I2C / Wire not reliable

[extrapixel]

after the other Wire/i2c issues have been closed ( #166 etc), and the speed got better, some I2C devices started working with the 101.

However, it's still far from really working. Basically, in practice I2C is not usable on the 101 - it's slow, unreliable. Code / hardware that works with all other Arduinos, does not work with the 101.

I have multiple I2C devices here which work on all official Arduinos, except the 101.

BTW I'm using the latest code from this repo.

We're the manufacturer of the Tentacle Shields, these are isolation shields for sensor circuits made by Atlas Scientific. E.g. the pH sensor circuit: https://www.atlas-scientific.com/_files/_datasheets/_circuit/pH_EZO_datasheet.pdf

Schematics of the shield mentioned above is here: https://github.com/whitebox-labs/tentacle-mini-oshw/blob/master/hardware/T2_mini_schematic.pdf

@bigdinotech your fix did help somewhat, but unfortunately Wire is still unreliable. If someone's working on this, I can send i2c devices to test, if needed.

[intel-larry]

Hi Patrick, I'm sorry to hear we're having issues interfacing with your devices. Of course, we do want everything to work seemlessly. I've personally tested a lot of I2C devices, ranging from various sensors to RTC boards and LCD displays and generally, they've been working. I looked on your website for example code and it looks mostly like plain-jane stuff that I'd think should work.

In 30 years of working with I2C, I can say, that it can be tricky to get things working together sometimes. We would really like to see your devices and would greatly appreciate a few samples so we can work on them in our lab.

[extrapixel]

@intel-larry thanks for looking into this, greatly appreciated! Could you use the contact form here https://www.whiteboxes.ch/contact/ to send me your shipping address?

[eriknyquist]

@extrapixel done!

[eriknyquist]

Hi @extrapixel, I received your I2C devices (and a lovely handwritten note!) this morning. Do you have a sketch that you could share with us, which demonstrates the performance difference between an Arduino 101 and an Arduino Uno?

[PaulStoffregen]

@eriknyquist - Did you see this? #166 (comment) (if this link doesn't take you right to my posts, scroll around looking for a photo of stack of various Arduino boards... it's the only actual photograph on that mammoth-length conversation which isn't a screenshot or diagram)

In the simple test I ran, Arduino Uno and other boards took approx 1.5 ms. Arduino 101 was ten time slower, due to some sort of stall in the Wire library. I posted the complete code I used for testing on all boards. You can see 4 oscilloscope screenshots I posted, showing the long delay. @mjs513 also posted several tests with another chip.

I'll make the same offer for you that I made for @bigdinotech and @piroozn - just email me your shipping address and I'll send you the hardware for free. You can just plug that shield into your Arduino 101 or any other Arduino board and run the simple sketch to measure the time taken to read 12 bytes from that EEPROM chip. Truly, that shield is just sitting here collecting dust and precious bench space, so you'd practically be doing me a favor taking it off my hands! If you want it, just email me directly (paul at pjrc dot com) and I'll put it in the mail to you - anywhere in the world the US postal service can send.

[eriknyquist]

Absolutely @PaulStoffregen, send us that shield, extra testing hardware is always appreciated! I'll send you an email with a postal address.

In this particular case, @extrapixel has also sent us some specific hardware, and it would help me greatly if he could provide a sketch for his hardware that demonstrates the slowness.

[extrapixel]

@eriknyquist yes, we've got some code. It's not made especially for testing/comparing, but it should demonstrate the issue clearly. There's the "Tentacle setup sketch" to interactively talk to the stamps (https://github.com/whitebox-labs/tentacle/tree/master/examples/tentacle-setup/tentacle_setup). That code is probably a bit convoluted due to the auto-detection routines, compatibility with UART devices etc (it's meant to be easy for the "user", not easy to understand code).

Then there is the "continuous reading" example which is kind of basic, but might need some adjustments of i2c addresses etc (https://github.com/whitebox-labs/tentacle/tree/master/examples/continuous/i2c_continuous)

If the above is too much a hassle, I can create some simple test case for you, but I can do this tomorrow earliest.

[PaulStoffregen]

@eriknyquist - It's going in the mail today. You should have it by the weekend. You can find the simple self-contained test program on #166 (comment)

[eriknyquist]

Brilliant, thanks @extrapixel and @PaulStoffregen

[mjs513]

I believe @bigdinotech already has a Mpu6050 chip accrording to tje last post on. isssue 166. If you need the libs and sketch let me know and I can post them. If you want the 6050 the me know. I think I have an extra one laying around. Pretty much gave up on getting the i2c issue fixed.

Mike

[PaulStoffregen]

@eriknyquist - Did the EEPROM test shield arrive safely?