STFM 4MB RAM EXPANSION SUCCESS

[gigi92]

I reassure you the switch on the box it was only for the photo. The idea was to be able to switch with the box closed (switch on the outside of the case). I haven't tested without the wires yet. The strangest thing is that it worked well when I made the connections two days ago.And now and it works for 4MB (wires close) and not for 1MB (wires open).
A time i have the switch just the wires (like other post).

[gigi92]

First, thank you for your really wise advice. By separating the wires from each other, it works! If I stick them together again, it crashes. I'm curious that you explain this phenomenon to me and why we don't have this with the connected cables (4Mo).
It's a first win

[exxos]

First, thank you for your really wise advice. By separating the wires from each other, it works!

I'm curious that you explain this phenomenon to me and why we don't have this with the connected cables (4Mo).

It's "Parasitic capacitive coupling" between the two wires. When 4MB is selected, this does not matter because each wire has the same signal and essentially shorted out.

I ran some quick numbers with a similar cable I presume you used..

My display on my meter is not great unfortunately. Measured almost 14pF which I was slightly surprised at ( I did zero the meter before the tests) . I did expect something there, but I thought it would be the order of magnitude lower than that.

IMG_2255.JPG (111.46 KiB) Viewed 3206 times

I quickly measured A9 on a simm and saw pulses up to 6MHz. It varies and I did not spend much time investigating. But it is not needed to be entirely accurate this quick test anyway.

So basically running some quick numbers works out at 1,894ohms.

This would be why when the other guy modified the 100K resistor to 10K it also failed. Because the capacitive coupling between the cables was less than 2K!

So basically that 100K resistor would probably have to be 1K or lower to work as intended. Or wire the switch to GND as in my previous post. Of course separating the wires de-couples that effect also, so it works as intended.

Welcome to the wonderful world of screwy electronics

[gigi92]

I have a new problem, I finally welded the two studs and now it only detects 1MB, I want to cry ....
What to test ?
Thank you in advance.

IMG_20240526_104552.jpg (392.05 KiB) Viewed 3142 times

[dgw]

IMHO it looks like too much solder on pads may have caused ground connection. I would remove excess, link only needs to be fine wire and little solder on pads. If that is the case hopefully MMU MAD9 sustained being grounded.

[gigi92]

I redid the soldering, we can't do it any finer.it didn't change anything.
I changed the MMU by another one and it's good now...
Strange but hey I stopped crying.
Thanks again for all the advice

IMG_20240526_180708.jpg (344.29 KiB) Viewed 3099 times