Stuart

Has anyone got the NanoPEB serial port working in assembler? I can't get it to write a character. The relevant snippets of my test code are as follows: CR9902B EQU >1340 CRU base address for RS-232 port on NanoPEB. LI R12,>1300 Switch on serial interface on NanoPEB. SBO 0 LI R12,CR9902B CRU base address of port. SBO 31 Reset 9902. LI R1,>6300 Control register: 2 stop bits, even parity, 8 data bits. LDCR R1,8 Load control register. SBZ 13 Disable loading of interval register. LI R1,>001A 19200 Baud. LDCR R1,12 Load transmit and receive data rate registers. ... ... ... LI R12,CR9902B CRU base address of 9902. XXX SBO 16 Set RTS on. LDCR R2,8 ******* SEE COMMENT BELOW ******* LI R0,>FFFF YYY DEC R0 JNE YYY SBZ 16 LI R0,>FFFF ZZZ DEC R0 JNE ZZZ JMP XXX With the "LDCR R2,8" line commented out, RTS is toggling on and off about once per second. With the "LDCR R2,8" line not commented, RTS is staying in a steady state. Any ideas? The code works fine on my TM990 system. I've tried adding a LIMI 0 in the code but that doesn't seem to help.

There are details on using the 9901 internal timer on Thiery's site (currently offline - you'll have to download a zip from one of the other threads).

In your last photo, second pin up from the bottom left - is that a solder blob up against the trace that runs past it? No other solder blobs - a bit difficult to tell from the photo?

Does the FDC LED flash when you power on the console? Have you got a Minimem cartridge, which can be used to check the DSR can be read?

Do a search on eBay for "Antex 18W", and see what sort of price they are in your part of the world. That's a pretty decent iron, with power and tip size suited for TI work.

What were you soldering with, a blacksmith's iron? Shorting pins 5 and 6 should have no ill effects other than keeping the processor in perpetual reset. Remove the solder blob and all should be well. Another place you could connect the reset button is across capacitor C606, which is the large capacitor near the grey crystal. Far easier to access. See this picture [http://www.avjd51.dsl.pipex.com/ti/images/console_32k_ram_modification_fitted.jpg] and follow the grey cable on the left, which goes to my reset button.

Two chips should be possible. A 128kx8 RAM (the next 'size up' from 32k, and only a couple of quid) and a 74LS138 address decoder (to enable the RAM only when addressing the 8K blocks occupied by the 32K RAM expansion). Plus four diodes to OR together the CS signals plus four pull up resistors.

I think MESS has the most accurate simulation of the speech chip at the moment? So MESS would be fine if you haven't got the 'real iron' set up and handy.

So for each frame, are you in effect cycling through all the coefficient values, seeing what sound they produce, and comparing that with the input? Could you post a recording of the audio to see what sort of quality you're getting?

Was there an Oh_Sh*t_Im_Bouncing word defined?

'378 is 6 latches, '379 is only 4 latches.

If you want to drop me your address in a PM, I've got several spares of those and you can have one. Stuart.

So if Michael could just repeat that another 7 times for the other 7 bits ...

Pretty sure the sound chip is (normally) socketed (along with the GROMs), so should be an easy swap.

But the failures seen tend to be quite specific, and as reported in this thread - the display is absolutely normal except that some characters are garbled. And that's down to one of the VRAMs. They're fairly easy to replace, if one knows which one of the eight is faulty. I might have heard of a VDP failing once or twice, but have never heard of a problem with any other component in the video subsystem.

This seems to be a common problem. A while ago Michael (I think it was) done a temporary tweak in MESS to replicate a failed VRAM and show how it looked on screen. Might someone take up the challenge of adding this as a 'feature' to an emulator to help diagnose "if your startup screen looks like this, replace VRAM chip Uxx"? Or do temporary tweaks and take a series of reference screenshots? Not entirely sure how practical this suggestion is - I'm guessing that it's probably only one bit in *part* of a VRAM failing, rather than the entire contents of one VRAM failing.

Thanks MarkO. I've sent you a PM. Stuart.

Anyone out there got a web server with PHP running that I can test a small file on? Stuart

Video of the browser running on my TM990 here: www.stuartconnerdownloads.me.uk/z_ti_browser_1.mpg [a stupidly large 35MB file for some reason; you might need to 'save target as' if it won't display in your browser]. This is edited to remove the periods where at the moment it has to keep retrying the connection to the server. To clarify a previous comment: "If I'm understanding this correctly, anyone with a website could hide these <99ml> tags into any existing HTML page, but only those of us with TI's would see whats meant for us. The browser would just ignore everything else...I like it!" No. Internet Explorer for example (don't know about other browsers) will display text that is not inside <html> and <p> tags. Take a look at http://www.avjd51.dsl.pipex.com/z_ti_test_1.htm for example. The program design at the moment also requires the whole downloaded page to be held in memory (because it is re-rendered when you scroll up or down), which limits the size of the page that the browser will work with.

Well it's actually pretty much finished running on my TM990 - will need porting to the TI-99/4A. And assuming I can get my nanoPEB talking to the Lantronix. First, just to clarify that it was never going to and never will work with any 'modern' web page - there are just too many HTML tags and other stuff for the TI to handle, and there's no way you could display much of it any way with the TI's low resolution screen. Instead, I have defined a set of tags that are supported - within an outer set of <99ml></99ml> tags of course. Tags currently supported are: <p></p> [paragraph; text outside these tags is considered a comment and is not displayed] <br> [line break] <cdef:xx:yyyyyyyyyyyyyyyy> [redefine character xx; any character 00-127 can be redefined] <clr:bf></clr> [set foreground and background colours for text] <chr:xx> [display character with hex ASCII code xx, for characters with ASCII codes less than >20] <a></a> [hyperlink] <u></u> [underline] The character set that is loaded includes characters for table border lines, so tables can be drawn using these. Simple graphics can be drawn by redefining a number of characters and drawing it using those. There's a sample file here: http://www.avjd51.dsl.pipex.com/z_ti_test_1.htm. You should be able to load this in your browser (where it WON'T be displayed correctly) then do 'view source' to see the actual tags. I'm not at home at the moment so can't do a screenshot. The screen gives a 42 x 24 line display that can be scrolled up/down. Because characters are not all aligned on 8 pixel boundaries, there are some restrictions on changing foreground/background colours mid-line. A pointer sprite, controlled through the keyboard, enables hyperlinks to be selected. To resolve addresses, the program first uses the Lantronix to connect to a one-liner PHP script on my website (which has a pretty stable IP address) which returns the numeric IP address, then uses that to download the web page. I have a problem with the IP lookup at the moment - it normally works the first time but then the Lantronix won't connect to the web server again for another minute or so. The program has built-in retrying until a connection is made, but I need to look into this. No download functionality at the moment, but I think it should be possible. There will probably be a download size limit according to how much RAM is free, but should be able to maximise this by using a small standalone module to handle the download which then reloads the main browser program when it has finished. Hope to get a video of it running on my TM990 around the end of the week. Stuart.

F18 Mk II.

Try the attached, where I've hacked the .exe so that the [Add Source File] dialog defaults to .txt format. Everything else appears to work OK. WinAsm99 - Modifed to Default to txt Files.zip

Maybe ... but don't confuse the speed with which the device can send or receive a single character with the speed with which the rest of the system can send characters to it or read characters from it.

If you store the calculated value in the word after a LWPI instruction, and then execute that instruction, I think that would do it?

Electrically compatible yes, I believe, but you need software to understand and drive the device. OPEN 1,8,15, ...