Make Nissan Leaf Gen3 adapter board

[johu]

Ah yes, missed the comment about the scope of the switches. I'd say don't worry about it, I can test the detection with a breadboard or something. Just pick one resistor pair and roll with it. We should have a wiki page to keep track of them.

Or is this output held low until brakes turn on? Strong enough to pull the divider below fet turn on?

The brake output it actively pulled low when the brake light is not active. The 220R series resistor will always "win" against the 47k pull-up and the at least 2k7 pull-down.
Plus, to be honest I don't know if anyone even uses the output.

[jrbe]

The brake output it actively pulled low when the brake light is not active. The 220R series resistor will always "win" against the 47k pull-up and the at least 2k7 pull-down.
Plus, to be honest I don't know if anyone even uses the output.

STM32 has around 45-65 ohms when driving a gpio high depending on current. I'll do a simulation of what it's looking like at the mosfet with this stuff.

What's the 220r, a resistor inline to the gpio?

How do you want me to post this when I'm done?

[jrbe]

What is OUT_OUVTG? Pin 19, JP2 on mini main board. This is labeled CS_ADC on the i3 adapter board. Google site search only finds the wiki pages, no info on either there.

Should the 3 SOT-353 buffers from the i3 board be on this leaf board as well?
If so, C85397?

[johu]

What is OUT_OUVTG? Pin 19, JP2 on mini main board. This is labeled CS_ADC on the i3 adapter board. Google site search only finds the wiki pages, no info on either there.

Should the 3 SOT-353 buffers from the i3 board be on this leaf board as well?
If so, C85397?

See here: https://openinverter.org/wiki/Mini_Mainboard

It can be used to talk to SPI chips but since there are none in the Nissan inverter you can ignore that.

[jrbe]

I attached a pdf of the schematic if anyone is willing to look at it.

I decided to link the voltage divider to 5v for addressing. Thanks for mentioning 5V johu. I'm not sure all adapter / OpenInverter boards have 3.3v on them.

Info:

• I used 0805 sized components. I prefer to not hand solder anything smaller.

• I picked 4.7K for this board address (R18.) Next to R18 are 2 vias (blue lines) to pop R18 off and experiment with different resistor values if needed. You wouldn't do this normally, just for assigning & testing addresses for different boards and measuring what the brake circuitry does to it. The vias leave nice spots to check the address voltage.

• SJ4 is a shorted solder jumper that connects JP2, pin 18 (MISO / OUT_BRAKE) to the Nissan connector labeled here as OUT_BRAKE (not sure what Nissan calls it.) If the addressing works out it will be MISO / OUT_BRAKE / ADDRESS. This can be cut for testing if needed.

• D11 I'll call the oops diode. If someone forgets to put a pull down resistor in and has R15 in place it pulls R15 down to around 3.3v.

• 5V is there to remind people that this is a 5V divider, not a 3.3V as they would maybe expect for a STM32.

• Added pull downs for the Pre and DCDC contactor mosfets, 10K, R7 & R8.

• added a SMB TVS diode (D1 & D2) across the source / drain in case the diode in the mosfet can't take the flyback hit from a contactor releasing. Its a unidirectional 18v. johu mentioned some diodes can reverse power the board on some installs, should this go to a bidirectional 18v TVS? Currently LCSC C80394

• T_SINK got a 3.3k pull down, mentioned in johu's notes getting the previous to me version of this running. Not sure if this needs to be optimized.

• R24 is in the datasheet for the ESP32 (inline in USART_RX.) Says it suppresses 80MHz harmonics.

• QR code goes to the Gen 3 board wiki page (parked for now.) Hopefully this will help eliminate some questions once people get them in their hands.

• SJ1 & SJ2 are solder jumpers to enable a 120 ohm CAN resistor. Its technically 2x 60ohm resistors in series with a capacitor in the middle to ground to filter noise. Can be soldered / enabled if this is an end of the CAN bus. There is ESD protection inline right at the Nissan connector for the CAN lines (D6) just in case.

There are a few things I'm unsure of:

• J4 is the ESP32 programming header. This connects to 5V instead of 3.3v as I'd expect. I copied from the updated (V6 & V8) Tesla boards - I assume those are correct but not sure why.

