ROM dump tutorial for Pocket Pikachu & co.?

[vladkorotnev]

I would like to try my hand at getting this to run Pocket Sakura and Pocket Hello Kitty, as they seem to be a similar hardware platform to the Pocket Pikachu.

Is there any tutorial on how to remove the epoxy and scan the chip correctly, and any tools to turn it into a ROM file?

Thanks!

[azya52]

Hello Kitty probably uses the same chip, but Pocket Sakura is definitely different, although it is likely somewhat related.

I remove the compound with a scalpel and a hot air gun at around 150°C. If the scalpel does not remove it completely, I additionally boil it in rosin.

Next, the metallization and polysilicon layers need to be removed. For this, I use aluminum soldering flux. The key point here is not to overexpose the chip and to wash off the flux immediately after the polysilicon has been etched.

After that, the bits need to be developed for visual reading. This is the most problematic step, since I am not very strong in chemistry, and achieving uniform development is fairly rare for me. At the moment, I use a mixture of the same flux with 6% hydrogen peroxide (I think 3% would also work) in an approximate 1:5 ratio as the developer.

(!) All of this chemistry can be extremely harmful, so proceed at your own risk. At the very least, the workspace should be well ventilated.

Once the bits become visible, the ROM is photographed under a microscope, resulting in something like this:
https://commons.wikimedia.org/wiki/File:PP_ROM_DASH_ETCH.jpg

Unfortunately, it is difficult to automatically read this with an acceptable errors, so I manually marked all the bits and then converted them into a binary file using a Python script. For a ROM as large as the one in PP, this step is very time-consuming.

If you really want to do this, I can describe each step in more detail and answer any questions.

[dev-zzo]

Thank you for sharing your process! I'd really like to learn more on bit staining and try to reproduce it: which specific flux do you use for this? I take it is mixed as 1 part flux and 5 part peroxide?

[azya52]

I tried two different aluminum soldering fluxes available in Russia. One had no name or ingredient list at all, while the other was called "Ф-64" and listed "Triethanolamine (TEA), fluorides, corrosion inhibitors, wetting agents (surfactants), and deionized water". It's likely that the fluxes from AliExpress have roughly the same ingredients. You're right about the ratio, but the mixture actually works over a fairly wide range, and I'm not convinced 1:5 is optimal.

[dev-zzo]

Is there any heat applied as well, or does the mix work at room temp?

[dev-zzo]

Hi again! I am sorry to bother you this much @azya52 but could you give any indication for temperature and timing of the staining process? So far I did not get much in terms of results...