J1850-VPW schematic question

General Discussion
#1

Hi everyone…

Im working on a project thats only going to be for GM vehicles. It needs to be as small as possible, and I dont care about having support for the older Ford vehicles…so I want to slim down the J1850 circuit to be VPW only…cut out the extra PWM fat.

Following the M2 schematic, this is what I came up with. Anyone have any input on if this will work properly and is acceptable? I know J1850VPW software libraries dont really exist yet, but I just want to check my hardware design.

Using an L7808 8v fixed regulator instead of adjustable, because VPW only requires 8v supply.

Combining D8 and D10 on the macchina schematic into one BAT54-S.

Replacing the LM393 dual comparitor with a TL331 single comparitor. I assume I dont need the second comparitor anymore because im eliminating J1850PWM…the TL331 single comparator package is obviously smaller than an LM393 dual package and both components’ specs look similar on their datasheets

One question I did have is why is the “-” input to U6B on the schematic pulled up to 5v with the 2k R117 resistor and not 3.3v?

Everything else ok? I wish there were J1850VPW libraries available to test the circuit, but hopefully maybe the Macchina guys or someone else who is an electrical engineer can give me the “yeah that’ll work fine”. :grin:

Thanks
Ben

J1850_VPW_ckt.jpg697×468 56.7 KB

#2

Lots of work to try and reinvent the wheel…

Check this out…
MC33390D/DR2

Looks like should connect directly to the Due as it has a 5 volt processor interface (I believe Due is a 5 volt chip?) and just requires a couple components to connect to your network.

Would also likely require a different library to be written.

#3

Ahhhh yes, wouldnt it be nice if we could just use that…I found that chip a couple years ago, but when I looked into it, it seems like all manufacturers discontinued J1850VPW transceiver chips 5+ years ago. :sob: :rage:

Phillips/NXP, Motorola/Freescale, and one other company all made their own J1850VPW transceiver chips way back when.

ALL THREE mfg’s have discontinued production, the bone heads.

Hell, the corvette still used J1850VPW for interior comfort/convenience stuff up until 2013 model year. Seems premature to discontinue something so soon.

Ben

#4

The Phillips/NXP equivalent part is AU5780.

You can get them on Aliexpress/Alibaba…but who knows if theyre old stock, fakes, poor reproductions etc…

#5

I got that data sheet from the Chinese based simulator I have. I have two of them now but have been too busy to try connecting them together and see if I can get them to talk. Going to make a 2 node network with them so I can try to get a datastream going. Will be helpful in the projects I hope to accomplish. I had the doggon thing working at one point in time. Must have messed up the wiring or something somehow. Hopefully will have time soon to sort it out.

Rodney

#6

I bought some chinese clone AU5780 chips a long time ago, I have them right here…but I never had any good J1850VPW libraries to actually see if they worked or did anything when hooked to an arduino.

Ben

#7

Ok I was able to dig up those MCZ33990’s that I bought years ago…does anyone have any ideas at all for a simple arduino program I can make to see if these chips even work or output something usable thats able to be interpreted by the regular 5v digital input on an arduino Uno board Ive got here???

#8

Same here! Maybe we make some kind of bounty for this? Who’s with me?!

This thread had some links to existing projects that might be a good starting point:

#9

Ok I setup my bench with an 03-06 GM truck/SUV Body Control Module, an arduino Uno, the chinese MCZ33990, and supporting circuitry.

I also have a separate J1850 VPW bus analyzer program hooked up, so I can watch the actual traffic on my PC…and eventually maybe corroborate it with what the Arduino receives.

It seems to work!

I basically just wrote a super basic sketch that outputs “0” to the serial terminal when the “Tx” output from the MCZ33990 is low, and “1” when its high.

When I send messages out on the bus, the serial terminal window shows the Tx output going high low high low etc.

So thats theoretically the RAW binary stream from the bus…now we just have to take that raw stream of 1’s and 0’s, measure how long the highs and lows are, and decrypt it into actual data packets

#10

Hey @dmaxben, can you help me get my lab going? I have a ~2002 S10 Blazer BCM and trying to sort out how to get it working. From previous discussions I am assuming that they should be the same BCM.

Would help dramatically. Should be simple but I am missing something somewhere.

