jaluino
Sebastien Lelong
@development board: What about this one then:
http://jallib.blogspot.com/2009/06/pic-18f4k50-usb-interface-board-part-1.html.
You've responded to that post already with some feedback I see :-)
What I think that would be useful is to make this development board
pincompatible with the arduino, so you can use their shields too. It
would help such board a whole lot further!
Make the board compatible with already existing Arduino shield, that's an idea !
But I wonder how possible it is: pin count wouldn't be the same, and wouldn't have the same functionality. This would make the board wiring complex maybe. More, not all arduino boards have the same number of pins. Maybe having a "converter" shield, from jaluino to *duino would do the trick :)
Cheers,
Seb
Sébastien Lelong
http://www.sirloon.net
http://sirbot.org
phalox
of specific pins (like there are 3-6 PWM pins on an atmega168), there
are only 2 on a 16F877.
I think this is the main problem. Next to that the routing will be
difficult. which means there will be some dual layer work :)
It was just an idea I was playing with.
I think too that such a board has to be kept really cheap to be
successful. But also an easy programming language ( could be a problem
too with JAL)
On 13 jul, 18:40, Sebastien Lelong <sebastien.lel...@gmail.com> wrote:
> > @development board: What about this one then:
>
> > .
Sebastien Lelong
phalox
There's some error, here it is:
/////////////////////
We're sorry, but we were unable to complete your request at this
time.
////////////////////
And here is my reply:
With all this quoting, it isn't as easy as I would want to, but
well... Here's my response :-D
I agree with you, Single sided is the best for a first version. When
prints will be made at a professional service, they could be double
layered.
About the price, I can get a Freeduino for just €16 (http://
www.samenkopen.net/action_product/795718/462690 - dutch website)
The next version is a lot more expensive, but also SMD. I won't go for
that just yet :)
The problem with JAL is for starters.. They have to know just too much
settings and stuff. An extra library to set those basic settings would
be usefull. Especially since the basic package of hardware is the
same.
The arduino language is really easy because some of the more
'difficult' settings are already set for you. Te knife cut's both
sides -> easy for starters, limiting for the more advanced people.
But I think this can be fixed by making a custom library for this
board.
Joep Suijs
I'll be the devil's advocate on this one ;)
>
> The arduino language is really easy because some of the more
> 'difficult' settings are already set for you. Te knife cut's both
> sides -> easy for starters, limiting for the more advanced people.
>
Like the starter pack.
>
> But I think this can be fixed by making a custom library for this
> board.
>
Like jallib!
>> Pin count doesn't really matter in my oppinion.
It does. I have the arduino with an atmega168 and ran out of pins
before my project was completed. I'd go for a 40 pin chip or larger
(smd gives you more pins at the small form factor, which is nice. But
it is less easy to build and repair).
>> It's more the number
>> of specific pins (like there are 3-6 PWM pins on an atmega168), there
>> are only 2 on a 16F877.
Now you get close to the most anoyning thing of 'doning the best you
can for a use that does not know what he is doing'. In arduino, all IO
pins seem to be equal. So on a 20 pin board, you can configure 20 pwm
signals. But when you use them, you might notice that they don't all
behave like expected. Depending on the chip, some pin's support pwm
and work fine. But others don't support pwm and will be set either
high or low (depending of the pwm value).
And similar, you might need two pwm's, say for an h-bridge. If you
read the datasheet, you will choose two related one's, so you only
need one timer and limited setup code. But with arduino, you try (!!)
until you have two pins working like you want to. And probably make
less efficient use of you hardware (two timers etc).
Similar issue's are there for use digital ports in the way that you
are not encouraged to put related pins on the same port etc.
This kind of issue's made me switch halfway my arduino project:
dismantle the arduino framework, only keeping the required libraries
as includes for c++ and work form there. Much better! (and you regain
some options from C++ that are lost with arduino).
Or to make a long story short: IMO it is a bad idea to shield the
hardware features from the user. It's like designing a car for drivers
that do not know it is common to use the left front seat. It creates a
lot of issues to be handled and a lot of work while it is not
unreasonable to expect a driver to know which seat to take.
>> > I think too that such a board has to be kept really cheap to be
>> > successful. But also an easy programming language ( could be a problem
>> > too with JAL)
I realy wonder why you state this. What makes Arduino more easy to use
then Jal?
