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

Status
Not open for further replies.
Lets see, onehorse has 10 shields that mention Teensy in the header (and some that can mount on a Teensy, without it being in the header): https://www.tindie.com/stores/onehorse/

Tall Dog's teensy carrier that allows access to the bottom pads: https://www.tindie.com/products/loglow/teensy-31-breakout/?pt=full_prod_search

Petit Studio's carrier that allows access to the bottom pads and also breaks out a Teensy to an Arduino shield: https://www.tindie.com/products/freto/teensy-3-breakout-board-and-shield/?pt=full_prod_search

A pre-made Teensy, wifi, WS2812b setup: https://www.tindie.com/products/zedled/ws2812b-teensywifi-led-controller/?pt=full_prod_search

Smart Matrix controller for 32x32 RGB panels: https://www.tindie.com/products/Pixelmatix/smartmatrix-sd-kit/?pt=full_prod_search

Pico's shield that brings out nRFL2401+ radios as well as an Arduino Shield: https://www.tindie.com/products/pico/rfx-teensy-3x-nrf24l01-carrier-board-w-prototyping-area-/

If you go over to OSH park, and do a search for Teensy, you will get 10 pages of PCB designs (including many of the above products, and presumably some test boards).
 
Add-on boards. Each unto itself- generally, need a way to coexist - and how to get an Erector Set - is the point.
 
Last edited:
I believe Onehorse will appear any moment, begging to differ about how many add-on boards exist.
;)

...Really, my intention it to let others build lots of interesting add-ons rather than trying to make everything.
Hear hear. Bottom line is that the board can be made by those who feel strongly enough about it. Put in the sweat, reap the rewards. I have a fistful of inferior Teensy 'clones' that reflect my skills as a ME doing CS work. However, my hope is that they'll do a great job of measuring power or communicating via GPRS.
 
I've not made myself understood.. or I'm the dummy.
The point is to get a capability to add-on/stack 2, 3, 4 boards with signal availability so that most combinations of boards can coexist in a stack. Rather than making a mainboard or breadboard. Better than Arduino's AVR-centric stack.
 
The MikroElektronika plug-in board stack for a Cortex M4 is a good idea. I think Teensy could do much better - more pins, more CS, more GPIO. The MikroElektronika bus socket is find for a university lab project where only one plug-in is present.
But enabling a max of 4 or 5 would be so much better. several GPIOs, CS's, 2+ SPI ports, 1+ I2S, 3+ UARTs, on the bus.

An octopus as in the photo here isn't what I had in mind. .
http://www.seeedstudio.com/depot/images/product/Grove I2C FM_03.jpg

Think physically small. teensy. Teensy. Teensyduino unchanged.
like these stackables
https://www.microduino.cc/module but for ARM Cortex!
https://www.microduino.cc/
https://www.microduino.cc/module/view?id=53d89d10c69eee000055f536

Indeed, some of the above could work with the ARM.
Just need a board with the socket as in the above, and it slides onto Teensy 3.1 (or LC?) pins.
Footprint stays Teensy sized.

Make one of these that uses the Teensy 3.1's Freescale
https://www.microduino.cc/module/view?id=53e9d9563f6ada00004fb0c3
 
Last edited:
pin count - should be what it needs to be to get the I/O flexibility - not implying a DIP type layout.

I'd love to see, and I think it would take Teensy to a broader audience, is stackable boards as shown by example in post #31.
The stackables from MikroeElektronika are nice (e.g., the STM MCU) but not enough pins in theirs.
The pin count is higher in the microduino even though they are smaller.

A stackable arrangement makes this more solderless plug and play with add-on boards.
The trick is to get the right pins in the stack - e.g., if one wants two SPI or 3 UARTs (and the current or future MCU can do so) then all those pins have to be present up/down the stack. And so on. It's a guess what most people will want.

One could do a passive device PCB that creates such as stack with sockets, and connects to the existing T3.1 and another for the LC.
No new PJRC MCU board - just a clever choice of pins and board interconnects. I like the small pins of the microduino boards.

Or the existing collection of microduino boards might well work as is (some are N/A, like the MCU boards that source signals).
 
Last edited:
My post #8 referred to a making a way to plug to get extra I/O - this shows such a plug: https://tiny-circuits.com/ Suggestion was for the TBD Teensy ++3.1++ version - not going back and re-working the T_3.1.