I think I have your email and we can take it there but if I wasn’t wasting so much of my time trying to get this stupid thing working I could make some progress.

#11

@josh I am sure a promise of some sort of bounty would make getting time I can dedicate to this project easier after the semester is over… (Promising GF a nice time out would go a long ways I can assure you…)

I plan to work on stuff regardless as I have time. Just getting time crunch right now with being scheduled to work 120 hours next two weeks and have 2 major assignments due for school then finals. Once I am over that I will have more free time for about a month. Hopefully can make major advancement during that time on building the J1850 stuff and building the ELM/STN library.

Rodney

#12

Sent you an email Rod

#13

Josh, any thoughts on my trimmed down “J1850VPW-only” circuit schematic?

Do you think it is electrically equivalent to the Macchina’s J1850 circuit, just with the PWM side cut out?

#14

@dmaxben Personally don’t believe there is nothing wrong with the M2 implementation. It should work fine.

The chip I mentioned that you already have works as a transceiver and handles the actual electrical connection and signals. You just have to build a controller to handle those signals.

I think if someone were building a stand alone VPW device this would be the way to go. (Maybe even worth taking the M2 processor board and building a separate interface board for it for JUST the VPW)

But realistically the M2 can handle the work.

What your likely looking at is the transceiver will just return the 1’s and 0’s and you have to sort them out. It looks like its a simple UART connection. The controller has to be able to manipulate the bits and figure out the frame. So you look for specific patterns as outlined in the J1850 documentation and go that route.

With the M2 you have to actually do the transceiver work too. Since there are some extra electronics on the pins they have them on it shouldn’t take long to actually accomplish it. You can build a blocking interface first to work out the details and once you have that working go back and refine it to not be blocking so it can be used with the M2RET interface.

It would be reasonable to expect to be able to build a library that could use the chip you have as the transceiver and also build one without it controlling the pins directly. Might not be a bad way to go actually to build the library from this chip then build a transceiver in the M2.

I do have a Due board here with a prototype shield…

What did you actually have to use to make it work for additional hardware?

Probably wouldn’t use it in my truck directly but worth working towards the ultimate goal.

#15

I ordered those. They came in. They do not work. I ordered from eBay, as well. Same story. I am guessing they are some other chip, and have been mistakenly identified as MC33390 or MC33290’s.

#16

I guess I got lucky…mine seem to work fine.

I have more on the way, just for grins…

#17

Note that if you pull a data sheet on it there are some requirements to protect the chip. If you do not follow them the chip can be damaged.

The Battery line needs a Diode in place to protect from reverse voltage, the load needs a resistor to the databus line to protect from loss of ground and a capacitor.

@dmaxben, You mentioned you used a load resistor on the circuit. Was this an additional resistor?

If you look up datasheets for these chips they are out there.

But this is why I asked exactly what Ben used to connect his chip up with.

“I connected the 4 MCU lines to the Arduino Directly, Connected a #### ohm resistor between pins # and # then connected that to …” Sort of thing.

#18

Well, I have about 50 of them left. I would definitely be interested to see what I may be doing wrong, but even following more than one published configuration (one from the data sheets themselves), I’ve gotten zilch in or out of them. FWIW, this is a different project than M2. I have a project running off of a Teensy 3.2. The CAN side of things seems to be working just fine. Yes, I did level shift for the 3.3v Teensy, and yes, I did have the load resistor, diode and inductors.

#19

Also, per the earlier post, I have had luck getting onto the J1850 bus just using discretes. Having a little trouble with the software side of things, but it does seem to work.

#20

All I did was:

MCZ33990,

Tied the BATT pin to +12v

Tied the GROUND pin to ground.

Tied the “SLEEP” pin to 5v.

Tied the 4x/LB pin to ground.

Tied the Rx and Tx lines directly to the Arduino Uno

10k resistor between the BUS pin and LOAD pin.

10k resistor between the BUS pin and ground.

BUS pin directly to the J1850VPW bus line.

+12v battery supply also tied to the “Vin” pin on the Arduino Uno so the battery power to Arduino and BCM/ECU is common.

Ground of the Arduino, MCZ33990 chip, and BCM/ECU all tied to common ground.

Thats it…

No diodes, no capacitors, no inductors. Extremely simple…and it works, at least for receiving. Havent tried transmitting yet.

1 Like