Call to arms | Teensy + WiFi = true

Do you really need that many ports as seen on the last picture?
Wlan is just something like a tool.
Pico does only need one port as far as I know!
If this is meant for general use, then using only a minimum of ports would be better than top speed!
Sorry, Christof
 
cebersp said:
Do you really need that many ports as seen on the last picture?
Wlan is just something like a tool.
Pico does only need one port as far as I know!
If this is meant for general use, then using only a minimum of ports would be better than top speed!
Sorry, Christof
Click to expand...
Not sure what you mean by "ports"? Do mean pins? As far as top speed usage, that depends on the application and is subjective. The pico is using one SPI port and the Teensy 4.1 is using one SDIO port out of two;)
 
Meant port pins, but it is 4 of them, while there seem to be 8 port pins blocked in the picture above.
Of course this is my very personal opinion, and this is not at all my project. I only chimed in, because I have been one of those, who voted for WLAN, but this is going into a wrong direction for me. More free port pins are more valuable than the additional speed.


1774079263326.png
 
cebersp said:
Meant port pins, but it is 4 of them, while there seem to be 8 port pins blocked in the picture above.
Of course this is my very personal opinion, and this is not at all my project. I only chimed in, because I have been one of those, who voted for WLAN, but this is going into a wrong direction for me. More free port pins are more valuable than the additional speed.


View attachment 39039
Click to expand...
SDIO has 6 pins - 4 data line, a clock line and a command line.
Two additional GPIOs are needed to control the sleep state and an interrupt trigger.

IIRC, the murata chip variant that was used for this project does not support SPI.
Those that do from the w43434 series are discontinued.
This is based on closed conversations we had back in 2024/2025
 
Last edited:
@shawn - I have been playing with QNEthernet and T4_WIFI_CYW4343W. Started off with an initial setup in "driver_select.h":
Code: 
// SPDX-FileCopyrightText: (c) 2024-2025 Shawn Silverman <shawn@pobox.com>
// SPDX-License-Identifier: AGPL-3.0-or-later

// driver_select.h chooses a driver header to include.
// This file is part of the QNEthernet library.

#pragma once

#include "qnethernet_opts.h"

// --------------------------------------------------------------------------
//  External Driver
// --------------------------------------------------------------------------

// First check for the existence of an external driver
// https://forum.pjrc.com/index.php?threads/new-lwip-based-ethernet-library-for-teensy-4-1.68066/post-345539
#if defined(__has_include)
// https://gcc.gnu.org/onlinedocs/cpp/_005f_005fhas_005finclude.html
#if __has_include(<qnethernet_external_driver.h>)
#include "qnethernet/drivers/qnethernet_external_driver.h"
#define QNETHERNET_INTERNAL_DRIVER_EXTERNAL
#endif  // __has_include(<qnethernet_external_driver.h>)
#endif  // defined(__has_include)

// --------------------------------------------------------------------------
//  No External Driver
// --------------------------------------------------------------------------

// Don't include anything else if an external driver has been included
#ifndef QNETHERNET_INTERNAL_DRIVER_EXTERNAL

#if defined(QNETHERNET_DRIVER_W5500)
#include "qnethernet/drivers/driver_w5500.h"
#define QNETHERNET_INTERNAL_DRIVER_W5500
#endif
#if ARDUINO_TEENSY41_CYW4343W 
#include "qnethernet/drivers/driver_CYW4343W.h"
#define QNETHERNET_INTERNAL_DRIVER_CYW4343W
#elif defined(ARDUINO_TEENSY41)
#include "qnethernet/drivers/driver_teensy41.h"
#define QNETHERNET_INTERNAL_DRIVER_TEENSY41
#else
#include "qnethernet/drivers/driver_unsupported.h"
#define QNETHERNET_INTERNAL_DRIVER_UNSUPPORTED
#endif  // Driver selection

#endif  // QNETHERNET_INTERNAL_DRIVER_EXTERNAL

Then added this to the beginning of "qnethernet_opts.h":
Code: 
// Disables T41 native ethernet IF and enables T41 and CYW4343W IF.
#ifndef ARDUINO_TEENSY41_CYW4343W
#define ARDUINO_TEENSY41_CYW4343W 1
#endif
This prevents any conflict with "ARDUINO_TEENSY41" and makes the CYW4343W driver version unique. Set to 0 for native ethernet an set to 1 for CYW4343W usage. I tested both.

I tried going the external driver route but that failed for me.
I then tracked down the initialization sequence and started setting up the QNEthernet driver, "driver_CYW4343W.cpp" and "driver_CYW4343W.h". For right now I want to keep the T4_WIFI_CYW4343W library separate from the QNEthernet library until it is trimmed down to the bare minimum needed.
There is one more layer that has to be added to use the CYW4343W and that is joining the network which also sets up the link status. This uses a poll function in the events class to monitor the state of the link and can be checked any time using "link_check()". I have added evt.pollEvents() to the QNEthernet "driver_poll()" function.

