- OSI 541 Board.jpg (222.88 KiB) Viewed 4098 times
OSI 541 Board
OSI 541 Board
Hi Jeff,
I thought this is so interesting, it should be in a new topic.
I would like to see a higher-resolution image of the board to be able to make out all the chip numbers with greater confidence. The board reminds me very much of a couple of different modifications I made to my C2-4P 540 board, so I'll offer a couple of possibilities:
1) A programmable character generator. If that's the case, you'll be looking for a grouping of 3 74LS157 chips side by side, with lines going to the blank socket next to the chargen, and other lines going to the 8T26 chips near the left lower corner of the photo. At least 2k would be dedicated to the programmable character set, but this card has 6k. (Edit: These are actually 2102 chips, so the board really has 24K. The discussion that follows is therefore off base.) The extra 4k could be for multiple banks of character sets, or for mode bits, indicating which character set is to be used, of how the sets are to be combined (Chargen, programmable, Char (AND|OR|XOR) Programmable, inverse, blinking, etc.) or one extra charset bank and one mode bit bank. In my circuit, I had 2K for the char gen, which plugged into the chargen socket on the 540 card (a 540A in my case). In my hack, I didn't have a separate card--I had a bunch of wires going to the address and data lines on the 540 board, which is why I later had to remove it to do another mod (below). This card looks a lot like that circuit would look if it weren't parasitic from the 540, and just connected to the 540 with a 24-pin cable.
2) Bitmapped graphics display. I think this is less likely, but I'd like to see a better close-up of the board. Without hooking into the 540 timing counters more directly, you'd expect to see a set of counters to track ROM accesses and advance the board address. You'd expect to see a bank of 4 74LS157 chips to switch betwen system bus access and 540 board access. The 6K would be odd for a bitmap graphics circuit. Even in 32x32 mode (256x256) 8k would be required, unless there were something like 256x192 over the upper part of the screen, and 8 lines x 32 of regular characters from the 540board memory at the bottom of the screen. Using the 540 board memory for part of a hi-res display would require complicated latching and counting circuitry and a lot more chips.
For now, I'd try leaving that board out. Once you have the system up and running, it would be helpful trying to map out a few of the connections to that empty socket. That should help clarify the purpose of the board.
Dave
I thought this is so interesting, it should be in a new topic.
I would like to see a higher-resolution image of the board to be able to make out all the chip numbers with greater confidence. The board reminds me very much of a couple of different modifications I made to my C2-4P 540 board, so I'll offer a couple of possibilities:
1) A programmable character generator. If that's the case, you'll be looking for a grouping of 3 74LS157 chips side by side, with lines going to the blank socket next to the chargen, and other lines going to the 8T26 chips near the left lower corner of the photo. At least 2k would be dedicated to the programmable character set, but this card has 6k. (Edit: These are actually 2102 chips, so the board really has 24K. The discussion that follows is therefore off base.) The extra 4k could be for multiple banks of character sets, or for mode bits, indicating which character set is to be used, of how the sets are to be combined (Chargen, programmable, Char (AND|OR|XOR) Programmable, inverse, blinking, etc.) or one extra charset bank and one mode bit bank. In my circuit, I had 2K for the char gen, which plugged into the chargen socket on the 540 card (a 540A in my case). In my hack, I didn't have a separate card--I had a bunch of wires going to the address and data lines on the 540 board, which is why I later had to remove it to do another mod (below). This card looks a lot like that circuit would look if it weren't parasitic from the 540, and just connected to the 540 with a 24-pin cable.
2) Bitmapped graphics display. I think this is less likely, but I'd like to see a better close-up of the board. Without hooking into the 540 timing counters more directly, you'd expect to see a set of counters to track ROM accesses and advance the board address. You'd expect to see a bank of 4 74LS157 chips to switch betwen system bus access and 540 board access. The 6K would be odd for a bitmap graphics circuit. Even in 32x32 mode (256x256) 8k would be required, unless there were something like 256x192 over the upper part of the screen, and 8 lines x 32 of regular characters from the 540board memory at the bottom of the screen. Using the 540 board memory for part of a hi-res display would require complicated latching and counting circuitry and a lot more chips.
For now, I'd try leaving that board out. Once you have the system up and running, it would be helpful trying to map out a few of the connections to that empty socket. That should help clarify the purpose of the board.
Dave
Re: OSI 541 Board
Hi Dave,
Thanks for your reply. I did some more investigation last night and discovered that the story isn't complete without examining the 540 video board. The chargen was removed from the 540 and transplanted onto the 541. A 24 pin cable connects from the chargen socket on the 540 to the 541. Also, a 16 pin cable transplants one of the 2114 video memory chips from the 540 to the 541. There are a handful of mod jumpers on the back side of the 540 that (I think) share some additional signals from the 540 to the 541. I don't think the system can operate correctly with just the 540 installed, unless I undo all the jumper mods on it and replace the chargen and 2114 onto it.
I will email you large versions. I hope you can handle a few megabytes in your email inbox! Feel free to post them to the site if you desire.
