Forum Rule: Always post complete source code & details to reproduce any issue!
Results 1 to 8 of 8

Thread: Availability of Teensy 3.5 in the future

  1. #1
    Junior Member
    Join Date
    Jun 2018
    Posts
    4

    Availability of Teensy 3.5 in the future

    This is not a technical question, but more of a question about using Teensy in projects that, perhaps, might be still reproduced in ten years time.

    The Atmega328p is underpowered by modern standards, but it has the advantage of being so ubiquitous that I am fairly certain that in 10 years time, they will still be available. I have implemented projects with 32-bit ARM processors (see http://jumentum.sourceforge.net/) and my frustration is that there is such a proliferation of 32-bit processors, that few are likely to be available in the future. Each processor has its own quirks and changing it over to a new one would likely be a large hassle.

    Is the processor on the Teensy 3.5 going to be available into the future? Is there any indication that the company selling it is going to keep producing it? One of the benefits of open-source software is the "recyclability" of its components, and if I put in a big effort, I would like the effort to be good for many years for others who follow. One of the benefits of the AVR is that it is so well established it seems like its not going to go anywhere, like the 8051.

    Thanks,

    Dan

  2. #2
    Administrator Paul's Avatar
    Join Date
    Oct 2012
    Posts
    334

  3. #3
    Junior Member
    Join Date
    Jun 2018
    Posts
    4

    NXP Longevity List

    Quote Originally Posted by Paul View Post
    Thank you for responding to my question. I have seen so many of these 32-bit dev boards in the past 10 years or so, and many of them are not available anymore. I have used the LPC2148, LPC2368, and LPC1768, and the problem is that the support for these chips seems to be pretty brief. The problem for open source hardware I think is that if a chip is not going to be available for a long time, the project keeps having to be ported to new microcontrollers every few years, and each microcontroller has its own quirks and headaches. This is not a problem for consumer electronics but open source hardware, it's as much about the design itself as the product.

    I don't even see the MK microcontroller series at all on the product longevity list. Is this microcontroller series too new to be listed? It seems strange that none of these microcontrollers would be considered important enough to put on the longevity list.

    Dan

  4. #4
    Senior Member
    Join Date
    Dec 2016
    Location
    Montreal, Canada
    Posts
    3,025
    Code:
    Microcontrollers 	Kinetis L 	- 	KL26 	October 2013 (Teensy LC)
    Microcontrollers 	Kinetis K 	- 	K20 	July 2011 (Teensy 3.1/3.2)
    Microcontrollers 	Kinetis K 	- 	K64 	January 2014 (Teensy 3.5)
    Microcontrollers 	Kinetis K 	- 	K66 	May 2015 (Teensy 3.6)
    They're listed.

  5. #5
    Junior Member
    Join Date
    Jun 2018
    Posts
    4

    K64 series in availability program

    I was looking at their web page, and it looks like that the K64 series is in the program. Do you know if that means all of the processors including the MK64FX512VMD12 are going to be produced for 10 years?

    Click image for larger version. 

Name:	longevity-program.jpg 
Views:	38 
Size:	66.3 KB 
ID:	13996

    I can email NXP and ask them but I was wondering if this a consideration when selecting processors for Teensys.

    Thank you,

    Dan

  6. #6
    Senior Member
    Join Date
    Dec 2016
    Location
    Montreal, Canada
    Posts
    3,025
    If thats what they say then yes. I don't think Paul would develop on a platform where it had a short lifecycle.

  7. #7
    Junior Member
    Join Date
    Jun 2018
    Posts
    4
    Thank you for answering my question. This is looking promising. I am working on a project for amateur radio and would like it to be at least possible for ten years for others to built it without substantial modifications.

    Dan

  8. #8
    Senior Member PaulStoffregen's Avatar
    Join Date
    Nov 2012
    Posts
    20,327
    Quote Originally Posted by profdc9 View Post
    wondering if this a consideration when selecting processors for Teensys.
    No, not really. I'm kinda cynical about longevity statements. I mostly ignore them, kinda like buzzword-of-the-day talk.

    I chose Freescale Kinetis back in late 2010 time frame based on its technical merit and pricing (or at least a budgetary quote from a Freescale rep). I got the first samples in 2011, and it wasn't until September 2012 that we could actually buy them.

    I *really* wanted DMA. Back then almost no microcontrollers had it, and the few that did were prohibitively expensive. Even back in 2010 I was dreaming of making the audio library. The M4's DSP extensions were also highly desired.

    Back in those days most of the 32 bit microcontrollers were still ARM7TMDI. As some Cortex-M parts started to arrive, most of the silicon vendors took a very traditional "market segmentation" approach to their product lines. Freescale decided to buck that trend with very foward-looking thinking. The decided to put M4 into every Kinetis chip and put the same DMA engine and the same 16-priority NVIC and other key infrastructure. Their idea was to maximize software portability across the entire product line (which back then was before M0+ existed), whereas pretty much everyone else viewed M4 as only something to offer on their most expensive parts.

    Freescale Kinetis was chosen for Teensy because it was the first (and for years the only) chip to offer powerful DMA and M4 DSP extensions at a good price.

    The software portability worked out very well. Compare with the route Arduino took. Nearly all the engineering time they poured into Arduino Due (SAM3X) did them little good when they made Arduino Zero (SAMD), because Atmel took the market segmentation approach with no effort for compatibility across their 32 bit Cortex-M product lines.

    As far as PJRC's business is concerned, I don't pay any attention to those longevity guarantees. They're meant for risk-adverse "enterprise" customers. The honest reality of making dev boards with substantial independent engineering & development (as opposed to substantially cloning published designs, which applies to most Sparkfun & Seeed & some of Adafruit's boards) is pretty much a world of high risk and about as far from "enterprise" scale as anyone could image! If you're an executive at General Motors, those sorts of assurances give you something to show to the CEO & CFO, who in turn answer to a board of directors and public investors. If Freescale (now NXP) doesn't honor the deal, you've got the purchasing clout to get them to change a major decision, at least temporarily, and in the unlikely event they won't listen you've got an in house legal team! If you're me & Robin with 2 employees and a couple local contact manufacturers, you're about as far away from the world of "enterprise" as anyone can get.

    Still, perhaps these longevity statements are reassuring if you're a small startup. Still, I'd at whether a company's product line is actually selling in good volume to a diverse customer base and it's profitable for them. That's how you can pick out products like PIC, AVR, Kinetis, STM32 that are likely to be around, and products like AVR32 and AVR XMEGA and (maybe) PIC32, and especially products like Intel Currie. When a product line isn't making a huge corporation money, a former longevity statement isn't going to be worth the pixels it's printed upon.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •