to pick up).

That being said if you can build on the work of others with the arduino then go for it. The cost of any of these microcontrollers is relatively small in comparison to the cost of the sensors for instance. You can always add or change microcontrollers when you have something working.

[Tim McNamara]

On 27 September 2012 08:23, Tristan Eversole

<custome...@trioptimum.com> wrote:
>
> On Sep 25, 2012, at 3:42 PM, Simon Quellen Field wrote:
>
>> Because the Pi is basically a laptop computer without the screen and keyboard,
>> what you know about the computers you use for home, school, and work will
>> apply to it. The Arduino is nothing like what you are used to, and will take
>> longer to learn, even though it is very simple.
>
>
>> The advantage of a multitasking web connected controller is that you can monitor
>> it from anywhere on your phone or laptop, and it can send you alerts if something
>> goes wrong with pumps, temperature, power failure (use a UPS for your router),
>> oxygen, pH, or anything you'd like to monitor.
>>
>> It can keep a record of all the sensors on the hard drive, so when you kill your
>> first batch, you can find out what went wrong and fix it before trying again.
>>
>> You can watch graphs of all the things you are measuring, and adjust things in
>> response over the web, without having to be home watching it all the time.
>
> You make a compelling argument for the Raspberry Pi. I hadn't spent much time looking at the Pi before, but it's clearly warranted now.
>
> Nonetheless, I think you really hit the main issue on the head in that first paragraph above. The Pi is basically a laptop computer without the screen and keyboard, but that doesn't help me much:
>
> 1) I have no familiarity with Linux, Python, PHP, Apache, bash, etc. I currently have no idea how to set up THIS computer as a web server, or have it display continuously-monitored data as a graph. (Can I use R?) Maybe this is much, much easier than I think it is, or maybe I can get someone to help me, but I wouldn't count on it. I'm hardly averse to learning this stuff, either, but I have a very hard time not exploding into twenty different topics as it is, and I need to watch that.

If you're interested, I am happy to gather some content together for a
tutorial. I'll put it up on data.geek.nz. It can be intensive to write
these kinds of tutorials well though, and would only bother if you're
genuinely interested. If you just want graphs from a data stream, I
recommend Cosm (https://cosm.com/about_us). That way you don't need to
know about the web server component, just how to send data to Cosm.
Perhaps part 2 could be a completely self-service model.

[John Griessen]

On 09/26/2012 03:23 PM, Tristan Eversole wrote:
> it may be possible to design a board or something with general utility,
one that could be used in everything from reef tanks to
> chemostats. (I guess this is the goal of the BioBoard project, really.
Maybe I should talk to them, because they're roughly in
> the area and all.

Taking all you've said about your own experience level, this is sounding the most
results-getting direction yet. Does the bioboard have cost levels you can fit to your want?

Going with low performance Arduinos because they're easy,
yet not too too expensive, but more expensive than some,
sounds like a miss. If you try microchip you'd need to learn c programming
and reading long datasheets. If you try this ARM chip board from Olimex,
you could get a development method similar to arduino...not sure:
https://www.olimex.com/Products/Duino/STM32/OLIMEXINO-STM32/

The arduino-like development code comes from: http://leaflabs.com/docs/ http://leaflabs.com/devices/

(I've not tried it) but... it sounds useful... and you get access to rapid prototyping
shield hardware addons with sensors on them...

[Dakota]

Would this http://atlas-scientific.com/product_pages/kits/ph-kit.html

Work with this?
http://www.digikey.com/product-detail/en/MSP-EXP430G2/296-27570-ND/2331789?cur=USD

If so I'm ordering one, my arduino is tied up in something else and
it'd be great to not have to spend another $30 on one.

[Simon Quellen Field]

Yes, the Launchpad has serial I/O and will work with that hardware.

One of the more modern I/O options would be better, sich as I2C or

SPI, but old fashioned serial is certainly one of the MSP430's capabilities.