• J3, J5, and J6 are there to allow easy break out of these different dead ends in case someone wants to use them. J5 I setup to a typical 2x03 programming header pin out, not sure that will work with other stuff going on. If this is a bad idea let me know and I'll rearrange it.

• The inductor footprints around the outside will be unpopulated. There is a ground shield plane that connects to the mounting holes on the bottom. There's a ring around the outside edges on all layers up to the top. These inductors footprints would connect the 2 planes if needed, and could be capacitors if that worked better.

• OUT_OVTMP on J6 seems to be a Nissan thing. Not sure if anyone uses this. If anyone knows a use / where it should connect please share.

• OUT_TEMP on J6 seems to be a Nissan thing. Not sure if anyone uses this. If anyone knows a use / where it should connect please share.

• CS_ADC / AKA OUT_OUVTG Isn't connected anywhere. Not sure if it should be or where / how. Couldn't find info on it. For SPI, not needed for this Nissan inverter (johu's previous post.)

• R2 isn't connected, not sure it's needed.

• IL1 & IL2 are connected to 2 terminals each. This seems wrong. Might be one of each needs ground or? Unfortunately the original Nissan board has almost all the connector traces under the connector bodies for our inconvenience.

• There are a few 0R / jumper resistors in these unknown connections to allow pulling the resistor then connecting a bodge wire if needed.

• Not sure the Nissan connector's mechanical pins matter if they connect to ground or not. Connecting to ground helped the ground plane flow so its setup that way for now. I'm assuming they don't really go anywhere and just hold the plastic of the connector. Would be great if someone could confirm.

• Need to confirm the Nissan connector footprints are correct. The pads are long by around 4mm to help with hand soldering.

• Boot delay for the ESP32 is 1uF (C13) / 10K (R25.) Its the typical values in the datasheet. The Tesla boards use different values, not sure if it matters.

• C3 & C4 are currently 100nF to debounce the contactor outputs. There was mention of contators chattering or pulsing at power up, not sure if this was resolved. Either way these should help eliminate contactor chatter (if sized right..) These can be adjusted if needed. Left room around them and did a small keepout area for the ground plane to help swapping them if needed.

*edit* clarified the bottom ground plane info. Also I moved the Q1 and Q2 gate traces to the signal layer to allow better ground plane flow around the contactor mosfets.

[johu]

Thank you very much for your hard work.

I looked through the schematic, only thing that popped up was the pull-up of mprot. That should be hard wired to 12V without the 4k7 resistor because there is one on board already.

added a SMB TVS diode (D1 & D2) across the source / drain in case the diode in the mosfet can't take the flyback hit from a contactor releasing. Its a unidirectional 18v. johu mentioned some diodes can reverse power the board on some installs, should this go to a bidirectional 18v TVS? Currently LCSC C80394

Yes perfect as it is. Reverse powering only happens with a regular diode with cathode tied to 12V.

T_SINK got a 3.3k pull down, mentioned in johu's notes getting the previous to me version of this running. Not sure if this needs to be optimized.

Not sure why I picked 3k3 there, probably had one handy. "Standard" is 1k2. Still need to map out the sensor.

J4 is the ESP32 programming header. This connects to 5V instead of 3.3v as I'd expect. I copied from the updated (V6 & V8) Tesla boards

Most USB uart adapters have a 5V outlet but no 3V3

OUT_OVTMP on J6 seems to be a Nissan thing. Not sure if anyone uses this. If anyone knows a use / where it should connect please share.

OUT_TEMP on J6 seems to be a Nissan thing. Not sure if anyone uses this. If anyone knows a use / where it should connect please share.

That's just general purpose outputs of all OI boards but I think they are not brought out to the round connector anyway. If exported the should go through a similar circuitry as the other outputs, maybe with smaller FETs 2N7002 or so.

R2 isn't connected, not sure it's needed.

With the older mainboard R2 was GND. Now it needs to be connected to R2 of the main board (pin 20 of J2 presumably)

IL1 & IL2 are connected to 2 terminals each. This seems wrong. Might be one of each needs ground or?

It seems to be some automotive standard that current sensors have redundant outputs. I was never sure how to make use of these so one pair remains disconnected

Not sure the Nissan connector's mechanical pins matter if they connect to ground or not. Connecting to ground helped the ground plane flow so its setup that way for now. I'm assuming they don't really go anywhere and just hold the plastic of the connector

They are indeed plainly mechanical and can be connected to anything

[jrbe]

Thank you very much for your hard work.

