HP Forum Archive 14

[ Return to Index | Top of Index ]

HP-41 HEPAX poll (for the NoVRAM module) Message #1 Posted by Diego Diaz on 14 Apr 2004, 9:32 a.m. Hi all, I have to make some decissions regarding the final design of the NoVRAM module, intended to minimize the parts count... Just to summarize for those not yet aware: NoVRAM is a 16k (for now) RAM box in a standard HP-41 module enclosure. It fully emulates the Advenced HEPAX module (including auto allocation in the lowest available page and RAM write protection) while provides two extra ROM pages, in a Clonix-like way The question I'd like to poll you is about the automatic mapping ability of the Adv HEPAX. i.e. the 4 RAM pages are mapped to 8 - B if you plug the module into ports 1 or 2; or mapped to C - F if the module is plugged into ports 3 or 4. I've been trying to figure out a good reason for keeping this feature in the NovRAM emulation (NoVRAM prototypes actually have it implemented), but so far it only takes some code space and requires two extra resistor for the B4 line sensing, and a somewhat more complicated PCB routing. I've prepared a new prototype without B4 sensing (it always placed its memory in 8 - B pages, and have been unable to find any illegal configuration (appart from those stated in the HEPAX manual, which BTW have nothing to do with the mapping feature) As I'm not one of those fortunate "real" HEPAX owners/users I'd like to ask you all if is there any reason for what such a feature should be preserved or not? Thanks in advance. Diego. PS. I'll leave for holidays in a few hours, but will take a look to the 'net everynow and then. Oh, Just a good news I've managed to re-shrink the code so the 6th page is again available!!

Re: HP-41 HEPAX poll (for the NoVRAM module) Message #2 Posted by 聲gel Martin on 14 Apr 2004, 10:34 a.m., in response to message #1 by Diego Diaz Diego, The RAM mapping is in my opinion a convenience thing. It is much better to have the RAM mapped to the position where the module is plugged in (and its adjacent one). When writing MCODE or uploading programs to HEPAX RAM it's always important to know where you are (address-wise). Consider otherwise the confusion that it can create, when actual ram location has no linking to its physical position. Say that you want to plug in a module into an apparently "empty" slot (nothing there or aside), but it turns out to be "occupied" by the module *underneath* it ! Having said that, a couple of differences between HEPAX and NoVRAM (BTW, how will you christen it?) apply: 1. NoVRAM isn't volatile when unplugged. This can advise for a flexible mapping as well. 2. HEPAX could add more RAM, up to 32 kB. In this case it was imperative that the addressing was linked to the actual port location! So my conservative 2 cents is: well worth the admission price, go ahead ad add those 2 transistors! Best wishes and enjoy la feria! 聲gel

Re: HP-41 HEPAX poll (for the NoVRAM module) Message #3 Posted by Bill Holes on 14 Apr 2004, 6:37 p.m., in response to message #1 by Diego Diaz Hi Diego, Like you, I'm not one of those fortunate few that own a real HEPAX... However, I believe this will all change after the release of NoVRAM, which in my opinion will be far better than owning a real HEPAX. Having said this I must agree with Angel's vote to go ahead with adding the "two extra resistors". I see no sense in limiting any of it's functionality after all the hard work you've put into this effort. I doubt the extra two resistors and additional board traces will increase the cost beyond an affordability level for anyone who understands it's true value. Enjoy your very much deserved holidays!!! Best Regards, Bill

Re: HP-41 HEPAX poll (for the NoVRAM module) Message #4 Posted by Lawrence Hood on 14 Apr 2004, 7:59 p.m., in response to message #1 by Diego Diaz Diego, As a proud owner of a STD HEPAX I recommend the inclusion of the additional resistors. I'm looking forward to purchasing your latest creation!! Thank you for your continuing efforts, Lawrence Hood

Re: HP-41 HEPAX poll (for the NoVRAM module) Message #5 Posted by Christoph Klug on 15 Apr 2004, 5:24 a.m., in response to message #4 by Lawrence Hood Dear Diego, yes - please add the two transistors to your hardware design. May be you are able to use a small double transistor array or two transistors housed in one case... For easy installation we want two adjacent ports, like a 16KByte straight ZEPROM... Regards - Christoph

