open-source teensy-compatible - what features do you want?

[eugene]

I have used and loved Teensies for many years, however, I have have to admit that in no case did I actually need all of the performance and peripherals. Further, an open-source Teensy seems attractive, but for me to consider it an alternative is would have to have at least the same performance, preferably more. How about some of the newer i.MX RT11xx series MCU's with dual cores?

I have also become fond of the RP2xxx MCU's. And really they're about all I ever really need, but I can't consider them legitimate Teensy compatible devices.

Phil, I am truly sorry that you and sparkfun have had a rift, but I think that you might need to take a step back along with a deep breath and think about what it is that you want to sell. You don't actually NEED to sell Teensies, and you already have a selection of RP2350 options. Focus on your new baby and its mother, and let this issue chill for a while. That's my unsolicited advice.

[EEVblog]

There's another major factor though: shortcutting the effort of getting up and running on a complex microcontroller, including all of the libraries.  One could just use the STM32 and CUBE IDE to do development, but the learning curve is quite high here.  Teensy provides hardware libraries and tons of other useful libraries and it has been designed to get one up and running quickly with the Arduino environment or other simplified development tools.  That is one of its main selling points.  If performance is completely not an issue, then using the standard ATMEGA328P is useful because of its ubiquity and because the Arduino environment comes with completely open source libraries, and I do use it still for many projects because it is adequate.  The Pico 2, while not having the performance of the Teensy, already has an open source ecosystem around it that could still use simplification to be more like the ATMEGA328P-based Arduino which is about the simplest thing you can set up as microcontrollers go.  Adafruit is in a good position to do this.  Given Qualcomm's acquisition of Arduino, this is more than just an academic exercise; this is something that holds Qualcomm's feet to the fire because there would be an easily available alternative in case Qualcomm tries to "monetize" Arduino by, for example, creating proprietary only extensions of Arduino or trying to shove AI down our throats.

Of course, there's nothing stopping Adafruit from using the most advanced and fastest processor available now, but that could have repercussions when such parts are discontinued.  The ubiquity of the RP2040/RP2350 chips makes their continued availability much more likely as well.

FYI, the Teensy creator replied here on the effort required to make the Teensy what it is.
https://www.eevblog.com/forum/chat/sparkfun-splits-with-adafruit/msg6162691/#msg6162691

From what I gather there is no way that Adafruit can easily match this level of support with a clone?

[SiliconWizard]

There's another major factor though: shortcutting the effort of getting up and running on a complex microcontroller, including all of the libraries.  One could just use the STM32 and CUBE IDE to do development, but the learning curve is quite high here.  Teensy provides hardware libraries and tons of other useful libraries and it has been designed to get one up and running quickly with the Arduino environment or other simplified development tools.  That is one of its main selling points.  If performance is completely not an issue, then using the standard ATMEGA328P is useful because of its ubiquity and because the Arduino environment comes with completely open source libraries, and I do use it still for many projects because it is adequate.  The Pico 2, while not having the performance of the Teensy, already has an open source ecosystem around it that could still use simplification to be more like the ATMEGA328P-based Arduino which is about the simplest thing you can set up as microcontrollers go.  Adafruit is in a good position to do this.  Given Qualcomm's acquisition of Arduino, this is more than just an academic exercise; this is something that holds Qualcomm's feet to the fire because there would be an easily available alternative in case Qualcomm tries to "monetize" Arduino by, for example, creating proprietary only extensions of Arduino or trying to shove AI down our throats.

Of course, there's nothing stopping Adafruit from using the most advanced and fastest processor available now, but that could have repercussions when such parts are discontinued.  The ubiquity of the RP2040/RP2350 chips makes their continued availability much more likely as well.

FYI, the Teensy creator replied here on the effort required to make the Teensy what it is.
https://www.eevblog.com/forum/chat/sparkfun-splits-with-adafruit/msg6162691/#msg6162691

From what I gather there is no way that Adafruit can easily match this level of support with a clone?

