DAQ questions

[Gordonious]

I am looking to do some measuring on my PC via probes connected to my aquarium (as well as my room and terrariums). While looking at different DAQ devices I have come up with a couple of questions you can see below. There is a great deal of things I would want to measure if I won the lotto, but currently I am just hoping to get my bank account into positive numbers before the end of the month. Once I purchase my DAQ I would like to measure temperature of the room, tank temperatures with two probes, and pH. With in the next year I would like to add additional probes to also measure salinity, ORP, conductivity, and dissolved oxygen. (In an ideal world I would also measure Calcium, Ammonia, Nitrate, flow, and more but that is obviously out of budget.)


What range of voltage measurements will I likely need?
Do I want to measure process current signals (4 to 20 mA)?
If I were to get DAQ that works for a serial port and then needed to make it run through USB via a converter would it be likely any programming I had done would need to be done over again?
Are most of the probes I am getting going to require Analog Input channels?
Would Digital I/O Channels be better?
Can you convert between Analog and Digital easily?

I am currently looking at a device that has 8 Analog inputs, and 6 digital I/O ports. I will later on upgrade to a device that is capable of more once I have had a chance to play around with an affordable DAQ.


BEFORE YOU POST please do not even mention the words: aquacontroller, neptunesystems, or anything of the sort. I am asking about DAQs and probes. If you want to ask me why PM me. That is not the subject of the thread.

Jon

 

[BeanAnimal]

There are dozens of devices out there to choose from.

You can use a DAQ based system and by the appropriate probes (expensive) and use their software to read and log.

You can also roll your own using any of a number of USB, PCI, SERIAL or ISA devices.

Stuff the comes to mind:
ActiveWire USB
Phidgets.com
ANy of hte ATMEGA or PIC micro products (basic stamps, demo boards etc).
Dallas 1-wire stuff

With these items you will be writing your own software to talk to the devices and log the data. Some require more skill than others. Temperature is easy, you can use Digital or Analog temp sensors. There are dozens of model and they are very cheap. Some will hook right to the parallel port, others to the USB, others to any of the DALLAS 1-wire bridges.

Your questions is rather open ended and that means that so are the answers.

You need to ADC to get the data into the computer. You will either be measuring small voltages or small currents. If you don't want to build your own OP-AMP circuits, you will need to purchase PROBE amplifiers (eBay is a good place to look).

What are your skillsets and what is your true budget?

 

[bromion]

I'll give my thoughts on a few of them:

USB<->Serial converters are seamless, so you would not have to re-program anything.

Nearly all the probes you will get will produce an analog signal, except maybe temperature. I'd go with more analog ports.

The point of the DAQ is to convert this analog signal into a digital form readable by the computer. This is done by sampling the analog signal 1000s of times per second (or more) and returning some number of bits which represents the input voltage at that time. Since the probes won't be producing fast-changing (high frequency) signals, you do not need a fast-sampling DAQ. Slower sampling DAQs are cheaper. The number of bits the DAQ uses for sampling determines how fine-grained the resultant digital signal will be. An 8-bit DAQ will allow for 256 levels of resolution (say, 0mV-256mV, with 1mV spacing, in a simple case). A 10-bit DAQ will have 1024 levels. I do not yet know what resolution is appropriate, but lower is cheaper.

When you get a DAQ it will come with software to program it, or you will use some standard library (Comedi on Linux comes to mind). You will not need to program at the low-level of the serial port, but rather give commands like read_data() and write_data() to/from the DAQ. Most likely it will be in the C language.

Hope this makes sense!

 

[bromion]

Also, as BeanAnimal suggests, you can use a dedicated microprocessor instead of a full computer and DAQ. For example, MicroARM, TinyARM, MiniPod(?), have built-in analog-digial conversion circuits. If you learn how to program these devices, you can make a compact, dedicated piece of electronics measure the probes rather than a PC with DAQ. This will require quite a bit more time to learn and get working, in my experience, but would be more robust for long-term use.

 

[Gordonious]

Thanks a lot, both of you. It’s hard to get my whole situation typed out there with out typing a novel and driving 99% of people to skip the thread as soon as they see the amount of reading they would have to do. As far as skills I am a student at the University of Delaware with several programming courses behind me, mostly in C, but has been a little while since I've messed with it. bromion it does give me comfort to hear they will mostly be coming with some sort of software. I was starting to get the idea most of you just bought parts of ebay and the PC wouldn't even recognize the device. If I am given commands like read and write I can work from there. I have been thinking I will set up a text documents to read the data into, one for each day and run all the raw data into them and back up the files. Then later on sweep the files with another app to round numbers and export data into excel documents. (all the programming end details will come much later, for now I need to get hardware I can work with)

As far as buget I am hoping to pull out about $20 a week from my tiny pay check. I am hoping I will also be able to get about $120 sometime later this week to get my started.

