Is there a market for a Teensy 3.1 48 pin ARM stamp?

Status
Not open for further replies.
We launched Teensy 3.0 with a Kickstarter campaign, which was funded on September 16, 2012. The campaign made this promise to backers:

(when reading this quote, keep in mind the context of 2012, when no 32 bit boards worked at all on Arduino, and the tutorials that existed usually had dozens of pages about installing toolchains and all sorts of complex setup)

Long-Term Project Goals:

With the initial release, Teensyduino will support the normal Arduino functions, digitalWrite, pinMode, analogRead, and so on.

However, there are many libraries for Arduino(R)** which are not designed to be cross-platform. Paul Stoffregen has personally ported dozens of libraries to Teensy 2.0 (and in the process, also Arduino Mega and now Arduino Leonardo). Firmata, Servo, IRremote and OneWire are just a few of the well known Arduino libraries which work across many boards, due to Teensy's porting project. Teensy 3.0 will continue this process, for most major Arduino libraries, and in the process probably pave the way for easy porting to other 32 bit boards, like the upcoming Arduino Due.

Paul has also contributed numerous features and bug fixes to Arduino. Teensy 3.0 will continue this tradition. Here is a list of Arduino contributions to date:

http://www.pjrc.com/teensy/arduino_contrib.html

This kickstarter, and continuing sales of Teensy afterwards, keep funding Paul's work on Arduino and low-cost, easy-to-use hardware.

The long term goal is not merely an ARM-based board, but a modern platform that works with the extreme ease-of-use you'd expect from the Arduino environment.

Half of the rewards were promised to ship in September and the rest in October. Like many projects, we had a setback. It involved the USB connector soldering. Most of the September rewards shipped in the final week, but about 200 of them shipped on October 1. All of the October rewards were shipped by the 2nd week of October. We were left with about 100 rewards lacking shipping info or with incorrect payment. Back then Kickstarter didn't factor in shipping, so international backers were supposed to voluntarily & manually add extra to their pledge, which was of course a logistical nightmare. By January 2013, we has resolved and shipped all but 3 rewards. Amazon Payments helped us to send refunds to those final few backers we never managed to contact.

For the first few months, many beta releases were made as the most popular libraries were ported and bugs were fixed. The first non-beta version to support Teensy 3.0 was Teensyduino 1.12, released in late January 2013. By that point, all of the libraries that come with Arduino and a few of the most popular ones like IRremote and Encoder had been ported.

In mid-2012, the WS2812 LEDs appeared on the market and started to gain popularity, since they were much less expensive than the then-common LPD8806. Due to some previous work I'd published for LPD8806, and some awesome projects people built with it, support for "LED video wall" support with the new, cheaper LEDs rapidly became the most common request. So I took something of a detour from porting the lesser-requested libraries and developed OctoWS2811. Since it's initial release in March 2013, an incredible number of really spectacular LED projects have used it.

But porting libraries and improving Arduino compatibility didn't take a back seat. Not for long anyway! Teensyduino 1.14 was published in May 2013. If you view the release notes on the download page (scroll down), you can see that 1.14 fixed an incredible number of finer points to Arduino compatibility and made most of the widely used libraries compatible. 1.14 also added new features that I believe still do not exist on any other boards, like analogWriteFrequency() and highly optimized Serial.readBytes().

If you view the rest of the release history, you can see the versions that followed in the rest of 2013 continued adding more and more compatibility with AVR-centric code (a feature no other boards support, even today).

In late 2013, I started really working on audio support. This had been one of my major long-term goals even in the Kickstarter days. Audio took a LOT more work than I had imagined. It dragged on for nearly a year before the 1.0 release, which finally happened in September 2014.

Teensy 3.1 was released in January 2014. Many other Arduino improvements also happened during the audio development.

Somewhere in mid-2014, I took a break from audio stuff to port the last of the libraries that Teensy 2.0 supported. Most of those last libraries had few, if any requests from people, but it was a final push to get all the libraries provided with Teensyduino to work on Teensy 3.0 and 3.1. Really, the 2 libraries that has any substantial amount of requests were my own: FreqMeasure and FreqCount. The others, like AltSoftSerial and TimerOne and MsTimer2, weren't really "necessary". I did them more of the sake of completeness.

That last round of library porting was published in Teensyduino 1.20. Most had been on Github for some months. But if you measure from the day the Kickstarter was funded to the release of version 1.20, it really was about 2 years until all the libraries included with Teensyduino (and a LOT more on other sites, especially Adafruit) were all ported.

Even now, plenty of software development is ongoing. Some features I'd planned since before the Kickstarter campaign have been chronically pushed onto the back burner. So in some sense, the software _still_ isn't 100% complete. It may never be. Really, it's all a matter of perspective!
 