That's sure a lot of effort, but from what I gathered now, the only thing that seems closed source is the bootloader? (Whereas I previously thought the hardware was "closed" as well.)
If the bootloader is the only thing that one can't copy, the effort to make a clone of the Teensy and support it shouldn't be that big a deal. The bootloader will have to be rewritten and at worst it won't be compatible, but I don't think that's the bootloader that makes the real difference here, the libraries do and they are all open source? Or aren't they?

But if the question is doing the same with a different MCU, then there's definitely a lot of work ahead to get to the same level of support, and then, using a RP2350 or something beefier doesn't make much difference. The low cost and guarantee of availability are probably more the key factor here.

[phil from seattle]

I have a product that is based on the Teensy 4.1.  It sells pretty well - multiple 1000s todate.  I also have a similar product that is based on the RP2350B.  It sells about 1/2 the rate of the Teensy based one.  It has a few better features and yet people prefer the Teensy one because of the performance.

I really like the RP23xx family. In particular, the PIO feature allows it to punch well above it's M33 weight and allows really complex timing signals with super low jitter. Another really nice aspect is the ability to update firmware with a simple drag and drop action - no programming HW or SW needed.  Super customer friendly. (I've got love/hate with teensy.exe.)  And, in-circuit debugging is easy, unlike the Teensy that doesn't have anything in that regard. However, an RP2350 based competitor to the Teensy is simply not interesting to me - it brings little to the table that you can't get elsewhere.

If you want to compete with the Teensy, here's what would get me interested enough to spin another board:
- long term stability of supply/design.  (the various *pills don't cut it, no real standardization)
- high performance: M7 or better, 600 mhz or greater, good to great floating point
- D&D firmware updating
- PIO or similar capability
- 48 pins or greater (ie, lots of I/O)
- in circuit debugging (SWD would be great)
- runs cooler than the iMXRT 1062 (it cranks out >20C above ambient - 26C in one setup I'm testing right now)
- good Arduino core support (like the Philhower RP2xxx core)

[SiliconWizard]

Yes Cortex-M7 MCUs tend to run pretty warm at > about 400 MHz. I used a STM32H7 in a design and it got pretty hot while running as well.

[nctnico]

In the meantime the topic starter went on a holiday or something 

[Kean]

In the meantime the topic starter went on a holiday or something 

Phil and Limor recently had a baby, so I'm sure they have family stuff keeping them busy.

[Simmed]

i tried this question on claude for fun

[AuricG]

In the meantime the topic starter went on a holiday or something 

The topic starter went on all popular websites and forums to market his new product that is (in my opinion) unrelated to Teensy and also not even available. Teensy is just a scapegoat used for blatant marketing.
I'm sure anyone can do the same on Adafruit forum/discord with no repercussions... 

Phil and Limor recently had a baby, so I'm sure they have family stuff keeping them busy.

It's a very busy time for sure and I hope they are doing well and enjoying some family time.
But honestly I have to say that I'm disgusted with what appears to me as an attempt to use his family to generate pity and sympathy... (especially on other sites)
Phil should realize that there are many parents out there fighting the battle of balancing the work life with family life and everything that comes with that.
Please don't use this as an excuse or a free pass for what you (Phil) are doing.

[AuricG]

@ptorrone
Adafruit has been selling Teensy for years now, right? And I'm sure it's not news to you that this is a partially closed project which clearly stands firmly against your "open" ethos.
So why rile up the community now, and why not make a Freensy years ago?
Because as of today; The Teensy is still available and supported, you can buy it, you can make awesome projects with it and do all the things you could yesterday.
The only difference compared to yesterday is that Adafruit can't sell them anymore and make a profit from that, but everything else remained the same, right?

Sure you can say something like "We didn't want to make a competing product to Teensy because we 'love the guys' and we were selling them and supporting the maker community bla bla bla".
But it's not clear to me how this became a huge community problem literally over night and coincidentally at the _exact_ same time when you realized _you_ are no longer able to make a profit. And not when you learned that this product is partially closed source.
Awfully convenient, don't you think?

[PlainName]

open-source teensy-compatible - what features do you want?

