[Adria]

Ah, this already exists and is configurable as needed. Lalomo Ethernet RJ45 to OBD2 connector

Amazon.com

same on a non bezos site

Lalomo Ethernet RJ45 to OBD2 16 Pin Connector Adapter for All F Series and Late E Series E-SYS ICOM Coding Reader Diagnostic Tools: Buy Online in Dominica at desertcart

Buy latest Lalomo Ethernet RJ45 to OBD2 16 Pin Connector Adapter for All F Series and Late E Series E-SYS ICOM Coding Reader Diagnostic Tools online at best prices at desertcart. ✓FREE Delivery Across Dominica. ✓EASY Returns & Exchange.

dominica.desertcart.com

 

[retepsnikrep]

Where are the details for the RJ45 pinout on Natalya's PCB?

 

[Natalya]

@retepsnikrep

Going by this diagram, with the pins as written in the bottom right of the photo, at the OBDIIC&C gauge I use the following pinout:
1 - Orange/White - Optional - this is RB1 on the Pic
2 - Orange - DLC Pin 16 +12V
3 - Green/White - DLC Pin 14 H-Line
4 - Blue - CAN High
5 - Blue/White - CAN Low
6 - Green - DLC Pin 7 K-Line
7 - Brown/White - DLC Pin 9 - SCS on the Pic
8 - Brown - DLC Pin 5 Ground

So, at the OBDII connector, Orange/White, Blue, and Blue/White don't get wired to anything..

This ensures compatibility with my newer CANBUS integration wire harnesses. I think you've got the older version (green PCB) which was a prototype and the PCB should say V0.00 on it.

Earlier this year I developed a new, complete kit, the point of which was to greatly simplify the wiring required to have the OBDIIC&C hook up to the LTO BMS modules. Around the same time is when Adria's OBDIIC&C went bad, so I decided to try to solve a bunch of issues all at once by attempting to improve the OBDIIC&C PCB and attempting to improve last year's version of the LTO BMS wiring. Half-way through this effort Greentec sold out of the LTO batteries, so further progress on this is now pointless, but I had a good time trying to get better at reading wiring diagrams and designing PCBs.

There are three PCB components for my system:

IPU CAN Integration Module V1.0.0 PCB <-- This is the updated version of the pcb you've got at the end of the LTO BMS pigtail thing I mailed you
Under-Console CAN Integration Module V1.1.0 or higher PCB
OBDIIC&C 1.1.0 or higher PCB

(Click for full size)


When I started this and mailed you the IPU LTO BMS connector you've got I didn't know about semantic versioning, but now everything uses it. I have the new PCBs for those two modules printed blue, to indicate incompatibility with the green ones. There are a few more improvements that can be made, mainly with the module that goes under the console, but because the LTO supply is gone it'll probably never get done.

I also added pins for IMAC&C switch wiring to the 1.1.1 version of the OBDIIC&C PCB, but those would need their own separate dedicated wire. I've never seen an IMAC&C in person so I'm not familiar with their wiring requirements.

I know this is all probably really confusing, but when this system is wired up properly, all you need is 1 single connection to the car's factory wiring, which was the primary goal. Secondarily, the OBDIIC&C can now easily be removed from the vehicle by simply unplugging it at its on-board RJ45 jack. Adria now has this new system in her car. Jeff has an older 1.00 OBDIIC&C that I retrofitted with an RJ45 jack to this specification, so should down the road he ever go LTO (probably unlikely) his gauge will be compatible. In a couple weeks Lily is leaving the country for an extended period, so I'll be taking care of her Insight, during which time I'll also upgrade her LTO CAN wiring to this new standard.

 

[retepsnikrep]

Thanks for that Natalya.

Here is a possible OBDIIC&C V1.5 rough schematic from me.

The wiring is different but it's the same basic design and the current OBDIIC&C firmware would adapt reasonably easily once I work out how to use the Nextion.

It uses the Nextion display and a CAN enabled PIC18F26K80 at 64mhz.
Crystal oscillator control should make it much more resilient in hot climates and less prone to heat related comms issues.

