That sucks - literally.DJ Sures wrote: ↑Thu May 11, 2023 11:16 amyou won't regret it for the price.
One bit of warning. When you press the button, it creates a sucking action. If you release the button, any solder in the tube falls out. So you do NOT want to release the button while the tube is over the circuit board Lesson learned!
What I do is suck the solder, and keep the button held. Move the sucker over to an old container, release the button and the solder falls out into the container.
Nabu video
Re: Nabu video
Re: Nabu video
Da dum Ching!MagnumPEI wrote: ↑Thu May 11, 2023 12:10 pmThat sucks - literally.DJ Sures wrote: ↑Thu May 11, 2023 11:16 amyou won't regret it for the price.
One bit of warning. When you press the button, it creates a sucking action. If you release the button, any solder in the tube falls out. So you do NOT want to release the button while the tube is over the circuit board Lesson learned!
What I do is suck the solder, and keep the button held. Move the sucker over to an old container, release the button and the solder falls out into the container.
- productiondave
- Posts: 117
- Joined: Tue Mar 28, 2023 10:01 pm
Re: Nabu video
Two snares and a cymbal fall off a cliff.DJ Sures wrote: ↑Thu May 11, 2023 4:42 pmDa dum Ching!MagnumPEI wrote: ↑Thu May 11, 2023 12:10 pmThat sucks - literally.DJ Sures wrote: ↑Thu May 11, 2023 11:16 amyou won't regret it for the price.
One bit of warning. When you press the button, it creates a sucking action. If you release the button, any solder in the tube falls out. So you do NOT want to release the button while the tube is over the circuit board Lesson learned!
What I do is suck the solder, and keep the button held. Move the sucker over to an old container, release the button and the solder falls out into the container.
Rock On!
~ productiondave
https://github.com/linuxplayground/nabu-games
https://www.reddit.com/r/beneater/
~ productiondave
https://github.com/linuxplayground/nabu-games
https://www.reddit.com/r/beneater/
-
- Posts: 19
- Joined: Sun Apr 30, 2023 3:39 pm
Re: Nabu video
Well the de-solder-er arrives tomorrow, and just in time. Powered up the Nabu and I see this now....
I accidentally left it on overnight, I wonder if this is just an early death of a DRAM chip, I'm tempted to replace them all if this is the case. I'll report on the effectiveness of the tool after I fix the Nabu boo-boo.
Which according to Rudy's handy dandy debug manual looks like U46 - HM4864P is bad. Will do some probing around to see and pick up some replacement parts & sockets.I accidentally left it on overnight, I wonder if this is just an early death of a DRAM chip, I'm tempted to replace them all if this is the case. I'll report on the effectiveness of the tool after I fix the Nabu boo-boo.
Re: Nabu video
good thing you left it on to give it the "burn in" test or in this case "burn out" haha
-
- Posts: 19
- Joined: Sun Apr 30, 2023 3:39 pm
Re: Nabu video
No joy in Nabu land so far. I was checking all the system DRAM outputs and they all look similar, which I assume is OK. So my initial guess of the U47, or one of the HM4864P, being bad based on the diagnostic guide and image on screen was incorrect.
So I went about creating some VDP test code to see if I can pin point what is going on. It is banging away on the VDP RAM and checking all bit patterns at every location. So far looks like random data read back.
BTW, to do this I am hijacking the serial out from HCCA and banging out text to see what is happening. Since I don't have any EEPROM's yet to program I am doing this by starting up the IA and the Nabu which luckily still loads the test code and runs it. Once serial data starts spitting out the HCCA the IA freaks out a bit, i just close it and start minicom (any terminal would work), and capture the output from my test code.
I'll do more probing around VDP DRAM with specific patters to try and isolate. My fear is the 9918a is dead. Those little buggers are hard to find, was thinking of the F18a. Any recommendations?
Cheers
So I went about creating some VDP test code to see if I can pin point what is going on. It is banging away on the VDP RAM and checking all bit patterns at every location. So far looks like random data read back.
BTW, to do this I am hijacking the serial out from HCCA and banging out text to see what is happening. Since I don't have any EEPROM's yet to program I am doing this by starting up the IA and the Nabu which luckily still loads the test code and runs it. Once serial data starts spitting out the HCCA the IA freaks out a bit, i just close it and start minicom (any terminal would work), and capture the output from my test code.