Looks to me like you didn't actually want to know since in 12 days you couldn't be arsed to even come and look at the responses, never mind engage with anyone. This post looks very much like:

1. Spreading your anger at Sparkfun across as many people as you can reach.

and

2. Hoping to get more traffic (and spending) at Adafruit.

I had, and have, no skin in this game. I've bought a couple or few things from both companies, but I am very sure that in future I would be unlikely to buy anything from someone with the business acumen demonstrated here.

[PlainName]

@ptorrone

You joined up today and have commented only on this fracas. Strange how this place was never useful to you before now.

[AuricG]

You joined up today and have commented only on this fracas. Strange how this place was never useful to you before now.

That's a very fair point. Maybe I only watch Dave's videos, maybe I was lurking on the forum for years, maybe it's an alt account or maybe I'm just a troll. Maybe all of the above or maybe none... That's for you to decide.
I found this "issue" polarizing enough that I wanted to share my opinion (even though why should anyone really care about it?).

[Simmed]

hi
this is just an idea

why dont you create a contest
"design the next "blablabla-sy"
prize money example 100k ?

conditions example :
1) no more than 3 persons in design group ?
2) you can use ai assisted design, you can be sponsored by PCBway, JLCPCB, mouser, etc
3) physical limit, min xxxMHz, must include x number of min working libraries
4) must have interfaces : USB, bla bla .. ..
5) cost below $xx, must submit by xxx date
6) optional : subject a working sample to some kind of speed/cpu benchmark
7) optional : pass / fail condition, the working sample must do certain types of tasks very well

so on and so forth .. ..

[EEVblog]

You joined up today and have commented only on this fracas. Strange how this place was never useful to you before now.

That's a very fair point. Maybe I only watch Dave's videos, maybe I was lurking on the forum for years, maybe it's an alt account or maybe I'm just a troll. Maybe all of the above or maybe none... That's for you to decide.
I found this "issue" polarizing enough that I wanted to share my opinion (even though why should anyone really care about it?).

That's how most people join the forum really. There is one thing they wanted to comment on or ask.

[Doctorandus_P]

If you want a "teensy compatible", then use the same processor and pin layout.

I don't like the teensy much. I don't care about bootloaders and dislike closed source stuff, but I do have a few teensy's mainly for CNC stuff. One for GRBL and another for LinuxCNC (Remora / Expatria). So I would be looking for software and pin compatibility.

I also have a bunch of STM32 boards. The "blue pills" are dead because of the many fakes (with various bugs and differences I have no interest in figuring out). I also do not have any of the Nucleo boards. I really don't like integrating an extra chip and circuitry on each board just for programming. I do have a bunch of STM32 boards from WeAct. (From small G030 to big F411) Newer versions with more modern processors then the "blue pill" (not pin compatible) but also with CLK and Data pins for the programming.

If you do not care about pin compatibility with the Teensy (4.1 I assume) Then do something like the Raspi uC's, which have a GND pin on the PCB every 5 pins or so. This is good for signal integrity and EMC compatibility. I also like the idea of using a 4 layer PCB with a full GND plane. Especially if the uC has ADC and DAC pins. Apparently the original Maple Mini was a 4 layer design, but I only got to know that name after the Chinese clones had already taken over the Maple Mini boards, and the original company (leaf labs?) no longer made them.

Apparently the Teensy's are strong in audio processing, and there are a lot of libraries around that. I'm not into audio myself, but maintaining compatibility with those libraries (licencing issues?) may be an important factor.

[ramon]

this is just an idea

why dont you create a contest

I really don't see the OP interested to get feedback ... or maybe too busy?

Related to this, I saw that jlcpcb made a 2026 contest through one of its sister/child websites (called OSWHLAB). TOP prize is US$30,000. The website says that one member of the jury is Dave itself.

Would anyone be willing to make some Teensy-like (small size, high performance) board through the contest? With NXP or STM32 >550Mhz variants (like STM32H7. STM32N6, STM32V8)?

Also, the breadboard footprint size (while very convenient for prototyping and appealing for hobbyist) is limiting. I actually would prefer mSATA/M.2 connector (38pin?), or pcie (64 pin) or even SO-DIMM module connector. But others might disagree.

[zapta]

Having a teensy like open source product that doesn't use a licensing chip would be great.

[ebastler]

Having a teensy like open source product that doesn't use a licensing chip would be great.