Converting the PIC C code or Arduino code to the Launchpad is also easy.

Converting it to honest-to-god C for the Raspberry Pi is even easier,

since the Pi has an actual operating system to do most of the work for

you in setting up the serial port and handling interrupts.

But you can get free samples of a nice chip that handles pH electrodes

and sends I2C or SPI data to the microcontroller. That would be a much

more robust system, and you could sample a lot faster, and chain a whole

bunch of them on the same SPI or I2C bus. Probably cheaper than $95 too.

Moreover, TI will give you code for the Launchpad for it.

Or you can just use an op-amp, or plug one of these directly into one of the

analog input pins on the Launchpad (but an amplifier is best).

There's a pH control system based on it here.

And you can get another Arduino for $15.25 at Hobby King.

Even though I have been suggesting more powerful, cheaper boards for serious

work, the Arduino is very fun to play with and I love all of mine.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[Simon Quellen Field]

The Arduino has been around since 2005. Linux has been around since 1991.

If seniority is something you think gives an advantage, you'd want to know that.

There are less than a million Arduinos in the world. There are hundreds of

millions of Linux boxes out there, including the one in your router, and all

of the ones that run Google and Amazon data centers.

If you needed some help with programming, there are tens of millions of Linux

programmers out there, and many of them would be happy to help.

If you are going to go to the trouble of learning a computer system at the level

required to do data collection, learning how to do it on Linux is a much more

transferable skill than learning on an Arduino. In your professional life you will

see far more Linux machines than Arduinos.

But both of them are easy to learn at the level you require for this project.

But if the project grows, the Linux board will handle it, and the software will

be there to support it.

In an Arduino based system, anything that needs substantial data analysis

would be sent to a more capable computer, one that could do floating point

and could handle large buffers. On the Raspberry Pi, you could do all that on

the same board, automatically.

All of these boards have parallel I/O, serial I/O, SPI, and I2C. Some have USB,

HDMI, RCA video, stereo sound, and Ethernet as well. It is simple to connect

any of them to sensors that use any of those I/O methods.

And yes, Linux supports R, unlike the other boards and chips we've discussed.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[Nathan McCorkle]

I agree with Simon on looking for I2C and SPI sensors,  but any of these platforms can handle these as well as parallel.  In Linux opening a serial port is via the fopen command,  which is the same command to open files as well (*everything is a file on Linux*).

If you want defined timing go with a microcontroller because real-time Linux is a few steps harder (and I'm not sure it actually gives filly determined timings) .

You probably don't need defined timing, especially with I2C and SPI  communications . If you can't tolerate microseconds (my guess on rasPi is 1-1000s) of random sampling timestamps, and have less than 8 sensors, I say go with the $20 propeller board.

[Nathan McCorkle]

And if you don't need defined timing,  Raspberrypi sounds pretty attractive to be honest.  You could even easily use it to try that eulerian feature detector I mentioned on the biocurious list, if you wanted to archive it as well record h.264 (high-quality, highly compressed video) via the raspberrypi's hardware encoder.

[Tristan Eversole]

On Sep 26, 2012, at 6:06 PM, Simon Quellen Field wrote:

> All of these boards have parallel I/O, serial I/O, SPI, and I2C. Some have USB,
> HDMI, RCA video, stereo sound, and Ethernet as well. It is simple to connect
> any of them to sensors that use any of those I/O methods.

Hot damn, that's good news. Thanks, Simon, that's very important information; if I can get these sensors to work with the Raspberry Pi without much trouble, then there'e no good reason not to use the Pi.

> And yes, Linux supports R, unlike the other boards and chips we've discussed.

Yeah, I doubt R would run too well on an 8-bit microcontroller. :)

--T.

[Dakota]

Thanks for the informative replies, I like that cheap OP amp pH meter.