It has DIP switchable CAN termination and three 200ma open collector outputs including one PWM enabled for external device/relay/charger control.

It has an SPI/I2C interface/spare I/O.

Two diagnostic leds and piezo buzzer.

It uses the RJ45 plug/socket and Natalya pinout for the basic car connections.

It has more resilient CAN filtering/protection.

It has a dedicated cheap IMAC&C Analog Joystick Port (You still need the small IMAC&C pcb inside the MCM)

I could work to get a SMD PCB made that fits exactly behind the Nextion display.
Then we can rescale/design the 3D printed OBDIIC&C case to fit the Nextion.
It would be slight less tall, narrower and thinner.

Note. I might well move to the 3.3V PIC 18LF26K80 and a 3.3V CAN Transceiver so this gets rid of the Nextion TTL Level translation issue. I am just checking out parts.

Comments???

 

[joeaax1j]

Peter, there are many links to various versions of the ODBIIC&C firmware throughout this thread. And some, like the one you have just posted, probably shouldn't be flashed on most of our units. I'm wondering if you might consider adding in your signature a link to the version you would recommend for most people?

Thanks,
Bryan

 

[retepsnikrep]

Bryan

What I posted is a proposed schematic not firmware.

I take your point about firmware but there are lots of variants due to country and usage variations.
I can't keep them all in my signature block.

 

[joeaax1j]

Understood. Thanks.

 

[Natalya]

@retepsnikrep
This looks pretty cool. Could a 3.3v CAN tranceiver take data from the LTO BMS boards?

I'm going to promote adoption of semantic versioning for OBDIIC&C, as described here:
www.semver.org

If you're not already familiar, the basics of it, are as follows:
1.2.3
First digit gets incremented when there are major changes that are not backwards compatible
Second digit gets incremented when functionality is added but it IS backwards compatible
Third digit gets incremented when there is only a small change, usually a patch or bug fix
If first or 2nd digit are incremented then the ones after it go to 0.

This would put us into 2.0.0, which is an issue because I know you've been referring to the new gauge as version 2. A solution could be to consider that as-yet-undeveloped gauge as a 3.0.0, or maybe you could consider it as a different, successor product because it's got the ECM integration and its scope is much larger than regular OBDIIC&C?

 

[Natalya]

Also, I'll have more time to look over the schematic you posted tomorrow and can make a more complete response. And I hope I'm not stepping on toes with the versioning thoughts, I just hope to make the different versions relationships with each other more clear.

 

[retepsnikrep]

The CAN transceiver function is not affected by the supply voltage change.

 

[retepsnikrep]

I sort of have the Nextion talking to my 18F2680 OBDIIC&C pcb now.
Not quite right but at least it's doing something..

 

[retepsnikrep]

Basic demo... Two way comms.

Doing stuff on OBDIIC&C appears on Nextion and doing stuff on Nextion appears on OBDIIC&C..

OK Let's call this version 2.0.0

 

[retepsnikrep]

This is the current V2.0.0 schematic minus the TTL level translator.

Note. Isaac and I are reviewing the RJ45 pinout for best signal practice.

I'm happy to use an RJ45 connector but the pinout might change slightly to improve resistance to interference etc.

 

[Natalya]

I haven't observed any issues with the gauge that would suggest a data transmission problem.

I paired the CAN wires because they're supposed to be a twisted pair.

Two other colours were chosen based on colour. Orange for +12V, Brown for Ground. The other 4 are more-or-less arbitrary

 

[retepsnikrep]

I think the power wires would be a bit better as a twisted pair.

Just trying to make it as good as possible..

 

[Natalya]

So, I thought about that at first, but the reason I went with the colours for them instead is because I wanted them further away from eachother at the jack.

Let's say you had the OBDIIC&C plugged in, but you had the case off for some reason. I'm not sure why, but someone might do that. If the pins were for example 1 and 2 or 7 and 8 they'd be very close to eachother on the PCB and it'd be really easy to accidentally short them. Having them at 2 and 8 puts them almost as far away from eachother as possible, making a short that much less likely.

