I2C

[Brain Higgins]

Will I2C code run on a Ardunio pro micro controller board or does it need Arduino Uno and a computer ?

Brian
Sent from my iPhone XX

[andy]

The Hookup diagram seems to show a SDA and SCL on pins 2,3.

https://cdn.sparkfun.com/assets/f/d/8/0/d/ProMicro16MHzv2.pdf

The Wire documents talk about Leonardo.

https://www.arduino.cc/en/reference/wire

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/7FEAEC2C-F30E-4713-B9CD-041A91671023%40comcast.net.

[Richard Chycoski]

Hi, Brian:

I haven't used the Arduino pro micro controller myself, but according to
this:

https://learn.sparkfun.com/tutorials/pro-micro--fio-v3-hookup-guide/hardware-overview-pro-micro

pins 2 and 3 are SDA and SCL.

- Richard, VE7CVS

[Brain Higgins]

Cool, that’s big info in my quest

Sent from my iPhone XX

> On Jul 24, 2020, at 5:46 PM, Richard Chycoski <ve7...@gmail.com> wrote:
>
> Hi, Brian:

> --
> You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/84ed18d4-d308-3440-8fbd-bc40cde6076a%40gmail.com.

[Chris Albertson]

You do have to match voltages.   Both sides need to be 5 volts or both sides 3.3 volts.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/F818874F-F68D-484D-BAC1-61701802D527%40comcast.net.

--

Chris Albertson
Redondo Beach, California

[Brain Higgins]

The I2C diagram showed a capacitor instead of a resistor on the PWM

Sent from my iPhone XX

On Jul 24, 2020, at 11:12 PM, Chris Albertson <alberts...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CABbxVHvAfAFyn1ew2ExsiN0be%2BwWDT2wX1Jq-Fau79AOhYp3Tg%40mail.gmail.com.

[Mark Johnston]

For cases where you need to run I2C with one side 3.3V and the other side 5V I have used a few different types of level converters.

Refer to:   https://www.adafruit.com/product/757

I have also used the 2N7000 based converter for same but the little boards with converter on them are easier to wire up (less wires and thus less possible mistakes).

If you want to test your basic knowledge of how n-channel mosfets work look for   '2N7000 level shifter' for a nice diversion.  Fiddlerbot used that approach but droidbot uses the simpler Adafruit (or ebay) 4 line board.

Mark

[andy]

Hey Brian,

There seems to be a note in the Garmin sketch for the I2C.

(Capacitor recommended to mitigate inrush current when device is enabled)

680uF capacitor (+) to Arduino 5v

680uF capacitor (-) to Arduino GND https://github.com/garmin/LIDARLite_Arduino_Library/blob/master/examples/v3/GetDistanceI2c/GetDistanceI2c.ino

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CBD0FEDD-6316-443D-9CBB-CE7DF7C074CB%40comcast.net.

[Ralph Hipps]

I have a Pro Micro but haven’t used it much, does show I2C on the pinout. I did have some odd issues getting it to launch programs, not sure why yet.

You might consider the new Arduino Nano 33 BLE, much faster, more RAM & flash, integrated IMU.

These are level shifters I have used extensively:

https://www.sparkfun.com/products/12009

[andy]

Hey Brian,

It looks like from Hackaday that it is a bit more involved to debug the I2C problems.

In addition to software debugging the hardware debugging might need an oscilloscope.

https://hackaday.com/2016/07/19/what-could-go-wrong-i2c-edition/

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/7FEAEC2C-F30E-4713-B9CD-041A91671023%40comcast.net.

[Chris Albertson]

On Sat, Jul 25, 2020 at 5:07 PM andy <aj48...@gmail.com> wrote:

Hey Brian,

It looks like from Hackaday that it is a bit more involved to debug the I2C problems.

In addition to software debugging the hardware debugging might need an oscilloscope.

Not a 'scope, but a Logic Analyser.  These will decode the I2C signal and let you real what is on the wire.  These can cost under $20 and are well worth it.  In fact debugging is very hard without one.

The "Salese Logic 8" will do it.  "Real"ones are expensive but Chinese clones cost under $10 and work as well

Here is an example.

ebay.com/itm/USB-SALEAE-24M-8CH-Logic...

