Clock compatible programs

[+InsaneMultitasker]

I just think that, well, what else do you expect from a RTC? If you did not have a clock on the desktop of your PC you would use your watch to get the current time; but you are obviously very happy to have a time stamp on your files.

 

I think only some disk controllers actually do time stamping, at least the HFDC and I think also BwG. It's clear that if have to do extra struggle to see the time stamp, you don't use it. So you need the proper controller and a suitable output.

 

Excellent point. I added time-stamping to the Horizon ROS via the MBP card, primarily because it is memory-mapped and didn't require DSR gymnastics. Without any programs to truly see the time stamp, I found the feature pretty much useless. (I also no longer have my MBP). On the other hand, I would be "lost" without the timestamps displayed and used by the Geneve.

 

I believe the IDE card also time-stamps the files.

[+mizapf]

I'm not planning to build such a controller ... just thinking ... what controller chip could be used? NEC 765? It's getting increasingly difficult to find ANY disk controller chip, since the PC floppy controller is integrated into the chipset of the main board. If there is a floppy drive at all.

 

Compatibility mode would in particular mean to be able to read and write PC-formatted 512 byte sectors and to map the FAT file system on the TI file system. These are all matters of the DSR, and it is already possible today. At least on the PC side you can easily read and write TI floppy disks directly (I have a C program in Linux for that); the other direction with using 1.44 MB formatted disks on the TI is the real challenge.

 

The FD1771 / WD1770 / WD1773 chips in our controllers have a minimum cell size of 2µs on the floppy disk; independent of any formatting magic you cannot get more than 0.2 sec / 2µs = 100000 cells on a track, and as you need 16 cells per byte, this yields 6250 byte unformatted, so you cannot squeeze 36 sectors into the track. You need a controller which can handle 1µs cells. The HFDC could do that with the HDC9234 controller, but it has a FDC9216 data separator which is not intended for that rate; the 9216B would have been a better choice. For that reason we would need a redesign. For those who want actually want to use floppies, or just for the sake of doing it.

[+OLD CS1]

We could find a bunch of old PC motherboards and steal the floppy controller from them

[+InsaneMultitasker]

The FD1771 / WD1770 / WD1773 chips in our controllers have a minimum cell size of 2µs on the floppy disk; independent of any formatting magic you cannot get more than 0.2 sec / 2µs = 100000 cells on a track, and as you need 16 cells per byte, this yields 6250 byte unformatted, so you cannot squeeze 36 sectors into the track. You need a controller which can handle 1µs cells. The HFDC could do that with the HDC9234 controller, but it has a FDC9216 data separator which is not intended for that rate; the 9216B would have been a better choice. For that reason we would need a redesign. For those who want actually want to use floppies, or just for the sake of doing it.

 

During my employment at Cecure, Don and I upgraded many HFDCs to the 9216B and 32K, specifically to make use of 36 sectors/track high density and to extend the hard drive format from 32 to 34 sectors using Mike Maksimik's CFORM program. The upgrade consisted of replacing these two chips with no other hardware modifications required. Some HFDCs were originally populated with the 9216B though these were rare finds.

 

Unfortunately, PC Transfer does not work with the HFDC nor is the HFDC DSR equipped to handle the 36 sector/track TI format. Only the Geneve can make proper use of the High Density capacity.

[Omega-TI]

Hey listen, if one of you hardware guys wants to make his very own BwG type FDC, << THIS AUCTION >> on eBait has those WD1773 chips. GOOD LUCK!

 

BTW - As of this post there are TWELVE available!

[+mizapf]

If there were some real intentions for a new floppy controller, I would not recommend a WD17xx chip but another one that allows for 1.44 MB disks. As I said, NEC µPD765, Intel 82077AA etc.

[+OLD CS1]

Do any FPGA implementations of these floppy controllers exist?

[+Schmitzi]

Summary of Clock compatible programs (and Hardware):

 

 

TI-99-4A-CLOCK-Compa-v1.00.pdf

 

 

Screenshot:

 

[Omega-TI]

+ The 9640 Menu System brought to us by our own insane multitasker.

[Omega-TI]

Oh, and we cannot forget The Great Gazoo's program to set the clock << HERE >>

[+Schmitzi]

.

>maybe< one more for the list

 

F-Date by Wolfgang Bertsch, errorfree.de, for BwG and other controllers

But I have to find out if it uses the systemclock, or only manipulates the file´s date on the disk...

 

[+9640News]

Got a question regarding the CorComp TripleTech card and the date functionality.

 

This more pertains to use from Extended Basic.

 

If a user requests the date, do they get a dd/mm/yy response, or do they get a dd/mm/yyyy response?

 

And, if they do get a yyyy response, does the card know it is 2021 or did it roll back to 1921?

 

I'm trying to ascertain the most basic i/o response the card gets back without the program adding the YY to YYyy so that Extended Basic and Geneve Advanced Basic report a similar date format.

[+arcadeshopper]

13 minutes ago, 9640News said:

Got a question regarding the CorComp TripleTech card and the date functionality.

 

This more pertains to use from Extended Basic.

 

If a user requests the date, do they get a dd/mm/yy response, or do they get a dd/mm/yyyy response?

 

And, if they do get a yyyy response, does the card know it is 2021 or did it roll back to 1921?

 

I'm trying to ascertain the most basic i/o response the card gets back without the program adding the YY to YYyy so that Extended Basic and Geneve Advanced Basic report a similar date format.

yy

http://ftp.whtech.com/datasheets and manuals/Hardware/CorComp/triple tech manual.pdf

 

[+9640News]

32 minutes ago, arcadeshopper said:

yy

http://ftp.whtech.com/datasheets and manuals/Hardware/CorComp/triple tech manual.pdf

 

Thanks.  That will be my reference point which leads me to think the direction I am going with the change in ABASIC will be consistent.

[SkyPilot]

11 hours ago, arcadeshopper said:

yy

http://ftp.whtech.com/datasheets and manuals/Hardware/CorComp/triple tech manual.pdf

 

I've been collecting TI books and found this one with links and a correction.  I bought a TI today too.

CorComp Clock Updated Manual.pdf

[apersson850]

I mainly used the p-system for my program development on the TI-99/4A. The p-system by default includes date-stamping of files. Without a clock in the system, this requires the user to remember to input the correct date each day, or it fails.

That was one reason for me adding a battery-backed clock to my system. When I did that, none existed to purchase, so I designed my own.

I developed a driver for it, so the p-system could use it and automatically set the date. Later, I added a different driver, so the p-system could do the same with the CorComp triple tech card.

My clock can be configured to generate interrupts on any counter roll-over. Thus I could make a driver which would automatically update the p-system's date when the date changed in the clock circuit. All interrupt driven.

I also added the ability to constantly display the clock in the corner of the screen, if desired, when using the p-system.

 

Only one of my clock cards were ever built. It still works.

[+arcadeshopper]

On 7/28/2021 at 12:15 AM, apersson850 said:

I mainly used the p-system for my program development on the TI-99/4A. The p-system by default includes date-stamping of files. Without a clock in the system, this requires the user to remember to input the correct date each day, or it fails.

That was one reason for me adding a battery-backed clock to my system. When I did that, none existed to purchase, so I designed my own.

I developed a driver for it, so the p-system could use it and automatically set the date. Later, I added a different driver, so the p-system could do the same with the CorComp triple tech card.

My clock can be configured to generate interrupts on any counter roll-over. Thus I could make a driver which would automatically update the p-system's date when the date changed in the clock circuit. All interrupt driven.

I also added the ability to constantly display the clock in the corner of the screen, if desired, when using the p-system.

 

Only one of my clock cards were ever built. It still works.

 

is this stuff somewhere archived?  if not i can put things on whtech 

[apersson850]

No, it isn't. Since it's my own clock design, nobody else had any use for it. I do have some wiring diagrams and software myself, though. If I remember correctly now, I had the problem that a couple of disks got ruined, so some of my source code is missing, I think. I only have the executables left.

[+arcadeshopper]

8 hours ago, apersson850 said:

No, it isn't. Since it's my own clock design, nobody else had any use for it. I do have some wiring diagrams and software myself, though. If I remember correctly now, I had the problem that a couple of disks got ruined, so some of my source code is missing, I think. I only have the executables left.

how about the driver for the cc?

 

[apersson850]

Don't remember. Can check when I'm back home.