[Tristan Eversole]

That would be great! However, it would probably be best for me to get my mitts on a Pi and some sensors first, and that will take some time and money. So don't write a tutorial on my account, at least not yet; however, there might be other people on this list, or other amateur scientists in general, who could really benefit from one.

--T.

[Marnia Johnston]

for those in the SF Bay Area Noisebridge (a hacker-space in San Francisco) is hosting Raspberry Pi this Sat. I've pasted the info below from Mich Altman (founder of Noisebridge).

Best, Marnia

Come to Noisebridge on Saturday, 29-September, and play with Raspberry Pi!

When:

      Workshop / Tech Demos: 11am

      Talk: 3pm

      Show & Tell / Prizes: 4pm

Where:

      Noisebridge, 2169 Mission St., 3rd floor (1.5 blocks from 16th St. Mission BART Station)

Cost:

      FREE

Rob Bishop from the RaspberryPi Foundation is touring popular hackspaces in the US throughout September 2012 with the aim of giving talks and workshops about the RaspberryPi to both hackspace members and also RaspberryPi users in the local commun! ity.

He'll be coming to visit Noisebridge on Saturday, 29-September-2012 at 11am - 5pm. The event is free, and open to all!

The Raspberry Pi Foundation is a charitable organisation founded with the aim of promoting the study of computer science and related topics, especially at school level. The Foundation is responsible for the design and sales of the popular RaspberryPi single-board computer. You can find out more about the Foundation and the RaspberryPi at the Raspberry Pi website.

The tour will be blogged/vlogged on the RaspberryPi website and we hope to attract RaspberryPi enthusiasts and hackers/makers from across the areas we will be visiting, allowing us to meet and support our community.

The Noisebridge event will be very informal and will consist of the following; 

Workshop / Tech Demos (11am -3pm)

Rob will be giving a demonstration of th! e latest developments with the Pi as well as bringing a number of Pi's for people to play with hands-on.

This will give people a great opportunity to get advice from the Foundation and get help kickstarting Pi-based projects.

There will be a limited number of Raspberry Pi's for sale at Noisebridge. But to ensure you're not disappointed by lack of availability, you can buy a Raspberry Pi in advance -- Rob says that there is now plenty of stock at MCM Electronics (http://www.mcmelectronics.com/content/en-US/raspberry-pi).

Talk (3pm)

RaspberryPi: Past, Present & Future - An introduction to the RaspberryPi, including an overview of its history and development, details on the technical specification and an outline of future developments. Fo! llowed by a Q&A session.

Show & Tell / Prizes (4pm)

A chance for people who have started/completed projects based on the Pi to bring them along to do a show and tell, with prizes being given out for the most awesome ones shown.


Noisebridge wiki page: https://noisebridge.net/wiki/RaspberryPiEvent


Mitch.

[Patrik D'haeseleer]

On Wednesday, September 26, 2012 4:26:32 PM UTC-7, John Griessen wrote:

Taking all you've said about your own experience level, this is sounding the most
results-getting direction yet.  Does the bioboard have cost levels you can fit to your want?

Actually, the BioBoard *is* an Arduino platform. And it is far from a well-developed product - it's a couple of people at Noisebridge hacking some sensors. Really cool work though - they were building their own pH and DO probes.

Given Tristan's lack of experience with Linux, OS's, programming, and electronics, I still think the fastest way to get him up and running would be Arduino. He'll have his hands plenty full setting up the aquarium, rigging the sensors, etc.

I'd say get to the point where you can collect all the data you need on an SD card, and switch some relays on an Arduino. Pretty much everything you've learned up to that point will translate over directly if you do decide to move to a RasPi platform.

[John Griessen]

On 09/26/2012 06:45 PM, Dakota wrote:
> Would this http://atlas-scientific.com/product_pages/kits/ph-kit.html
>
> Work with this?
> http://www.digikey.com/product-detail/en/MSP-EXP430G2/296-27570-ND/2331789?cur=USD

No code mentioned from them, but they do mention Microchip PIC c code.

