Image Map Image Map
Page 19 of 25 FirstFirst ... 9151617181920212223 ... LastLast
Results 181 to 190 of 244

Thread: I wish to create a new DMA/RAM expansion card for the Tandy 1000 line.

  1. #181
    Join Date
    Feb 2017
    Location
    Chilliwack, BC, Canada
    Posts
    376

    Default

    Quote Originally Posted by Eudimorphodon View Post
    Yeah. I have this bad feeling I might have screwed this up even if I had breadboarded the design, because there's a fair chance it's an error I introduced while running the tracks on the board. (It took me several false starts to come up with a board that didn't have a million awkward vias in it, so I was trying to follow some advice about paying attention to the physical layout of the chips instead of the schematic view when choosing how to route signals. In the course of that I switched the order of the data lines through the buffer a couple times, both directionally and end-to-end, I might have just lost track or flipped the direction hi-low in my head at some point.)

    At least it should be an easy fix for rev 2, if that ever happens.
    Yeah, that's something I found when designing my boards too. If it makes more sense for routing to rearrange the gates on a logic IC or a '688 comparitor, etc, from what's in the schematics, then that's what you end up doing, but you gotta make sure you update the schematic at the same time, otherwise you'll tie yourself in knots quickly. It helps having multiple monitors for that, so you can have the schematics and board layout side by side.
    My vintage systems: Tandy 1000 HX, Tandy 1100FD, Tandy 1000 RSX, and some random Pentium in a Hewitt Rand chassis...

    Some people keep a classic car in their garage. Some people keep vintage computers. The latter hobby is cheaper, usually takes less space, and is less likely to lead to a fatal accident.

  2. #182
    Join Date
    May 2011
    Location
    Outer Mongolia
    Posts
    1,263

    Default

    In other news, I installed Check-It 3.0 onto a disk image, and according to its memory mapping function it sees 128k of high memory in addition to the expected "624k" of base RAM.

    memmap.jpg

    I've set it to continuously run the rigorous RAM test for a while just for laughs. (It passed the quick test fine.) The only thing separating this from the brain-dead 74x00 one-chip version is a little more gate delay in the decoding (and the one address line that'd be a little delayed compared to the others) so I don't expect any issues, but we'll see.

    Also just for the sake of testing I've plugged my modem card into one of the pass-through bus slots, and I can confirm that the spacing is right. Check-it sees the serial port integrated into the card so I'll take it as read pass-through works.

    modem.jpg

    Looking at how I laid things out if I end up offering the rest of my proto run to intrepid DIY-ers I might just recommend skipping the buffer and soldering wires across the socket. (As you said it's not great, but boards like the Lo-tech RAM card get away with it, I guess.) Unfortunately the MEMW trace that goes to the memory chips goes *through* the land on the buffer socket on its way to the bus connector, so to sever it free would take two jumpers/trace cuts to free up the pin and bridge the two subsystems to their different bus points. I suppose another workaround would be to just bend up pin 1 of the '245 and run a flying wire to MEMR; I could anchor it to the backside of the down-facing tandy bus connector. Guess that's going to be my test strategy.

  3. #183
    Join Date
    May 2011
    Location
    Outer Mongolia
    Posts
    1,263

    Default

    Hah. Earlier in the thread we brought up the "mempdrvr.exe" file on the oldschool.org Tandy 1000 archive that can create a RAM disk on the (rare?) 512k version of the official Plus memory board. Apparently it's compatible with the UMBs on my card.

    ramdisk.jpg

    It initialized and seemed to work fine. (I could copy files to it, etc, and they survived when I ran memory-intensive programs in base RAM. Amazing it's only 921 byte binary.) Now I'm curious again what the memory mapping for the original card was like when it was installed in an HX. That omnibus collection of Tandy jumper settings said it had different settings depending on whether it was installed in an HX or an EX; obviously the E000 page is out, did you only get 64k in an HX or did it take the C000 page?

  4. #184
    Join Date
    May 2011
    Location
    Outer Mongolia
    Posts
    1,263

    Default

    And it's confirmed, bad direction signal. Cobbled up this solderless jumper arrangement to test and it works

    20190719_174836.jpg

    Interesting side observation: I used the jumper for the E000 page to construct this, and the RAMdisk software refused to initialize with only 64k.

  5. #185
    Join Date
    Feb 2017
    Location
    Chilliwack, BC, Canada
    Posts
    376

    Default

    This is progress! Congrats!
    Still gonna build up that board when it arrives though. I'm curious to see it.
    My vintage systems: Tandy 1000 HX, Tandy 1100FD, Tandy 1000 RSX, and some random Pentium in a Hewitt Rand chassis...

    Some people keep a classic car in their garage. Some people keep vintage computers. The latter hobby is cheaper, usually takes less space, and is less likely to lead to a fatal accident.

  6. #186
    Join Date
    May 2011
    Location
    Outer Mongolia
    Posts
    1,263

    Default

    If nothing else it might be interesting to run a few benchmarks, since you have access to the DMA equipped board, and see if it makes any difference. Topbench scores my EX a bit slower than the SX in its database, I'm curious if that's normal.

  7. #187
    Join Date
    Feb 2017
    Location
    Chilliwack, BC, Canada
    Posts
    376

    Default

    Yeah, I'll give that a go.
    My vintage systems: Tandy 1000 HX, Tandy 1100FD, Tandy 1000 RSX, and some random Pentium in a Hewitt Rand chassis...

    Some people keep a classic car in their garage. Some people keep vintage computers. The latter hobby is cheaper, usually takes less space, and is less likely to lead to a fatal accident.

  8. #188

    Default

    Quote Originally Posted by Eudimorphodon View Post
    If nothing else it might be interesting to run a few benchmarks, since you have access to the DMA equipped board, and see if it makes any difference. Topbench scores my EX a bit slower than the SX in its database, I'm curious if that's normal.
    Topbench scores my SX a point slower than the SX in its database
    Looking for:
    A foot from a Tandy 1000/SX/TX
    Tandy 1000SX expansion bay support bar
    That little tab from the upper right corner of a Tandy 1000 keyboard or EX that holds the function key labels in place

  9. #189
    Join Date
    May 2011
    Location
    Outer Mongolia
    Posts
    1,263

    Default

    Quote Originally Posted by GrandPoobah View Post
    Topbench scores my SX a point slower than the SX in its database
    Looking at the screenshot of my scores I do wonder if this is a case of being right on top of the rounding mark.

  10. #190
    Join Date
    May 2011
    Location
    Outer Mongolia
    Posts
    1,263

    Default

    Quote Originally Posted by blackepyon View Post
    Yeah, I'll give that a go.
    I have a few ideas of hacks that might be interesting to try if you get bored with it, too. (in particular I've been wondering if it would be legit to map the calendar chip at D000 along with the lower UMB, in order to incorporate the calendar into a Flash-less board. I think you could test that by bending up the CE lines for the DS1315 and hacking some wires to the enable jumper.)

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •