will this be a milestone change in embedded tinkering?

Status
Not open for further replies.
Generally I consider putting engineering time into supporting new platforms when there's at least several independent requests and some indication the platform really is gaining a substantial user base.

From their forum activity, it looks like 98 threads since June. Judging from the forum activity log, looks like under 100 registered users (just 3 pages, mostly registrations, approx 30 per page).

Their release & updates shows many in June & July, but they did release updates in August and October, which is a good sign.

Looks like it's off a decent start. Hardly seems worth investing a lot of work at this very early stage.

Realistically, this sort of thing probably also makes more sense on the upcoming high-end Teensy, where we'll have 256K of RAM. I saw a mention they only need 32K of RAM, but that's at a very early stage of development. Long-term, I suspect this platform is going to require more RAM.
 
Support for this platform would also automatically mean free publicity and perhaps more spread for teensy :)

Personally, i hav'nt use python. But it seems to be the new "BASIC".
I would be interested to help, and maybe my son is interested to learn a bit python(?)
However, I have a little problem , because I do not know python :)
 
well, for me, the RPi was the right tool for driving output to VGA and monitors, and it was just so easy to use a camera with. As for Python, I had not used it before, but it really is as simple as allowing yourself to accept indentation rules...
 
Seems like a different platform for sure - maybe good - but unless fully done to the extent the Arduino stuff is there would always be something you couldn't do.

I hit the KStart and other links last night - went by the DEV and tutorial - seems "BASIC" enough - interpreted but advanced, seem to have not implemented GOTO? Of course it isn't BASIC, but other note said the language was mostly well ported.

I found a HACKADAY saying it was ported to esp8266, just missing a little part called WiFi from 12 months ago.
Micro Python Now Runs on the ESP8266 – Contributors Wanted to get Wifi Working :: November 29, 2014

Then looking in http://doc.viperize.it/0.2.0.0009/supported_boards.html I saw the support is Broadcom based [Particle plus DUE] and no hardware listing for any ESP8266 devices - so I didn't bother to post. The http://doc.viperize.it/0.2.0.0009/board_udoo.html#udoo is a 96KB RAM DUE work alike - it can
With Udoo and Viper is possible to develop in a few clicks awesome interactive products where devices and sensors are controlled through the Viperized Arduino side while the User Interface is developed in Android or Udoobuntu

Looking again - just as a quick check on alternate platforms - ESP8266 isn't a good indicator - if that is any benchmark for moving beyond the initial devices:
GitHub 8266 readme update 2 months back says this: "Rudimentary WiFi support in station mode."
The ESP8266 unit has 64KB RAM and runs with 24KB free per that doc.

The Drivers directory lists: cc3000 cc3100 nrf24l01 onewire sdcard wiznet5k. So for IOT usage some connectivity would be needed - or it would just be a cool way to program - without all the 'Native bare metal Arduino' support"?
 
For me the Teensy is the most versatile SoC board in the moment. And if the Raspberrry will not speed up it's GPIO ports, it's no competition in some special field. E.g. try to control a RGB LED 32x32 matrix with a Raspi or with the Teensy (both with buffered adapters/hats/shields) - the Raspi is so slow you have no chance to create a stable, non flickering picture with this machine. The Teensy on the other side controls all LEDs in a rock solid manner - no flickering, no "after glow" or "ghost flashes" ... works all as wished. And also you always have the OS overhead with the Raspi, which is not needed at all when developoing specialized embedded functionalities. So from my point of view I believe this will add some new attention to the Raspberry, but to be honest - what's so amazing about this little machine? :cool: See? ;-)
 

Yes, I have one and I'm using it, recently. It's the most fun of all the new-gen boards I've played with (putting RPi in a different category, as it uses Linux). Viper is a port of microPython's interpreter with improvements to the Particle.io Photon and 2 or 3 other platforms, all ARM M3 or M4. The Viper guys encourage adaptations for more board types. And Viper has a nice browser-based IDE (some people object to such but I'm growing to really like it, for avocational software). The Python bytecode system even has simplified exception processing and user defined exceptions. And modules for multi-file projects. There are lots of great libraries and device drivers. But, Viper so far lacks over-WiFi downloads; it must be USB DFU which is USB speed, not UART speed. Viper is young. No revenue source so it has to be a labor of love which can be risky long term. But so too this is the case of GCC and Eclipse and others.
C code in-lined with Python - said to be supported coming soon. Viper is based on the ChbiOS RTOS and Python uses it for threading.
WiFi with Broadcom's chip is uber better than ESP8266. MCU is better by far.