But, like Simon says, "use a sensor chip" like LMP91200 since you are going to
digitize and input into a computer control system, not read an analog scale
and write it on paper.

[John Griessen]

On 09/27/2012 02:09 AM, Patrik D'haeseleer wrote:
> On Wednesday, September 26, 2012 4:26:32 PM UTC-7, John Griessen wrote:
>
> Taking all you've said about your own experience level, this is sounding the most
> results-getting direction yet. Does the bioboard have cost levels you can fit to your want?
>
>
> Actually, the BioBoard *is* an Arduino platform. And it is far from a well-developed product -

it's a couple of people at
> Noisebridge hacking some sensors. Really cool work though - they were building their own pH and DO probes.

What I was recommending is for him to find a project, with experts to help him.

>
> Given Tristan's lack of experience with Linux, OS's, programming, and electronics, I still think the fastest way to get him up and
> running would be Arduino.

Me too. Partly since he wants low cost, and they cover how to make low cost sensors. The computer board
can become low cost later with design effort. His wish that it be low cost to start, easy with
many project members to enlist, and runs R stat. software is a tall order -- not happening without effort.

Will it become a product ever? Don't know, could be -- either as low low cost based on $5 chips
on a single custom board, or a R-Pi plus add-on sensor boards...but I'm not seeing many
clamoring for its development. What else would you want to control besides pH and temperature?
How about light input for plants growing? Control could be by moving sun shades
or turning on grow lamps for small scale.

[Dakota]

I honestly get lost on TI's page for it
http://www.ti.com/product/lmp91200#samplebuy especially since I lack
experience in this particular area of more advanced electronic choices
and buying just IC's to setup.

It looks like the LMP91200 is just a little black integrated circuit
that I'd have to prototype a board for and solder up a bunch of other
parts, unless you get it in some standalone version like this
http://www.ti.com/tool/lmp91200eval which is $100

http://www.national.com/popr/LMP91200.html#Order

That's the buy page...still tons of options I'd have no idea which to chose.

It seemed like a neat project idea at first until I checked prices on
digital meters and found something like this

http://www.amazon.com/COM100-Waterproof-Temperature-Combo-Meter/dp/B0045LQFTK/ref=pd_sbs_lg_2

which would probably be just fine for use in adjusting buffers.

[Patrik D'haeseleer]

Speaking of TI microcontrollers - if you like the MSP430 Launchpad that Simon mentioned, you may also want to have a look at their new Stellaris LM4F120 LaunchPad Evaluation Kit, currently availabe at a promotional price of $4.99!

First units are shipping (eagerly awaiting mine any day now), and Hackaday just had a nice post on it with a little unboxing video.

Not that I would recommend this one in Tristan's case, but it's quite a steal if you like to live on the cutting edge of cheap prosumer microcontroller platforms.

[Tristan Eversole]

On Sep 27, 2012, at 6:46 AM, John Griessen wrote:

> On 09/27/2012 02:09 AM, Patrik D'haeseleer wrote:
>> On Wednesday, September 26, 2012 4:26:32 PM UTC-7, John Griessen wrote:
>>
>> Taking all you've said about your own experience level, this is sounding the most
>> results-getting direction yet. Does the bioboard have cost levels you can fit to your want?
>>
>>
>> Actually, the BioBoard *is* an Arduino platform. And it is far from a well-developed product -
>
> it's a couple of people at
>> Noisebridge hacking some sensors. Really cool work though - they were building their own pH and DO probes.
>
> What I was recommending is for him to find a project, with experts to help him.
>>
>> Given Tristan's lack of experience with Linux, OS's, programming, and electronics, I still think the fastest way to get him up and
>> running would be Arduino.
>
> Me too. Partly since he wants low cost, and they cover how to make low cost sensors. The computer board
> can become low cost later with design effort. His wish that it be low cost to start, easy with
> many project members to enlist, and runs R stat. software is a tall order -- not happening without effort.

