RomRaider

I rewired "upper address lines to power\ground instead of programmer outputs."
Upper addresses in dump goes to zero, but all other data is also corrupted.
Perhaps all lines needed for correct reading. Unfortunately, this method not working((

I have found opensource CLI for my programmer https://gitlab.com/DavidGriffith/minipro
Perhaps it could be used to slow down programmer operations?

Can't compile minipro for win32, perhaps cygwin for 32bit systems is too old...

Compiled it under Ubuntu. After changing buffer size from 0x001 to 0x800, in rom settings, maximum time for reading is 17 minutes. Minimum 3sec.
But nothing changed... fck...

Overclocked it to 17mhz, vcc up to 5.5v. And get a half of code, from 0 to 20000. Interesting, where a bootloader there?

There is no isolated bootloader code.
All the ROM will be completely erased before flashing.
And you will have a brick if something goes wrong while flashing.
Completely similar k-line Nissan ecu's.

For this reason I do not propose customers to reflash those ecu's on their own.

I mean instructions or CID's, to read all rom)

Few times mcu was started at 20mhz. But still can't get stable reading higher than 0x20000.

Ordered old programmer Willem from China, works with LPT port. But they on holidays now. Need to wait...
It should be slower, then my USB Minipro. If this not help... then i will capitulate.

Some company has programmers which can read this MCU. https://www.mqp.com/hitachi.htm
But they do not produce programmer stuff anymore.
In his software presented needed options to slow down access time:


Pin-Master 48 programmer. Found one on eBay. But it will be too expensive.
https://www.ebay.com/itm/MQP-PIN-MASTER ... 4497716021

Victory! Old LPT programmer, with some modifications, did the trick!))
Don't need to rebuild PCB at all... And tones of decoupling capacitors, also don't needed.

Final look, before desoldering.

Yikes, no wonder the capacitors you added didn't help. They're much too far away from the uC, and also not the right type (you really want small ceramic capacitors for this to minimize series inductance).

In case you ever make another PCB for a microcontroller, make sure to add in plenty of decoupling immediately surrounding it, like this:

Now it reads fine, even without capacitors. I don't need them anymore...
Problem was in programmer, not in elements or PCB.

Yes, likely because the newer programmer is running slower than your previous one, so the transient current requirements of the uC are less demanding.

I was just saying that the capacitors you added to your board weren't actually doing anything for decoupling (so they won't make any difference for the problems you were seeing). Something to be aware of if you put together another PCB in the future, is all.

Got full dumps from 1.5 AT & MT (EJ152), 2.0 AT & MT (EJ202).
EGR and Rear O2 disabling tested. Swap from EJ152 to EJ202 tested.
Remote flashing is possible with simple VAG KKL aka K-Line cable.

Just photo, ecu's after experiments and readings. Few of them survived after soldering mcu back on pcb.