Last edited:
Teensy does have far more AVR emulation on the ARM than any other I've seen.
But there are 2+ ARM boards that provide Wiring support (pin # concept, digital write/read, analog write/read), USB/serial. And some do better at some of these functions, often because of better MCU peripheral sets.

There too, some of these boards are focused on industrial applications and support both Wiring-like support and MCU vendor I/O libraries as a primary focus - business case driven.
There's an up-turn in the IOT craze, with some good boards with (Broadcom) WiFi aboard at amazing prices. That's a world of early adopters that Teensy would well address.

There's no Teensy-match I've seen for the ready-to-go apps on top of the I/O libraries, for video walls, audio, etc.
 
Last edited:
Exactly. I would never expect a hardware manufacturing concern to provide library support for libraries that are unrelated to the basic hardware itself. That Paul has done so much for both Teensy and arduino alike speaks more about how Paul embraces the maker community than anything else. He didn't have to submit fixes to the broader arduino community for issues that have been around for over a decade (ie SPI) and arguably that work detracts from his teensy pursuits.

The technical support here and the overall core unit support is second to none. I don't expect Paul to port third party libraries but he does frequently. In fact, it would be great if more teensy users would / could step up to the plate and submit a library or two instead of expecting Paul to deliver revenue-earning solutions for free. The price of admission for Teensy tech support is so laughably low, I cannot believe that anyone would even begin to complain about it.

Bottom line for me is that teensy is an amazing ecosystem, one that eclipses the original mother. My issue is rather to find the time to move my projects ahead. But that's my fault alone.
 
Flexible circuit for extending underside pins

Does anyone have any experience with flexible circuits?
Maybe a flexible extender could be used to bring out the bottom pin connections.
Adding pin headers would allow attaching to breadboard, or it could be directly soldered to a motherboard.
It might be a little tricky to solder to the T3.1.

Here's a concept sketch

Flex Circuit.jpg
 
Last edited:
Quote Originally Posted by t3andy View Post

We waited 1.5 - 2 years for the Teensy 3.1 software libraries to be in decent shape so we could finish our projects.
Why? Which libs exactly?

One library was the "low power library" and it was not done by PJRC. Most of the work was done by the fine gentleman from State College, Pa - Duff.
 
Last edited:
One libraray was the "low power library" and it was not done by PJRC. Most of the work was done by the fine gentleman from State College, Pa - Duff.

Thanks for nice description of me but it wouldn't really amount to anything if paul didn't put the mini54 chip into low power also.
 
Remembering way back, we had to shame PJRC in doing any type of work for low power. We always said "Apple computer chose the ARM chip for low power - why doesn't PJRC have low power for their Teensy ARM chip?"
:rolleyes:
 
t3andy: you are implying that you have contributed something to the world by shaming. Paul and others here have been very generous to you. You have contributed nothing. You complain and brag with little justification for the former and no justification for the latter. Maybe someday you'll grow up enough to appreciate the value of being respectful. I hope so. Meanwhile, I wish you would find something better to do than trolling us on this forum.
 
You have contributed nothing

You must be new to this board. Please do your homework before speaking! In fact, several of our non-commercial projects are even showcased on the PJRC projects!!!!:confused:
 
Please, remember to be nice and civil! and that over text there are no visual cues about sarcasm, irony, exaggeration etc.
 
I'm not going to comment on the last few pages, it has diverted quite a bit from the original topic. A few days ago I had an idea might make the bottom pins more easily accessible; OSHPark seems to support internal cutouts; http://support.oshpark.com/support/solutions/articles/134889-internal-cutouts-and-slots. Perhaps it would be nice to create a small PCB which extends the Teensy form factor and makes the bottom pads more easily available (the Teensy is placed on top of this PCB). With an internal cutout we can ensure that these bottom pads are easy to solder. As a picture says more than a thousand words, I made a quick mockup for it. I'm sure someone can think of a nice pin arrangement, make a nice board with silkscreen and upload it to OSHPark. If you are willing to drop the program and ground pin a bit thicker traces could be used. Save for the cutout, this design passes OSHPark's design rules, so it is doable. Just my two cents.
mockup.jpg
 
The mockup is 2.2 by 0.7 inch, OSHPark has a flat fee of 5$ per square inch, that gives you three boards. So somewhere in the order of: ((2.2 * 0.7) * 5$) / 3 = 2.56$ per board? Time will probably have to go into deciding on a pinout for the extension part, and a few prototypes runs are likely to be necessary, the OSHPark guidelines state that via's are never a problem, so perhaps it's better to replace the current pinheader holes with via's in the Teensy board. I've never had need for the bottom pads, I'd like to try the internal cutout on a PCB, but I'm swamped at the moment (and roundtrip time to Europe is quite long...) Anyone who can chime in on the internal cutouts? Would this work?
 