Really, I can run R on my laptop, and this sort of information isn't likely to need any particularly sophisticated analysis. Certainly not on the controller side. It was a sort of semi-joke.

I'm not sure what you mean by "easy with many project members to enlist". I wanted to know whether other people would find a sort of general-purpose environmental board useful, definitely; if so, that would justify spending a lot of time on it. Otherwise-- well, I'm not necessarily counting on anybody caring about brachiopods. There's a reason the last (and possibly only) publication on keeping them in closed aquaria was written in 1973. I was somewhat concerned that I might be making more work for myself by opting for the (more powerful) Pi over the (simpler) Arduino.

> Will it become a product ever? Don't know, could be -- either as low low cost based on $5 chips
> on a single custom board, or a R-Pi plus add-on sensor boards...but I'm not seeing many
> clamoring for its development. What else would you want to control besides pH and temperature?
> How about light input for plants growing? Control could be by moving sun shades
> or turning on grow lamps for small scale.

I suspect that it depends heavily on what you want to do. In my case, water flow, temperature and dissolved oxygen are critical; although it's important to maintain a stable pH in a saltwater aquarium, you don't really need a microcontroller to do it. (That said, T. transversa could turn out to be bizarrely pH sensitive, or something-- I imagine more monitoring capability will always be better if you're trying to keep something that nobody knows how to keep.) However, if you're building a fermentation tank, say, pH starts to be an extremely important parameter to monitor, and the alcohol sensor I mentioned before probably couldn't hurt; similarly, a digester might want that methane sensor...

I like the idea of creating an open platform for doing this kind of monitoring and control.

--Tristan

[Tristan Eversole]

Ironically, that's the one I ordered previously, after that email you sent to the BioCurious list. :)

It'll find a use.

> --
> 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.

> To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/u5qMvz2duogJ.

[Simon Quellen Field]

The TI site is only letting you order two at a time.

If you want more, Digikey is selling them for $8.36 each, and they don't ask

weird questions about whether you are using them to build military radars.

:-)

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

--
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.

To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/u5qMvz2duogJ.

[Tristan Eversole]

The critical point, which I had not realized before Simon replied, is that my choice of microcontroller does not strongly constrain my choice of sensors. If it did, that would be a powerful argument in favor of the Arduino: I can type "arduino carbon dioxide sensor" into Google, and bang, there's one on eBay. (http://www.ebay.com/itm/MG811-MG-811-CO2-Carbon-Dioxide-Sensor-Module-for-Arduino-and-other-MCUs-/120804868793 ) However, if it's not that hard to connect the aforementioned carbon dioxide sensor (for example) to a Launchpad or a Raspberry Pi or a Propellor, then the choice of microcontroller is no longer any sort of huge limiting design decision-- particularly because the microcontroller is going to be quite cheap relative to the tank, the sensors, the refrigeration system, and the brachiopods themselves.

So I really owe you guys for pointing this out.

--T.

> --
> 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.

> To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/nfjV5XtR9XMJ.

[John Griessen]

On 09/27/2012 11:37 AM, Dakota wrote:
> seemed like a neat project idea at first until I checked prices on
> digital meters and found something like this
>
> http://www.amazon.com/COM100-Waterproof-Temperature-Combo-Meter/dp/B0045LQFTK/ref=pd_sbs_lg_2

Are you saying "oh anything will do"? Do you do a lookup of conductivity in the presence of a buffer
to get pH, then write it down?

Sounds full manual labor.

[John Griessen]

On 09/27/2012 02:13 PM, Tristan Eversole wrote:
>
> On Sep 27, 2012, at 6:46 AM, John Griessen wrote:
>

<snip>

> I'm not sure what you mean by "easy with many project members to enlist".

I meant as in "enlist help of bioboard developers."

