Exploring the VT241

[FuzzyLogic]

I happened across a VR241-A monitor awhile back, and recently found a fairly cheap VS240 unit (and an even cheaper broken one too). They looked like they might be quite fun to play with. (I'm actually extremely new to hardware terminals, but quickly grew fond of my first when I acquired it a few months ago.)

This past weekend I worked out how to adapt a cable to connect the monitor to the unit (if "linking two breakout connectors with breadboard wires" counts as "adapting a cable" anyway! ). Since I couldn't find many images of a VT241 in action here in the forum, I thought I'd share a few.

The following digital art is mostly the work of DEC, really. I wrote a program to throw lists of random screen coordinates at the terminal, with instructions to "connect the dots in a smooth closed curve" and then "fill the polygon", and these are examples of what it produced. (I also randomized the colors. It's fun to watch.)




Intriguingly, although every list has exactly the same number of random coordinates, there is a very noticeable variation in processing time before the terminal starts to draw. Some curves must be more complicated to work out.

I also noticed an open question in an earlier thread as to whether the VT240 / VT241 can display sixel graphics. The answer is Yes*:

*But honestly with only 4 colors, most images look better in grayscale, and the rather extreme 800×240 aspect ratio tends to distort photos (at least as produced by img2sixel).

Finally, here's the interior of the broken unit. It very clearly experienced a case of RIFA-go-boom, but my hope is that I can repair it (once I learn how). Many curious things stand out to me, especially the empty sockets (ROM expansion, I think?) and the NEC-branded 8085.


Has anyone else played around with one of these?

 

[leegleason]

Far more interesting than the 8085 is the DEC 310ES chip on that board. It's a DEC T11, the PDP11 on a chip. I've been thinking about getting a VT241 system unit to try and hack into a more general purpose 11. There's bound to be some extra space in those ROMs for more code.

 

[FuzzyLogic]

Far more interesting than the 8085 is the DEC 310ES chip on that board. It's a DEC T11, the PDP11 on a chip. I've been thinking about getting a VT241 system unit to try and hack into a more general purpose 11. There's bound to be some extra space in those ROMs for more code.

That's certainly neat!

Probing around with a multimeter, it seems like the two empty sockets to the right of the 8085 belong to the same bank as the four populated chips just below that DEC 310ES. (The one on the left is connected to the 8085.) Probably plenty of room for new code.

By way of hacking, the expansion header for the integrated modem option seems like maybe an interesting target too.

 

[Radix]

I'm not sure why you're not referring to the actual print set and other docs for the terminal! - It's easier than buzzing it out...

Index of /pdf/dec/terminal/vt240

 

[daver2]

The print set is much more informative...

E131 is additional ROM for the 8085 CPU whilst E85/E39 are additional ROM for the T11.

Depending upon what size the existing T11 ROMs are, you may be able to 'double the size' of them by changing a link.

Dave

 

[snuci]

I also noticed an open question in an earlier thread as to whether the VT240 / VT241 can display sixel graphics. The answer is Yes*:

Where might one find this image file? I wonder of the GIGI would display this?

 

[FuzzyLogic]

Where might one find this image file? I wonder of the GIGI would display this?

It's just a photo I took, as converted by the img2sixel utility on Linux. Because I was converting for the VT241, I invoked it like this: img2sixel -p 4 -h 240 -w auto my_photo.jpg which specifies a 4-color palette, 240 pixel height, and auto-scaled width.

I'm less familiar with the GIGI; different arguments to img2sixel might make more sense for it. But you can use any image you like!

 

[FuzzyLogic]

A few curious anecdotes as I've poked around with ReGIS on the VT241:

1. The "connect dots in a smooth curve and fill the polygon" program shown earlier works well for curves with up to about 27 points in their list. Beyond that, graphics are drawn incorrectly. I was able to replicate the behavior (with the same apparent limit of 27 points) on my VT330. This seems strange.

2. I do not seem to have a full range of 64 colors, at least if I specify palette registers by RGB (as percentages) using DEC command strings in text mode. Instead, it seems like the transitions are around 33% / 67% for each component, leaving only 27 colors to choose from. This unit is running version 2.2 of the firmware, FWIW. I haven't tried specifying colors as HLS yet, nor counting colors in ReGIS mode.

3. The set-up menu temporarily scrolls screen contents upward (including graphics), and also temporarily changes colors back to default palette values. It restores everything on exit. I thought that was neat, and also pragmatic; there's no way to make the set-up menu unreadable by accident.