[Jacques] sent us this little project he has been working on. It is an amazingly cheap USB 2 chanel scope. The total cost of the project was around 5 Euros. It is based off of an Atmel Tiny45, has 2 analog inputs, and can supply 5 Volts to a breadboard. He has listed the bill of materials as well as downloads for the source code, plans, and display software.
thats epicly awesome.
im going to have a look at the source and see if i can make a pic version :P
Posted at 3:56 pm on Oct 2nd, 2008 by charlie webb
Now if someone can go and build something like this for pennies on the dollar of what a commercially manufactured unit costs, why can we not get the commercial scopes for cheap too? A guy could sure use a dirt cheap oscope\spectrum analyzer with a nice big color lcd.
Posted at 7:19 pm on Oct 2nd, 2008 by mike
@mike – because, as td-er pointed out, the sample rate of the attiny is really slow. Even with some interpolation, you still cant push many micros beyond a gigahertz sample rate. they are usefull for slow speed circuits, and basic sensor checking chores, but not much else.
dont expect to poke around a computer logic board, or even 74ls logic with this thing.
this all brings to mind the question “what qualifies something to be called an o-scope?”
analog to digital converters have been around since the first logic chips, yet were never called o-scopes. plug one into a display and all of a sudden its an o-scope?! no. an O-scope has a very specific set of minimum hardware performance specs and user controls.
you wouldnt call an atari2600 with paddles attached an o-scope, but it absolutely no different than what we have here.. except someone would need to code an `o-scope cartridge`
“ohh sure, the sample rate is only 30 hertz.. but its an o-scope damnit because I say it is, and see the waveform on the screen?!?!”
rediculous
So, I pose the question: “in your mind, why is this an o-scope, and not just another a-d converter attached to something that can display the data in a waveform like fashion?”
Posted at 8:47 pm on Oct 2nd, 2008 by MRE
not to mention, this sort of ambiguity confuses the hell out of newbs..
“why should I pay $500 for a scope the size as my desktop computer, when I could build one for $5?!”
a.la mike..
(no offense there mike)
for a seasond pro, the answer is obvious.. for the naive initiate, its confusing to label something like this ‘an o-scope’
Posted at 8:54 pm on Oct 2nd, 2008 by MRE
I personally would call it a data acquisition card rather than a scope (you know what with the low bandwidth, no trigger, no display, decent resolution, etc) but either way its a useful little tool. Its not new, but its certainly useful if you are looking at up to moderately high speed events (up into low audio frequencies are probably possible, but I would try it over about a KHz due to sample rates probably about 5ksps for a single channel) and don’t want to use your sound card for whatever reason (ie, you need to look at dc signals). I would be a little wary of the form factor sticking out of a breadboard like that (not using a right angle header would solve that) but since most people don’t need to scope mhz (capital m) signals it would work well.
This is interesting for education purposes. See the measure project at olpc :
http://wiki.laptop.org/go/Measure/Projects
I added a page on our olpc-france wiki.
http://olpc-france.org/wiki/index.php?title=Explorations#M.C3.A9trologie
Feel free to comment and update!
Samy
Posted at 12:30 am on Oct 3rd, 2008 by samy
Yep i totally agree with you…
Not everyone has the luck or money to have a scope at his workbench (like me)… And not everyone needs high sample rates for small projects… So this nice little thing made by [jaques] is just perfect… It should be obvious for everyone that you can’t get the performance of > 500€ products ;)
An another interesting point in this project is that it connects as a HID USB device – when i’ve first seen the project on the main page, i thought: Uhm, it reads ADC, sends the values to UART und then theres this little USB UART chip like in many other projects. But especially the fact that USB is actually done in the Tiny makes this nice project very interesting. To also put up some critical statemens, the software needs some work to be good…
Posted at 2:13 am on Oct 3rd, 2008 by stealthmonkey
mre: I don’t think I’m following your criteria for what makes an oscilloscope. I’ve got scopes that are likely older than you are that use tubes and are little more than frequency generators and amplifiers attached to a CRT. That is **the** minimum requirement for a scope. Your other criteria are for more professional and modern models but that does not necessarily exclude this idea as a scope.
