Anyone considering a Prop/Mikrobus development?
Mickster
Posts: 2,693
in Propeller 1
Lots of modules available....would be very cool.
Comments
I've got to hand it to those guys at MikroElektronic -- they churn out a lot of bits.
I wish the Prop had similar plug-in modules.
Pretty good idea and simple enough to implement with the propeller. Also leaves enough propeller pins free for additional interfacing.
Just so I understand the question, are you asking if someone is considering a Propeller clicker board?
More of a Prop-based host, actually. One socket per cog?
Put as many sockets as you want, but it makes more sense (to me at least) to have one cog for each of the three signal busses (spi, i2c, rx/tx). Of course implementing CS (more or less the board select signal) needs a bit of thought. They could be "or'ed" together like the propeller does for I/O
"Click" = peripheral daughter board.
The existing development kits are extremely limited, with just 1 or 2 MikroBus sockets.
Propeller already has (RX/TX/SCL/SDA/RST) defined. Assign dedicated pins for (MOSI/MISO/SCK). You have enough pins remaining to allocate unique PWM/INT/CS for each of 8 sockets.
The analog pin is a whole other story. The Propeller host PCB would need an ADC to be fully compatible.
That would make for a pretty powerful development system, and there is certainly enough room under/between the connectors for the propeller, an 8 channel adc, 8 channel dac, and any other peripheral hardware you may need.
With all the Click module boards;
http://www.mikroe.com/click/
I think there potential for a neat and comprehensive Prop-based modular system.
I actually previously asked the MikroElecronika guys if they would consider a Prop board, but they said it was not in their plans.
+1
From a commercial standpoint, being based on an established, supported architecture is always an advantage.
The range of modules is already impressive and will continue to grow.
I think with some care you could cover the following plug in modules all in one board
- up to 8 Digilent Pmod's
- up to 8 MikroBus modules
- up to 4 Arduino shields (plus 3 M-F jumper wires to link the supply across to the power supply pins on the shield. )
- breadboard in the middle for other modules
Plugging in an arduino shield would knock out 3 of the smaller modules.
For the 'host' this would work primarily with
- the Prop123-A7 or A9 boards (P2 development),
- DE0-Nano (P1V or P2), via the twin 40 pin headers, but you could also fit
- one or two DIP props for a low cost and familiar option
This can be (just) made to fit in a 4x4" sized PCB, $9.90 for 10 from Seeed or other manufacturers
And there is even an octagonal case available from Hammond
[url="http://"]http://www.hammondmfg.com/pdf/1590STPC.pdf[/url]
But routing wouldn't be fun (not impossible, just not fun)
Yes, but while MikroElektronika's modules are more or less for prototyping and development, Tibbo has made a similar thing for low volume production series (because they are not really cheap). But they have also enclosures even with display:
http://tibbo.com/tps.html
Very, very nice....but where are the performance spec's?
http://tibbo.com/programmable/modules.html
http://tibbo.com/programmable/boards.html
And here a sort of specifications (for the main T1000 chip/module)
http://docs.tibbo.com/phm/em1000.htm
They are a bit unconventional regarding datasheets and traditional specs, because it is a SoC with their OS inside. The T1000 is an ASIC I think probably based on an ARM core. Their OS runs a bunch of core functions isolated in the background while at the same time executing your code. You can choose a firmware with fixed function (eg ethernet rs232 converter) or a programmable one (thet also allows for converter function and more)
I've switched to them from Lantronix for the sole function of ethernet interface and ethernet/rs232 converter.
When I only need a converter or a bit more I use the EM500. When I need mor power/functions or eg field upgrades programming of propeller/pic chips I usually go with the EM1206. The EM500/1206 are P1/PIC companion modules.
For stand-alone HMI interfaces I use the EM1000 because of the higher IO available.
Lastly, for quick prototyping and/or really low volume products I've used the new EM1001 up-side-down directly on my pcb.
Except for EM500 the others (EM1000/1001/1206) have the same CPU so the same performances.
Their TPS system with TIBITS is basically an EM1000 module on their PCB where you can add pre-built external interfaces, the tibits. Basically the TIBITS are relays, buffers, ADC/DAC converters with the right fiedl protection and interface that are controlled sometimes as IO but most of the time through SPI/I2C interface from the central unit (T1000).
If you need only the field interface you can use them with any microcontroller.
I do not use them because they are a bit expensive, but they allow you to configure your hardware from the CPU through the interfaces ending with connectors and right voltage levels. Yuo have only to think about software.
Perhaps building such PCB but with P1/P2 on it will allow to use your microcontroller with their field interfaces.
If you look in the middle of this picture
There is a certain elegance, purpose, and value to each. Yet I was still compelled to make my own system to satisfy my immediate needs. My only point is that tinkerers are rarely satisfied. Attempting to perfect a tinkering product will cause it to be P2'd (the verb that describes an endless refinement that never comes to fruition).
For the MikroBus it would be better to use 2x8 connectors on the "Clicker" and right angle connectors on the "Click" plug in boards. Makes for a smaller and lower cost system.
That is the basic image that is conjured in my mind when discussing peripheral buses.
The Clicker + Click model, though, seems to be small-shields-within-a-bigger-shield. Need to add more click boards? Within reason, you can just build vertically.
http://www.mikroe.com/mikromedia/shields/mikrobus/
Yes, vertical stacking is an option but it has some shortcomings. Main one is it makes accessing boards more difficult. Another is it requires stackable connectors which add cost to a project and may not be available on the board you need.
Amen.