It also shows another way to abuse electrons in a stackable way as stevech suggests, not that I look forward to stacking.

embed_tinycircuits.jpg

https://s3.amazonaws.com/grandst/images/products/embed_tinycircuits_gif.jpg
 
Last edited:
At first it seemed you wanted a 48 pin form factor Teensy. Now your talking about stackable shields. I'm struggling to see how they're related, or what this is really about?
If they agree on a form factor and then manage to agree where to map signals to pads, agree that it is primarily JUST a Teensy 3.1 clone and then promise to be patient then I *might* volunteer to be the dummy that has a slash at laying out their PCB - I doubt I'll ever get to practise enough to start thinking I might even get close to as good as you seem to be at it, Paul, but I am not completely useless and have even completed a few designs for work which weren't riddled with cut-tracks and link wires by the time they were finished functional.

No promises, bit on my plate (and a blasted flu or something) atm, but if you manage a dazzlingly clear picture of what you want you may even get much better volunteers than me - whoever volunteers (if really anybody) should do their best to make a reasonably easy layout to verify and then interested parties should review it carefully before anybody has any cut.

That was silly of me to think I might not be able to find signals to route to 48 (+ 5) pins because I can find at least 50 pads on Teensy 3.1 which all (at least mostly) connect back to the MCU.
 
If you want to come up with a shield form factor for a stackable experience, I suggest some serious research in advance re: what people want to attach to the thing before committing time and energy to designing it.

For example, should it feature a grove-like pinout option in addition to a stackable pin header configuration? Should all the analog stuff come out on one side the way that Paul has it or should it be interleaved to allow the grove-like connectors to feature a mix of analog and digital outputs/inputs? The latter option is a lot easier to lay out as the Analog and digital pins are all over the place on the MK20. If it was up to me, I'd design the thing to be rectangular with the stackable pins inset from the edge and right angle headers on the edges to allow grove- or jeelab nodes to be attached to the edges.

Another suggestion: Using a standard 6-pin or 8-pin header is your best option to allow others to easily come up with stackable shields since 6-pin and 8-pin extended pin headers are easy to come by. For the Teensy, that might mean using 2x8pin headers and 4x6pin headers to allow you to bring out at least 40 pins. Or 8x6pin for a 48-pin shield.

I'd keep the shield dimensions to something that can be designed on Eagle with the free license (i.e. 3x4, IIRC). That makes adoption easier. USB right-angle micro B with through holes (harder to damage) and a PGM tactile switch that is PTH as well and accessible from the side.

Pull up at least 1 or 2 power pins, in addition to a number of ground pins. In particular, I'd look to create a bus power trunk with Schottky's feeding into to it to allow the shields to feature their own power supplies.

Well, that's more than a few suggestions! Sorry!:rolleyes:
 
In terms of stackable systems, it reminds me of smARtDUINo (https://www.kickstarter.com/project...ystem-by-former-arduinos-manufact/description).

Now given these jokers are the fine (other adjectives may apply) folk behind arduino.org I wouldn't give them much credo in terms of design and I would probably count my fingers if I shook hands with them, but at least initially they did think about stacking designs.

Defragster mentioned tinydunio, which seems to have created a market niche (when I go to Micro Center, I see quite a few of their little packets): https://tiny-circuits.com/

Or you have the Raspberry Pi model, where you have 2x20 row of 0.1" pins, and everybody builds a hat that stacks on top of that (including specified holes for standoffs, and a dedicated i2c bus for communicating the shield features).

The problem with a lot of shields is every shield tries to use the same pins, and even if the LC has alternate i2c/spi buses, the shield usually uses the standard ones. I've seen Arduino shields that tried to address this by allowing you to remap pins, but you tend to end up with big boards.
 
Last edited:
It is extremely hard for me to say this but I think I found a new religion in stevech's suggestion of a modular, stackable, bus for the Teensy 3.x. Maybe a 48 pin dip is not necessary but a square with "all" the GPIO properly engineered for a common stackable bus.
 
Excellent! The more people agree on form factor, pin-outs and 'extra bits' (like regs, level shifting(?), grove stuff etc) the more likely I will volunteer - I would be doing it in Protel99SE btw, pretty sure Eagle will import files from it.

I'd never volunteer to design the darn thing but if enough is defined it becomes much more a matter of basic schematic capture and PCB layout and those I can volunteer at least a bit of my time to quite happily.
 
As more of a mc product user than a mc product developer, small form factor for embedding, powerful processing, good support, and a reasonably well developed interface were what drew me to Teensy. Cost, well yes, to some degree, but a robust small product was key. Size matters ...
 