Unfortunately the only "teensy like" aspects of Adafruit's proposed design were the footprint and the name... But maybe someone else comes up with an open design based on the NXP i.MX RT1062.

[phil from seattle]

this is just an idea

why dont you create a contest

I really don't see the OP interested to get feedback ... or maybe too busy?

Related to this, I saw that jlcpcb made a 2026 contest through one of its sister/child websites (called OSWHLAB). TOP prize is US$30,000. The website says that one member of the jury is Dave itself.

Would anyone be willing to make some Teensy-like (small size, high performance) board through the contest? With NXP or STM32 >550Mhz variants (like STM32H7. STM32N6, STM32V8)?

Also, the breadboard footprint size (while very convenient for prototyping and appealing for hobbyist) is limiting. I actually would prefer mSATA/M.2 connector (38pin?), or pcie (64 pin) or even SO-DIMM module connector. But others might disagree.

Part of the the reason why you don't see iMXRT 1062 competitors to Teensy is the difficulty of a small outfit working with NXP.  Even PJRC had problems during chipagedon because they are relatively small (I'd estimate Teensy 4.1 does <50K units a year).  NXP is not hobbyist friendly.

The M7 STM32 chips are pretty pricey in moderate volume. I regularly consider building with one of them but realize I'd have to price my product higher than the market will bear (based on LCSC prices).  I'm not sure how Teensy 4.1 is so cheap.

[SiliconWizard]

The RT1062 is a few bucks in moderate quantities, it's completely doable to release a board at around $30 which is what the Teensy 4.1 is (Sparkfun retail prices).

The STM32H7 - depends on the model. The H743 is pretty "cheap" when buying from asian distributors.

But even so, Adafruit said they would have about $10 as a target retail price, so that makes it impossible to do this with a Cortex-M7 CPU and be profitable. Whether the $10 tag is reasonable for a Teensy "replacement" is theirs to decide. It's their business. It's a business decision, not a technical one.

[thm_w]

Related to this, I saw that jlcpcb made a 2026 contest through one of its sister/child websites (called OSWHLAB). TOP prize is US$30,000. The website says that one member of the jury is Dave itself.

I don't think a module would win 1st, but good idea and interesting contest. Added a thread here: https://www.eevblog.com/forum/contests/oshwlab-stars-open-source-pcb-design-contest-$85k-prize-pool/msg6172705/#msg6172705

[ramon]

Initially, I thought that it could be easier with STM32H7 to have a lower cost board.

While investigating about i.MX RT (Crossover MCU) series, I found that there is already a low-cost full open source board at US $15.00. The Olimex RT1010-Py.

It's using 500MHz MIMXRT1011DAE5A (LQFP-80 0.5pitch 12x12mm). MCU is US $4.31 single qty and $2.72 @100 PCS. Worth looking.

Sure, RT1062 or STM32H7 will have much more peripherals. But this is 32-bit MCU 500MHz LQFP-80 for less than $3. Amazing.

[Doctorandus_P]

I just had a (very) brief look at /stm32h7r3a8 at TME (in the EU, I think the're in Poland)
The chip costs EUR 5.45 (100+)

600MHz
Ethernet
2x USB
"Graphics functions" (Scaling, rotation, JPEG, LCD controller, Camera interface)

Downside is that it only has 64kB Flash. It's got plenty of RAM (610kB) I guess it would not be difficult to create an eco system around it to put the main program in some external memory chip and load it during booting. This also reminds me of "Beginning STM32 with ..." from Warren Gay ISBN-13 (pbk): 978-1-4842-3623-9       ISBN-13 (electronic): 978-1-4842-3624-6 It has a chapter over overlays (loading software from external memory into RAM. It's an old technique from the "640k should be enough for everyone PC era).

https://www.tme.eu/Document/1427818b7db5b20cd14303ecb79c584a/stm32h7r3a8.pdf

Every now and then I'm looking at STM32 development kits with an "on board" TFT screen, and they seem to be much more expensive then for other uC's.

[brucehoult]

Have you looked at CV1800B and SG2000/2002? Around $6 each in qty 5.

https://arace.tech/products/sophon-cv1800b-5pcs