Cheers
/Jeff
Thanks for your reply. I did some more investigation last night and discovered that the story isn't complete without examining the 540 video board. The chargen was removed from the 540 and transplanted onto the 541. A 24 pin cable connects from the chargen socket on the 540 to the 541. Also, a 16 pin cable transplants one of the 2114 video memory chips from the 540 to the 541. There are a handful of mod jumpers on the back side of the 540 that (I think) share some additional signals from the 540 to the 541. I don't think the system can operate correctly with just the 540 installed, unless I undo all the jumper mods on it and replace the chargen and 2114 onto it.
- 540topsmall.jpg (236.07 KiB) Viewed 4060 times
- 540bottomsmall.jpg (241.38 KiB) Viewed 4060 times
Cheers
/Jeff
Re: OSI 541 Board
I posted the large images to Photobucket.
541 board
540 board top
540 board bottom
I hope the community finds this interesting.
Cheers
/Jeff
541 board
540 board top
540 board bottom
I hope the community finds this interesting.
Cheers
/Jeff
Last edited by Jeff on Sun May 31, 2015 1:50 am, edited 1 time in total.
Re: OSI 541 Board
Jeff,
I still think it's one of the above guesses. The fact that one color 2114 is replaced suggests that only 1K of text is to be combined with whatever the 541 board generates, suggesting it's in 32x32 char mode, would favor a black&white bitmapped graphics overlay, with some different modes for combining the text and graphics. I think that tracing out a bit of the circuit to see what feeds the data pins on the socket would be helpful to tease out the function of this board.
A photo of your startup screen would be interesting, and perhaps helpful. If you cycle on and off and get different screens, pictures of those may be helpful to see as well.
Dave
I still think it's one of the above guesses. The fact that one color 2114 is replaced suggests that only 1K of text is to be combined with whatever the 541 board generates, suggesting it's in 32x32 char mode, would favor a black&white bitmapped graphics overlay, with some different modes for combining the text and graphics. I think that tracing out a bit of the circuit to see what feeds the data pins on the socket would be helpful to tease out the function of this board.
A photo of your startup screen would be interesting, and perhaps helpful. If you cycle on and off and get different screens, pictures of those may be helpful to see as well.
Dave
Re: OSI 541 Board
Hi Dave,
What I find curious about this board is the 24K of RAM! I sent you a component layout with all the chip numbers. I have downloaded the data sheets for each of the chips on this board and will start to trace the circuit.
Any thoughts on the 24K? Do you think it is video ram? Or do you think that this board might have multiple functions?
I will fire up the machine and take some photos of the screen and post them shortly.
Cheers
/Jeff
What I find curious about this board is the 24K of RAM! I sent you a component layout with all the chip numbers. I have downloaded the data sheets for each of the chips on this board and will start to trace the circuit.
Any thoughts on the 24K? Do you think it is video ram? Or do you think that this board might have multiple functions?
I will fire up the machine and take some photos of the screen and post them shortly.
Cheers
/Jeff
Re: OSI 541 Board
When I first turn on the machine I get this:
If I toggle the power, I get this:
I have the C4P hooked to a Samsung LED TV. No colour is coming through.
Your thoughts?
/Jeff
- Screen Shot 2014-03-24 at 11.17.41 AM.jpg (203.74 KiB) Viewed 4049 times
- Screen Shot 2014-03-24 at 11.18.00 AM.JPG (205.76 KiB) Viewed 4049 times
Your thoughts?
/Jeff
Re: OSI 541 Board
Hi Jeff,
Hard to say--this looks like a typical power-up garbage screen. Some characters are inverted. I think that the color-capable boards (540b) could produce inverse characters. This looks pretty much like black&white; I'm not sure if the inverse video is happening on the 540 or 541 board. However, I think the 541 is doing something more interesting that is not showing up here. I was hoping to see random scattered dots in there as well, but no such luck. I think your options are to start poking around in high memory and see what happens, or start tracing out some of the signals. If you can trace out the address decode signals, it may give you some idea where to poke, if you go that route. If you start tracing out signals, I'd start at the ribbon cable socket and work backwards.
Dave
Hard to say--this looks like a typical power-up garbage screen. Some characters are inverted. I think that the color-capable boards (540b) could produce inverse characters. This looks pretty much like black&white; I'm not sure if the inverse video is happening on the 540 or 541 board. However, I think the 541 is doing something more interesting that is not showing up here. I was hoping to see random scattered dots in there as well, but no such luck. I think your options are to start poking around in high memory and see what happens, or start tracing out some of the signals. If you can trace out the address decode signals, it may give you some idea where to poke, if you go that route. If you start tracing out signals, I'd start at the ribbon cable socket and work backwards.
Dave
Re: OSI 541 Board
Hi Jeff,
For some reason, I originally thought those were 2102 chips for a total of 6K, but as you point out, they are 2114s for a total of 24K. That puts a high-res option back on the table. The question is, do the address and data lines go anywhere else besides the system bus?
I believe there is at least one ex-OSI insider here on the board. I'm hoping maybe someone in the know will post about this most interesting board.
Dave
For some reason, I originally thought those were 2102 chips for a total of 6K, but as you point out, they are 2114s for a total of 24K. That puts a high-res option back on the table. The question is, do the address and data lines go anywhere else besides the system bus?
I believe there is at least one ex-OSI insider here on the board. I'm hoping maybe someone in the know will post about this most interesting board.
Dave
Who is online
Users browsing this forum: No registered users and 1 guest