¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io

Alternative to Teensy 4


 

Greetings, I've purchased the SDRT book and I'm currently waiting on delivery.
In the meantime, I'm hunting around trying to source components I know I'll need, starting with the Teensy 4x.
One thing that strikes me straight away is the steep price of the Teensy compared with other Arduino platforms I'm familiar with (Nano, etc) and that's before you include the sound processing extension.
So, the obvious question would be: has anybody tried to replace the Teensy with a comparable (caveat!) platform, like the STM32 (BluePill)?
The caveat being that from what I understand, the Teensy has more grunt.
Since my plan at this stage is to treat the T41 as a great example of how to build an SDT in general and perhaps actually building something slightly simpler (an 80m SSB mono-bander would do me just fine!) I wonder then if you can loosen some of the constraints (starting with the Teensy).
I may be way off the beam, of course.? Regards.


 

¿ªÔÆÌåÓý

Best you start with the Teensy and work back to the STM32 because there are a number of things the STM 32 will not be able to do that the Teensy does. ?The audio hat is a good place to start.

?

?

Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ VP2EHZ

?

Owner - Operator

Big Signal Ranch ¨C K9ZC

Staunton, Illinois

?

Owner ¨C Operator

Villa Grand Piton ¨C J68HZ

Soufriere, St. Lucia W.I.

Rent it:

?

Moderator: North American QRO Group at Groups.IO.

Moderator: Amateur Radio Builders Group at Groups.IO.

?

email:? bill@...

?

?

From: [email protected] <[email protected]> On Behalf Of picoguy via groups.io
Sent: Tuesday, September 19, 2023 2:32 AM
To: [email protected]
Subject: [SoftwareControlledHamRadio] Alternative to Teensy 4

?

Greetings, I've purchased the SDRT book and I'm currently waiting on delivery.
In the meantime, I'm hunting around trying to source components I know I'll need, starting with the Teensy 4x.
One thing that strikes me straight away is the steep price of the Teensy compared with other Arduino platforms I'm familiar with (Nano, etc) and that's before you include the sound processing extension.
So, the obvious question would be: has anybody tried to replace the Teensy with a comparable (caveat!) platform, like the STM32 (BluePill)?
The caveat being that from what I understand, the Teensy has more grunt.
Since my plan at this stage is to treat the T41 as a great example of how to build an SDT in general and perhaps actually building something slightly simpler (an 80m SSB mono-bander would do me just fine!) I wonder then if you can loosen some of the constraints (starting with the Teensy).
I may be way off the beam, of course.? Regards.


 

First, thanks for buying our book. I hope it's been a useful read.

I've worked with most of the common microprocessors and considered most before I settled on the Teensy 4.1. There are a bunch of reasons why I chose the Teensy. First, it has the resource depth (e.g., memory and I/O pins) necessary to do a "real" SDT. Second, it has the requisite horsepower: a 600MHz clock and floating-point processor. Third, it is supported by a robust library that will do almost all major tasks and I "trust" its libraries. Fourth, the Teensy 4.1 board is well-defined while others are not (e.g., the ESP32 boards have dozens of variations and not all ESP32 libraries work with all boards--indeed, some boards have 32 pins, others have 38). Fifth, it works in the current IDE and PRJC makes a real effort to keep it and the supporting libraries current with the IDE. And sixth, it supports the RA8875 display, which gives us a choice of 5", 7", or 9" display. All of the above combine to give the T41 a feature set that is unlike other SDTs (e.g., we can show a 192KHz bandwidth).

That said, the purpose of the book is to bring much of the DSP theory into one place, discuss it, and show why we elected to use the circuits we do. We've tried to write the book in an "approachable" manner, making the reading a bit easier than most DSP books, but still giving the reader what they need to know. The book, and the T41, are a jump-off point and we encourage experimentation. Indeed, much of the improvements in the T41 over the past two years are from posters to this site. Developments in the STM32, ESP32, and RP families continue to improve their performance stats to the point where they become viable alternatives to the Teensy 4.1,. We hope others pick up the essence of the T41 and continue to run with it.