Software is available in the Saleae web site

[Brain Higgins]

Cool a new device just ordered it

Sent from my iPhone XX

On Jul 25, 2020, at 6:21 PM, Chris Albertson <alberts...@gmail.com> wrote:



--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CABbxVHt%3Dv1s8BDqDVvguWX%3DXLJzvRUyUr0dRN-%3DmFWOQm8ia3g%40mail.gmail.com.

[Chris Albertson]

You need the software.  It runs on Linux/mac/Windows and also has a simulation mode where you can play with it with no device connected to it.   It might take an hour to learn to use.   What you get is a very good digital 'scope but with the limitation that it ONLY works on digital signals.

The software can do things like measure the duty cycle and frequency of a PWM signal and show you the data on a serial pin.

In fact a good debug technique is to write debug data to an unused serial pin, then figure you can always look at it with a logic analyzer

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/66ED67E0-7B7C-424C-911A-9B817D3E83B7%40comcast.net.

[andy]

Yes, Chris is absolutely right I was thinking in terms of the Hantek that

 has a usb o-scope and logic analyzer.

[gmail only gives me 60 seconds to roll back 😀]

https://support.saleae.com/tutorials/example-projects/how-to-analyze-i2c

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CABbxVHt%3Dv1s8BDqDVvguWX%3DXLJzvRUyUr0dRN-%3DmFWOQm8ia3g%40mail.gmail.com.

[Mark Johnston]

I have a Rigol that does I2C decoding but for Brian right now I would try a couple 'best guesses' and it may start working

The way the address is specified can be confusing.   A real I2C address is 7 bits long but often you see it as an 8-bit hex value which is basically shift left one.

So tip #1:   Read your data sheet carefully for the part and know if your driver wants to see a 7-bit address specified or an 8-bit in the call to do the read or write with the driver.

If you convince yourself on the datasheet the address is 0x20 in 7-bit form and you have no clue what the driver expects try address 0x20 and also try 0x40

Messed up address is super common and really easy to do.

2nd tip is I2C   lines of   SCLK and SDA both require a pullup.  If you have no pullup that is your problem for sure.

Tip 2.1  is to realize the actual pullup used can vary.   Try values about 4.7k first but then try 2.2k as lowest pullup

Tip 3 is if you have a scope just look at the signals of SCLK and SDA and if they are not 'reasonabally' fair rise time your pullup is too high.    

This is a little tricky.

My point here is do not blindly trust some logic analyzer.   I2C is impacted by RC so if there is too much line capacitance in any way that will slow the edge way too much.

I was not reading the earlier comment on some sort of capacitor but if your circuit shows any sort of capacitor from ground to either SCLK or SDA it is a kludge to hide use of super long leads and bad transmission line.

Do your tests to get your I2C chips working with the SCLK and SDA lines max of 6 inches long.  After your SOFTWARE works then move on to debug transmission over longer lines.  Basically try like heck to never run I2C over very long lines.

Good luck,

Mark

On Friday, July 24, 2020 at 4:30:18 PM UTC-7, Brain Higgins wrote:

[Brain Higgins]

What I’m trying to do here is convert the code of PWM to I2C.  Which may not be that easy.  On the pro Micro I’m driving a piezo speaker and a lily pad vibrator. I’m not sure a what range the piezo starts to buzz. And same with vibrator. 

Brian

Sent from my iPhone XX

On Jul 25, 2020, at 11:24 PM, Mark Johnston <mjst...@gmail.com> wrote:



--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/8f84aed1-491c-403e-87ca-13a9edea5de2o%40googlegroups.com.

[Chris Albertson]

The Saleae unit is only $8 shipped.     I have a rigol scope that decodes i2c also.   For looking at digital lines the low cost analyzer is a little bit easier to use and will capture data over a longer period of time.  Then you can scroll around and zoom over many seconds' worth of data.   

The larger Saleae analyzers can't be cloned so easy and are expensive but the 8-pin version is made in China

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/8f84aed1-491c-403e-87ca-13a9edea5de2o%40googlegroups.com.

[andy]

Hey Brian,

I have not used the Lidar Lite 3 with I2C but looking at the Github code

 it does not look bad.