I did realize that a lot of the more expensive units were working with very fast streaming data that I would not need to worry about. Glad to hear I was right on that part.

I am taking a computer engineering course this semester and I plan on bugging the heck out of my prof. He works with DAQs a lot. I went him an e-mail asking about the sort of DAQ devices he is familiar with and I get this response:

"For misc work, I've used this inexpensive unit:
http://www.dataq.com/products/startkit/di194.htm

for research quality units, National Instruments has a robust hardware/software
combination. http://www.ni.com
"

I think it is likely I will start with an 8 input USB from the first site, and then once I've played around a bit and have a better idea what I am doing look st the second site maybe next year sometime. I hate to spend money on two devices. I usually like to get the best to begin with instead of losing money, slowly working my way up, but I really have no experience actually working with this stuff.

I am currently thinking to start I will get something like the <A HREF="http://www.dataq.com/products/startkit/di148.htm">DI-148U</A>. What do you think of this one?

If this one works will, what would be the next step to get pH and temp through this device?

BTW I am willing to work with a micro processor if it is an assured thing that I could get it to work. I do not want to run into the same dead end many people who have started on similar projects have come to buying parts off ebay that not a soul has experience with and have to learn everything about them and code from bottom up.

Hit me with questions, opinions, anything.

 

[BeanAnimal]

In your situation I would suggest maybe looking at doing some DIY stuff.

Here is a great DIY project that will log more temp sensors than you will ever want!

http://www.anotherurl.com/therm/logger.htm

I built the basic "arbiter" but made some modifications to it and wrote my serial port and CRC routines to interface it into my homebrew controller software. You can however use it as is. Very cheap project! You can breadboard it and do point to point wiring if you want.

Also those same DALLAS 1-Wire digital temp sensors can be plugged into the DALLAS 1-Wire bridges (serial or USB). You know how to code, so writing a small VB, C, C++ or similar app would be easy.

Also check out http://www.riccibitti.com/pc_therm.htm this guy built a pretty nifty circuit to use with the analog temp sensors and hook them to the PCs serial port. Very straight forward and all you need to do is bang the serial port for the data when you want it.

You could write a small VB or C app to do that in about 20 mins. Dump the data into a texst file or access database.

The phidgets stuff will do ADC and you can get an electrode amp off of ebay and plug it into that. Your same simple C or VB code could log that data as well.

That is again if you want to go the DIY route.

If your really frisky, you could build some PC or ATMEL based units with an ADC chip and a serial chip... let them do all the work (will reuire a little bit of C or Basic code) and dump that to the PC over serial (via a USB to serial bridge if you want) Each unit would cost maybe $12

The reason I am suggesting DIY is because you seem to want to learn about data the process and are in a computer related area. Using an off the shelf DAQ and modules is not going to teach you anything other than how to plug stuff into an expensive unit. If you still want to go that route... look at the verneir brand products http://www.vernier.com/

Like I said there are a hundred ways to skin this cat. Maybe get the small phidgets device and play with it and some code.

 

[Gordonious]

If I am spending closer to $100 on any probe I would like it to work with any system and don't like the idea of only have one, or just a few, places to buy my probes. I was originally excited about vernier, but they are not the way I am going to go.

I definitely would like to learn about just about every little bit of it and seriously if you could tell me three website and 5 books to read and I could pause time and go through and study it all I would. Problem is I just spent a small fortune on text books for this semester so the amount of time I have for studying and reading is limited. (This semester: Organic Chemistry, Micro Biology, General Ecology, Intro to Digital systems, entomology lab)

I think it is time to draw some pictures. BRB

 

[bromion]

The DI-148 looks good to me!

 

[Gordonious]

Could I use the <A HREF="http://www.dataq.com/products/startkit/di148.htm">DI-148U</A> and then use the same probes some of the same code with a micro processor later or to go between the two would I be starting just about from scratch?

I have been looking for a way to make temp probes like this for years: http://www.anotherurl.com/therm/logger.htm BeanAnimal
your awesome, thanks much. I will probably play around with these soon.

