Itsyitsy M4 I2C Not Working for Project that works on Feathe

[RocketScientist]

Tried double reset. Nothing. Never used for anything. Took it out and attached headers purely to do this test.

Also ... just did the following:

1. Reflowed all solder joints on first Itsy and on the MCP9808 breakout.
2. Replaced all breadboard wires with new ones out of the package.
3. Redid the MCP9808 test on first Itsy and it again failed the MCP9808 test.
4. Took a fresh Feather M0 out of its package, added headers and tried the same MCP9808 test with same wires and same breakout as just used with the Itsy. It works. See below.

K

[adafruit_support_carter]

Tried double reset. Nothing.

Nothing = what? No LEDs come on at all? Like you were getting for the other Itsy:

1. With original M4 board with LiPo backpack - boots into M4 LED flashing purple. Double reset puts board into M4 green/red state allowing download.

[RocketScientist]

Nothing. Nada. Not a flash. Not a hint of light or life.

Only hint of life I get with second Itsy is when I plug it into USB into laptop ... a quick flash of "white" on Dotstar. Nothing afterwards from anything. Dead as the proverbial Monty Python parrot.

[adafruit_support_carter]

OK, so sounds like something is wrong with that Itsy.

How about the 2nd Itsy. Does it talk to the MCP OK? It's working with the Feather M0, so we know the MCP is OK.

[RocketScientist]

The 2nd Itsy, fresh out of the package does nothing but the white Dotstar flash when powered via USB. Does not program, not functional.

The first Itsy does program, sometimes successfully accesses the I2C bus (at least successfully does the .begin method) but never successfully accesses the MCP which we know works.

We know the wiring is OK. And we know the MCP is OK since both work with the M0. Using the same program.

I think both Itsys are bad somehow and in different ways. I see no other explanation.

K

[RocketScientist]

I am going to take a break for lunch ... and now that I have an M0 configured ... going to test all the I2C devices I have with both the M0 and the sorta working Itsy.

My suspicion is that the I2C logic of the (this?) Itsy is faulty but we'll see.

K

[RocketScientist]

Using same wiring, same Adafruit MPL3115a2 breakout board and same Adafruit test program - works on Feather M0, does not work on my "working" ItsyBitsy M4.

[RocketScientist]

Using same wiring, same software, the SparkFun MLX90640 IR Array breakout board work works on M0, not on the Itsy.

[RocketScientist]

Using same wiring, same breakout, same software succeeds with M0, fails with M4 for Adafruit 9DOF FXOS8700+FXAS21002 breakout.

K

[adafruit_support_carter]

I2C on Itsy M4 works fine in my testing. And *lots* of other previous testing.

The 2nd Itsy, fresh out of the package does nothing but the white Dotstar flash when powered via USB. Does not program, not functional.

This sounds like it might have missed factory programming or some other issue. We can replace it.

The first Itsy does program, sometimes successfully accesses the I2C bus (at least successfully does the .begin method) but never successfully accesses the MCP which we know works.

This is more of a mystery and since you soldered on a LiPo backpack, there's really no way to determine if it arrived like this.

[RocketScientist]

OK. Send a replacement Itsy and let's see if it can work. I will pay for a second to have a backup (are they back in stock?). Delivery speed is a good thing right now - heading to a NASA event next week at Marshall Space Flight Center where I was planning on showing this project off. Right now I don't have an Itsy that works with any I2C sensor (just tried an LSM9DS1 and MS5611 with same result) that works with Feather M0.

I understand and appreciate the mystery. I really like the Itsy M4 ... but the fact that it all these sensors work with the M0 is very frustrating. I have an inventory of about 20 Feather M0 kits (multiple boards per kit) that I intend to replace with an Itsy implementation to fit in a smaller size. Never had problem with the M0 ... and proceeded with PCB layout (and new 3D printed enclosure) for M4 based on success with M0.

So ... quite frustrating.

How do we expedite this?

[adafruit_support_carter]

Let's just go ahead and replace both Itsy's. When you get them, please verify basic functionality before doing anything else. Just upload the blink sketch, change blink rate, upload again. That way you'll know the Itsy's are functional.

