1 Meg Super AMS Discussion Thread

[+TheBF]

3 hours ago, MikeV said:

Circuits are generally on or off. When SAMS memory is accessed, does the identical memory address (and its contents) in SAMS overwrite its counterpart in the 32K, or is the circuit (pathway) simply rerouted to its SAMS counterpart and the 32K address then ignored? What do the 'mappers' actually do?

 

ChrisB. said more or less the same thing about a DSR approach. However, over the last 30 years or so, ramdisks (which have DSRs) have been used many thousands of times in conjunction with TIs. SAMS - not so much. I am puzzled why this flexibility has not proven itself and become commonplace by now.

It does not overwrite the information, but the same addresses are used for a different chunk of memory.

It really is like replacing all the houses on a city block with different houses and therefore different families inside them, but the house numbers are all the still the same.

 

You can for example take the memory at location HEX 2000...2FFF and replace it with one of 256 different chunks of memory with 1Mbyte SAMS card. The addresses are still 2000..2FFF but the contents will depend on the SAMS piece you swapped in to that space.

 

You could also think of it like DIM A(4096)   but you can replace A() with a different 4096 numbers at the flip of a switch.  Same array indexes, different numbers inside.

 

A DSR is fine for a device with a finite set of functions.  RAM is way more general purpose and getting everyone to be happy with one way would be problematic. Every programming environment on the TI-99 is a bit different.

 

 

[GDMike]

Hmmm.. change out the people who live there....ok...I'm listening 

?

[MikeV]

Quote

Ahh I never had the pbox card version, the sidecar did not provide 32k..  based on the manual here: http://ftp.whtech.com/datasheets and manuals/Hardware/CorComp/corcomp 256-512k memory plus card manual.pdf   that TAKES AWAY the 32k from the memory card memory and dedicates it to 32k expansion. ALSO it does NOT increase program space beyond 32k.   It's only a RAM disk.. PERIOD

Edited November 25, 2020 by MikeV
deleted text

[MikeV]

18 hours ago, arcadeshopper said:

Ahh I never had the pbox card version, the sidecar did not provide 32k..  based on the manual here: http://ftp.whtech.com/datasheets and manuals/Hardware/CorComp/corcomp 256-512k memory plus card manual.pdf   that TAKES AWAY the 32k from the memory card memory and dedicates it to 32k expansion. ALSO it does NOT increase program space beyond 32k.   It's only a RAM disk.. PERIOD   

 

 

That went well!!!

 

I have the sidecar SAU as well. The SAU will provide 32K if another memory card is not present. In addition try this:

 

Have the SAU powered, switch up. Turn on the TI without powering up the PEB. Go to XB and do a SIZE command and you will get 24,488 bytes of Program Space and 13,918 bytes of Stack Free. If you work from TI Basic in this setting you can load and run the large TI Basic programs originally written for cassette, e.g. Starship Pegasus, from the ramdisk and still have sufficient memory.

 

As you pointed out it does not provide additional RAM beyond 32K to my knowledge. My earlier post was in regard to its ability to sense another card, act accordingly and why couldn't the SAMS do the same. I believe Asmusr answered that question, though I'm still mulling that over...

 

 

[MikeV]

16 hours ago, RXB said:

CALL FILES is only needed to fit more program into a space taken up by files access buffers. 

Thus when you do a CALL FILES(1) you are deleting the two extra buffers not used by the current program you are loading.

See the normal default set up for XB or Basic is CALL FILES(3) so more memory is available using CALL FILES(1) 

RXB 2020 has a new feature of CALL FILES(0) that opens up even more memory, of course you can not use OPEN any more.

(Well you can but you will get some strange results as it messes up temporary variables sometimes.)

That will have its uses. I've done it from Classic as well.

[MikeV]

16 hours ago, TheBF said:

It does not overwrite the information, but the same addresses are used for a different chunk of memory.

It really is like replacing all the houses on a city block with different houses and therefore different families inside them, but the house numbers are all the still the same.

 

 

Thank you for the assistance. I think I'm finally getting there. Everything is actually still there...

 

Just for grins I asked the wife yesterday, "What do you know about memory maps?" She replied, "It's what you use when you are trying to find things, doesn't seem to work very well at all."  My question was apparently successful as she was grinning from ear to ear!

[RXB]

6 hours ago, MikeV said:

That will have its uses. I've done it from Classic as well.