I'm happy to contribute. Thanks for helping make all of this possible and thanks for your patience with my million questions.

I looked through the schematic, only thing that popped up was the pull-up of mprot. That should be hard wired to 12V without the 4k7 resistor because there is one on board already.

I removed the 4K7 and wired it straight to 12v but something seems off here.
On the mini mainboard 12v goes through a voltage divider and connects to PA3 / pin 17 of STM32. It does not go to either 20 pos. headers though so the STM32 has no feedback on this line from the adapter board, seems like its just an override on the mini mainboard.
On the Nissan 40 pos. connector this is the wiring that jumps out of the inverter box. Seems like this is not required / maybe mixed up with the original mainboard functions? (or I'm missing something..) I'd rather not output 12v through the pc board back out the inverter connector unless it's necessary.

Not sure why I picked 3k3 there, probably had one handy. "Standard" is 1k2. Still need to map out the sensor.

I added a reminder note in the schematic to verify it. Its the annoying one that's going off the page in the pdf. Was this post that mentioned it, viewtopic.php?p=23175&hilit=3k3#p23175


..OUT_OVTMP & OUT_TEMP..

That's just general purpose outputs of all OI boards but I think they are not brought out to the round connector anyway. If exported the should go through a similar circuitry as the other outputs, maybe with smaller FETs 2N7002 or so.

These are currently only labeled on the 40 position Nissan connector that I'm seeing.
Pin 32 of the 40 pos. connector is OUT_TEMP.
Pin 27 of the 40 pos. connector is _OUT_OVTMP.
I couldn't find where they would connect on the mini mainboard. Happy to add some small fets for them but not sure what mini mainboard outputs are supposed to drive them.
OUTVTG could be one of these but I have not found any info pointing to that / linking the labels together.

In the same thinking I added 2N7002's with 10k gate pull downs for:
- PWM_USER output (JP2, pos 20) thats going to dead end header J5 pos 8 currently.
- corrected OUT_ERR output (JP2, pos 17) - removed from the dead end header (J5.)
- corrected OUTVTG (AKA CS_ADC - JP2 pos19) and have it going to the dead end header (J5) after the fet.
- - Should any of these get solder jumpers that can be cut?

With the older mainboard R2 was GND. Now it needs to be connected to R2 of the main board (pin 20 of J2 presumably)

I think I see this now. Pin 20 of the Nissan 40 pos. connector is labeled as R2 but goes to GND in Kelju's previous work. I removed that 0R to GND and connected it to JP1, pos. 12 / R2.

It seems to be some automotive standard that current sensors have redundant outputs. I was never sure how to make use of these so one pair remains disconnected

I wonder if this is a connector resistance / connector mechanical failure redundancy? I left one of each with a 0R footprint as DNP, removed the other footprint on each and ran a trace across the missing spots.

I added 2 ceramic 10uF caps at the 4x +5V lines on the Nissan 32 pin connector, they're C5 & C6. I don't think this would be an issue with the 5 volt regulator and should help keep the 5V for different sensors stable / clean. Let me know if those should get swapped / removed.

I attached V0.0.1 schematic pdf. Going with small version #s while we work through this.

Gen 3 Leaf Adapter V0.0.1.pdf

(171.99 KiB) Downloaded 62 times

Updated board image for reference.

[johu]

Oh yes you're right, they mprot signal just no longer exists. You can omit all components.
I also found the OI pin mapping doesn't quite line up with the OEM pin mapping, for some reason emcystop is connected to resolver S1 (this should connect to mini mainboard S4S1) just like Pin 15. You could also swap R1 and R2 to line up with the OEM board.
I think the traditional output signals won't be used by many. If you prefer you could even omit them for now. I omitted them on the Gen2 board as well. Anyway, OUT_TEMP has been renamed to PWM_USER.
Likewise OUT_BRAKE doesn't make much sense to be exported without a FET. I guess omit it completely on the outside connector.

Caps on 5V should be ok to have.

I wonder should we just add a table with assigned voltage levels/resistor values on the mini mainboard page?

[jrbe]

Oh yes you're right, they mprot signal just no longer exists. You can omit all components.