Posted at 4:43 am on Oct 3rd, 2008 by eric
This is probably nice for analog experimentation. It would be even better if it had a simple function generator. The theory I learned in my EE analog courses didn’t click until I saw them in action in the associated labs with RLC networks and op-amps with a function generator and scope.
That being said, a “real” o-scope is more than just an ADC and sampler. It needs some sort of ranging and triggering otherwise you can’t truly visualize a periodic waveform. As a previous poster said, it’s just a data acquisition device.
Posted at 6:54 am on Oct 3rd, 2008 by pokey
charlie: Have a look at the Microchip PicKit2 – it already has the hardware required to make this, and full schematics and source code are available on the net, even for the windows app. As an added bonus, it can even program pics, but if you don’t need that functionality you can probably simplify the circuit diagram to <10 components.
Adding simple ranging to this should be pretty easy – just regulate the Vref+ and Vref- pins to “bound” the signal you want to measure.
As far as sample rate goes, I think PICs can sample at full resolution to near 100ksps, and triggering can be simulated in software on the PIC. (or even in software on the PC, if you can stream the sample data to the PC fast enough)
This is good enough for most people, since most digital circuits can be slowed down by a factor of say 100 for easy debugging with an o-scope like this. For high speed analogue circuits you might need something better though.
Posted at 7:37 am on Oct 3rd, 2008 by O Mattos
Hello, i am the designer, i agree with everybody that it is not a real scope,too slow (but 10bits!) i should have call this a VuMeter with graphic functions, or realtime data logger,
but please, look at the simplicity !
This was also to demonstrate how today in electronic project the parts count has been replaced by the “lines of code” count (quite big in this case).
ps: pic vs AVR ? hot news:microchip offer to buy Atmel !
many thanks to Hack a Day for publishing
@ all the haters
I am sorry to burst your snobby engineering bubble(s), but this IS a scope and it is useful.
Have you ever tried to diagnose a modern car? (some with up to 30 separate simple computers).
Just being able to log injector pulse-widths or check the crank signal shouldn’t cost $3000, but it does.
Think outside the box people. I know some of you at least must be interested in cars.
Posted at 9:04 am on Oct 3rd, 2008 by nubie
@nubie: Agree that this is useful for some applications, especially automotive ones. Just being able to see O2 sensor cycling is handy, and that only happens a few times per second.
That said, there’s no need to pay $3000 for a ’scope. There are tons of old analog oscilloscopes out there for under $100.
Posted at 9:32 am on Oct 3rd, 2008 by Orv
Interesting… I just bought something a little more sophistaced (with LCD): http://www.dpcav.com/xcart/product.php?productid=16330&cat=0&page=1&featured for US$ 48
I don’t want this to look like advertising… please remove if not appropiate.
Posted at 11:09 am on Oct 3rd, 2008 by JD
@ orv, you are right, but I was looking for something more like a digital storage oscilloscope, many times you are on your own so you need to save the data to look over while you concentrate on driving the car.
@jd, that looks pretty cool.
Posted at 5:34 pm on Oct 3rd, 2008 by nubie
It graphs voltage differences over time – It’s an o-scope. Whether it’s good enough for what you need is another story, but that doesn’t change the fact that it’s an o-scope…
I really want to build this…
Posted at 6:01 pm on Oct 4th, 2008 by Bryan
how do i program the Atmel Tiny45 mircocontroller?
do i need a special programming jig for this?
where should i buy it?
Posted at 8:52 am on Aug 30th, 2009 by vladimir T
@vladimir T
http://www.adafruit.com/index.php?main_page=product_info&cPath=16&products_id=46
Posted at 2:42 pm on Sep 10th, 2009 by prog
Hack a Day serves up fresh hacks each day, every day from around the web and a special How-To hack each week.
Powered by WordPress
According to the datasheet (page 126) of the ATTINY45, the max sample-speed is “up to 15 ksps at maximum resolution”
Datasheet: http://www.atmel.com/dyn/resources/prod_documents/doc2586.pdf
Still a very nice project to use in all kinds of experiments.
Posted at 3:55 pm on Oct 2nd, 2008 by TD-er