Hmmm if you do a CALL FILES(0) in normal XB or any other XB other then RXB 2 things happen:

1. NOTHING I mean nothing do a size and nothing has changed.

2. SYNTAX ERROR as 0 is not allowed.

 

These two results depend on you DISK or RAMDISK or Hard Drive card used.

 

As RXB is in GROM and over rides the DSR for these cards it does the CALL FILES(0) before the DSR can do anything.

[RXB]

6 hours ago, MikeV said:

Thank you for the assistance. I think I'm finally getting there. Everything is actually still there...

 

Just for grins I asked the wife yesterday, "What do you know about memory maps?" She replied, "It's what you use when you are trying to find things, doesn't seem to work very well at all."  My question was apparently successful as she was grinning from ear to ear!

One crazy thing you can do with SAMS is put the same exact memory into 2 or more locations and any change to any of them is duplicated in all.

Example: >2000, >3000  and >A000, >B000 and >C000, >D000 all have the same pages of 20,21 and 20,21 and 20,21

So in RXB I had a sector copies program that would read DSK1 into >2000, >3000 (you know 20, 21) and DSK2 would be copy.

But to add fun to it I also made another copy in DSK3 while the entire thing ran from RXB in >E000 and >F000

 

This was fun so I made a Track Copies version, and could modify the copy disk before written to get around copy protection schemes.

[Tursi]

20 hours ago, MikeV said:

That will have its uses. I've done it from Classic as well.

Just to note, this only works because Classic99 has its own DSR - technically its own disk device. Otherwise it would be impossible to do many of the tricks that it does.

 

 

[Sinphaltimus]

Am I missing something? I've searched and cannot find my particular issue listed or any notice about it.

I purchased a sideport 1MB SAMS to replace my 32K sideport. 

My setup has always been Console->SpeechSynth->32k->TIPI.

So I went with  Console->SpeechSynth->SAMS->TIPI

but when I went to connect the SAMS/TIPI duo to my speech synth I discovered that the SAMS sideport connector is too short.

I even compared it to my 32k and in fact, it is too short.

I ordered this preassembled and it simply does not connect to the sideport of the speechsynth.

 

What are folks doing in this case? Are they modifying their speech synth case (seems a shame, destructive modification) or are they removing the speech synth case altogether and leaving it exposed?

 

Was it not possible to to make the connector a tad longer to fit the speech synth sideport?

Edited August 30, 2021 by Sinphaltimus
Typo as usual.

[xahmol]

1 hour ago, Sinphaltimus said:

the SAMS sideport connector is too short.

Following this with great interest as I have an order in processing for a SAMS at Arcadeshopper and use the exact same setup now: console -> speech synth -> 32k -> TiPi.

And also ordered the SAMS to replace the 32k in this setup.

 

So if this really is a generic issue this will also be a problem for me.

Edited August 30, 2021 by xahmol

[GDMike]

I've yet to order one, because of low funds, but intend to one day. I currently have a Sam's peb card that works well. I'll watch for answers here.

[+arcadeshopper]

Am I missing something? I've searched and cannot find my particular issue listed or any notice about it.

I purchased a sideport 1MB SAMS to replace my 32K sideport. 

My setup has always been Console->SpeechSynth->32k->TIPI.
So I went with  Console->SpeechSynth->SAMS->TIPI
but when I went to connect the SAMS/TIPI duo to my speech synth I discovered that the SAMS sideport connector is too short.
I even compared it to my 32k and in fact, it is too short.
I ordered this preassembled and it simply does not connect to the sideport of the speechsynth.
 
What are folks doing in this case? Are they modifying their speech synth case (seems a shame, destructive modification) or are they removing the speech synth case altogether and leaving it exposed?
 
Was it not possible to to make the connector a tad longer to fit the speech synth sideport?

Any reason you didn't contact me directly? I know the first batch that were made for me have shorter extensions on the side port connector, though it should still be long enough to work, it may be close and require you to make an adjustment to make it fit like possibly removing the door on the SS (non destructive) send me an email with a photo

Greg

Sent from my LM-V600 using Tapatalk

[+arcadeshopper]

Following this with great interest as I have an order in processing for a SAMS at Arcadeshopper and use the exact same setup now: console -> speech synth -> 32k -> TiPi.
And also ordered the SAMS to replace the 32k in this setup.
 
So if this really is a generic issue this will also be a problem for me.