Are this the super-include and function (names) provided??
Joep
Sebastien Lelong
It does. I have the arduino with an atmega168 and ran out of pins
before my project was completed. I'd go for a 40 pin chip or larger
(smd gives you more pins at the small form factor, which is nice. But
it is less easy to build and repair).
This comforts me that a huge 18F4550 would be a good choice
Or to make a long story short: IMO it is a bad idea to shield the
hardware features from the user. It's like designing a car for drivers
that do not know it is common to use the left front seat. It creates a
lot of issues to be handled and a lot of work while it is not
unreasonable to expect a driver to know which seat to take.
By no mean I want to hide anything, any complexity to users. I just want to build a board everyone could build, and could experiment with it. A common hardware. This is great tool IMO for beginners (and I'm pretty sure this is why Arduino also rocks: people can just plug it, program it, and observe the results -- and we can do this without hiding anything --, they just have to focus on the code, without having to deal with hardware troubles). And this is also a nice tool for advanced users. A common hardware, a common base to build over. (sounds like a slogan :))
Seb
phalox
D)
At the Arduino center, you can read about 3 guides, and you know how
to control a driving robot. Steering left and right with pwm, using
LED's, checking switches.
IMO this IS about the same with the _bert libraries, but not with
JALlib.
I think levels in advancement would be great!
* For starters you have all those parameters set for you (which off
coarse limit you, but I don't think that's the problem for a starter)
* Medium levels would have to set their own pragma's, or do the setup
of (e.g. PWM with all the timers)
* Last we have the standard level, which just includes the standard
libraries
On 14 jul, 08:12, Joep Suijs <jsu...@gmail.com> wrote:
> Hi Guys,
>
> I'll be the devil's advocate on this one ;)
>
>
> It does. I have the arduino with an atmega168 and ran out of pins
> before my project was completed. I'd go for a 40 pin chip or larger
> (smd gives you more pins at the small form factor, which is nice. But
> it is less easy to build and repair).
What I meant by this, was that you can have more pins than a arduino,
but still be pin compatible. (just split up the pin headers, and put
them somewhere you can access them while having a shield attached.
---
This knife cuts both sides too :-)
For starters it might be easy (but still confusing) that they aren't
entirely dependent of the datasheet setup. But hell, use the pins what
they're made for!!
Again, what I meant with this: if a shield uses a hardware
communications port to communicate with a shield attached, and we want
it to be compatible with JALuino, it's probably the best to have this
done at "the same" hardware port.
---
> >> > I think too that such a board has to be kept really cheap to be
> >> > successful. But also an easy programming language ( could be a problem
> >> > too with JAL)
>
> I realy wonder why you state this. What makes Arduino more easy to use
> then Jal?
> Are this the super-include and function (names) provided??
As stated above, you don't have to know all the harder hardware
settings, to get started at first. One with just some programming
experience in HTML could perfectly build, and program a circuit that
blinks the led in all kinds of patterns. That's the power of
arduino... But next to that this is also the group that will abandon
this piece of hardware sooner than more experienced will!
---
I'm sorry if this all sounds like b*llsh*t, it's just my narrow vision
at this case :-)
Toon
funlw65
will start using it when I will buy a 18f2550 or 18f4550). And you can
have up to 16 PWM outputs if you want.
Jaluino it will be not Jaluino if don't provide the same programing
language (as Pinguino is doing!!).
But as a new project (which can use same shields as for Arduino - or
maybe only the shield principle) will be interested. I think is easy
for you to build another layer over jallib similar to bert's starter
pack for beginners.
In fact, IMO is the next step for jallib and another good reason for
keeping this project alive from a user view: his own board. A reason
good enough also for jal compiler himself.
Another step is a math library (is a must in windmill controllers,
loggers, weather stations) - don't say is not an efficient way to use
a microcontroller, PICAXE have this requirement and is intensively
used. Believe me or not, IMO is the main reason of loosing users... Of
course, this is for JAL compiler developer.
Joep Suijs
>
> I just saw Pinguino.
Looks good indeed! Couldn't we 'reuse' there hardware?
> Another step is a math library (is a must in windmill controllers,
> loggers, weather stations) - don't say is not an efficient way to use
> a microcontroller, PICAXE have this requirement and is intensively
> used. Believe me or not, IMO is the main reason of loosing users... Of
> course, this is for JAL compiler developer.
