Genmod questions

[+mizapf]

There are only few people who have a Genmod; maybe someone can give me more information.

 

I found the description of the Genmod on WHTech, but it is just a setup instruction manual, not a technical manual that describes what actually happens in the Genmod. I suppose we do not have those details anymore ...

 

Essentially, the Genmod cannot influence the operation of the Gate Array itself. It is piggy-backed on it, with some more lines connected, and creates the respective signals, while some traces have to be cut so that the Gate Array signalling is inhibited, and the newly created signals may be effective. The Genmod has no idea of memory pages because those are purely internal to the Gate Array, so it must conclude from the selector lines what is about to happen.

 

As I learned from the manual, the 32K expansion must be removed because the RAMENX* cable is a physical hindrance to the mounting of the Genmod daughter board. I have some more questions on that:

 

- Does the Memex provide the SRAM in pages ec-cf, or is it the on-board SRAM? I did not see any instruction to cut the trace of the RAMEN signal. Accordingly, the Gate Array would then turn on the SRAM on the board. If the SRAM chip were socketed, this could be easily checked by pulling it out, but I think it is soldered.

 

- How is the DRAM turned off? As I saw on the schematics, the DRAM does not have a OE* (output enable) signal but delivers the data as soon as the CAS* signal is active. This would mean that the two CAS lines (for both arrays of DRAM) would require cutting, which is not mentioned. The only other way that I could think of is to remove the power from the DRAMs. The "TI Mode switch" is wired to the Genbod board, so this does not reveal much.

 

- I read about the necessary AMA/B/C decoding on the peripheral cards. Genmod introduces two more lines, AMD and AME so that we have a 21 bit address width in the P-Box. I did not read anything about a AMD/AME decoding on those cards. This would again mean mirroring; I guess this is the reason why some pages are blocked.

[+9640News]

I don't know how, but I can tell you a few things here.

 

One switch on the GenMod switches memory pages >00 to >3F from the Geneve to use of the memory on the Memex.  I think this was called TIMODE ON/OFF.  This is only available if you a GenMod system.  To do this, requires having a GenMod Eprom, otherwise the memory scan fails at page >3A and would also fail at page >3C if the memory scan proceeded if it had a Rave Speech card, or the old A/D card in the PEBox.  Not sure if there were other cards that mapped into those page boundaries.  If TIMODE was OFF, then one could not go into GPL mode because the gate array required page mapping to be on the Geneve rather than using the same memory pages on the Memex.

 

Unmodified cards, which I think were frequently the older Horizon ramdisks, and some Myarc and CorComp cards also required modification as they would also answer on pages >3A, >7A, and >BA.  By default, DSR access maps in page >BA which is generally not present in an unexpanded system.  Pages >F0 to >FF are already unuseable on the Memex as the Eprom occupies two of those pages.

 

The AMD/AME necessary decoding is shown on directions for use cards not fully decoded for use with the Memex/GenMod.  I'm pretty sure they are on whtech, if not, I can scan them again.  There is also a switch on the Memex to allow you to use those cards without modifying them, but it blocks out all pages with a >xA above page >40 so you lose a good chunk of memory.  I know memory testing pages >4A, >5A, .... >EA shows no ram.  I don't know if it turns off power to a bank of chips, or blocks the signal somewhere. 

 

As I recall, if you had added the extra 32K ram to a stock Geneve, the addition of the GenMod required the extra 32K to be removed.