Depending on the current code compare this to the Garmin example code for

 the PWM:

// https://github.com/garmin/LIDARLite_Arduino_Library/blob/master/examples/v3/GetDistancePwm/GetDistancePwm.ino

void loop()

{

pulseWidth = pulseIn(3, HIGH); // Count how long the pulse is high in microseconds

// If we get a reading that isn't zero, let's print it

if(pulseWidth != 0)

{

pulseWidth = pulseWidth / 10; // 10usec = 1 cm of distance

Serial.println(pulseWidth); // Print the distance

}

}

// end of loop

Perhaps encapsulate the other devices in functions and pass a Distance value.

There probably is some added complexity to using the I2C so that must be considered.

The example code for I2C is not bad and looks like this:

// https://github.com/garmin/LIDARLite_Arduino_Library/blob/master/examples/v3/GetDistanceI2c/GetDistanceI2c.ino

void loop()

{

/*

distance(bool biasCorrection, char lidarliteAddress)

Take a distance measurement and read the result.

Parameters

----------------------------------------------------------------------------

biasCorrection: Default true. Take aquisition with receiver bias

correction. If set to false measurements will be faster. Receiver bias

correction must be performed periodically. (e.g. 1 out of every 100

readings).

lidarliteAddress: Default 0x62. Fill in new address here if changed. See

operating manual for instructions.

*/

// Take a measurement with receiver bias correction and print to serial terminal

Serial.println(myLidarLite.distance());

// Take 99 measurements without receiver bias correction and print to serial terminal

for(int i = 0; i < 99; i++)

{

Serial.println(myLidarLite.distance(false));

}

}

// end of loop

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/29D7CE80-E9B0-4A89-AE67-9B3E92E7E59C%40comcast.net.

[Brain Higgins]

Then I will need to add write to pin, whatever the pin for the piezo speakers hooked to, or  read / write however that works. And the same for the Lily pad vibrator

In my slideshow Wednesday night I have a picture showing when I had this hooked up to using I2C with a capacitor and a bread board and Arduino Uno, and a computer gave distance readings.

Brian 

Sent from my iPhone XX

On Jul 26, 2020, at 7:58 PM, andy <aj48...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAJZ%2Bfr1edLBiMPUf-1jOzMgHqVLUTVuTP%3DgPe0DeL%3DOXNDQL9g%40mail.gmail.com.

[Rafael Skodlar]

Addressing clarification is needed here. I2C device address is the starting point to communicate between the master/slave units.

Based on this page https://i2c.info/i2c-bus-specification

"Each slave device on the bus should have a unique 7-bit address. The communication starts with the Start condition, followed by the 7-bit slave address and the data direction bit. If this bit is 0 then the master will write to the slave device. Otherwise, if the data direction bit is 1, the master will read from slave device. After the slave address and the data direction is sent, the master can continue with reading or writing. The communication is ended with the Stop condition which also signals that the I2C bus is free. If the master needs to communicate with other slaves it can generate a repeated start with another slave address without generation Stop condition. All the bytes are transferred with the MSB bit shifted first."

Further down on that page:

"10-bit addressing can be used together with 7-bit addressing since a special 7-bit address (1111 0XX) is used to signal 10-bit I2C address. When a master wants to address a slave device using 10-bit addressing, it generates a start condition, then it sends 5 bits signaling 10-bit addressing (1111 0), followed by the first two bits of the I2C address and then the standard read/write bit."

As far as wiring goes, I would use twisted pair which limits crossover talk.

Rafael

On Sat, Jul 25, 2020 at 11:24 PM Mark Johnston <mjst...@gmail.com> wrote:

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/8f84aed1-491c-403e-87ca-13a9edea5de2o%40googlegroups.com.

--

Rafael

[Chris Albertson]

On Sun, Jul 26, 2020 at 10:29 PM Rafael Skodlar <kc6...@gmail.com> wrote:

As far as wiring goes, I would use twisted pair which limits crossover talk.

Which two wires would you twist together?  Twisting SDA and SLC would be a bad idea.   Twisting works with differential signals but I2C is not normally differential.

