My Arduino for 8$ or Sunduino

[blue_17]

Hi

This is my version Arduino
This version it is compatibility with ArduinoUno and shields



Download files on the bottom of the article

http://www.blue17.elektroda.eu/mikrokontrolery-avr/1424

Regards All

[Gallymimu]

Very cool!

How is the analog sampling precision?  I'd imagine the board is pretty noisy (electrically) without a ground plane.

Good job on this!

[solutions]

You don't need a ground plane to keep the noise down on the analog sampler.

[CocaCola]

Cool to see it done primarily with single sided layout...  Always fun to see a design come to reality and I remember when it was fun to etch your own boards...  I'm not a fan of Arduinos to start with and when you can get perfect knock offs for about $15, the idea of making my own really diminishes, but hats off to you for doing it...

[blue_17]

Thanks for comment but is my first arduino, I am very pleased with the device because is very small and useful with breadboard.

Regards all

[Gallymimu]

You don't need a ground plane to keep the noise down on the analog sampler.

That is completely untrue, especially if you want to get a decent bandwidth or bit depth out of it.  Single sided boards are always plagued with ground bounce associated with switching currents in the digital portions of the chips.  Also any current drawn that has a return path not through a ground plane will produce some inductive level shift.

Now if you just want 8 bits plus or minus a bit and are oversampling and just want 100Hz it's probably fine.  If you want 10 or 12 bits at 100KHz then it COULD be pretty poor.

Rather than just make assertions though here is some relevant reference material:

http://www.analog.com/static/imported-files/tutorials/MT-031.pdf
http://www.analog.com/library/analogdialogue/archives/43-09/EDch%2012%20pc%20issues.pdf
http://www.nxp.com/documents/application_note/AN10974.pdf

And then of course right from the horse's mouth for the ATMEGA
http://www.atmel.com/Images/doc4278.pdf

And Blue_17, I'm not trying to knock your design, just curious how it performs.

[Sideshow Bob]

You don't need a ground plane to keep the noise down on the analog sampler.

And at least not then we are talking about 10 bit, and low speed AD. I do not think the "real" Arduino use a 4 layer board either. I am not a Arduino person my self. But to me it looks like the OP have given the circuit board design some thought. The layout is nice and tidy as far as I can see
This has often been the topic on free lunch seminars. I have been on. Microchip have often claimed that the internal noise in a micro. Make it hard to go higher than 10 bit. So it is a question of cost also. We all want cheap micros. And the AD function is not that important. So they keep sticking to 10 bit. To keep the cost down.

[blue_17]

Arduino is only toy for fast prototyping not for professional devices, in my opinion.

If you want to high precision use external ADC

[CocaCola]

I do not think the "real" Arduino use a 4 layer board either.

Just your ordinary 2 layer on the Arduino, the official board files are open source (in Eagle format) so anyone can view or make changes to suit their needs on a custom build...

That is completely untrue, especially if you want to get a decent bandwidth or bit depth out of it.

The Arduino is simply a small general purpose AVR developer board, it's not suited or optimized to any particular application...  If you have specific and demanding requirements than don't do it on a general purpose developer board like the Arduino, you would build a proprietary circuit with a suitable layout for your application...

**EDIT blue_17 post crossed mine, but what he said is the same as above, the Arduinio is just a quickie general purpose AVR developer platform nothing more...

[metal]

It is hard to make oversampling work. A friend of mine uses LTC1286/98 ADC chip, and he told me on Thursday (it was just a question I asked him by coincidence, which ADC do you like to use for industrial applications) that they are very good and kinda cheap for what they can do. Myself I am planning to buy one and experiment with it to see how it performs. Thu I still think that ADCs from microchip are much cheaper, I will have to compare and see which is better.

Something else, as long as I am going to use an external ADC, isn't it a waste of time and money to use PIC or even AVR that has internal 10 bit ADC? IMO, yes. SM5964 is a nice option to start with, I worked with these chips for quite some time, and I like them.

Which external ADC you are using is not something to debate about, but, it irritates me to use an external ADC with a MCU that has an internal ADC. In this case I am talking about Arduino, it is just a waste of time being tied to it for all projects.

small interfacing example or this for C users. Note how PIC16F84 was used in the latter project. Sure there are other - better - examples

[solutions]

That is completely untrue, especially if you want to get a decent bandwidth or bit depth out of it.  Single sided boards are always plagued with ground bounce associated with switching currents in the digital portions of the chips.  Also any current drawn that has a return path not through a ground plane will produce some inductive level shift.

I'm not sure you know what you are talking about.

What does ground bounce have to do with an ADC that has a separate analog supply and ground? Do you even know what causes noise? Do you know what affects bandwidth (hint, starts with a "C"...something you get with a multilayer structure)?

Having had applications engineers working for me at a large analog semiconductor company, I can tell you that I've seen a complete MESS in a multilayer board, and at the same time, I've seen GHZ designs on a single layer. Though some are, many are not, gods.