Email [email protected] with a link to this thread and your order number(s) and they can send you 2 new Itsy Bitsy M4's (PID 3800). In terms of being out of stock and expedite options, they will also be the ones that can help you with that.

[RocketScientist]

Sent.

Will do on checking them out before touching them otherwise.

Thank you for help.

I hope the mystery clears up.

K

[bcochran1]

I have not had any problems with my BME280 sensor on a brand new ItsyBitsy M4 Express board. Here is the procedure I followed.

I do this each time I receive a new ItsyBitsy or other "Express" series board.

• Take the board out of its packaging immediately upon receipt, and connect it to my MacBook Pro laptop over USB.
• A small hard drive should show up on my MacOS desktop shortly after connecting the board. This hard drive should be labelled "CIRCUITPY".
• Double-tap the reset button on the board. I should get a hard drive labelled "[PRODUCT]BOOT". For example, "ITSYM4BOOT".
• Check the bootloader version on this board. If it is not up-to-date, follow the procedure for updating the bootloader to the latest released version.
• Start the Arduino IDE, select the board as per instructions in the tutorial, and try out the blink sketch. It should upload and cause the red LED to blink.
• Change the blink sketch in a simple way like increasing the time the LED is on. Upload it. This should work.

If the above works as expected, then I solder headers on the board. For ItrsyBitsy parts, I like to have male headers on the long sides so I can plug the device into a breadboard, and female headers for the pins 2 through RX on the short side.

I've never had a bad board from Adafruit. They are rock solid. I do have a fair number of parts from past orders. Not just ItsyBitsy devices.

In this picture, an ItsyBitsy M4 Express is connected to a BME280 sensor in an I2C configuration. I followed the tutorial for the BME280 as nearly as possible. This sensor is brand new: I received it earlier this week. For purposes of testing the BME280, I power the ItsyBitsy with a USB cable. The BME280 is then powered with 3.3v sourced from the ItsyBitsy. GND, SCK and SDI on the BME280 are also connected per the tutorial for I2C connections. The Arduino example code provided for BME280_Test seems to run just fine in this configuration. Although the test code doesn't say so explicitly, it defaults to an expected I2C hardware configuration.

Note the short connections. Note that no part of the sensor board other than the circuit connections is in physical contact with the ItsyBitsy board. My guess is you have everything jammed together in your "S4Egg" configuration.

The "BAT" pin on the Itsy is connected to give me the option of using an external 5v power source, but it doesn't make a lot of sense. I really need a USB cable connection since I need the Arduino IDE's serial monitor to be active so I can see the sensor output. You know -- when trying to add an SD card capability to the ItsyBitsy your total package is probably more bulky than the corresponding Feather AdaLogger board you are trying to replace.

Thanks so much

Bob

bme_280_on_itsybitsy_m4_small.jpg (157.39 KiB) Viewed 128 times

[/color][/size]

[bcochran1]

....have an inventory of about 20 Feather M0 kits (multiple boards per kit) that I intend to replace with an Itsy implementation to fit in a smaller size. Never had problem with the M0 ... and proceeded with PCB layout (and new 3D printed enclosure) for M4 based on success with M0.

Your ultimate purpose seems to be to shoot these parts into space using a rocket. The goal is to record environmental information such as temperature at high altitudes. My personal opinion is that the ItsyBitsy platform is unlikely to work for you. When you add in separate battery connections, a microSD card breakout board for data logging, the battery itself, from 1 to n sensor breakout boards, and a micro SD card, and add material to protect these parts from electrically shorting each other out and also protect them from shock damage, I think you will end up with more bulk and greatly increased risks of parts not working. Then throw in the gravitational forces involved in shooting these parts in the air and I think something is sure to break. Physical damage to a lipo battery as from the stress of launching or a crash landing invites a fire which will at the very least destroy all the parts...not to mention the other things a flaming battery can destroy. To summarize, I don't see how an Itsy Bitsy with a large number of additional parts and wire connections can survive a rocket launch into space and subsequent landing. The Feather platform is more of a "possible" because it includes parts for data logging and battery operation, but still would need very extensive physical protection and testing. I've never done this type of rocketry work and have no expertise. I'm expressing doubts about the viability of the ItsyBitsy for your intended use.

Thanks so much

Bob