Among the example programs and drivers, there's one for the serial LEDs. Not my thing but it's likely a subset of what T3 has. Might be a demand for reuse of Paul's work on this, and audio.

Three years ago when I started with Python, coming from asm, Pascal, C, C++, javascript, I though Python was crazy. Now, I still use it a lot, professionally and for hobby stuff. On PCs and now on micros. On PCs, used with Qt Designer and pyQt for GUIs. I really like it - but it one has to be open to doing things differently. Huge improvement in functionality per line of code and time spent.

I see no evidence of Adafruit doing any development work for Viper. Adafruit has done a nice getting-started web site for Viper-supported boards that they sell. But there are a small and growing list of supported boards. Adafruit is just a reseller it would appear.

I could make a demo video and post it somewhere.
 
Last edited:
I'm more excited about the ESP32 than any RPi product.
I just never really got into RPi, though I can see potential uses, its not interesting for me.
I also like how the Teensy community is smaller and things are easier to find, whereas it seems to me a lot of the programs for RPi lack support, whereas people here keep working on their projects and are willing ro help.
 
I don't see anything that can replace my teensy's anytime soon. They are just too good for realtime processing. However an RPI can be a very nice complement to the teensy. I have been using a tool set that makes it extremely easy to display and control what my teensy's are spitting out...HMI the easy way. I'm running Processing 3.0.1 on a Raspberry Pi 2 to communicate with teensy 3.xx boards and it almost seems too easy. With the G4P libray and G4P gui builder for Processing......GUI programming doesn't get any easier than this.
https://www.raspberrypi.org/blog/now-available-for-download-processing/
 
Last edited:
Maybe there's a business model that can sustain such losses? Long-term, if the market for independent hardware (like Teensy) dries up, I'm going to have to look pretty seriously at alternate business models!

My suspicion is this is to kill off competitors like the $9 computer, orange pi, etc. It could also be broadcom trying to empty their excessive stock.

My less cynical thought is that the $25-$35 price tag might be a little much for parents getting something that could end up in the back of a drawer. The RPI folks saw that and cut a deal (maybe using the more expensive models to subsidize the cost).
$5 is that sweet spot where a kid can go to a summer camp or a electronics fair get excited and go home with one. It's also not a unreasonable fee for schools to charge for a class.
 
My suspicion is this is to kill off competitors like the $9 computer, orange pi, etc. It could also be broadcom trying to empty their excessive stock.

My less cynical thought is that the $25-$35 price tag might be a little much for parents getting something that could end up in the back of a drawer. The RPI folks saw that and cut a deal (maybe using the more expensive models to subsidize the cost).
$5 is that sweet spot where a kid can go to a summer camp or a electronics fair get excited and go home with one. It's also not a unreasonable fee for schools to charge for a class.
I don't think it's competitive warfare at all. From the recent interview I saw with the RPi Zero announcement, it seems that they decided "this can be done in our non-profit business model". As was the original RPi which caused exactly the same reaction - "are they selling below cost?". Not at 5 million units!

I'll bet too, that the RPi brought a cheap PC to schools in poverty-stricken areas where they were happy to have them - and some cast-off old monitors with HDMI.

Mine is in the mail.
 
Last edited:
Update... 6 December 2015
Adafruit's $5 RPi zero arrived yesterday.
With a micro-USB to USB-A jack cable adapter plugged into a USB hub, keyboard, mouse and WiFi dongle plugged into hub, and mini-HDMI to monitor... I backed up then booted the uSD card with RPi "Jessie" distribution.
Came up right away, correct resolution (1920x1080) on my monitor! No struggles to get the video mode correct.
Took me a few minutes to realize that the WiFi driver is already installed and there's an icon for it, top right of desktop. Click that, choose my SSID, enter password and ta-da, I'm on yahoo.com with graphics.

Now, on to installing VNC server on the RPi zero so I can disconnect the keyboard/mouse/monitor and run headless. Remote connect using VNC client on whatever PC. I've done this for a long time with the older RPi's.

I noticed that this kernel has drivers for SPI, I2C and other stuff built-in.

Nice.
 