The initialization output:
Code: 
Starting...

[Start]
Starting Ethernet with DHCP...

driver_has_hardware

===========================
CYW4343W Card::begin: SDIO2
===========================
Attaching OOB interrupt to pin 34
Enabled CYW4343W bus high speed interface
BUS_IORDY_REG (Ready indication) returned OK
SDHC bus set to 4-bit, speed set to 33MHz
*************
CardID: 43430
*************
Set SRAM_IOCTRL_REG validated
Set SRAM_IOCTRL_REG second validation
Set SRAM_RESETCTRL_REG validated
Uploading firmware data
Uploaded firmware, 419798 of 419798 bytes
Uploading NVRAM data
Uploaded NVRAM, 680 of 680 bytes
Set BAK_CHIP_CLOCK_CSR_REG validated
Set BUS_IORDY_REG validated
Configuring WL_IRQ OOB
==============================
End W4343WCard::begin: SDIO2
==============================
EthernetClass::start()
enet_init()
driver_init
Uploading CLM, size: 7222
CLM uploaded 7222/7222 result: 1
CLM version API: 12.2
Data: 9.10.39
Compiler: 1.29.4
ClmImport: 1.36.3
Creation: 2020-02-16 22:32:13

Initialization Done
In joinNetworks
WiFi CPU running
Set country succesfully
Join events enabled
Joining network wswn
driver_set_mac
driver_get_mac
driver_get_mac
MAC = 70:87:a7:10:64:b5

********************* LINK Established *********************

********************* Joined Network ***********************

The MAC address is hard coded in the CYW4343W chip and is only available after the chip has been initialized, but as you can see it is working.
There is a lot more to learn and setup in the rest of the driver and I am not sure about how to do it :confused:

I'll keep playing with it:D
The library I am using for the CYW4343W is here...
 
@wwatson thanks for starting that. One note is that the initialization for Wi-Fi might be a little different because there's an extra connection step with SSID's and things like that. I think I'm going to have to modify how that works in the library.

At some point soon, I'll have a look at what an API might look like if Wi-Fi is added.

There's also some assumptions I might be making regarding the order of chip initialization and MAC address detection. I'll have to look into that too.
 
Last edited:
I'd love to help more, but I think I need some hardware. Could someone send me something or tell me what to acquire? (If I don't already have what's needed.) Here's what I already have: (image attached). One board has the chip on one side, and the other board has the chip on the other side; the silver chip is labelled SS2211007 1DX. There's also an antenna. Is this the correct stuff?
 

Attachments

  • IMG_0369.jpg
    IMG_0369.jpg
    164 KB · Views: 37
Last edited:
shawn said:
correct stuff
Click to expand...
Yes, that is the right '1DX' chip. You have TWO variants it seems. One wired to SD Socket adapter - ignore that for now. [this could work but would require altering the code to move to alternate SDIO and adding the INT wire IIRC]

With the Other look to post #73 : https://forum.pjrc.com/index.php?threads/call-to-arms-teensy-wifi-true.77099/post-366611

Photo there and wiring notes should find matching small print names on the silkscreen bottom side with no components.

Code: 
 // wiring used from prior post: Viewed chipside up header connector to the left
INT (34)    WL_ON (33)
CMD (22)    CLK (23)
D0 (40)     D2 (17)
D1 (41)     D3 (16)
GND         3.3v

In that photo is a set of 10 4" jumper wires MALE on one end and Female on the other. Using that Male '1DX' board socket the Female wire ends went to Male pins on an otherwise bare T_4.1.

Progress down the T_4.1 from GND, then 3.3V and so on connecting wires to the indicated pins.

Then one wire at a time connect the Male wire ends into the properly identified Female header pins as indicated.

Carefully attach/Press the Antenna on the small round connector.

With that complete: The @wwatson github code examples should run as indicated in posts on the page prior to #73.

The only added hardware parts you might need for that are a Male pinned T_4.1 and the M<>F jumper wires - shorter is better.
Of course, a Female headered T_4.1 works as well with M<>M set of jumper wires.
 
Last edited:
I think I sent you the one with the SDcard connector, allows you to use SDHC1 via the SDcard slot on a T41 instead of wiring SDHC2 to pins. Just needs one wire in the header to a pin for sleep, IIRC. That's the device I used to get the first one of these things to work. But @defragster seems to have all the info you need for the other one, you have all the hardware you need, assuming you have a T41 and M->F jumpers!
 