I'll do more probing around VDP DRAM with specific patters to try and isolate. My fear is the 9918a is dead. Those little buggers are hard to find, was thinking of the F18a. Any recommendations?
Cheers
Re: Nabu video
So you can boot into text mode? It just goes funky in mode G2?
If that's the case, you could put the tms99 in bitmap mode, meaning it splits the memory into 3rds. You can use NABULIB for that. A good test could be to simply draw a horizontal line across the center of each tile. Use this project: https://github.com/DJSures/NABU-LIB/tre ... c%20Bitmap
Essentially just loop from and draw a line from the top left to the top right. Move down one Y pixel and do it again. Keep doing that until you see where it starts to go funky. That will tell you approx what memory area is messed up. Get the approx memory address, divide it by the number of vdp ram chips and bob's your uncle.
If that's the case, you could put the tms99 in bitmap mode, meaning it splits the memory into 3rds. You can use NABULIB for that. A good test could be to simply draw a horizontal line across the center of each tile. Use this project: https://github.com/DJSures/NABU-LIB/tre ... c%20Bitmap
Essentially just loop from and draw a line from the top left to the top right. Move down one Y pixel and do it again. Keep doing that until you see where it starts to go funky. That will tell you approx what memory area is messed up. Get the approx memory address, divide it by the number of vdp ram chips and bob's your uncle.
Re: Nabu video
This is exactly one of the uses for the 8251A hack. You could use this for serial out for debugging by writing to 90h and 91h to display debug info. No need to switch the console using redirect.com just write to the keyboard controller to send chars out. Make sure you preserve RX enable in 91h so the key board will still work. Looks like DJ got a good suggestion otherwise but if you need some debug info too then this would be the way.TorontoNabu wrote: ↑Mon May 15, 2023 5:07 am BTW, to do this I am hijacking the serial out from HCCA and banging out text to see what is happening. Since I don't have any EEPROM's yet to program I am doing this by starting up the IA and the Nabu which luckily still loads the test code and runs it. Once serial data starts spitting out the HCCA the IA freaks out a bit, i just close it and start minicom (any terminal would work), and capture the output from my test code.
Cheers
See: viewtopic.php?t=12
Greg
-
- Posts: 19
- Joined: Sun Apr 30, 2023 3:39 pm
Re: Nabu video
Thanks DJ, that's a good suggestion. I'm also using this as a Z80 assembly training exercise so sticking with stand alone for now. I will switch to C and use the library soon, and leverage all your good work.DJ Sures wrote: ↑Mon May 15, 2023 7:32 am So you can boot into text mode? It just goes funky in mode G2?
If that's the case, you could put the tms99 in bitmap mode, meaning it splits the memory into 3rds. You can use NABULIB for that. A good test could be to simply draw a horizontal line across the center of each tile. Use this project: https://github.com/DJSures/NABU-LIB/tre ... c%20Bitmap
Essentially just loop from and draw a line from the top left to the top right. Move down one Y pixel and do it again. Keep doing that until you see where it starts to go funky. That will tell you approx what memory area is messed up. Get the approx memory address, divide it by the number of vdp ram chips and bob's your uncle.
I'm planning to alternate one data bit on/off (or one chip) at a time continuously to ensure RAS, CAS, & R/W, & address & data are correct for each chip. I'm fairly sure the 9918 is gone, writing & reading VDP RAM now produces random result which kind of aligns with the picture I posted of kind of random noise on the display. Does anyone knows a source for TMS9918 in Canada or even Toronto for these? I can always roll the dice and go ebay or overseas.
-
- Posts: 19
- Joined: Sun Apr 30, 2023 3:39 pm
Re: Nabu video
I'll take a closer look! Thanks Greg. I'll need to dig up my USB/RS232 (5V level) adaptor.greghol wrote: ↑Mon May 15, 2023 4:27 pm This is exactly one of the uses for the 8251A hack. You could use this for serial out for debugging by writing to 90h and 91h to display debug info. No need to switch the console using redirect.com just write to the keyboard controller to send chars out. Make sure you preserve RX enable in 91h so the key board will still work. Looks like DJ got a good suggestion otherwise but if you need some debug info too then this would be the way.
See: viewtopic.php?t=12
Greg
Cheers