Last edited:
I'm out of topic, I know but can't resist!
An old friend at DFRobot send me (as xmas present) a preview version of this one, nothing to do with embedded but really impressive for 70Us, I was able (out of the box) to install and run even photoshop!
It's a real computer, runs Windows10 really smooth, I added a 1Tb HD to the USB3 and it looks like my desktop computer!. Btw it's overheating, I would prefere 4Giga ram version and it's very dangerous play with boot firmware (install unix can be seriously difficult) and it has a ridicolous name.
I have installed Arduino IDE and Teensyduino and using to program Teensy with no problem, serial works better than my Win10 notebook.
 
a few days now fiddling with the RPi Zero board, $5.
Some impressions...
I have its Linux running headless with VNC, networked with a $7 WiFi dongle. Also can use telnet and SSH to access.
Have SMB running so Rpi can access terabytes on the NAS.
The GPU has its own 2 cores... I moved from HDMI out at 1920x1080 monitor to using VNC via WiFi. Works fine for interactive editing using VNC and editor on the RPi, versus SSH login and edit without the GPU carrying the load for VNC.

The 1GHz ARM11 is oh so fast for the RPi Zero's cost.
Single and double precision hardware floating point (vector). RAM is its own package piggybacked with the ARM chip's package. Clever. Called "stacked packaging" or some such. Versus stacked die - it makes sense for using standard packaged parts and not wafers and die. And saves board space. The Zero has all components on one side of the board.

I just noted too that besides ARM and Thumb instruction set modes, it has a Java bytecode mode, as if it can speedup a JVM.

Loaded Python wrapper library for the C/C++ drivers for the GPIO, PWM, timers, SPI, I2C.
Just 3 lines of code to create a PWM or a <= 40MHz square wave on 1-n GPIO pins. Hardware based so CPU utilization is low.
Haven't tried input capture yet. Nor SPI.
Has DMA for most I/O as well.

pretty neat.
I haven't seen wave 2 of RPi Zero's hit the market yet. I wonder, and doubt that they will raise the $5 price.
 
Last edited:
To me, this is starting to sound like a very interesting iOT hub to tie various sensors together and then either host or upload the data to the web. The CPU certainly has the horsepower to do both, the OS gives it the potential to make use of a lot of inexpensive hardware. If the MCU is more accessible re programming, all the better. Is it a replacement for Teensy? Not at all, at least for me.

There is no way that you can run anything but casual DAQ on that platform because the timing cannot be controlled the way it can on a close-to-metal MCU. The ADC on the Teensy is flexible, versatile, offers 2 simultaneous inputs, etc. A single high-resolution DAC, well-defined PWM outputs, and so on further features allow for a pretty bomb-proof swiss-army knife re: data acquisition.

But, does the average consumer of this hardware care about achieving precise timing, good ADC results, and so on? It wouldn't be the first time that inferior hardware reigned as the market went on to adopt a less expensive platform. If I was beagle bone, I'd be much more worried about the Pi because I'd argue that the Beaglebone is much harder to differentiate to the average consumer. Similarly, the folks at Propeller may have a very interesting architecture but the maker market settled on AVR instead. From a manufacturer point of view, that is what these Pi projects are about - capturing market share and hopefully becoming a dominant platform for that chip maker to milk in the future, short term profits be damned.

Pis may be a competitive threat to Teensy or they may be a potential symbiotic partner, where the Pi takes care of some of the really hard to implement OS features (i.e. USB host mode, Wifi access, web server, etc) that do not require precise timing but a very flexible OS that benefits from decades of software development to allow robust performance, while the Teensy takes care of the very-had to implement things on a PI - precise DAQ / ADC, etc. What would worry me about a symbiotic relationship is whether that market (DAQ, etc) is a big enough / profitable enough one to make the work worthwhile. Consumer expectations may become totally warped by hardware prices that can only be described as unsustainable.

However, the sheer scale of Pi sales is pretty eye-opening. The question going forward is how many will still be in use in a year instead of lying in a drawer like my Maple board. Dumping tasty hardware on the market is one thing, putting the time into making said hardware accessible, reliable, etc. is an entirely different task and one that hardware manufacturers typically fail miserably at. Whether it's the lack of longevity (i.e. see the multitudes of 'paradigm-shifting' kickstarter projects whose makers have moved on to other things) or the lack of supporting its users (mbed, maple, diligent anyone?) Teensy is differentiated majorly by its ease of use and its support.