What is it you're missing?
And needs it to be part of the compiler or could a library do the job?
(specs, specs!)
Joep
Sebastien Lelong
> I just saw Pinguino.
Looks good indeed! Couldn't we 'reuse' there hardware?
This is a breadboard version, I'd like to have shield version too :)
But that's a nice board, very small. No MAX232 chip, everything seems to go through USB, like bootloader. USB Guru Albert has made a lib for bootloader over USB. We might be able to reproduce this in jal world.
phalox
The 'programmer' should be some DOS program that does the serial
communication.
But I do like the (2/4)550's capability to do usb communication
without any extra hardware! This makes the FT232 superfluous! No more
SMD for this board :-)
I think we should make somekind of an index of the arduino shields...
then we know what pins we really have to take into account
PS: their programming interface kinda st*nks :-D arduino rip-off!
funlw65
Here ( http://sites.google.com/site/funlw65/electronics/jaluino-pinguino-28-pins-starting-bo
) is a starting Jaluino / Pinguino board (eagle files right now).
Sebastien Lelong
Looks nice, can't wait for the 18F4550 version :)
I would too prefer the PicKit2 connector, looks like more widely used
in in PIC's world (right ?)
About the schematics, I've seen there are currently only connectors.
This means we can still do what we want right ? :)
So, back on this post,
http://groups.google.com/group/jallib/browse_thread/thread/a1649a8b27a771ec/f09d680ac611d75c?hl=en
there's a "wishlist". Here I recall some bullet points, with more details...
- we should have one USB connector, connector to PIC's USB module.
This is to experiment USB.
- we should also have a connector for serial comms. Here I can see 3 options:
1. using a DB9 connector, and MAX232 chip. Big.
2. using a FTDI converter. Small. Expensive.
3. as Albert Faber suggested in the post, using a 18F14K50 to
build this converter. Cheaper than FTDI. (as far as I'm concerned, I
can't easily find this chip, except with high shipping rate, making it
almost as expensive as a FTDI chip).
4. No other connector. We use the first, and use a USB bootloader.
Guru Albert already developed one with jal (but bigger than tiny
bootloader, see later). Serial comms. are done through this USB
connector. I'd personally prefer to have a connector wired to TX/RX,
in addition to USB connector.
Maybe we could provide the three/four options... There can be 2
USB connectors, it would be nice if both would allow to power the
board through them
- I²C: alsways a must. An "I²C ready" board allows many experiments
and is opened to many 3rdparty chips.
- few LEDs for status (RX, TX, I²C, USB ?)
- power sections: 5V, 9V (original power), ideally 3.3V
- bootloader activated via serial (no need to manually reset the
board). I like Tiny bootloader, as many people here. Very small,
reliable, widely used, many clients (win, linux), never had any
problem with it. Resetting PICs requires some more parts, see
"Options" in http://www.etc.ugal.ro/cchiculita/software/tinybldusage.htm.
- we could also provide different version of Jaluino, as for
Arduino: with shields, for breadboard. Tiny Jaluino (18F14K50 ?),
Medium Jaluino (18F4550), BigFat Jaluino (18f6680...). We definitively
need to find some better marketing names for tiny, medium, bigfat...
What do you guys think about this ? Any more ideas ?
I'm willing to help, but I'm not an electronic Guru... I'll soon
received some FTDI chips + breakout boards, so I'll soon be able to
prototype and try few things in this wishlist.
PS 1: I wasn't aware of Fritzing, looks nice, I definitively need to try it !
PS 2: it seems your blog is powered by Google Site. How is it ? Do you
think it's an option as a website for jalv2/jallib ?
Cheers,
Joep Suijs
>
> - we should also have a connector for serial comms. Here I can see 3 options:
> 1. using a DB9 connector, and MAX232 chip. Big.
> 2. using a FTDI converter. Small. Expensive.
> 3. as Albert Faber suggested in the post, using a 18F14K50 to
> build this converter. Cheaper than FTDI. (as far as I'm concerned, I
> can't easily find this chip, except with high shipping rate, making it
> almost as expensive as a FTDI chip).
> 4. No other connector. We use the first, and use a USB bootloader.
> Guru Albert already developed one with jal (but bigger than tiny
> bootloader, see later). Serial comms. are done through this USB
> connector. I'd personally prefer to have a connector wired to TX/RX,
> in addition to USB connector.
