HP 41CV memory module

[Pieroz]

Dear hp 41 friends best regards!
Does anyone know if the data saved in the x-function/x-memory module are kept in case of "memory lost" condition?
Thank you for your answer.

[Thomas Okken]

I don't think the contents of extended memory are preserved. It's not quite the same, but I just tried ON + Back Arrow on my HP-41CX, and it caused MEMORY LOST, and wiped extended memory as well.

[rprosperi]

Definitely not. When you encounter a MEMORY LOST, all data, programs and settings are lost, including ExtMem, Key Assignments, etc. I don't know if all RAM contents are in fact set to zero (I suspect they are) but all pointers and settings are reset to the same as initial power-on condition.

[JMBaillard]

However, all the data, programs in HEPAX or NoVRAM modules are saved.
Best regards,
Jean-Marc.

[rprosperi]

Important (!) clarification Jean-Marc, that should have been mentioned. Thanks for adding that.

[hth]

(11-30-2020 10:50 PM)rprosperi Wrote:  Definitely not. When you encounter a MEMORY LOST, all data, programs and settings are lost, including ExtMem, Key Assignments, etc. I don't know if all RAM contents are in fact set to zero (I suspect they are) but all pointers and settings are reset to the same as initial power-on condition.

They are reset to zero. However, here is where it gets interesting. It is noted in a comment in the firmware that they do not need to be reset the whole memory range, but HP did not dare to do it.

When I made my update for larger extended memory, I faced the problem that it takes a good amount of time to clear 15-16 pages of memory compared to 4 (which also takes a moment or two). I decided to change it to only clear page 0 and 1 (as suggested in the comment). This clears the base module of extended memory and all normal data registers, but not the additional extended memory modules. This makes master clear twice as fast compared to earlier.

This works the same way as when you insert an extended memory module, which will contain garbage in it once inserted and there is no code to clear such module (apart from the one doing the master clear). Pointers in the base module tells if the extra modules has been initialized and it never peeks into them to see if they contain anything interesting (garbage). It will always write something before it reads anything in these additional extended memory modules.

[rprosperi]

Thanks Håkan, for that explanation; I suspected the 41 OS did write all zeros, but wasn't sure. The HP design team were clearly conservative... I agree the first 2 pages are enough, as I don't believe there is any means (except curious MCODErs) to access any of that RAM except as you write files and expand into it.

[Massimo Gnerucci]

Karl, in the 4th post in the thread here suggests that if you quickly swap a X-functions module between two C or CV, its X-memory contents are preserved.

Never tried this myself.

[rprosperi]

(12-01-2020 07:24 AM)Massimo Gnerucci Wrote:  Karl, in the 4th post in the thread here suggests that if you quickly swap a X-functions module between two C or CV, its X-memory contents are preserved.

Never tried this myself.

I can confirm that this sometimes does work, though usually not. I've had success a few times and whether or not it works seems to be dependent on multiple factors, not only speed. But it's worth giving it a try.... Good luck.

[Pieroz]

Than k you for all the answers, very interesting.
At this point i think that the safest solution to saving data is the card reader; other operations using xfunc/xmemory module it could be risky.

[Sylvain Cote]

(12-01-2020 01:34 PM)rprosperi Wrote:  

(12-01-2020 07:24 AM)Massimo Gnerucci Wrote:  Karl, in the 4th post in the thread here suggests that if you quickly swap a X-functions module between two C or CV, its X-memory contents are preserved.

I can confirm that this sometimes does work, though usually not. I've had success a few times and whether or not it works seems to be dependent on multiple factors, not only speed. But it's worth giving it a try.... Good luck.

It was very popular in the 1980s to add a capacitor in a XFM module to allow longer data retention when outside the calculator.