What would I need to go between a pH probe and either the DAQ or the micro processor?(probe amplifier?) Would I need a different one for each other sensor? Where would you suggest getting the pH probe from?
I am thinking that if I can get the probes for less then $10 then I will get what ever cheap probe I can find from ebay or whatever, but if I have to spend $60-70 for one I would rather spend a little bit more and get one from a company with some kind of support and warranty for a least a week or so.

Ok no laughing this isn't art work it is just sketches for talk about some ideas.

<IMG SRC="http://www.gordonious.com/RC/plan1.bmp">
<IMG SRC="http://www.gordonious.com/RC/plan2.bmp">

 

[Gordonious]

Sorry I had a line between the two plans, but edited them several times and split the image trying to get it on RCs gallery. Ended up still being to large and I uploaded it to my website.

 

[Gordonious]

I wouldn't ever need a thermocoupler would I? The ability to use one of those as well as faster speeds seems to be the biggest different between the DI-148U and others. Could I use the DS18S20 1-Wire Digital Thermometer with the DI-148U? Would it be to picky to ask for something with a better Accuracy (±°C) then .5 or do you think that is fine? I was hoping for a little better for my reef tank, but at that price I suppose I could settle till more $ comes around.

 

[BeanAnimal]

the 1-wire stuff has to be used with a 1-wire interface, the output is digital (not a simple binary number, but a string of binary data).

I don't see why you would need anything with more precision than .5 degrees....

To get yourself started, why not build either the arbiter project or the Ricci Bitti Therm project. Very cheap and will give you an insight into Micro code, PC interfacing via the serial port and similar concepts.

 

[pista01]

<a href=showthread.php?s=&postid=8117462#post8117462 target=_blank>Originally posted</a> by Gordonious
Would it be to picky to ask for something with a better Accuracy (±°C) then .5 or do you think that is fine? I was hoping for a little better for my reef tank, but at that price I suppose I could settle till more $ comes around.

Personally, I think the DS18S20 is more than accurate enough. If you really want a PC to log data and you don't want to spend a lot of $$$, consider going 100% 1-wire. A USB interface can be had for $26 here:
http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=1503

You can get some CAT5 cable and solder a DS18S20 to the end of it. Of course you'd have to seal it in something to make it water tight. You could also use the HVAC Monitor from the same site above (ask them to make one that reads 0-5v instead of 28v) with a pH transmitter and probe. Some Google searches would have to be made in order to understand how to convert 4-20mA into voltage for the DS2450 to read. The 1-wire java API comes with enough samples that you can modify to write the sensor values to a text file every so often without writing anything from scratch. Load the file into a spreadsheet program and chart it from there. You would have to do some coding to calibrate th pH probe.
eBay is a good source for some really nice pH transmitters. At the very least, you can be logging temperature on your PC for less than $40, maybe $50 if you have to buy a short CAT5 cable.

 

[Gordonious]

Woah, that 1-Wire USB Adaptor actually looks really expensive. How much is actually in that thing? There has to be an easy way to build one of those. I would think all it would have would be a couple of resistors anyone know?

Jon

 

[BeanAnimal]

You can get a Dallas 1-wire USB adaptor for about 20-35 bucks... the serial version is a little cheaper. If you really want you can build one with the the dallas components (not worth the trouble really). I am not sure what you mean by expensive... but $30 is cheap for what it does. You could put thousands into a DAQ package like a LABJACK or similar.

Bean

 

[pista01]

There are schematics and parts lists on the HobbyBoard web site if you really want to build it yourself.

 

[Gordonious]

Sorry I got a bit long winded here. I am a little eager to get started with this. I finally with have a tiny bit of loan money I can use in a couple of hours.

Summary of questions below:
<DD>Would these three options accomplish the same thing? What are the benefits from one to another.(keeping in mind I want to expand to more in the future, but I am just starting and have a limited budget.)
<DD>A. <A HREF="http://www.hobby-boards.com/catalog/advanced_search_result.php?search_in_description=1&keywords=HVAC+Monitor">HVAC Monitor</A> and the <A HREF="http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=1503">USB interface</A> with DS1621 and a ph probe.
<DD>B. <A HREF="http://www.dataq.com/products/startkit/di148.htm"> DI-148U Starter Kit Low-Cost USB Data Acquisition System</A>
<DD>C. <A HREF="http://sine.ni.com/nifn/cds/view/comp/p/sn/n24:USB/pg/2/lang/en/nid/1036/ap/daq/sd/9e899a63b22b2010VgnVCM100000940aa482RCRD">USB-6501</A>

pista01 do you mean that I could just DS18S20 and that USB interface?(I have enough Cat5 to tie down a Dinosaur). This would give me similar results or better then using the <A HREF="http://www.riccibitti.com/pc_therm.htm"> DS1621 pc thermometer</A> or the <A HREF="http://www.anotherurl.com/therm/logger.htm"> Temperature Logger</A>. With using the USB interface what is the limitation as far as the number of sensors?

