John Meng's 9980/81 modules and Power Basic

[+FarmerPotato]

I found another paper by John D. Meng, the author of "Making Your TI-980 Play Your TMS9900":

 

Power Basic and the 9980/9981

 

Presented at 10th annual TI-MIX, New Orleans, LA, 1981.

 

 

 

Owners of TM990/101 CPU modules could write applications in Power Basic.  John Meng's laboratory did that, and then transferred them to purpose-built modules. The module was a 2.5 x 6 inch board,  containing a 9902, 4 EPROMs, 1K RAM, and the (much less expensive!) TMS9981.  

The paper describes 4 of the module's applications in the laboratory. 

Along the way, they encountered bugs.  The way they progressed is entertaining.  John Meng is a lively writer!  I recommend you read the paper. 

Excerpt:  "Only One Problem"

 

Quote

 

When we first looked into the possibility of utilizing the 9980 running Power Basic, only one problem was immediately evident. The 9980 was capable of operating only up to 2.5 MHz, slower than the 3 MHz which was used for TMS 9900 based systems.   Consequently, a baud rate code sent to the TMS 9902 RS-232 chip would not set the same rate on a 9980 system that it would set on a 9900 system operating at 3 MHz.   

 

Not having access to Power Basic source code, it was not initially practical to think of modifying the Power Basic EPROMs to accommodate this difference. As a result, at first we operated our 9980/9981 controllers at 1.5 MHz, exactly half the frequency of a 9900 system.  At start-up, Power Basic apparently detects the baud rate [during] the first bit input.  ... Executing at half-speed, [it reports] that the baud rate is twice what it actually is. The resulting baud rate should in fact be correct. 

 

Three things complicate this apparently happy conclusion, however. 

 

First, Power BASIC does not support 600 baud. Consequently, a 300 baud terminal connected does not respond properly.  Unfortunately, 300 baud is very common in printing terminals, and this faulty response is a severe handicap. 

 

Second, a 1200 baud terminal will respond correctly, but the power basic start up routine believes it to be 2400 baud.  The result is that the end-of-line and inter-line pad characters required for some 1200 baud printing terminals (such as the silent 700) are missing, although the character transmission rate is correct.

 

Our original solution for these unfortunate characteristics was to start the system with a 9600 board video terminal and execute:

       BASE 80H:: CRB(11) = 1 :: CRB(12) = 1:: CRF(13) = 468H

 

Thereby switching the terminal 9902 to 300 baud, leaving the 9980 thinking it was running at 9600 baud.  The lack of end-of-line spacers still proved an annoyance, however.  We found that by changing the last part of the above line to 

 

       ::CRF(13) = 49CH

 

we could operate satisfactorily at 200 baud.

 

The story continues with a bug discovered in Power Basic, occurring only on the 9980. 
 

Enjoy!

 

Reference

 

Meng, John D. (1981) . POWER BASIC AND THE 9980/9981.  Lawrence Berkekey Laboratory, University of California at Berkeley.  March 1, 1981.  To be presented at TI-MIX, New Orleans, LA, on 3/8/1981.   Permalink: https://escholarship.org/uc/item/1pn4c94k

 

[jbdigriz]

Another nice find. Wish there was an online archive of TI-MIX and other TI sponsored newsletters from that era, as well. I'd be happy to help if anyone has a collection that needs hosting or posting to bitsavers, where there are a few at present. Whtech, too, iirc.  I'm backed up on scanning right now, though.

Edited January 29, 2023 by jbdigriz