If you know what you are doing in analog, you don't need to throw money at a problem. If you do not, then you throw everything at it, and cross your fingers. But you can still have a mess, albeit a more expensive one, if you don't know what you are doing.

[Ichan]

A good planned ground plane is essential on reducing (or eliminating) current loops, but for this board no way to do it unless adding one more layer - a ground jumper probably will, but it must be very tricky.

-ichan

[blue_17]

If you want programming PICs in Arduino mode is Pinguino

supported PIC18F2550, PIC18F4550 and PIC18F26J50

in  PIC18F4550 you have ADC   with 13 channel and 10-bit precision


Regards all

[solutions]

A good planned ground plane is essential on reducing (or eliminating) current loops,

-ichan

The ground plane will not necessarily reduce current loops, though most times it will through overkill. It merely shields, lowers current density, maybe reduces antenna loop cross section, and lowers inductance. You can completely eliminate current loops in a single layer, and should, even in a multilayer design.

[Gallymimu]

I'm not sure you know what you are talking about.

What does ground bounce have to do with an ADC that has a separate analog supply and ground? Do you even know what causes noise? Do you know what affects bandwidth (hint, starts with a "C"...something you get with a multilayer structure)?

Having had applications engineers working for me at a large analog semiconductor company, I can tell you that I've seen a complete MESS in a multilayer board, and at the same time, I've seen GHZ designs on a single layer. Though some are, many are not, gods.

If you know what you are doing in analog, you don't need to throw money at a problem. If you do not, then you throw everything at it, and cross your fingers. But you can still have a mess, albeit a more expensive one, if you don't know what you are doing.

C word... um.. cookies, no, condescending? No. Oh Capacitance!!!! I have heard of capacitance.  If you put a ground plane 3 mil from a 50 mil trace you can create 10s of picofarads of capacitance (depending on track length) and with high impedance lines yes it would limit bandwidth.  I guess you didn't understand my meaning (and I wasn't clear).  I meant usable bandwidth and therefore SNR which is seriously impacted by noise which relates to my other comments.

I may have no idea what I'm talking about.  I could be a complete buffoon.  But I bet everyone here believes Linear, Analog Devices, Microchip, National, and TI which is why I attached app notes from where my comments are derived.

Please share some credible tech notes describing good ADC/analog techniques for 1 layer boards rather than "claiming" my points are invalid.  Honestly I'd be delighted to learn good analog techniques for low layer count boards.  It would be appreciated.

And sorry for drifting off the thread topic!

[~5]

Cool project!

In your project, the USB port is only for power supply, isn't it?

As a single side Arduino-like board, this is also interesting:
http://vonkonow.com/wordpress/2012/10/nanino-the-diy-friendly-arduino/

[blue_17]

Thanks

Yes in my PCB USB port is only for power supply.

As a single side Arduino-like board, this is also interesting:
http://vonkonow.com/wordpress/2012/10/nanino-the-diy-friendly-arduino/

Yes but my board is in Eagle format

Regards

[Mozo1971]

It´s a very nice selfmade arduino board for the 28pin µC´s. Good job 

Do you plan a board for the bigger 40pin µC´s like Mega1284?

Because the 32k Flash with some libraries are to less so 128k much better
and the 40pin DIL µC is very easy for a selfmade board instead of SMD packages.

[blue_17]

I made big development board for AVR PIC or MSP so I don t to need small board but I like connecting to my development board others development board example ARM Discovery board or Texas Instruments LunchPad Board witch 8 and 32 bits microkontrollers


http://tnij.org/syp3


http://tnij.org/syp6

soon present details on the forum

Regards

[~5]

Yes but my board is in Eagle format

Fair point.

To be honest, I'm surprised they released the project in Rhino format instead Eagle.

[~5]

(...)
Yes in my PCB USB port is only for power supply.
(...)

I don't know if it is of any interest but I discovered this. The USB is implemented via firmware (no USB driver required). As far as I can understand, it also requires a change in the Adruino IDE configuration file but apart from that, it shouldn't require anything else.

It might be intersting add the functionality to your board, it requires only 2 diodes and a few resistors.

[Gallymimu]

I don't know if it is of any interest but I discovered this. The USB is implemented via firmware (no USB driver required). As far as I can understand, it also requires a change in the Adruino IDE configuration file but apart from that, it shouldn't require anything else.

It might be intersting add the functionality to your board, it requires only 2 diodes and a few resistors.

according to this site:
http://www.atmel.com/devices/atmega168.aspx

The ATMEGA168 used in this metaboard project doesn't have a USB hardware port.  I don't see how this could work without a USB transceiver.  Am I misunderstanding the project?

[Alex5532]

@Gallymimu
I suppose that the Metaboard use V-USB, you can fine more information at http://www.obdev.at/products/vusb/index.html

Alex

[~5]

@Gallymimu
I suppose that the Metaboard use V-USB, you can fine more information at http://www.obdev.at/products/vusb/index.html

Alex

I think it is exacly what it does.

There is another board (MHVBoard) that does something similar. My understanding is that it very similar to the Mataboard but it has some extra funcionality (e.g. +3.3V out)