[andy]

Hey Brian,

Shout out to people like Chris & Mark that have shared their expertise on I2C.

I am also learning along and will hook up my accelerometer with I2C to the

 Arduino soon.

Have a decent book that goes into hardware theory for the RPi but was browsing

 the internet and found this book has a decent chapter on Arduino I2C.

I may not cover all of the corner cases for your particular case so I defer to others

 whether this is a buy.

Exploring Arduino: Tools and Techniques for Engineering Wizardry 2nd Edition, Kindle Edition

On Sun, Jul 26, 2020 at 3:08 AM Brain Higgins <see...@comcast.net> wrote:

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/29D7CE80-E9B0-4A89-AE67-9B3E92E7E59C%40comcast.net.

[andy]

Under the Github for LLv3 there is a document about a Custom I2C used in newer versions:

https://github.com/garmin/LIDARLite_Arduino_Library/blob/master/doc/LLv3-I2C-Guidelines.pdf

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/7BBA98CF-62B0-4A36-A0E2-FD86E5F0B41A%40comcast.net.

[Ralph Hipps]

here's another low cost option:

https://www.sparkfun.com/products/12942  

[Mark Johnston]

As if I don't have enough test equipment to fill a small truck I just HAD to get one of those Salese Logic 8

Seems to have it's uses and is very small to store so what they hey!    Good tip Chris

I would like to find extended software to run on my DS212  really handy dual trace pocket scope that could also do I2C decode but that too is not really 'required' but I bet it does or will exist someday soon.

There is one by same people called something like LA104 or similar but that is separate unit.

Mark

[andy]

Hey Mark,

I like the Salese Logic 8 too because of the elegance.

Found a link on the Seed Studio about I2C and SPI:

https://www.seeedstudio.com/blog/2019/09/25/uart-vs-i2c-vs-spi-communication-protocols-and-uses/comment-page-1/

The seem to have an LCD kit that breaks out the signals to a little screen:

https://www.seeedstudio.com/I2C-Driver-Adapter-Easily-Driver-I2C-Devices-p-4022.html

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/511fd350-3e1a-44ec-ba12-a192f9df1764o%40googlegroups.com.

[andy]

I also could not resist the value of a SL 8. Some of the blogs indicate that the Hantek

 works for periodic signals in the o-scope and logic analyzer mode. But the triggering

 of one shot signals can be problematic. The quality scopes often have a plugin to

 capture a one shot. The blogs mentioned the sigrok software along with SL 8

https://sigrok.org/wiki/Getting_started_with_a_logic_analyzer

On Mon, Jul 27, 2020 at 9:01 PM Mark Johnston <mjst...@gmail.com> wrote:

[Rafael Skodlar]

One wire in the twisted pair can be used as a shield. On the other hand you can use more advanced circuit for longer distance where needed

https://www.nxp.com/documents/application_note/AN11075.pdf

https://www.hackster.io/chipmc/extend-the-reach-of-your-i2c-sensor-simply-and-inexpensively-ea6b53

That is more critical for IoT projects where the sensors are relatively far from processing systems.

If not shielded all wires pickup noise from electro-magnetic sources which interfere with data signals. On the other hand they transmit also so that needs to be taken into consideration one way or the other.

Rafael

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CABbxVHuzMZZ869-eF2c%2BTvfkZisJYoKKNicnac6bFP-c2mnnOA%40mail.gmail.com.

--

Rafael

[andy]

Hey Rafael,

They also have shielded 2 wire in 22 to 26 awg.

Would the 2nd wire affect the capacitance or inductance ?

I wonder if shielded coaxial would work too.

https://hackaday.com/2017/02/08/taking-the-leap-off-board-an-introduction-to-i2c-over-long-wires/

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAB4wyt0EPjKBH5RN4f9Dnf_RRtf8rNXogyt6g1wngyUH%2BpfLhw%40mail.gmail.com.

[Richard Chycoski]

A twisted pair only works well when driven by a balanced driver, which is what the PCA9605 does. You aren't using the second wire as a shield, you drive the pair of wires with a balanced signal (each in opposite polarity), which helps it ignore outside noise (and reduces noise emission as well).

https://en.wikipedia.org/wiki/Twisted_pair

Just using a twisted pair without a proper driver will not likely extend your range and will not give you any real noise immunity.

