OneROM

[antus]

Well its just 2 serial channels put over the top of each other on 1 wire. Cant beat that for simplicity. And its 8192 because of the (lack of) CPU speed and so they could handle that cleanly with the speed the OS ran in the early PCMs. It was never originally intended to be a car bus protocol, ALDL is Assembly Line Diagnostic Link and it was originally just to do something like read the DTCs and make sure the car was working in the factory. Only it grew and grew until it was the vehicle bus and they couldn't fit any more data on it!

[jxx]

Yeah, that's where it's getting me, I know it's simple, I know it can be linked with a resistor and a diode into 1 wire, just for the life of me can't get a reliable clean signal over 2 wires using seperate lines from the ecu to the mcu.

Or when you decide to check against another ECU and find out the one you have been slaving over for hours is faulty.....

[jxx]

Updated available revisions in post viewtopic.php?p=138791#p138791

[antus]

Nice! My one roms arrived today. Have other projects on the go so not ready to look a them right now, but will do at some stage. I'd like to have a nvram for tuneing and a onerom as a memcal replacement. When I get the power hooked up and heat proofed in my project car it'll be ready for tuneing and that'll be the time. Got mine from https://jcm-1.com/ and it looks like the real deal, not a clone.

[jxx]

Nice, hopefully I have this firmware sorted and as reliable and bullet proof as possible for ecu use by the time you get around to using them.
Some extra testers with suggestions and feedback would be great.

I really don't think clones in the bare stock onerom hardware is really an issue, you'd have to be pretty determined to find the cheapest nastiest pcb manufacturer to undercut costs enough...
And i've looked at plenty of different manufacturers including the bargain basement cheapest ones online and the end cost really isn't much of a saving compared to using jlc.

I would suggest at least making sure they are functional instead of putting them on a shelf for weeks/months, the odd faulty one isn't unheard of.
I've had a few myself in the batches i've had made up, if it atleast turns on and connects to the computer they're usually good.

1 24 pin version i have connects and programs up but does not function correctly when in place on a board, this is the rare excepting though.


New REV H files uploaded viewtopic.php?p=138791#p138791
Quite a few improvements.

02_sml.png

[jxx]

Moved entire post to the first post of the thread.

EMU.jpg

Moved Download onerom-emu-flashtool_v0.1.1.zip to first post.
viewtopic.php?p=138305#p138305
Flashtool does everything (except the ALDL) automatically instead of manually uploading all the different bin/uf2 files in this thread.

[vlad01]

So does RT work over ALDL?

[jxx]

This i cannot answer.
It should work but i don't have a working 808 to bench test.
Hopefully someone gets a one rom and the ALDL firmware and can actually test and tell me what needs to be changed/fixed/or if it works.

Currently the emulator is working as a memcal replacement with RT tuning.
The data channel really was just optional instead of having to use a data cable for feedback.

[jxx]

Here's a test beta version enabling ALDL 8192 8N1
Added switches that Invert TX/RX for testing purposes, doesn't need all the options but to make it easier to flash the onerom for bench testing i've shoved them all in.

Best to use a schmitt-trigger buffer between the onerom and the ecu.

Functionally the same as v0.1.2 in the first post for normal and disabled modes.

****REMOVED FOUND BUG IN UART CODE****onerom-emu-flashtool_0.1.3-beta.zip

[antus]

So does RT work over ALDL?

If you mean with just an ALDL link and OSE plugin I think it wont, because then it'd need to be emulating the DS1245Y which has a larger pin count and the write pin, and unless that has been added in both hardware and software then it cant work. Assuming its the 28 pin footprint then it can't at the hardware level as it is just emulating 27C EPROM hardware. It would need the header to be lengthened to accept the write pin, and then connect that in to a GPIO and have software support for it added too.