HP-41CL & NoVRAM/NoV-32 issue fixed... ;-)

[Diego Diaz]

Hi all,

After some diving into NoVRAM and NoV-32 code, I've fixed the issues that prevented them to properly work on HP-41CL machines.

Details on previous thread: here.

The updated code is here, just replace the enclosed files in your working directory and reprogram your NoV's as usual.

Now the whole range of Clonix and NoV's modules are compatible with HP-41CL.

As a "collateral effect" I've also added the RAM shadowing feature to the NoVRAM and NoV-32 modules... ;-)

Please let me know of any issue you may find.

Hope this helps.

Enjoy your 41's and best wishes for the upcoming Xmas season.

Diego.

[Meindert Kuipers]

Could you shed a light on what the issue was? I expect my CL soon and could run into the same type of problem when it needs to work with the MLDL2000,

Meindert

[robertoataulfo]

Meindert I did not understand what would be the use of MLDL2000 with 41CL?

I have both but on different machines.

I do not see what use to use the two simultaneously.

Can you help me this question?

Edited: 21 Dec 2011, 6:20 a.m.

[Meindert Kuipers]

Obviously this type of knowledge may be interesting just for fun. And there may be several use cases for the MLDL2000, if only for easy transfer of ROM images from the PC to MLDL2000 to the 41CL.

Meindert

[Massimo Gnerucci (Italy)]

Thank you Diego! (as usual...)

Season's Greetings,
Massimo

[Ángel Martin]

Hola Diego, I tried my NoVRAM with the new configuration files quickly this morning and sure enough the start sequence works great, and the HEPAX relocates itself to page#6. I believe you said that page #8 must be free during the start-up process, but I might be wrong. Just in case the MMU configuration is YFNS on page#E and CLUTILS on page#F - nothing else so no conflcits.

There is however a (new) problem: the two non-hepax ROM images burned (the "back-seat passengers" so to speak) are not there!, Seemingly not on the VRAM at all, as they don't show up on the standard CX either.

I'm using pages #C and #D for the "passangers". I tried it twice, starting the burning from the scratch. Same result, and the disconcerting part is that those images are not shown in the non-CL machine either. Is it possible that the upgraded files have caused this?

I'll try this evening with other pages to see if the same issue occurs - it may be me making some mistake, who knows.

Saludos,
Ángel

[Diego Diaz]

Hi there,

Meindert, it was a clock phase missing issue. Since the code in the NoVRAM is so tight (just 2Kbytes) I have to run some "tricky" code in order to keep NoV "in time".

At some point it seems that the Phi pulse witdh on the CL is slightly narrower than the original 41 and the NoV lost its pace.

After moving a few instructions up and down a bit, (namely 166.66nS in two different Phases) NoV can now stay in sync as it should.

Due to the very different concept in your MLDL2000 (HW interfaced) I doubt this trouble could arise in your case, but it is always a pleasure to find someone who can listen to these "old techie tales"... ;-)

Thanks for your (as usual) kind words Massimo, and Merry Xmas to Italy!

Ángel, you're right (sorry) I switch the "back seats" OFF, to ease the testing procedure, and forgot to turn them back ON once the timing issue was fixed.

I've already uploaded the "fixed" update, it may take a few hours until my ISP will actually "update" it on-line.

In the meanwhile, you can use the NoV-32 files to configure your NoVRAM, it will work as expected (of course just one 16K RAM block)

Enjoy your toys!!

Diego.

[Eric Smith]

Does this apply to Nov64 also?

[Diego Diaz]

Hi Eric,

To your question, NoV-64 has a different microcontroller, different code, and duoble program space, so it works with HP-41CL.

Update is applicable only to NoVRAM and NoV-32 modules.

Cheers.

Diego.