Shielded wire has varying characteristics - foil shields mess up higher frequencies (due to mutual transconductance), and you do have to consider capacitance between the inner conductor and the shield. If noise is your problem, low capacitance shielded cable may extend your range. RG62 coax is one of the lower capacitance cables that you can probably find around (IBM computer terminals used a lot of it), RG6 is not too bad (commonly used for CATV and SATTV), RG174 is somewhat higher in capacitance but is small and flexible, and will be better than using an audio cable.

If you really want to extend your range, do get a driver like the PCA9605 and use the twisted pair cable that they recommend. CAT5e is what the PCA9605 application note mentions, at the frequencies involved in I2C I would suggest that CAT5 or better would work just fine.

If you just need to extend your range and/or reduce noise issues a bit, try a good quality RF coax cable first.

- Richard, VE7CVS

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAJZ%2Bfr3G4Cw6xu2WtdtVBZKavmMuJxO-OaOzah48j3soAeBJZA%40mail.gmail.com.

[Rafael Skodlar]

Andy,

Second wire will make a difference but I can't tell for sure what kind. The best thing is to try and use an oscilloscope to see the signal. AM radio receivers are also good for checking  electric noise. Sometimes just bringing your hands close to the wires changes the signal because your body is a moving capacitor :-)   Remember musical instrument Theremin? Clever people wrote computer programs to play music (computer electric noise) on AM radio :-)

I would not use a coax cable for I2C because of impedance and length size issues. Cat-5 is tempting but I haven't tried it for I2C yet.

For RS-232 connections I made an adapter with DB-9 to RJ-45 connectors. That way I can use Cat-5 or such cables to connect devices for troubleshooting purposes.

RS-232 can be looked at as "voltage driven", similar to I2C. As others have mentioned, you need special drivers for twisted pairs. That is "current driven" so voltage drop doesn't matter that much. In mainframe computer days we used to have a 60mA current loop connections for remote terminals in places where RS-232 was "not long enough". We learned soon that 60mA adapters were very sensitive to lightning.

You can't always carry around special cables if you work as a field service engineer. That's not my current occupation (unfortunately) but an old habit based on experience. I tend to think this way, let's say you are an astronaut away from industrial supplies and you need to maintain equipment to survive. Tools and other things need to be multipurpose as much as possible.

Fibre optics is another technology that could be used more in robotics IMO. Cables and laser diodes in connectors are becoming common in home multimedia boxes nowadays. Fibre optics cables are electrically isolated, flexible to a degree, and could be used for fast data transfer between IoT sensors etc. in robotics. If I'm not mistaken, there are cables with fibre optics and power wires embedded together but I don't have sources for this.

Sorry for straying far away from the subject matter.

Rafael

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAJZ%2Bfr3G4Cw6xu2WtdtVBZKavmMuJxO-OaOzah48j3soAeBJZA%40mail.gmail.com.

--

Rafael

[David Gustavson]

And note that you're not twisting the clock with the data, you're twisting the clock with its complement and the data with its complement.

-- 

  David Gustavson

  d...@SCIzzL.com

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/d178c89d-2026-95c8-91b8-1c6dbf7f3586%40gmail.com.

[Ralph Hipps]

If you're not sending the I2C a long ways, like 50 or 100 feet or longer, the balanced approach might not be necessary.

You can twist a grounded wire around the clock & another around the data to provide a low inductance return path and a bit of shielding. So a twisted pair for each.

I've used this approach for other single ended signals pretty successfully up to 1 MHz and 8 feet of twisted pair with zero data errors. I don't know how much faster & further it might go, that's the upper limit of my testing.

Certainly not as effective as a differential/balanced approach, but lower cost and may be good enough for your needs depending on rate & distance.

To unsubscribe from this group and stop receiving emails from it, send an email to hbrob...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CABbxVHuzMZZ869-eF2c%2BTvfkZisJYoKKNicnac6bFP-c2mnnOA%40mail.gmail.com.

--

Rafael

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.

