REAL TIME CLOCK FOR THE SPEECH SYNTHESIZER/NANO PEB

[Omega-TI]

I have a question for people running CF7’s or Nano-PEB’s…

Do any of you have a REAL-TIME CLOCK installed in your systems? Does one even exist?

 

If not… one of those “clocks-on-a-chip” could be mounted on a small PCB with a chip containing the DSR software and support circuitry, it could then be easily be soldered onto the speech synthesizer circuit board and reside within the same case! Since the speech synthesizer has that little door, it would also give easy access to a battery holder mounted right underneath. This is an option I WOULD BUY!

 

For some Uber programmers, this code would be childs play. If it could be compatible with say, the old CorComp Clock, people would be able to have full functionality in old very useful programs like “Remind-Me!”, use the “Super Clock Support utilities” or even new stuff that could be released in the future.

 

Does this type of thing interest any of you NANO-PEB or CF7 users?

 

 

P.S. -- One question added later... It's been 24 years now, so my memory is foggy, but didn't the BOOT MENU program also make use of the Real-Time Clock feature of the CorComp card?

[mnbvcxz]

I have one of the new nanopebs, I read that the speech synth does not work with the nanopeb, is this still the case or is that just the older nanopeb or even cf7?

[+OLD CS1]

Speech works fine with my CF7+.

[Omega-TI]

I have one of the new nanopebs, I read that the speech synth does not work with the nanopeb, is this still the case or is that just the older nanopeb or even cf7?

 

I cannot speak for anyone else, but I use the new version of the Nano-PEB and it works just fine plugged into the side of the speech synthesizer.

[mnbvcxz]

Thanks for the replies, I do not have my ti99/4A set up yet, I am still acquiring parts.

[Omega-TI]

I'm amazed this topic is getting so little activity. I thought for sure people would be interested in a real-time clock for their TI's. Considering how much smaller an inexpensive the technology has gotten since the 80's, this should be a cheap and easy project, compared to the F18A and Nano-PEB projects.

 

You know this does not have to be just for Nano-PEB users, ANY 'real iron' system could benefit from a real-time clock. It also makes me wonder if a clock project could just be soldered to V V THIS V V and mounted on the clamshell with some double back tape?

 

You have to admit if one could be soldered onto that, most anybody could feel comfortable soldering it on, as there is very little chance of damaging anything if they screwed it up.

[+OLD CS1]

For the TI, I think a RTC is a chicken-egg situation.

 

There is no native use for a RTC on the TI. It would be niche programs specifically designed for RTC usage which would benefit. I cannot say none exist, but most likely very few. I would say go ahead and design and implement one, then let the software follow. Otherwise you are preaching to a crowd which has very limited use for it.

 

My SCSI card has a RTC built-in, but again, for no real purpose. To my knowledge, the Geneve would best benefit from one.

[Willsy]

Yes. The Geneve would benefit from it, because its file system supports time-stamped files, so it would be very useful indeed. The TI doesn't, so I see it as less useful. Okay, I guess it's kind of neat to able to get an XB program to tell me the time, but there again, I have a watch for that!

 

Regarding soldering an RTC to the back of the cart connector, I don't think it'll fly. You probably don't have the address lines that you would need (IIRC only the address lines required for cart access i.e. >6000 to >7FFF are present). Additionally, you would still need to add some memory decoding logic to get the RTC to respond (rather than, say, RAM or ROM) to a specific address(es). Then you need a DSR for the RTC to allow you access it from BASIC. A DSR either has to go at >4000 (IIRC) if it's an assembly language DSR and are designed (at least, conceptually) to be fitted in a PEB card. You need CRU decoding logic to allow the DSR to be paged in and out of memory.

 

As you can see, the complexity rapidly escalates.

[+mizapf]

The Geneve has a RTC on board.

 

BTW, the RTC proved to be quite useful for me to calculate command execution and wait state times.

Edited September 5, 2013 by mizapf

[Asmusr]

Did you know Classic99 has a clock device? I didn't until a few days ago. This is how to use it:

 

10 OPEN #1:"CLOCK"

20 INPUT #1:A$,B$,C$

30 PRINT A$:B$:C$

[+OLD CS1]

I recall reading somewhere that the TI filesystem has space reserved in the file descriptor for time stamps, but no DSRs to date use it. Can someone confirm that before I have to go diving, again?

[+Lee Stewart]

I recall reading somewhere that the TI filesystem has space reserved in the file descriptor for time stamps, but no DSRs to date use it. Can someone confirm that before I have to go diving, again?

 

There are reserved spaces in the FDR; but, they were not reserved specifically for time stamps. They were simply co-opted by third-party DSR developers for that purpose.

 

...lee

[Tursi]

Did you know Classic99 has a clock device? I didn't until a few days ago. This is how to use it:

 

10 OPEN #1:"CLOCK"

20 INPUT #1:A$,B$,C$

30 PRINT A$:B$:C$

 

This was based on the spec for the Corcomp clock to be compatible with it.

[Gazoo]

I recall reading somewhere that the TI filesystem has space reserved in the file descriptor for time stamps, but no DSRs to date use it. Can someone confirm that before I have to go diving, again?

 

HFDC time stamps all files, creation and update times are used.

 

Gazoo

[+OLD CS1]

 

HFDC time stamps all files, creation and update times are used.

 

Gazoo

 

It has a clock built-in, correct? My WHT SCSI controller has a clock, too, but I have never checked to see if it time-stamps.

[+InsaneMultitasker]

The IDE card supports time stamping http://home.vodafonethuis.nl/fgkaal/

 

I also thought the BwG floppy controller had this feature.

 

The Geneve DSR time-stamps all floppy and hard disk (IDE/MFM/SCSI) files.

