![Very Happy :D](./images/smilies/icon_e_biggrin.gif)
I just wanted to write down a short summary for people who may have the same plan.
That was kind of a brutal job, to cut of the patchboard of my issue 2-PCB and to desolder all those chips and components, which do not like to give up easily. Most of them have ugly bent wires on the solder-side of the PCB and stick to the pads. But in the end, I managed to remove the chips without cutting of the wires. I'm pretty confident, that the 16k-DRAMs survived the procedure. The PCB did, the lines and pads did not come off immediately.
That's what it looked like: I decided not to use the patchboard any more, since I had to desolder the old DRAMs on the mainboard anyway - so way not use that space for the new chips? I ended up with: Final job: changing the decoder a little bit, adding a few wires and cutting one line (red cross): Before testing, the voltage jumpers (red circle, picture 2) had to be adjusted. Since I was lazy an hoped that the setting "d) Set the jumpers to l,4,5,8,l0", as mentioned in the article, would be ok, I did so and immediately fried two of the eight new DRAMS.
![Surprised :o](./images/smilies/icon_e_surprised.gif)
Now it's working and, for example, DUPLI-DISK2 recognizes a Dragon 64 and can therefore process bigger files, without overwriting it's own memory.
Finally: The Dragon-Dos Disk controller (LAFUMAT) is not affected by the changes of the decoder circuitry. Obviously it's only using IO-addresses of the top memory page.
All in all, the job took me about half a day.
By the way: Is there anybody out there, who is repairing Dragon 32s on a "regular basis" and needs a few spare parts (16K-DRAMS)?