The HVAC Monitor Kit does not look difficult to assemble could I possibly do that and just us a different part at some point to get 0-5v or would you say for surety that I should have them assemble it. Have you done this exact set up before pista01?

Would the combination of the HVAC Monitor and the USB interface give me the same abilities as the <A HREF="http://www.dataq.com/products/startkit/di148.htm"> DI-148U Starter Kit Low-Cost USB Data Acquisition System</A>? The DATAQ starter kit has software with it.

If I were to go with the site my professor said he would use if he wanted "for research quality units". What would you consider the ups and downs of going USB to PCI? I mean the USB could be used with a lap top, though I don't imagine I would be doing that. The USB could also be easier to make other hardware to work with it correct? I've never seen anyone solder anything to a PCI board to use it with anything else, but USB devices seem to be added onto and played around with all the time. I think the <A HREF="http://sine.ni.com/nifn/cds/view/comp/p/sn/n24:USB/pg/2/lang/en/nid/1036/ap/daq/sd/9e899a63b22b2010VgnVCM100000940aa482RCRD">USB-6501</A> would be the only thing near my range.

I have heard plenty of people say to buy a pH probe off ebay, but the thing that scares me is most of the auctions I have seen for them say they are used, come with nothing, and they are not even show how or if it they work. In the past I was also not sure what connection I would need on the end of it.

Speaking of ebay do you think <A HREF="http://cgi.ebay.com/Dallas-DS18S20-DS1820-Temperature-Sensor_W0QQitemZ7528380693">this</A> would be a good way to order the temp probes?

What exactly does it take to “You would have to do some coding to calibrate the pH probe.” Would this just be placing the probe in a calibration fluid to find out how far off the single is coming into the PC and then just telling the computer to take “input+ or â€"œ difference”? Or is this way over simplifying what I would need to do?

Regardless can you suggest a place where I could purchase most of the parts for the <A HREF="http://www.anotherurl.com/therm/schematic.htm">DS1820 Arbiter</A> with out searching through Google for deals for a month? I would like to try to build it, but I would enjoy building it not looking for the parts.


Thanks agian for all your help,
Jon

 

[pista01]

The DS18S20 can be soldered directly to the end of a CAT5 cable and the other end plugged into the USB (or serial) interface. You can use Java or C to read the temperature. I think .Net is an option too, but I stay away from that stuff.

I do have the HVAC kit working for pH, but I'm using a pH transmitter with a 12volt source and a 250 ohm resistor. The HVAC then reads the voltage of the 4-20mA loop. You would need to calculate the pH based on the voltage of two known pH points, such as 7 and 10. If you know the voltage at 7 and 10, then you can calculate the pH at any voltage in between using a slope formula.

The 1-wire stuff does NOT equal a DAQ. I'm not a DAQ expert, but I do know that a DAQ is much faster than reading/writing 1-wire devices. Kind of like comparing a bicycle to a car. Both can get you there, but a car is a lot faster and more expensive.
I chose 1-wire because I didn't need real-time, the software was easy to use, it's cheap, I wanted a PC based platform, and I could write the software to run on either Windows or Linux.

I would go with whatever you are the most comfortable with, or willing to learn.

Go here to learn more about 1-wire -> http://www.maxim-ic.com/1-Wire.cfm

 

[Gordonious]

what are the speeds of the DAQ vs the 1-wire? The DS18S20 has Meas. Res. (Bits) 9 to 12 and the DI148U has a A/D resolution of 10bit? Which numbers would I be looking at to know they are different speeds. I've heard it a million times, but have also seen that this DAQ is no where near the speed of the faster DAQ and.... I dunno.

 

[bromion]

If you are talking about sampling speed, that will be measured in kHz or MHz. For an aquarium application, sampling speed is not really important, as the system changes much more slowly than even the cheapest DAQ will sample. The number of bits of resolution determines how fine a measurement you can get at any particular time (during any one sample). 8-bit will probably be perfectly fine for the aquarium monitoring.