Jack, W8TEE


 

Hi Picoguy,
?
Based on your post, it looks like you are new to microcontrollers and embedded systems in general.? Welcome!
?
The ¡°BluePill¡± is a Cortex-M3, which has neither hardware floating point operations nor hardware DSP.? The firmware for the T41-EP was born from the Teensy-ConvolutionSDR (DD4WH), which uses a Teensy 3.6 (a Cortex-M4 with hardware multiplication, hardware DSP, and a processor running at 180 MHz) but has support for a Teensy 4 (from looking at the source code).? Checkout:

If you want to stick with STM and hopefully a Cortex-M4 (with hardware floating point operations and hardware DSP), check out the ¡°BlackPill.¡±? You will also have to work out getting the BlackPill to work with the Teensy Audio Adaptor Board or roll your own.
?
The T41-EP firmware has been geared toward the Teensy 4.1 (where ¡°T41¡± comes from according to Jack and Al), which is a Cortex-M7 that has floating point operations, hardware DSP, and a superscalar processor (running at 600 MHz).? If you want to stick with STM and a Cortex-M7, check out the NUCLEO-H7A3ZI-Q, which is priced about the same as the Teensy 4.1, but you will have to work out getting the NUCLEO-H7A3ZI-Q to work with the Teensy Audio Adaptor Board or roll your own.
?
If you want to save a few dollars by using another embedded platform, one will probably have to spend many, many hours porting or adding support to the T41-EP firmware to get it to work with another embedded platform.? If one was planning on making thousands or millions of radios, spending that time may make sense.? If you are just looking for a challenge, more power to you ;)
?
I hope that helps =)
?
Ian


 

Thanks for the advice guys.? You make a convincing argument for the Teensy, so I'll be sticking between the lines on that one.

As an analogue transceiver builder, I'm really looking forward to learning something new.? Also, I'm enjoying reading "Digital Signal Processing A Practical Guide for Engineers and Scientists" by Smith as an accompaniment to your book.? Since I'm in the Lit / Research phase of the project, I might ask if you had any personal recommendations of books that offer intuitive insights into DSP (although it's hard to beat Smith in that regard!)

But perhaps that's the subject of another thread.
P.S. the book has just arrived.? It's going to be a long night!
73


 

Not a book, but an online course:



There are lots of exercises using Python.? This is one of my favorite technical courses.? I didn't take it for credit, only for learning.
This course covers the fundamental basis of DSP.

--
73 Greg KF5N


 

Thanks for the link. It looks like a useful course.

I'm curious as I started a Coursera course a couple of years ago. I won't say which university it was through, but I'm pretty sure it was in Colorado. It was on sensors and motors as in process control, and it was much more detailed than I expected. I'm not exactly a rookie, but it was going to take a lot of study. I reported some problems with equations being wrong (units didn't work out, etc.) and I got no response from the instructors. Several other students reported that quizes had some wrong answers and also got no resposne. I dropped out due to lack of instructor support.

How was your experience with this course?


 

It has been a few years since I took it, but it was top notch instruction.? Also very unique, since the data being analyzed is music.? A refreshing change from the usual dry engineering material.
I had better understanding of the DFT and data sampling than from any other source I had studied before.

--
73 Greg KF5N


 

Having worked with most of the Arduino line and several other microcontrollers, including the Blue Pill - the Teensy has a whole lot going for it, especially
for demanding applications like SDR. Some of the Blue pill brothers have the raw power, but the question is do they have the wide library support that Teensy has? There is also the question of a suitable higher res screen. You need at least the 800x480 pix to show what the T41 has. On the grand scale of things, you would not be saving much if you do want performance. - Note that even RASPI is much more expensive that Teensy 4,1.

A primary requirement of the T41 Sw is support for the ARM DSP library. Other micros do support it, but that is a basic need. Doing a port of T41 SW to the STM32 family would be a pretty large task, IMHO.