I asked Laen (OSH Park founder) about cutouts not long ago. He said it's unofficial but usually works fine. Apparently all you need to do is draw it on your outline gerber. I've been intending to give it a try, but just haven't had the need yet.
 
Talldog over at tindie.com who sells the current 3.1 shield to bring out the pins had an earlier version of the shield that did not add to the width of the Teensy like the current version does. I bought one and it along with the second generation is in my large pile of things to do: https://www.tindie.com/products/loglow/teensy-30-mini-breakout/

Note, this was designed for Teensy 3.0, so the DAC (A14 on 3.1, A12 on LC) pin will be labeled reset, the LC pin to translate A3 to VIN voltage will be labeled Vbat, and there won't be a pogo pin for the 3.1 reset pad.

One thing that I wish this and other similar shields did was put 1 or two rows of pin on the end with the USB connector, and in particular make Vusb and Vin next to each other, so that you can have a simple jumper between the two pins, instead of running a longer wire.

I think I may have seen this shield over at the shared designs at Oak Park, but I don't think I saved the link.
 
I asked Laen (OSH Park founder) about cutouts not long ago. He said it's unofficial but usually works fine. Apparently all you need to do is draw it on your outline gerber. I've been intending to give it a try, but just haven't had the need yet.

I've tried it and it works very well. Not sure what he means with "unofficial" as there are instructions on the OSH park website on how to draw it (in Eagle) so it will be recognized by the factory.
 
Yes! I ordered some of those. Javascript, simple IDE (uses Chrome scripts). Nice library. "Shims" (as in Arduino "shields") for WiFi, Etherent (Wiznet), and others.
This is a great example of more state of the art in software and hardware versus a breadboard-centric PCB. I hope a next-Gen Teensy effort will look earnestly at Espruino Pico, Particle.io's Photon, Microduino, micropython (board), and others.

Also, re Espurino Pico - the Javascript interpreter C code is open source - and it can be built for many ARM targets - mostly ST so far.
You can do the interpreter C build on Macs too. That's a lot of return for a small amount of time to migrate it rather than another burdensome DIY for bare metal.
There's a GUI-based programming tool for beginners/children too. Like RPi's "scratch".

Python on RPi has opened up programming / "coding" to a huge audience - whereas C++ in disguise (Arduino) can't do so.
Interpreter, yes. Wider appeal? Definitely. Let's give 'em libraries in C++ for performance when needed, and give the user a higher level language: Python, Javascript, GUI like Scratch, etc.

"To Infinity and Beyond!"
 
Last edited:
The word "Teensy" means small.

Now days, ARM chips have a very large count GPIO pins. Most 64 pin ARM chips could have counts of greater than 50 GPIO.

To be useful, the GPIO should be "easily accessible" so users can utilize all the GPIO for complex applications.

The word "Teensy" really can't be applied to ARM chips because to make a ARM stamp extremely useful, you would need to expand its physical size so users can take full advantage of all the extra GPIO an ARM stamp would offer. Of course, if you wanted to make a Teensy ARM stamp small, in physical size, then you design an ARM stamp which does not allow "easy access" to all the GPIO needed in complex applications. (Just hide the GPIO on the bottom of the stamp)

Look no further than what the Australians and New Zealanders are doing with their MicroMite+ Explore64 stamp module. It not an ARM but a PIC32 64 pin chip. They expanded all the GPIO out on the stamp module so it is easily accessible for complex applications.

Imagine a 51 pin stamp module driving a 7" TFT LCD color display with touch, keyboard input, SD memory storage. MMBasic is built into the chip, along with all the graphic, touch routines for the TFT display. In fact, you could have a complete standalone portable computer no IDE or PC required.

Great things happen when the GPIO is easily accessible!

:cool:
 
Last edited:
I'm fiddling with the Pico and now, the Particle.io Photon.
I like both. They differ.
The Photon is, IMO, where mbed wanted to go, but Mbed got stuck in a ditch.
Photon and it's easy infrastructure is AMAZING.
Lots of libraries. Compile and re-flash via Internet and WiFi, with the compiler/IDE web browser based.
And you can elect to DIY libraries and use SWD for debugging and a local standalone tool chain.

This is where the Make folks are going and its a much different path than Teensy 3++ is focused.
 
Nobody will buy it, it has not every single io broken out ;):D
Cortex m3 only??? Much less gpios than teensy..?
 
Last edited:
I wouldn't be surprised if mbed sold that many. The key issue is post sales support and uptake/support by the community. The concept certainly sounds very cool!
 
Status
Not open for further replies.
Back
Top