>
What else would you want to
>> control besides pH and temperature?
>

> I suspect that it depends heavily on what you want to do.

<snip>

>
> I like the idea of creating an open platform for doing this kind of monitoring and control.

But, the rhetorical question has been posed and there is little interest.
Most on this list seem to want to shop for bits and pieces to do experiments with, not design
experimental bits of instrumentation hardware.

[Simon Quellen Field]

This may be of interest to you then:

"https://projects.drogon.net/raspberry-pi/wiringpi/"

It basically replicates the Arduino I/O functions on the Raspberry Pi, to make it

easy to port Arduino software to the Pi without having to think too much.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[Dakota]

Oh wow woops, didn't realize that was just a conductivity meter,
thought it was pH as it came up under a quick amazon search.

Something like this then,
http://www.amazon.com/Milwaukee-Instruments-PH55-Waterproof-Tester/dp/B002SGKE8W/ref=sr_1_2?s=lawn-garden&ie=UTF8&qid=1348779199&sr=1-2&keywords=pH+meter

Even cheaper, and does pH and temp. For $40 I'd probably take that
for making buffers, but in the case of someone wanting to constantly
monitor an environment, I can see how the micro controller / pH board
with constant data logging would be superior to taking daily
measurements with a handheld.

[Simon Quellen Field]

One of us could probably open that up and get to the signals to

connect it to a microcontroller.

But pH meters need to be calibrated fairly often, so it would still not be

a build-and-forget type of sensor.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[Nathan McCorkle]

Would it work to just connect the LCD/LED matrix output lines and
monitor them? you could find the update time by titrating a solution
and oversampling the display screen to find the refresh frequency.

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

[Simon Quellen Field]

LCD displays aren't simple on/off lines from the microprocessor.

It isn't as simple as looking at the segment line.

But the inputs to the microprocessor can be tapped into.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[Patrik D'haeseleer]

On Thursday, September 27, 2012 2:12:51 PM UTC-7, Simon Field wrote:

But pH meters need to be calibrated fairly often, so it would still not be

a build-and-forget type of sensor.

Not to mention biofilms and algae starting to grow *on* the sensor...

[Cathal Garvey]

Small nitpick; gas sensors tend to draw a lot of power, and it's
challenging to get even enough power to drive modest USB equipment from
a RPi without a separate power hub. That's one nice feature of Arduino;
a really wide input voltage band, and the potential to draw larger
ampages straight from the voltage regulator than would otherwise be
avaialble on the pins.

Having said that, I'm not sure even the arduino can supply enough power
for some of the more popular gas sensors like those used in alcohol
breathalysers. It may or may not apply to your CO2 project.

--
www.indiebiotech.com
twitter.com/onetruecathal
joindiaspora.com/u/cathalgarvey
PGP Public Key: http://bit.ly/CathalGKey

[Tristan Eversole]

Well, the brachiopod project doesn't need a CO2 sensor, obviously. I mentioned it mainly to illustrate that there seems to be a wide variety of Arduino-compatible sensors out there, often with Arduino sample code. According to the datasheet, the Sandbox Electronics CO2 sensor (sold in a not-eBay fashion here: http://sandboxelectronics.com/store/index.php?main_page=product_info&cPath=66&products_id=197 ) uses 5.5V from the Arduino to condition the sensor signal, and requires a separate power supply for the sensor's heating circuit, so presumably that's how they get around the power issue.

Microcontrollers are not the big issue for a brachiopod project. The more I learn about the subject, the harder it looks. Even the first step, getting a laminar flow through a closed aquarium, is something that only became feasible relatively recently. I need a laminar flow system, a refrigeration system, and a moderately sophisticated filtration system, and you can get more complex from there (by using an algal turf scrubber, for example). One of the major problems we've encountered in keeping filter feeders is just knowing what they'll eat, and giving it to them-- you have to keep a culture of the plankton itself and introduce it to the tank without removing it via the filter. Progress on this problem has been made via the development of synthetic plankton substitutes(!!!), which I guess obviates the culture issue. Adding a microcontroller sensor system just makes the actual conditions in the tank more visible, and would hopefully make it less mysterious; eventually it would help maintain stable conditions.