And if we do none of the above, we can have a 4-pin header:
- ground
- 5v
- tx
- rx
Cheap, small and still ease to access serial port (with only one
convertor you like for all your boards).
> - I²C: alsways a must. An "I²C ready" board allows many experiments
> and is opened to many 3rdparty chips.
I propose a 5-pin header here with the following layout:
- ground
- scl
- sda
- 5v (= regulated ps of pic)
- 9v (= unregulated input ps)
And optional pull-up resistors.
Joep
Sebastien Lelong
>And if we do none of the above, we can have a 4-pin header:
> - we should also have a connector for serial comms. Here I can see 3 options:
> 1. using a DB9 connector, and MAX232 chip. Big.
> 2. using a FTDI converter. Small. Expensive.
> 3. as Albert Faber suggested in the post, using a 18F14K50 to
> build this converter. Cheaper than FTDI. (as far as I'm concerned, I
> can't easily find this chip, except with high shipping rate, making it
> almost as expensive as a FTDI chip).
> 4. No other connector. We use the first, and use a USB bootloader.
> Guru Albert already developed one with jal (but bigger than tiny
> bootloader, see later). Serial comms. are done through this USB
> connector. I'd personally prefer to have a connector wired to TX/RX,
> in addition to USB connector.
- ground
- 5v
- tx
- rx
Cheap, small and still ease to access serial port (with only one
convertor you like for all your boards).
Yes, you're right ! I have a bluetooth/serial converter (great), and such a converter could be used with what you're proposing.
Maybe we could provide mini-boards, ready-to-plug, for each of these options. Would be highly flexible, I'd like that !
I propose a 5-pin header here with the following layout:
> - I²C: alsways a must. An "I²C ready" board allows many experiments
> and is opened to many 3rdparty chips.
- ground
- scl
- sda
- 5v (= regulated ps of pic)
- 9v (= unregulated input ps)
Looks ok for me. With such a standardized connector, we'll able to easily plug multiple jaluino boards together...
And optional pull-up resistors.
You mean, with a jumper ?
Oh, I forgot one thing: the PCB should be a single-sided PCB...
funlw65
Google Sites have a tree structure regarding to static pages. As an
example, you can have few main categories, inside categories you can
have how many books you want. Then every book can have chapters, and
so on.. but I don't know how far you can go (how many branches). It
also have a news module (something like a blog).
At every page you can attach files.
You can use it for jallib but must organize it with a paper and a
pencil before. Create a test site before and explore it.
The news module can be used as blog but only Google Sites members can
comment... so, I'm still using Blogger: http://moriscanet.blogspot.com.
Also, you can't include javascript as in Blogger.
Is at no cost and is secure but can't be compared with Drupal, e107,
etc... Test it and decide what to do.
As for boards, I'm a beginner with limited knowledge and resources. I
just wanted to give an impulse to this team. I really need a general
good working board. A board which I can use it for a weather station,
windmill controller/logger, cnc board (as arduino is used for reprap),
etc. and all using USB for comm. and data transfer (because these
applications require mainly portable computers where you can't find
serial/parallel connectors anymore).
Joep Suijs
>
>> And if we do none of the above, we can have a 4-pin header:
>> - ground
>> - 5v
>> - tx
>> - rx
>> Cheap, small and still ease to access serial port (with only one
>> convertor you like for all your boards).
boards. The upper 8 are for ICP.
>> > - I²C: alsways a must. An "I²C ready" board allows many experiments
>> > and is opened to many 3rdparty chips.
>> I propose a 5-pin header here with the following layout:
>> - ground
>> - scl
>> - sda
>> - 5v (= regulated ps of pic)
>> - 9v (= unregulated input ps)
>
> Looks ok for me. With such a standardized connector, we'll able to easily
> plug multiple jaluino boards together...
Again, I have a lot of experience with this kind of layout. Especially
the combitation of the two power supplies well surprisingly. It can be
used for power distribution at both levels within the same system. And
since it are different pins, mistakes result in system parts not
having power, not releasing smoke ;)
>> And optional pull-up resistors.
>
> You mean, with a jumper ?
>
At least a place to put them. Jumpers (you need 2) may be overkill
(you can put a 2-pin header at the resistor's place and put the
resistors on there if you to switch often). Two soldering pads could
be usefull if the space is there.
Joep