Ok, I pulled the solder jumper and the connection to 12v. Just leaving a test point to the nissan connector.

I also found the OI pin mapping doesn't quite line up with the OEM pin mapping, for some reason emcystop is connected to resolver S1 (this should connect to mini mainboard S4S1) just like Pin 15. You could also swap R1 and R2 to line up with the OEM board.

Having a hard time following. Do you mean the schematic Kelju started is incorrect? Can you include a marked up picture or Nissan connector # to OI pin so I can follow better?
*edit and does this mean you can swap some pin functions around on the mini mainboard?*

I think the traditional output signals won't be used by many. If you prefer you could even omit them for now. I omitted them on the Gen2 board as well.

I'll leave them in for now, It will help test these theories as there's time or a user who wants to. Can just leave them unpopulated if decided later.

Anyway, OUT_TEMP has been renamed to PWM_USER. Likewise OUT_BRAKE doesn't make much sense to be exported without a FET. I guess omit it completely on the outside connector.

Thanks. I corrected PWM_USER <-> OUT_TEMP and have it driven by a 2N7002.
I swapped the SJ4 solder jumper to open and connected it to the EXT_BRAKE with a 2N7002 that has a 115ma constant current rating. But that wont drive brake lights by itself or even a relay. Whats the recommended way to use this brake output?

I wonder should we just add a table with assigned voltage levels/resistor values on the mini mainboard page?

I would love to see a table of pin mapping / info for all of the boards available. Alternative pin labeling should be part of this too I think. This would have helped making this adapter board a bunch. I think part of this table should be voltage level / address. Is there a spot in the wiki that exists already that you think this would fit best?
*edit I saw your mini mainboard info about addressing. Some of the voltages are close together. I think we need to define a target for each (you did already) and a voltage range for each. Likely need to keep these at 1% or better resistors but should define that as well.
I'll change 5v to 5.3v in the voltage divider area.
It might make sense to have a simple and deep dive version of info. Simple is I don't want to know why, just want it to work. Deep dive is I want to understand and maybe develop things further / help out. There should be a place for openinverter footprints, logos, common components, circuit references, etc. as well.*

[johu]

Having a hard time following. Do you mean the schematic Kelju started is incorrect? Can you include a marked up picture or Nissan connector # to OI pin so I can follow better?
*edit and does this mean you can swap some pin functions around on the mini mainboard?*

It is somewhat incorrect as he connected emcystop to the resolver. Check the pinout.pdf in the second post to see what I mean.

Thanks. I corrected PWM_USER <-> OUT_TEMP and have it driven by a 2N7002.
But that wont drive brake lights by itself or even a relay. Whats the recommended way to use this brake output?

You'd have to ask someone who actually uses it Normally they use switched 12V so a relay is needed. Like said I consider all the outputs as priority Z as the wire harness doesn't export them anyway.

I would love to see a table of pin mapping / info for all of the boards available. Alternative pin labeling should be part of this too I think. This would have helped making this adapter board a bunch. I think part of this table should be voltage level / address. Is there a spot in the wiki that exists already that you think this would fit best?

I guess this comes closest, even though it doesn't 100% apply to the mini mainboard. Could be repeated there. https://openinverter.org/wiki/Main_Board_Version_3

*edit I saw your mini mainboard info about addressing. Some of the voltages are close together. I think we need to define a target for each (you did already) and a voltage range for each. Likely need to keep these at 1% or better resistors but should define that as well.
I'll change 5v to 5.3v in the voltage divider area.
It might make sense to have a simple and deep dive version of info. Simple is I don't want to know why, just want it to work. Deep dive is I want to understand and maybe develop things further / help out. There should be a place for openinverter footprints, logos, common components, circuit references, etc. as well.*

I have updated the table with from-to values. Noticed with +-5% tolerance they overlap, 3% would work. So 1% resistors then.

[jrbe]

It is somewhat incorrect as he connected emcystop to the resolver. Check the pinout.pdf in the second post to see what I mean.

You mean this? (my quick photoshop)

I missed that wires / pins were repurposed..

• So from that it looks like R1 & R2 need to swap places (as you mentioned.) I'm broken, I needed the visual.

• And E-stop needs to be removed. But I'm unsure if S1 / SIN/S1&COS/S4 stays where it is (there's no ground there.) Do S1 & S4 wires connect together for S1 / SIN/S1&COS/S4?