All of this stands in stark contrast to the methods described in McCammon's paper, which sound pretty trivial by comparison. McCammon's methods might work! However, just GETTING brachiopods is not going to be easy. It's not going to be like buying a tiger shark or something, but we can't keep 'em, so if I go after this, someone has to get some. I'm not going to try McCammon's methods straight out when the creatures themselves are so hard to come by (commercially) and everything else I see says I'm in for a major challenge.

Evidently some of the progress in this area has been made by people who want to culture sponges to study their antibiotic properties?

This is a long-term project because I don't have the money to do it. It's not out of the question, I think-- I don't have to be the Monterey Bay Aquarium-- but it'd definitely require much more than I have.

I found a forum thread in which someone is attempting to use the Raspberry Pi and the Atlas Scientific sensors to make a tank monitor, so that's probably a thing you can do. (http://www.thereefuge.com.au/board/threads/raspberry-pi-tank-monitor-project.3475/ )

--T.

[Simon Quellen Field]

People who want more power from their Raspberry Pi (and don't want

to include a powered hub in their project for some reason) have simply

been soldering a wire from the +5 volt input power jack to the +5 volt

USB output connector. This bypasses two poly fuses on the board, and

allows the USB device to draw all the power it needs from the Pi's power

supply directly.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[Cathal Garvey]

Oh, that's clever. :) Is there any specific reason this wasn't in the
board design though? Blows up some USB devices, perhaps?

>> http://www.ebay.com/itm/MG811-MG-811-CO2-Carbon-Dioxide-Sensor-Module-for-Arduino-and-other-MCUs-/120804868793) However, if it's not that hard to connect the aforementioned carbon

[Simon Quellen Field]

> Is there any specific reason this wasn't in the board design though?

> Blows up some USB devices, perhaps?

I really don't know. I might guess that the fuses are needed to get certified

for USB, but I have no evidence to support that guess.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

[John Griessen]

On 10/02/12 09:32, Cathal Garvey wrote:
> Is there any specific reason this wasn't in the
> board design though? Blows up some USB devices, perhaps?

Good guess. It's against the USB spec to draw more than so much.
The wire allows more, so meltdowns of supplying USB hosts or hubs could ensue.

[Patrik D'haeseleer]

Here's a great overview of the issues with the limited 5V power supply on the RPi, and ways to bypass it:

http://raspberrypi.stackexchange.com/questions/340/how-much-power-can-be-provided-through-usb#342

[Simon Quellen Field]

I've found that simply using a powered 4 port hub to power my RPi and plugging

its input into one of the RPi's outputs gives me three high power ports and one

port left over on the RPi.

Why people worry about adding a hub is not really clear to me. You'll need power

anyway, and the extra ports are really handy.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

--

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.

To view this discussion on the web visit https://groups.google.com/d/msg/diybio/-/aKCVrOfND20J.

[Cathal Garvey]

I did the same; forked out for a swanky-ish hub with over 2A of supply,
and wired the RPi to draw power from the hub, while using it for input.
I was concerned that this ouroboros arrangement would result in board
death but it seems to work well.

My reluctance initially was because I just wanted a machine that could
simultaneously power a keyboard and a mouse at the same time, and was
disappointed that it could not do so with the hardware I had. And,
because the USB hub was more expensive than the computer; a unique gripe
to this device I guess.

Having accepted the situation and bought the hub, I don't look back. The
Pi is still fantastic for network-embedded tasks; perfect for a
power-failure-resistant mail server, WebDAV host, or other experimental
platform for convenient "personal cloud" stuff.