On Wed, Mar 27, 2024 at 5:18?AM Jorge Gonzalez <jorgegl50@...> wrote:
Why do you think the acquisition memory is the problem?? I am not able to see any pictures of the screen that are? related to the problem. Have you posted them? I have a TDS754D that had a memory?problem in channel 4 and, later, in channel 3. The memory modules in my acquisition board are?AS7C256_12JC (32K x 8 bits). (These allow?option 1M.? ?Your memory modules are 4 times larger, which?allows? option 2M very probably.)
You can access the acquisition memory from the console (serial, RS232 link using an adapter) and display the memory contents at the terminal. "Help" tells you the command to use:
???d adr[,nunits[,width]]
The? memBaseAdr 0x73c0020 above corresponds?to the 68040 processor address space. The correspondence between the acquisition processor space and the 68040 space can be inferred from the "TDS 520B Mod CM? Digitizing Oscilloscope Component Service Manual 070-9710-03" , Table 1-13, p.1-41. One acquisition space location corresponds to two 68040 locations.
I applied a triangular?signal to the bad channel, with an amplitude? as to fill 80-90% of the display. I set non-repetitive, single acquisition mode.Then I inspected the memory contents and observed?that one every 16 locations did not follow the linear pattern implied by the triangular waveform (16 memory modules are accessed in parallel?during acquisition).
Determining the memory mapping to modules is more tricky. I observed?the OE and data out signals from memory when reading single locations. I do not remember the details but I was able to?identify the bad modules precisely.
I used "ChipQuik REM4.5?Chip Removal Alloy" to remove the chips. This is a low temperature solder alloy that makes the removal of SMD chips easy and safe.