• On the mini mainboard connector for signal e-stop, that needs 12v, correct? If so, I'll likely add a shorted solder jumper and a via on each side in case someone wants to use it.

[johu]

• So from that it looks like R1 & R2 need to swap places (as you mentioned.) I'm broken, I needed the visual.

• And E-stop needs to be removed. But I'm unsure if S1 / SIN/S1&COS/S4 stays where it is (there's no ground there.) Do S1 & S4 wires connect together for S1 / SIN/S1&COS/S4?

• On the mini mainboard connector for signal e-stop, that needs 12v, correct? If so, I'll likely add a shorted solder jumper and a via on each side in case someone wants to use it.

Yes, yes and yes

S1 and S4 are connected to the same pin of the mini mainboard but individual pins on Nissans connector

[jrbe]

Attaching my recent work on pin out info. Seems either the Nissan connector / info or the info / image Kelju posted is *edit numbered backwards*, but even then theres conflicting info. Rotating it and flipping it still seems to not line up empty pins to wire colors.

Kelju's work top left
Damien's pin out text, and a rotated connector image he posted to the right (along with a nissan diagram someone posted.)

I added wire colors to Kelju's drawing of the connector *edit round colored dots are the wire colors from their work, wire colored stars are from the Nissan diagram.*
If someone could grab pictures of the nissan wiring harness from inside the inverter and the connector where it plugs into the first control board it would help digging through all of this a ton! This, with clear visuals of what wire colors go where on the connectors. I can guess at some from these images but most are not clear.

*edit, someone would just need to confirm star or dot colors. It would still be extremely helpful to get an image of the 40 pos. nissan connector wire colors, thats the most important part that I need for this.*

[Bratitude]

I’ll make some calls, I know some one with a gen 3

[jrbe]

I found 2 more Nissan pin outs Dala has on his github for his Gen 2 to Gen 3 inverter upgrade info. https://github.com/dalathegreat/Nissan- ... er-Upgrade

One for Gen 2,

one for Gen 3,

I really like this diagram, shows the + and - side of a feature, testing conditions, and a target value. A bit awkward until you realize they organized it this way and everything seems to be there if you look in both columns.

This is showing the wiring table that was previously assumed to be correct is quite off both in pin position and color to functions. Not sure what happened there but my confidence with that table is near 0 currently. Waiting on images of the short inverter wire colors (*edit, and actively searching for them*) in the connectors to use along with this Nissan terminal layout info and move forward to pin out the 40 pin connector. Then I can finish this adapter board.

[jrbe]

Looks like Nissan used different colors inside and outside of the inverter as a special kind of torture. That explains some of the wire color / conflicting info issues.

[jrbe]

I don't mean to hold this hostage. We can try to do this before getting boards made or take a chance and do it after when trying to get it up and running. My preference is to do it before getting boards made.

Seeing a bunch of things that I can not currently figure out for the pin out. I don't mind doing the work / research, I just don't have the images of the connectors or info to be able to tell if the 40 pin connector is remotely correct. I've probably spent 6 hours trying to hunt down images of these connectors.
These are the best images I could scrounge up of the 40 pin connector / wire colors.

Pin numbering based on how the pc board is labeled looks to start the upper right is pin 1, across to the left pin 20. Then bottom right is pin 21 across to the left pin 40. This is something I'd like to confirm, I drew up the Nissan connector footprints and this was the conclusion I came to.

I rotated the previous pin out work to line up to help work through this.

The right side has 2 reds, pin 1 & pin 21 - seems +12v would be sensible. Then there are 4 black wires in the connector, pins 2 & 22 are labeled as for the contactors. Then pins 3, 4, 23, & 24 are labeled as grounds but what looks to be pin 4 is red. I can't see Nissan using red and black as grounds but might be missing something..
The middle is too hard to tell what is what from these fuzzy and not oriented well for this work pictures.

On the left side of the connector pin 20 is shown as ground but is labeled incorrectly as R2 (supposed to be R1?)
Pin 40 shows as empty but seems the brown wire is bottom left.

Theres a good chance I'm missing spacing between positions, its just too hard to tell what's going on from these pictures. I'm not thrilled to post this where it is but will if that's what you guys want to do.