I know there's source code in various projects that show how to communicate with the CYW43439, but does anyone know of a protocol document I could read? I'm having trouble finding one. Note that the datasheet only shows a few "setup" and "test" registers.
 
@defragster - Do you still have all of the docs for the CYW4343W/9 that we collected on your GitHub? I am checking mine to see if I forked it from yours...
 
shawn said:
protocol document
Click to expand...
Finding docs isn't easy and the effort not entirely fruitful to date in any way - until @wwatson posted latest!
The WWatson github links to this: https://iosoft.blog/2022/12/06/picowi/
> there are lots of nested links from there that might get you to something usable?
wwatson said:
all of the docs for the CYW4343W/9 that we collected on your GitHub?
Click to expand...
Collection from prior efforts here in DOCS: https://github.com/Defragster/NXP_1062_MurataWiFi
That came from following nested links like above.
 
defragster said:
Finding docs isn't easy and the effort not entirely fruitful to date in any way - until @wwatson posted latest!
The WWatson github links to this: https://iosoft.blog/2022/12/06/picowi/
> there are lots of nested links from there that might get you to something usable?

Collection from prior efforts here in DOCS: https://github.com/Defragster/NXP_1062_MurataWiFi
That came from following nested links like above.
Click to expand...
Thanks @defragster. I found all of it on one of my desktops. It is a huge amount of useful and probably not so useful info. I'll try to get it all in a repo on GitHub...
 
Well, with the help of @shawn I have an initial working version of QNEthernet for the CYW4343x:) I only have the 1DX version of the chip to test with so it will be interesting to see if the 1YN? version works as well. I found the the QNEthernet version of "simplePing.ino" and "ping.ino" would not work as designed with the CW4343W. I determined that I would have to change code deep in the QNEthernet library. That did not make sense as I was able to modify and use the existing ping sketch from the CYW4343W library. I did the same thing with the scan sketch. Both of those are in the QNEthernet examples folder named "cyw43_ping.ino" and "cyw43_scan.ino". The ping sketch resembles the "simplePing.ino" in QNEthernet examples which just prints out the round trip time. Eventually I want to do a ping sketch that prints out more of the ICMP echo info. I have tested most all of the QNEthernet example sketches. Some of them I have no idea what they are trying to do. Those will have to be tested by others with that knowledge.

I have set this version of the library to be able to use wired Ethernet to WIFI by changing 1 define in "QNEthernet_opts.h":
Code: 
// 1 == Disable T41 native ethernet IF and enable T41 CYW4343W IF.
// 0 == Enable T41 native ethernet IF and disable T41 CYW4343W IF.
#ifndef ARDUINO_TEENSY41_CYW4343W
#define ARDUINO_TEENSY41_CYW4343W 1
#endif

To use the examples with the wifi card you must first edit the "QNEthernet/src/qnethernet/drivers/cyw4343w/src/secrets.h" file:
Code: 
// The secrets file

// SSID
#define MY_SSID            "testSSID"

//PASSPHRASE
#define MY_PASSPHRASE      "testPASSPHRASE"

// Security settings: 0 for none, 1 for WPA_TKIP, 2 for WPA2
// The hard-coded password is for test purposes only!!!
#define SECURITY        2
These are used to link and join the network (not part of QNEthernet).

In an effort to be the least invasive I could, there is only two modified QNEthernet files, two added files and one added directory. This makes it easier to upgrade to newer versions of the QNEthernet version updates. I know @shawn does regular updates to QNEthernet so this makes it easy to add in the WIFI portion. At least for now.

I probably will not put much more effort into this until there is more meaningful movement hardware wise and user wise. I have now spent months getting to this point with the CYW4343x software. Will be nice to see others with the various hardware doing some testing. The code is not optimized or clean and will probably remain that way until testing is done.

The library is found here...
 
wwatson said:
Well, with the help of @shawn I have an initial working version of QNEthernet for the CYW4343x:) I only have the 1DX version of the chip to test with so it will be interesting to see if the 1YN? version works as well. I found the the QNEthernet version of "simplePing.ino" and "ping.ino" would not work as designed with the CW4343W. I determined that I would have to change code deep in the QNEthernet library. That did not make sense as I was able to modify and use the existing ping sketch from the CYW4343W library. I did the same thing with the scan sketch. Both of those are in the QNEthernet examples folder named "cyw43_ping.ino" and "cyw43_scan.ino". The ping sketch resembles the "simplePing.ino" in QNEthernet examples which just prints out the round trip time. Eventually I want to do a ping sketch that prints out more of the ICMP echo info. I have tested most all of the QNEthernet example sketches. Some of them I have no idea what they are trying to do. Those will have to be tested by others with that knowledge.