Last edited:
Maybe i'm misinterpreting what people are requesting, but to me, right now the current Teensy 3.1 is fine.
If doing this wouldn't take resources away from (in my opinion) far more important things and was trivial, I would say sure.
It's been made clear by pjrc that's this is not feasible without a high cost in time and money.

In some sense we are lucky the pins were broken out at all.

Like others have said, other than being less elegant, if you need the extra IO you can simply solder right onto the board, buy that tindle thing, or design your own (either making your own pcb, or with protoboard)

What I think might be a more reasonable request though is that the next teensy whatever it is, not use this scheme and instead makes the board bigger if needed and put all the IO easily accessible. Perhaps laying things out in such a way one can cut the extra IO off if they aren't wanted or needed.
 
What I think might be a more reasonable request though is that the next teensy

The problem is on the next "future" Teensy, Teensyduino users will have to wait for another 1.5 to 2 years for the software to be created and debugged.
Right now, the Teensy 3.1 with resource heavy 256K flash, 64K Ram with loads of GPIO, would be the perfect candidate for a new GPIO pin/bus format since the software libraries has already been made and tested.
 
The problem is on the next "future" Teensy, Teensyduino users will have to wait for another 1.5 to 2 years for the software to be created and debugged.
What makes you think that?
The chips are compatible.

This whole discussion is a bit useless. Paul has published the schematics, and even boards. What do you want more ?
If you want bigger boards... why don't your make them ?

I'm sure, the majority (including me) a) dont want and b) dont need bigger boards.

Edit: Perhaps Onehorse can help your with your big boards ?
 
What makes you think that?
The chips are compatible.

In a sense, all ARM chips are compatible BUT how long did it take to port the software libraries, from the T3.0 to the "ARM compatible" Teensy-LC and T3.1? It was a great effort by many users/beta testers on this board to bring it to final fruition. (Along with Paul S.)

If you want bigger boards... why don't your make them ?

Correct, we have the schematics and several users are interested in providing/designing an alternative layout to the Teensy 3.1.

I'm sure, the majority (including me) a) dont want and b) dont need bigger boards.

Lets take a survey or poll?

In this beginning topic thread, I stated it would be convenient to have all the GPIO broken out "topside" in a 48 pin dip. Others, are looking to bus the GPIO in a square format. This discussion is about how it can be done and engineered properly since Paul S. has provided all the design docs for others to make their own design. It would be sad for PJRC bottom line if there was an alternative design that users would kick the door down to buy.
:cool:
 
without checking the actual details:
3.0 to 3.1, everything that works on 3.0 works on 3.1, only expanded capabilities, so no issue.
3.x to LC, platform with less capabilities and some code incompatibility.
Official relese, march this year . Status is that it mostly works real well, some issues do pop up
but it seems the libraries are real stable 3 months after official release. Far from 1.5 to 2 years until usability.

So, what will happen with the "super 3.1++" ?? Who knows, I guess Paul has some idea.
But, to keep the build format, a million new gpio pins are not wanted, but faster internal clock,
more and better peripherials. The only chip in current official portfolio that I can locate is
K26 180MHz, but it is a 144 pin chip, so lots will be unconnected(wasted).

There might be, surely is, other chips in the pipeline. M7 ?? . Probably a lot of the peripherial
controllers will reuse existing IP and be wery easy to adapt to.
 
I do indeed have my eye on the K26 for Teensy++ 3.x.

It has a lot more GPIO than can fit onto a 48 pin form factor, so of course I'll try to bring as many extra pins as I can to bottom-side pads. Maybe then we'll have a thread requesting a 100 pin form factor Teensy? (nearly the entire length of full size breadboard)
 
Well if you used a 3.0 and wired up the reset pin, it would not work the same in 3.1, since the reset pin was moved to make way for the new DAC/A14 pin. But that is a fairly minor difference.
 
I do indeed have my eye on the K26 for Teensy++ 3.x.

It has a lot more GPIO than can fit onto a 48 pin form factor, so of course I'll try to bring as many extra pins as I can to bottom-side pads. Maybe then we'll have a thread requesting a 100 pin form factor Teensy? (nearly the entire length of full size breadboard)

I kinda like the formfactor of these boards:
http://www.schmalzhaus.com/UBW32/
http://www.cypress.com/?rID=108038

Littlebit difficult to take them out of breadboard after plugging them though :)

But on the other hand, the Freedom boards, like http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=FRDM-K64F&nodeId=015210045A how far away from beeing usable with teensycode are these?
I guess they don't have the mini54 chip, but is it possible to bootstrap it otherwise?
 
Status
Not open for further replies.
Back
Top