New orders won't have an issue as I am building them and using the same extensions as we're used in the 32k sidecars

Sent from my LM-V600 using Tapatalk

[Sinphaltimus]

1 hour ago, arcadeshopper said:

 

 

Any reason you didn't contact me directly? I know the first batch that were made for me have shorter extensions on the side port connector, though it should still be long enough to work, it may be close and require you to make an adjustment to make it fit like possibly removing the door on the SS (non destructive) send me an email with a photo

Greg

Sent from my LM-V600 using Tapatalk
 

 

I didn't really think to contact you directly. I honestly thought this might be a known or already reported issue here in the thread but couldn't find anything. I did not suspect this was an assembly issue - I thought all assembled cards came from Ksarul (did I spell that right?). So figured it might be some limitation I did not know about.  I'll send a photo to your email as soon as I can. My apologies for not thinking to contact you first.

[Sinphaltimus]

1 hour ago, arcadeshopper said:

New orders won't have an issue as I am building them and using the same extensions as we're used in the 32k sidecars

Sent from my LM-V600 using Tapatalk
 

Greg, No worries. I am working from home today and quickly removed the side door from the SS and the SAMS/TIPI combo slid right into place nice and tight.

I'll have to wait about 8 hours before I can have time to power up and test.

 

Thank you.

Edited August 30, 2021 by Sinphaltimus
Gratitude for info.

[+Ksarul]

I build the PEB versions, @Sinphaltimus. The sidecar design was done by someone else (with my blessing too, as I didn't have time to design a sideport version between all of the other odd projects I'm working on). I provided some technical advice, but the design and construction is in the hands of @arcadeshopper and the board designer.

[+arcadeshopper]

28 minutes ago, Sinphaltimus said:

I didn't really think to contact you directly. I honestly thought this might be a known or already reported issue here in the thread but couldn't find anything. I did not suspect this was an assembly issue - I thought all assembled cards came from Ksarul (did I spell that right?). So figured it might be some limitation I did not know about.  I'll send a photo to your email as soon as I can. My apologies for not thinking to contact you first.

thanks! I am always available via email if anyone has any issues with anything I sell.  Ksarul has nothing to do with the sidecars they were designed for me by a guy in Europe (cant remember his aa handle) and he built the first batch using different extenders than I do.  Subsequent boards are being built by Pcbway for the smt and me for the rest..

 

Greg

[xahmol]

22 hours ago, arcadeshopper said:

New orders won't have an issue as I am building them and using the same extensions as we're used in the 32k sidecars

Very glad to hear that! Thanks for the answer. This thread made me worry for a moment

[+retroclouds]

What would it take to implement a 1MB SAMS hanging on the 16bit bus, so no wait-states?

[RXB]

Would speed up the TI99/4A immensely but would support almost no software at all. 

[Tursi]

7 hours ago, retroclouds said:

What would it take to implement a 1MB SAMS hanging on the 16bit bus, so no wait-states?

Pretty much the same mod as the 16-bit 32k, really. Most of the meat of that mod is the chip select and the patch to the multiplexer circuit. You'd use the same patch and just put the AMS circuitry in place of the chip select.

 

Edited September 4, 2021 by Tursi

[xahmol]

On 8/31/2021 at 2:16 PM, xahmol said:

Very glad to hear that! Thanks for the answer. This thread made me worry for a moment

Update: SAMS and TIPI received and working fine now, and also speech still working (needed another TIPI as well as the one I had, not sold by Arcade Shopper, had the connectors fitted the other way around as the SAMS card was expecting. So now have a spare TIPI for a future spare TI).

 

Edited November 12, 2021 by xahmol

[AltRN8]

I'm doing a SuperAMS build with a blank board I purchased from ArcadeShopper. I think the BOM list was off on the web page as it seems to be the list for the 2016 version and I received the 2017 version. I'm having trouble identifying one chip on the board and hoping someone knows what it is offhand. It's the chip in socket U17 which I think is 74LS373. Can anyone confirm?

[+Ksarul]

You asked in the right place, as I designed that board. . .it is a 74LS373. Here is the schematic for the board you have. One note: being semi-brain-dead when I was marking the IC numbers on the board mask, I listed U16 and U17 (schematic) as U17 and U18 (board). I'll fix that with the next board run, but is something to be aware of for now.

A3-SAMS-P1(R5a).pdf