There is a reason that so many folk gravitate toward Teensy - great hardware coupled with excellent product support (arguably, best in industry, at least at the maker price point).
 
Last edited:
Re timing cannot be controlled... Use of DMA?
Better DAC/ADC than built-in.. I see people using cheap I2C DIP chips for more bits, more channels.

But yes, there's a limit. But so too, on a Teensy or similar that has 1,000 interrupts/sec from systick, albeit just a few uSec. And other interrupt sources.

I succumb to, and observe, that a compelling low price draws buyers like moths to a flame.
 
Re timing cannot be controlled... Use of DMA?

Possible, for sure, but someone has to make it accessible to the average consumer. That's where most of these projects fall flat. For example, how many newbies like myself would be able to decipher / access the golden method to access simultaneous ADC measurements on the Teensy, or even just differential mode. I tried, I failed, I cried (a little). :p Bottom line is usually that the inferior chips can be made to do all sorts of exciting things with the right software support but that support tends to be non-existent when the hardware sales margins are zero - they count on volunteers to make it happen, with all the usual caveats.

Better DAC/ADC than built-in.. I see people using cheap I2C DIP chips for more bits, more channels.
All true, but that also raises the complexity of the rig, addressing those chips isn't necessarily easy, and so on. Pedevide's library is pretty darn amazing in terms of how accessible it makes the ADC in the Teensy.

I succumb to, and observe, that a compelling low price draws buyers like moths to a flame.
Ditto. Hence that Maple in my drawer. Though I now try to resist the urge to buy cheap since it usually means buying something twice. Only when the back end seems pretty much sorted does it make sense to try something like an ESP-12. I simply know too little to take on something much more complex.
 
I succumb to, and observe, that a compelling low price draws buyers like moths to a flame.

A couple years ago Texas Instruments sold an incredible number of powerful dev boards for $4.30 each. I believe they had a limit of 2 per person. Many people had their friends and family order pairs. The low price did indeed attracts hobbyist buyers like moths to a flame. But hardly any of those people actually used those $4.30 boards. Very few projects using those boards actually appeared on Makezine, Hackaday, Instructables and other sites over the next couple years, despite the low price and incredible number of them TI shipped.

Raspberry Pi is one of the few boards that's really been successful at real adoption, so odds are good these new $5 boards really will get used.

But history has show that just because something is really cheap and a huge number are sold doesn't necessarily mean people will actually use them.

History has also shown that hype also attracts buyers like moths to a flame. I believe anyone who's been around long enough can think of many example where something had huge hype and sold like crazy, but also faded away when nobody really used it.
 
History has also shown that hype also attracts buyers like moths to a flame. I believe anyone who's been around long enough can think of many example where something had huge hype and sold like crazy, but also faded away when nobody really used it.
Windows 95 was so hyped, even people who had no PC bought it. True story.:)
 
I'd not put the silicon vendors' (the mention of T.I. eval boards) loss leader/demo boards in the same category as Arduino and RPi et al. These boards don't have a philanthropic intent!
Teensy's forum is mostly "hard core" embedded people, I think, myself included.

We don't see far out of our box - to how many are doing interesting things with the RPi 2, and likely so too the RPi Zero. If nothing else, the secondary education kids get exposed to under-the-hood, somewhat, of computers. Rather than just more mindless, brain-fogging, violence-desensitizing video games.

(with apologies for leaking into philosophy here!).
 
Last edited:
Teensy's forum is mostly "hard core" embedded people, I think, myself included.
Well, some do play both kinds of music ... Country AND Western! ;-) E.g. I "live" in both worlds, because I allways was a technic afficionado and computer nerd - so both systems have my full interest. But you are right: the intersecting set of these two "worlds" is not very big, because the solutions on the raspi or the Teensy - even with the same goal - will differ a lot. I made this experience by myself when trying to control a RGB LED Matrix. First try with raspi and Adafruit hat - well, it works nice, but the raspi's slow GPIO makes this solution nearly unuseable, same goal aimed with teensy (and SmartMatrix shield v3) - works rock solid. The raspi solution flickers and sparkle, the Teensy solution drives the LEDs flawless.

So it depends what you want to do - with the raspi you have a versatile system with a huge amount of possibilities, and with the Teensy, you have all you need to really contol the devices you like to drive. Again: I believe both worlds are not comparable, you also won't compare a Tablet PC with an highly specialized SPC ... :cool:
 
Status
Not open for further replies.
Back
Top