I have set this version of the library to be able to use wired Ethernet to WIFI by changing 1 define in "QNEthernet_opts.h":
Code: 
// 1 == Disable T41 native ethernet IF and enable T41 CYW4343W IF.
// 0 == Enable T41 native ethernet IF and disable T41 CYW4343W IF.
#ifndef ARDUINO_TEENSY41_CYW4343W
#define ARDUINO_TEENSY41_CYW4343W 1
#endif

To use the examples with the wifi card you must first edit the "QNEthernet/src/qnethernet/drivers/cyw4343w/src/secrets.h" file:
Code: 
// The secrets file

// SSID
#define MY_SSID            "testSSID"

//PASSPHRASE
#define MY_PASSPHRASE      "testPASSPHRASE"

// Security settings: 0 for none, 1 for WPA_TKIP, 2 for WPA2
// The hard-coded password is for test purposes only!!!
#define SECURITY        2
These are used to link and join the network (not part of QNEthernet).

In an effort to be the least invasive I could, there is only two modified QNEthernet files, two added files and one added directory. This makes it easier to upgrade to newer versions of the QNEthernet version updates. I know @shawn does regular updates to QNEthernet so this makes it easy to add in the WIFI portion. At least for now.

I probably will not put much more effort into this until there is more meaningful movement hardware wise and user wise. I have now spent months getting to this point with the CYW4343x software. Will be nice to see others with the various hardware doing some testing. The code is not optimized or clean and will probably remain that way until testing is done.

The library is found here...
Click to expand...
Months of work you mention, that is amazing. I am so happy to see that this now works, I need to order WiFi chips, I sent all off to you guys. But I still have 1DX and 1YN in stock so I'll order in a few PCB's so I can test this out myself.

Hardware wise, I think it's good to start discussing what that should look like. an SDCard version is smart for Teensy users. It could also have headers for users that do not want to use the SDCard slot. Any other ideas?

It would be nice if PJRC or Sparkfun could sell modules. I will reach out to Nate at Sparkfun and see what's possible.
 
@Dogebone - Thanks for the input. I do not have the SDcard version so I could not easily test using uSDHC1. I think that using the SDcard version frees up a lot of the pins for I/O use but then again a lot of peoples program usage relies on fast SDcard access. Up till now I have been using a T41 for testing. I would like to revisit the DB4.5 or DB5.0 that I have using the WIFI adapter. Been a long time. Would have to revisit the pinouts and assignments...
 
Very cool @wwatson and @Dogbone06 !!! The ReadMe alone looks like an LARGE amount of work!

Indeed, hardware in short supply. @PaulStoffregen collected a series of commercial variants, not sure where that is in the queue,

Connections documented prior page of this thread as used on the T_4.1 with 'raw' @Dogbone06 board: https://forum.pjrc.com/index.php?threads/call-to-arms-teensy-wifi-true.77099/post-366611

Only the single raw, not the SD adapter, version of hardware here with the 1DX for any testing.
 
defragster said:
Very cool @wwatson and @Dogbone06 !!! The ReadMe alone looks like an LARGE amount of work!

Indeed, hardware in short supply. @PaulStoffregen collected a series of commercial variants, not sure where that is in the queue,

Connections documented prior page of this thread as used on the T_4.1 with 'raw' @Dogbone06 board: https://forum.pjrc.com/index.php?threads/call-to-arms-teensy-wifi-true.77099/post-366611

Only the single raw, not the SD adapter, version of hardware here with the 1DX for any testing.
Click to expand...
You also have a WiFi chip right? Would be nice if you could test it out. It will take weeks before I have one to test myself.
 
wwatson said:
ReadMe is @shawns ReadMe
Click to expand...
Okay - kudos to @shawn for that on QNEthernet then - that explains not seeing any hardware connect or details specific to 1DX :)
Dogbone06 said:
You also have a WiFi chip right?
Click to expand...
Yes, See link in p#96 that hardware is here and was used for early testing Mar 19, 2026.

Hoping the pins/names noted there match one or more of the hardware items @PaulStoffregen has on hand when he gets back to it for easy connection.
Code: 
INT (34)    WL_ON (33)
CMD (22)    CLK (23)
D0 (40)     D2 (17)
D1 (41)     D3 (16)
GND         3.3v
 
@defragster - Sorry about that. That info is in the previous library. I will add it to the QNEthernet_WIFI library ReadMe. For now this is the pinout:
Code: 
 ## PINOUT:
TEENSY 4.1   WIFI Board
- 23 --------> clk
- 22 --------> cmd
- 17 --------> D2
- 16 --------> D3
- 41 --------> D1
- 40 --------> D0
- 34 --------> INT
- 33 --------> WL_ON
 
You must log in or register to reply here.
Back
Top