To unsubscribe from this group and stop receiving emails from it, send an email to hbrob...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAB4wyt0EPjKBH5RN4f9Dnf_RRtf8rNXogyt6g1wngyUH%2BpfLhw%40mail.gmail.com.

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.

To unsubscribe from this group and stop receiving emails from it, send an email to hbrob...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAJZ%2Bfr3G4Cw6xu2WtdtVBZKavmMuJxO-OaOzah48j3soAeBJZA%40mail.gmail.com.

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.

To unsubscribe from this group and stop receiving emails from it, send an email to hbrob...@googlegroups.com.

[Chris Albertson]

I think if you are running I2c for more then about 18 inches you are abusing the technology.   It was not designed to ever leave a PCB, hence there is no standard I2C connecter.  It was designed to stay on a single PCB.

If you must jumper I2C between PCBs make sure bithe boards are the same voltage.  Both 3.3 or both 5.0 or you need to translate.  If noise is an issue SLOW the rate.  It is always best to run now faster than needed.

For cabled signals today all modern cables are differential with twisted pairs.  Ethernet, HDMI and USB are all much faster then any of the older technologies and very resistance to noise.

In any case, twisted pairs to be effective at all need equal current in each wire.  So it is pointless to try and use twisted pair for I2C or RS232,  Sheilds work for these.  

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAB4wyt032MUu%2BjN9W15E7-fjqFfkCrpThQ3xJQaRjE72CJ6cug%40mail.gmail.com.

[Ralph Hipps]

I should point out this was not I2C protocol or open drain signaling, so your mileage will vary, certainly should be carefully tested.

[David Hough]

If you use cat-5, use two twisted pairs, SDA/GND and SCL/GND. I2C is
generally slow enough that you can get away with all sorts of things
that might upset other physical layer protocols.

Dave

>>>> <https://groups.google.com/d/msgid/hbrobotics/CABbxVHuzMZZ869-eF2c%2BTvfkZisJYoKKNicnac6bFP-c2mnnOA%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>>> .

>>>>
>>>
>>>
>>> --
>>> Rafael
>>>
>>> --
>>> You received this message because you are subscribed to the Google Groups
>>> "HomeBrew Robotics Club" group.
>>> To unsubscribe from this group and stop receiving emails from it, send an
>>> email to hbrobotics+...@googlegroups.com.
>>> To view this discussion on the web visit
>>> https://groups.google.com/d/msgid/hbrobotics/CAB4wyt0EPjKBH5RN4f9Dnf_RRtf8rNXogyt6g1wngyUH%2BpfLhw%40mail.gmail.com

>>> <https://groups.google.com/d/msgid/hbrobotics/CAB4wyt0EPjKBH5RN4f9Dnf_RRtf8rNXogyt6g1wngyUH%2BpfLhw%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>> .

>>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "HomeBrew Robotics Club" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to hbrobotics+...@googlegroups.com.
>> To view this discussion on the web visit

>> https://groups.google.com/d/msgid/hbrobotics/CAJZ%2Bfr3G4Cw6xu2WtdtVBZKavmMuJxO-OaOzah48j3soAeBJZA%40mail.gmail.com
>> <https://groups.google.com/d/msgid/hbrobotics/CAJZ%2Bfr3G4Cw6xu2WtdtVBZKavmMuJxO-OaOzah48j3soAeBJZA%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> .
>>
>
>

[Sergei Grichine]

Just wanted to mention that there is a "kind of standard" for I2C in Autopilot/Drone community. 

You can buy connectors on Amazon and elsewhere - search for "pixhawk i2c splitter":

https://smile.amazon.com/Pixhawk-i2c-Splitter-trailer-Pixhawk-Controller/dp/B0188BHMIA  

These connectors are tiny, but work well. Wires are usually about 8 inches long.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/701ce964-d09c-38d1-6521-50e11b75e320%40uks.org.uk.

[Mark Johnston]

Got the Salese Logic 8 and setup for windows was painless.   

With no manual it was configured first just digital (easy) then setup analyzers for 2 serial and quick poke around to find it can act like a terminal nicely.

Then setup I2C and again very painless.

A nice little easy to use tool for an amazing $11.  The software is easy to understand and that is key for this sort of thing.

Thanks for the tip on this little piece of gear