[johu]

I have an inverter at home, will be back on Sunday. No rush at all, better get it confirmed

[Bratitude]

my guy feel through sadly

[jrbe]

my guy feel through sadly

Thanks for trying.

[johu]

So I can confirm the wire colours and blanked pins. They are counted looking into the pin side of the round connector
The 40-pin connector on the PCB matches the one they use on the BMS https://openinverter.org/wiki/Nissan_Le ... rt_numbers

I wonder if it would be more DIY friendly to use the through hole part? https://www.digikey.de/de/products/deta ... -2/2273492
EDIT: that said it has 0 availability. But so has the SMT part

[jrbe]

So I can confirm the wire colours and blanked pins. They are counted looking into the pin side of the round connector

Thanks for these.
That mean you did confirm the colors or will? To the Nissan diagram?
There's the inside and outside the inverter wire colors as well.

I can do this too but will need a few more pictures of each end of the harness of the hidden wire areas.
Some of the wires with same colors have colored dots on them. Well likely need to ring them out connector to connector or note the dot color as well.

I wonder if it would be more DIY friendly to use the through hole part? https://www.digikey.de/de/products/deta ... -2/2273492
EDIT: that said it has 0 availability. But so has the SMT part

I'd personally be fine with either smd or thru hole. There are some resellers I'd trust to actually have the thru hole version,
https://octopart.com/search?q=1318384-2 ... SD&specs=0

Do you know if the 32 position connector is the same as in the bms as well?

I'd prefer to buy new board side connectors vs struggling to desolder them but others might not. I'll see if both footprints can fit on the board.

[johu]

Thanks for these.
That mean you did confirm the colors or will? To the Nissan diagram?
There's the inside and outside the inverter wire colors as well.

I compared the actual connector at hand to the diagram Kelju had created that is attached to the second post. I don't know about the outside wire colours as I don't have the harness.
It makes sense that red and black are 12V and GND and the unicolor grey wires are the non-polar temp sensor. No need to distinguish here except between "ignition" and 12V.
I'm just thinking if Kelju got the diagram right that far, the functional mapping should be correct as well.

Do you know if the 32 position connector is the same as in the bms as well?

Yes, same family

I'd prefer to buy new board side connectors vs struggling to desolder them but others might not. I'll see if both footprints can fit on the board.

That would be genius. Otherwise not sure what to default to. THT is easier to solder for the "layman" while SMT could be harvested from the OEM board with a heat gun. Given the poor availability this might make sense...

[jrbe]

I compared the actual connector at hand to the diagram Kelju had created that is attached to the second post. I don't know about the outside wire colours as I don't have the harness.

Ok, I'm likely going to backtrack the wires because I'm broken. The orange dots are where terminals were added for open inverter features,

SMD / Thru hole connectors..

..not sure what to default to. THT is easier to solder for the "layman" while SMT could be harvested from the OEM board with a heat gun. Given the poor availability this might make sense...

Agreed. If one isn't available the other might be. I'll see if I can make a merged footprint to fit either smd or thru hole.

Thanks for helping fill in the gaps.

[jrbe]

Made a dual footprint layout. It works ok for signal level stuff but struggles trace width wise on the board edge side plated thru holes jumping through the first row. Those traces are .3mm and the thru hole rings are as narrow as I'm comfortable with. 2.2mm x1.475 oval pad with a 1.15 hole and .2mm clearance. Its within JLC's listed capabilities.

Downsides as is:
- only the plated holes near the smd pads can have high current traces, unless they can connect on different layer planes (GND, +12VSW do.)
- The DCSW or PRE output trace / plane to the thru hole will suffer.
- The recommended board edge for the thru hole version connector overhangs by 8.5mm as shown. I think we can add board material here without causing fitment issues. I'd like to add a bit more of a gap between the footprints to allow a thicker plane area for PRE.

I think we should do the dual footprints but rearrange some of the added feature pin / openinverter positions to utilize the inner row of thru holes for high current outputs. Any objections? I can also sequence them with the mini mainboard positions to help routing.

There are empty pin positions on the 40 position connector but not the input connector (only 5 empty spots there.) I think it would be a decent idea to standardize a second connector or maybe a replacement for the Nissan external inverter connector.

I also just noticed there is no 5v on the 40 pin connector (out of the inverter.) I just added that.