Re: HP-41 HEPAX poll (for the NoVRAM module) Message #6 Posted by Vieira, Luiz C. (Brazil) on 15 Apr 2004, 12:08 a.m., in response to message #1 by Diego Diaz Hi, Diego; Congrats; as many HP41 users, reading about your achievements leads us to "dreams-come-true" about to happen... Now for the "questions". My only experiences out of the HP41 Owner's Manual are related to Synth Programming (Bill Wickes) in the late 80's. After that I was known about the CCD, PPC and HEPAX ROM's after reading some posts here about three years ago. Yeas, a big lack in time. All I know about the inner HP41 is after that. What I want to know is mostly related to something I read (if I am not wrong) about the HEPAX modules. Is it a fact that you can plug two of them in one HP41 so one module uses the other one's RAM? I am not sure if this is actual information, but if it is, would one of the operation modes (with or without resistors) missallow this particular behavior? As I am not even sure if this is what I read about the HEPAX, please consider that this very information may also be unaccurate. I wish you "boas f廨ias" tambi幯, Diego. Cheers. Luiz (Brazil)

Re: HP-41 HEPAX poll (for the NoVRAM module) Message #7 Posted by 聲gel Martin on 15 Apr 2004, 3:15 a.m., in response to message #6 by Vieira, Luiz C. (Brazil) Luiz, There were (are?) two types of HEPAX modules. The first one has a combination of both ROM (its O.S. so to speak) and RAM. There are two sub-types, depending on whether there is 8k or 16k RAM included. The second type is RAM-only HEPAX, also with two modalities: 8k or 16k. And yes, you can plug one of each type together, for a total of 32k RAM plus its OS ROM, camuflaged in the internal ports (assuming space is available - it is a 16k ROM, but because it is heavily bank-switched, it uses only one page). You could also use the RAM-only HEPAX modules in conjunction with Eramco MLDL, RamBOXes, or any other "system" that can access Q-RAM space (i.e. RAM in the ROM addresses). For example, the David Assembler can be used to program that RAM. Plugging two of the first type together doesn't make sanse, and probably won't work at all. I'm almost sure that Diego's NoVRAM will also be capable of using the RAM-Only HEPAX, for those few guys who have them, but honestly this is not the main point here: the NoVRAM by itself will be a superb replacement of the HEPAX, with the added bonus of having two extra ROM pages available (Clonix-like). Best, 簍

Re: HP-41 HEPAX poll (for the NoVRAM module) Message #8 Posted by Werner (Germany) on 18 Apr 2004, 3:05 p.m., in response to message #7 by 聲gel Martin Dear Diego and dear 聲gel, first of all I would also be glad if you added the extra effort to having the opportunity to have a deterministic port occupation behaviour. As you may remember some time ago I had a severe problem with my W&W-RAMbox II but with the help of my ADV HEPAX I was able to fully recover it at last. This was only possible as the port occupation could always be taken into account because I had to use the HEPAX in conjunction with other ROM modules in order to mask some internal address space. Second, I do not know but I believe that you could use two HEPAX with RAM+ROM. Probably the ROM of one of them would cover the other identical ROM, and all RAM from both HEPAX modules would be acquired and mapped to different addresses. I cannot test this scenario but I believe it could work because my ADV HEPAX also acquires the RAM of the W&W-RAMbox if I plug it into my HP-41CY. It auto-detects all available Q-ROM and assigns all*) 4k blocks to different address pages. So, this seems to be a special property of HEPAX functionality. This is why I believe that a HEPAX-ROM/RAM-module will also detect an acquire the RAM pages of another HEPAX-ROM/RAM-module and not only of the special pure-RAM modules. *) of course not both parallel 32k-layers of a RAMbox II can be allocated. Best regards, Werner Edited: 18 Apr 2004, 3:07 p.m.

[ Return to Index | Top of Index ]

Go back to the main exhibit hall