Image Map Image Map
Page 7 of 18 FirstFirst ... 3456789101117 ... LastLast
Results 61 to 70 of 178

Thread: Commodore CBM 2001-16N - BIG MESS!

  1. #61

    Default

    Quote Originally Posted by Dwight Elvey View Post
    Hi Miata
    I don't think you can plug a 2716 or a 2732 directly into a 6332 socket without an adapter socket. Some of the pins will be wrong.
    Dwight
    It should work! I understood that my Board has original 2316 and 2332 ROM chips running. They can be replaced directly with 2716 or 2532. For the 2332 i can also use 2732 with an adapter. I did a lot of investigation in the internet and i am confident about it because my unit came with 2716 and 2732 EPROMs. In the photo you see the adapters underneath some EPROMs.

    Regarding my measurement - i guess i am running into the limit of my scope. I cannot get a better resolution to see the burst in the first part. Please have a look to the Youtube Video i uploaded: https://youtu.be/n2t903vwVnI

    And yes, the Trigger from the Pin20 of the ROMs (CH1 - not displayed) was stable.

  2. #62

    Default

    Once we show that the processor can read one location in a EPROM we can write some simple loops to test targeted locations with tight loops.
    Dwight

  3. #63

    Default

    Quote Originally Posted by miata View Post
    It should work! I understood that my Board has original 2316 and 2332 ROM chips running. They can be replaced directly with 2716 or 2532. For the 2332 i can also use 2732 with an adapter. I did a lot of investigation in the internet and i am confident about it because my unit came with 2716 and 2732 EPROMs. In the photo you see the adapters underneath some EPROMs.

    Regarding my measurement - i guess i am running into the limit of my scope. I cannot get a better resolution to see the burst in the first part. Please have a look to the Youtube Video i uploaded: https://youtu.be/n2t903vwVnI

    And yes, the Trigger from the Pin20 of the ROMs (CH1 - not displayed) was stable.
    You are turning the sweep the wrong way. Try something like .5 or .2 milliseconds and turn the intensity up. Also where are your 0V lines?
    The video is good as I can stop it at various places.
    Dwight

  4. #64
    Join Date
    Feb 2009
    Location
    Southern California, USA
    Posts
    2,833

    Default

    Quote Originally Posted by Dwight Elvey View Post
    Nope, 2716 should work without an adapter in the 2001N.
    Yes, the 2716 will work directly in the D8 2K EDITOR socket or as the Character Generator ROM. All other 4K ROM sockets can use a Texas Instruments TMS-2532 (or equiv) directly. The Intel 2732 will need an adapter.

    If it turns out the adapters are intermittent, I can send the OP some programmed 2532's if he cannot program them himself.

  5. #65

    Default

    If I stop the video right at the first part of the trace, I think I can see what I'm looking for. If it is just the ch2 on the screen, I can see highs and lows right at the start for about 1/3 of a division. I'd like to see it on a faster sweep though( 0.5ms or 0.2ms ). The video can be used to see more parts than the still shots.
    The later parts of the sweep are of little interest right now. The spikiness of the other parts of the trace are normal. Remember, TTL and NMOS can not pull the signal all the way to +5v. When driven high, the outputs tend to float when above 3.5V to 4V someplace. This is unlike CMOS that pulls up to the rail, with no loading. You get used to looking at such signals.
    Dwight

  6. #66

    Default

    Here is the math. Running at 1MHz, the full address space is 65536 adrresses. We are only interested in 1/16th of that space, while the EPROM pin 20 is low. Each full address pass should be 131.072 ms long. We are only looking at 1/16 of the address space ( SEL-E ). That would be 8.192 ms. If you set the scope at 5 ms/div, the entire event we want to see is less than 2 divisions at the left end of the scope.
    We'd like to see a square wave at 4 us.
    Dwight

    ( corrected my math )
    Last edited by Dwight Elvey; February 15th, 2020 at 12:46 PM.

  7. #67

    Default

    Hello Dwight, Thanks for your detailed instruction. I hope I can get a useful picture from my scope with your recommended settings. Unfortunately the picture contrast goes quiet down when I change from milliseconds to microseconds. The intensity of the CRT is already on maximum!

    At the moment I am on business travel until Thursday. I will have a chance to repeat the measurement on Friday then.

    Have a nice week.

  8. #68

    Default

    That may not be needed. I'm doing a little math on the image of the video we do have. The address should be sequencing sequentially. If the picture you have is correct, we should be looking at data while sync on the EPROM we created ( assuming you are sync'd on the negative edge of the pin 20 ). I'm also assuming you have the sweep at the calibrated 5 ms/div. Each select signal is 4096 bytes long ( 65536/16 from D2 75154, a 4 to 16 decoder ). Each 5 ms division is 2500 NOPs ( 0.005/0.000002 ) long. We can now measure across the screen to see where things start and stop.
    Now, each address is selected for 2 micro seconds. That means the E000-EFFF ( 4096 long ) would take. 4096/2500 is 1.6384 divisions. We know that 1/2 is not enabled being 0.8192 divisions. Looking at the scope picture, I'm assuming a little delay on the sweep start but settles into the 2500 NOPs per divisions. The second half of the 2716 isn't being read so should float until we get to the F000-FFFF EPROM.
    If I measure from the end of what looks likes our 1s and 0s to where I expect the F000-FFFF to start, I see about 0.8 divisions. That is from 0.4 to 1.2 divisions. I see 0.8 divisions of what looks like a float, then a change at 1.2 divisions to another float. That should be the F000-FFFF EPROM?? I don't see any data there??
    At 2.9, I begin to see data again. I suspect this is the RAM. That would tell me there is an issue reading D9. With the chan 2 sync'd to D8 pin 20, look at the selects for D9 EPROM. I suspect it isn't being enabled. With the same scope setup, I'd expect to see the SEL F to start there.
    Dwight

  9. #69
    Join Date
    Feb 2009
    Location
    Southern California, USA
    Posts
    2,833

    Default

    Dwight,
    That's a good hypothesis, no /SEL F pulse at D9-pin 20 would tell the story.

    My kingdom for a delayed sweep function for this exercise!
    -Dave

  10. #70

    Default

    Quote Originally Posted by dave_m View Post
    Dwight,
    That's a good hypothesis, no /SEL F pulse at D9-pin 20 would tell the story.

    My kingdom for a delayed sweep function for this exercise!
    -Dave
    The video actually works relatively well. It captures busts of video but since he'd only ran it for few seconds, I was able to piece things together. I'm not sure it is a missing strobe but that is the likely case. It could also be a missing ground from Pin 18. It is more likely the socket or the adapter. The D2 is also a potential failure point. In any case, it is missing data from D9. Even PETTEST won't work without the reset vector from D9 EPROM.
    We are narrowing in on the issue. It is always a little difficult debugging through the MB but even though slow, the pieces begin to fall into place.
    Anyway, this does show that there are creative things one can do with the NOP generator. If it were my tool, I might enhance it a little so I could tie address lines to 1, 0 or the NOP generator. One could do that with a double throw ON-OFF-ON switch and a pullup resistor. It would be useful on, at least, the top 4 to 5 address lines as it could speed the loop up 16 times. One could then target a narrow range of addresses. The only thing missing is a way to loop on writes. Still, once one has the reads working, one can write simple code to do writes to desired addresses in a loop. That is the great thing about being able to blow EPROMs.
    Dwight

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
  •