[Asmusr]

What would it take to implement a generic solution where a Raspberry Pi is connected to the expansion port, and the PPi is programmed to emulate any number of DSR ROMs + the other functions of the nanoPEB? The price of a RPi is only half of the price of a nanoPEB and it would be much more versatile solution.

 

I guess you would probably need one additional piece of hardware between the expansion port and the RPi. Perhaps one of the boards listed on this page could be used?

http://elinux.org/RPi_Expansion_Boards

 

I think the key is to avoid building custom hardware, if possible, but to focus on the software side that is much easier to share.

[Omega-TI]

 

HFDC time stamps all files, creation and update times are used.

 

Gazoo

 

Not only that, but the BOOT menu program that many people have been using for YEARS has features that they are probably unaware of.

 

1) RTC Support

Displays the time in the upper right-hand corner of the screen

 

2) 80 Column Support (Works with F18A too try it in Classic 99) (Just press "W" to enter 80 column mode)

When you do a directory in 80 column mode it will display file creation & update times if available.

 

There are programs out there already that will support a RTC, and in the case of the BOOT program even 80 columns. I just thought it would be nice to be able to fully use what we already have available.

[Asmusr]

How many of the 44 pins on the side port do you need to connect a device like the nanoPEB? If 32 pins are enough then this might work:

http://www.abelectronics.co.uk/products/3/Raspberry-Pi/18/IO-Pi-32-Channel-Port-Expander-for-the-Raspberry-Pi-computer-boards

 

From http://nouspikel.group.shef.ac.uk/ti99/pinouts.htm#Side :

 

# Name I/O Use

- ---- --- -----------

1 VCC +5 Volts power supply

2 SBE > High if addr in >9000-94xx (sound port)

3 RESET* > System reset (active low)

4 EXTINT* < External interrupt (active low)

5 A5 > Address bus, bit 5

6 A10 >

7 A4 >

8 A11 >

9 DBIN > Active high = read memory

10 A3 >

11 A12 >

12 READY < Active high = memory is ready

13 LOAD* < Unmaskable interrupt (=> BLWP @>FFFC)

14 A8 >

15 A13 >

16 A14 >

17 A7 >

18 A9 >

19 A15 > Address bus, lsb. Also CRU output bit.

20 A2 >

21 GND

22 CRUCLK* > Inversion of TMS9900 CRUCLOCK pin

23 GND

24 PHI3* > Inversion of phase 3 clock

25 GND

26 WE* > Write Enable (derived from TMS9900 WE* pin)

27 GND

28 MBE* > Active low if addr in >4000-5FFF (card ROMs)

29 A6 >

30 A1 >

31 A0 > Address bus, bit 0 (most significant)

32 MEMEN* > Memory access enable (active low)

33 CRUIN < CRU input bit to TMS9900

34 D7 <> Data bus, bit 7 (least significant)

35 D4 <>

36 D6 <>

37 D0 <> Data bus, bit 0 (most significant)

38 D5 <>

39 D2 <>

40 D1 <>

41 IAQ > Interrupt acknowledged by TMS9900

42 D3 <>

43 VDD -5 Volts power supply

44 AUDIOIN < To sound generator AUDIO IN pin

[jartz]

Way-back-when -- In 1990 or 1991, I ran a BBS on my TI 99/4a. I found a schematic on a BBS somewhere with plans to build a RTC that would fit into the speech synthesizer. I built it, and it worked great from what I remember. Fast-forward 20 years, and I no longer have that system, and can't remember the details of the RTC project, other than the fact that I built it...

 

If I could find the plans/schematics again, I'll do it with my newly-acquired TI 99/4a system. I don't remember what components were needed, so I'd have to google to see if I could find those original plans. Or maybe I could start from scratch, using a PIC or Arduino?!?

 

Heh, I wish I still had a copy of the BBS program - I had made many 'tweaks' to it to customize it to my liking... I think I had called my BBS "Jeff's 20 Megs" - or something similar - I actually saw it on a Google hit on a list of BBS numbers recently... It brought a smile to my face.

 

Jeff

[RXB]

It did not become popular or known by many but a guy (do not remember his name) at a TI Faire in Phoenix told me he had a PIO/2 talking to a clock device in XB for his BBS program.

 

He was very old compared to me so today would be in his 80's. Anyone remember someone that used this method?

[Omega-TI]

 

There are reserved spaces in the FDR; but, they were not reserved specifically for time stamps. They were simply co-opted by third-party DSR developers for that purpose.

 

 

 

Yep, it was used for time stamps, in fact the BOOT MENU program will use a clock (if one is avaliable).

Many people use this program and have no idea of all the features it has... including 80 column mode.

[Omega-TI]

Did you know Classic99 has a clock device? I didn't until a few days ago. This is how to use it:

 

10 OPEN #1:"CLOCK"

20 INPUT #1:A$,B$,C$

30 PRINT A$:B$:C$

 

 

 

This was based on the spec for the Corcomp clock to be compatible with it.

 

 

 

 

 

 

I love having access to the clock from Extended BASIC in Classic99. I wonder though, what's missing that prevents programs like "Remind-ME!" from showing the correct time and loading up on the proper date, or from BOOT where it normally shows the correct time in the upper right hand corner of the screen? Granted the last two programs are E/A 5, so I'm assuming some sort of DSR that was tapped from the programs themselves that is missing from emulation.

[Tursi]

I did no register-level emulation of the clock device, ONLY the DSR works. So anything that assumes what chipset lies behind it will fail. I don't intend to change that in the future.

[+InsaneMultitasker]

I did no register-level emulation of the clock device, ONLY the DSR works. So anything that assumes what chipset lies behind it will fail. I don't intend to change that in the future.

Could be the menu and other programs are coded to use the MBP clock card, which is just a simple memory mapped device. The specs for that are pretty simple if interested..