 

[Bull Dog]

@retepsnikrep

View attachment 91189

Going by this diagram, with the pins as written in the bottom right of the photo, at the OBDIIC&C gauge I use the following pinout:
1 - Orange/White - Optional - this is RB1 on the Pic
2 - Orange - DLC Pin 16 +12V
3 - Green/White - DLC Pin 14 H-Line
4 - Blue - CAN High
5 - Blue/White - CAN Low
6 - Green - DLC Pin 7 K-Line
7 - Brown/White - DLC Pin 9 - SCS on the Pic
8 - Brown - DLC Pin 5 Ground

...

So, I thought about that at first, but the reason I went with the colours for them instead is because I wanted them further away from eachother at the jack.

Let's say you had the OBDIIC&C plugged in, but you had the case off for some reason. I'm not sure why, but someone might do that. If the pins were for example 1 and 2 or 7 and 8 they'd be very close to eachother on the PCB and it'd be really easy to accidentally short them. Having them at 2 and 8 puts them almost as far away from eachother as possible, making a short that much less likely.

I don't see a dead short between power and ground being a huge problem. Replace the blown fuse and move on.

I am concerned about what happens if the +12V contact manages to short to the 5V H-Line that you have immediately adjacent to it on the male RJ45 plug. I don't know offhand how robust the H-Line, but accidentally dumping 14.4 Volts into a 5 volt signaling circuit seems like a bad idea.

To me, a pinout like this seems to be a safer option.

1 - Orange/White - DLC Pin 16 +12V
2 - Orange - DLC Pin 5 Ground
3 - Green/White - DLC Pin 14 H-Line
4 - Blue - CAN High
5 - Blue/White - CAN Low
6 - Green - DLC Pin 7 K-Line
7 - Brown/White - DLC Pin 9 - SCS on the Pic
8 - Brown - Optional - this is RB1 on the Pic

or alternatively put the SCS wire on the RJ45 pins 3 or 6 just so both the K-Line and H-Line aren't sharing the same twisted pair...which admittedly probably doesn't matter at the speeds they operate at.

1 - Orange/White - DLC Pin 16 +12V
2 - Orange - DLC Pin 5 Ground
3 - Green/White - DLC Pin 9 - SCS on the Pic
4 - Blue - CAN High
5 - Blue/White - CAN Low
6 - Green - DLC Pin 7 K-Line
7 - Brown/White - DLC Pin 14 H-Line
8 - Brown - Optional - this is RB1 on the Pic

 

[retepsnikrep]

or alternatively put the SCS wire on the RJ45 pins 3 or 6 just so both the K-Line and H-Line aren't sharing the same twisted pair...which admittedly probably doesn't matter at the speeds they operate at.

1 - Orange/White - DLC Pin 16 +12V
2 - Orange - DLC Pin 5 Ground
3 - Green/White - DLC Pin 9 - SCS on the Pic
4 - Blue - CAN High
5 - Blue/White - CAN Low
6 - Green - DLC Pin 7 K-Line
7 - Brown/White - DLC Pin 14 H-Line
8 - Brown - Optional - this is RB1 on the Pic

I prefer the alternate option as it minimises crosstalk between H & K lines..

 

[Natalya]

I think Isaac makes a fair point about the 12V being fused while the K and H lines are probably not. We can also get those lines out of a twisted pair.

I'm going to propose we use his alternate, but with one swap, which is to have 7 be the Optional line and 8 be the H-Line, to help me with how I have the traces on the PCB.


1 - Orange/White - DLC Pin 16 +12V
2 - Orange - DLC Pin 5 Ground
3 - Green/White - DLC Pin 9 - SCS on the Pic
4 - Blue - CAN High
5 - Blue/White - CAN Low
6 - Green - DLC Pin 7 K-Line
7 - Brown/White - Optional - this is RB1 on the Pic
8 - Brown - DLC Pin 14 H-Line

 

[retepsnikrep]

Sounds good let's go with that.