Just my 2 cents worth.
Al
AC8GY


 

On 2023-09-20 15:12, Albert Peter wrote:
Having worked with most of the Arduino line and several other
microcontrollers, including the Blue Pill - the Teensy has a whole lot
going for it,
I was wondering if there would be any point to ditching the Teensy and just
using its CPU, the IMXRT1062. Not to save money, but to have less pins, less
board real estate. AND to have genuine JTAG debug. And there are better IDE's
out there...

One problem with that, is that the exact CPU in the Teensy is a BGA ( Ball Grid
Array ). I'm damned good at soldering, but that's where I draw the line.

- Jerry, KF6VB


 

JTAG, or any debugger, would be nice, but don't forget this is for others who are not well-versed in either software or hardware development. The Arduino IDE continues to get better and I do think that someday we will have a symbolic debugger, but I don't see fewer pins or less real estate to be real problems. However, that's not to say that someone else can't do it.

Jack, W8TEE

On Wednesday, September 20, 2023 at 06:37:37 PM EDT, jerry-KF6VB <jerry@...> wrote:


On 2023-09-20 15:12, Albert Peter wrote:
> Having worked with most of the Arduino line and several other
> microcontrollers, including the Blue Pill - the Teensy has a whole lot
> going for it,

I was wondering if there would be any point to ditching the Teensy and
just
using its CPU, the IMXRT1062.? Not to save money, but to have less pins,
less
board real estate.? AND to have genuine JTAG debug.? And there are
better IDE's
out there...

? One problem with that, is that the exact CPU in the Teensy is a BGA (
Ball Grid
Array ).? I'm damned good at soldering, but that's where I draw the
line.

? ? ? ? ? ? ? ? ? ? ? ? ? - Jerry, KF6VB






--
Jack, W8TEE


 

Someone made an interesting FPCB to access JTAG for Teensy 4.1:


It looks like it has to be for a Teensy 4.1 without Ethernet.

I'm sure Paul (pjrc) has read the threads that ask for JTAG access or a "development" Teensy with JTAG access.? Maybe, one day.....

Ian?


 
Edited

I looked at the Debug Mod Project.? ?I believe that the Arduino IDE will no longer be able to work properly, so we'd have to move to Platform I/O, VSCode or eclipse.? Also libraries are added to the project in a different way.? In that way, it becomes more of a tradition embedded project, which I don't mind, but it may be more intimidating for others.

There is something called GDBStub that is supposed to work without hardware modification, but the last time I tried it, I didn't have much luck.

73,
Jack KC7VLO


 

It was not an Ethernet chip that was removed; it was the Cortex-M0+ (IMXRT1062), which is used to flash the Cortex-M7 (MIMXRT1062DVJ6B).? My mistake.? I have not used a Teensy for years, but it¡¯s coming back to me ;)
?
If one removes the Cortex-M0+ (IMXRT1062), the only thing that ¡°breaks¡± is using the Teensy flash application, which is separate from the Arduino IDE, at least on my development platform.? The Arduino IDE will continue to function normally and produce a file that can be used to flash the Cortex-M7 (MIMXRT1062DVJ6B).
?
If one¡¯s skills are up to (successfully) using a JTAG device, one can use the JTAG device to flash the Cortex-M7 (MIMXRT1062DVJ6B) with the file produced by the Arduino IDE.? So, there is no (good) reason to dump the Arduino IDE because some of us want to use JTAG, and anything else, in my humble opinion, would be throwing the baby out with the bathwater.? Further, continuing to use the Arduino IDE will make it easier for folks (new to microcontrollers) to begin to use and experiment with this project.
?
What would be really nice:? if someone(s) would write some JTAG (or some other debug) firmware for the onboard Cortex-M0+ (IMXRT1062).? STM and TI, to name two, have this type of functionality on their evaluation boards.? Well, it¡¯s good to have a dream, I guess ;)
?
Anyway, that¡¯s my $0.02.? Your mileage may vary.
?
Ian