So this board could potentially boot with these roms if I populated all the banks... I have a ton of DRAM here that came with the boards, maybe I'll try to breadboard one of those Arduino RAM testers.
You will be talking about a boot into Cassette BASIC.So this board could potentially boot with these roms if I populated all the banks...
Thanks acgs for commenting here on my behalf! I didn't see your post or I would have thanked you sooner!
As acgs said, I am working on a 5150 based project to redesign one, and a reference board would possibly be helpful to me!
To make sure everyone is on the same page, this is an address error, not a data error. An address related problem is broadly explained at [here]. The cause may or may not be a RAM chip.Ok, finally running Ruud's 4.3 With all RAM banks populated! I consistently get failure at Address: 15KB (exactly 03FFC [hex])
Thanks rodney, could use some! These 5150 bus protocols are nothing to sneeze at, thank god for modem7's diagrams and test code.It will be interesting to see your progress and what you discover VeryVon, good luck!
For those that have a screen display, note that the checkpoints also appear in the top right corner of the screen. Sometimes, that provides possibly relevant information. For example, the B7 shown in your screen shot within post #46, informs me that of the address test, it was the first part (of two parts) that failed.I also have the Mini Post Code card connected, it's really cool to see the checkpoints read out on the card
The ZIP download file of Rudd's Diagnostic ROM (RDR) includes the source code, and so if you really wanted to, you have the ability to predict the failure address before running RDR.Interesting what you have done in experimentation modem7. If I have the time, I may do this too just for fun. Having seen more examples of what the diagnostics show can increase diagnostic experience and help to recognize what is wrong when you see a certain type of test result, interesting.
I think it is rare (comparatively speaking). My last find was two years ago. I recorded it in the bottom table at [here]: "Discovered about 4 ohms between bits 2 and 1 of the external data bus. Tracked down to chip U10, a 74LS670, used for the DMA page registers."It's good to consider stuck bits as a possible cause because it can happen.
That 'Checkpoints' page that you point to includes:Thank you for the explanation on "address testing" it riddled my mind for a good day or two, and I didn't realize you have a page dedicated to that. Very well documented.
Just be careful about interpreting things using an oscilloscope.Data Lines: I'm getting a mixture of signals here that the logic probe often interprets as a "buzz" neither high nor low. Except for D3 which appears to be steady high. They don't look right on the scope either
In the ZIP download file is file TEST5065.TXT, and that has a 'WHAT YOU SHOULD OBSERVE' section.I went over the ram banks too. Not sure exactly what I should see there running TEST5065, but here are my observations: ...
So, activity on all 20 lines of the address bus. Good.... Address Lines: all active
No, you cannot say that they have good outputs. All you can say is that you observed activity on the output pins. Maybe the activity is 'bad'.I also checked U62, U79 and RN4 (U78) and all good input & outputs.
Indeed a good point, when I was writing about using the oscilloscope, I also remembered seeing some half voltage signals in my XT design. I forgot to comment about it so thanks modem7 for pointing to this possibility which I also have seen before while debugging my XT.Just be careful about interpreting things using an oscilloscope.
For example, the following is an oscilloscope capture of one of the RAM data pins on my good motherboard.
I think it is rare (comparatively speaking). My last find was two years ago. I recorded it in the bottom table at [here]: "Discovered about 4 ohms between bits 2 and 1 of the external data bus. Tracked down to chip U10, a 74LS670, used for the DMA page registers."
In diagnosis, I established that there was a problem on the external data bus, and eventually deduced that the same signal was always on data bits 1 and 2. That's when I checked using a multimeter and to my surprise, discovered the 4 ohms. I knew that the cause was either a chip or the PCB. In turn, I cut the data bit 1 (or maybe it was bit 2) pin on the chips on the external data bus, and the 4 ohms disappeared when I got to the 74LS670.Very interesting modem7 about the 74LS670 which failed with such a low resistance short.
No.Do you happen to remember who was the manufacturer of that 74LS670?
I used to have many 51xx motherboards, now down to about 8. These are boards that I have had for many years, and are stored well. I make sure they get periodic 'exercise'. Sometimes I bring out a motherboard (that was put away in working state), and discover that it no longer works. Of course, I do not know if the faulty component failed at application of power, or failed during storage.I also always wonder how ...
I guess that's where these five boards came from? I'll have to inquire with the original collector to see if he knows more.the board was unusable and would have been shelved
Which includes 'bank 0 populated only'. Maybe one (or more) of the chips presently fitted to bank 0 is causing the problem. A leg bent up under the chip is a possibility.I started removing banks and booting RDR. The same error appears each time.
Yes.I'm getting the gut feeling it's something with the data bus, but I guess that's a pretty broad statement haha. Should we test for "stuck bits" on the data bus?
Legs look fine, and I checked the underside of the board under the scope again, can't find any obvious problems. However looking at the chips in the sockets under the scope, some of the leg's aren't touching the outside socket wipes. They're double-wipe but it's not giving me super confidence, I wonder if these are the cheaper type of sockets one gets on amazon. I can see why people switch exclusively to machine pin sockets, and I've considered it myself. I'll do some more recon to verify conductivity, I've seen more than one war story out there about flakey sockets causing people to pull their hair out.A leg bent up under the chip is a possibility
I'd certainly like to be. I downloaded nasm-2.16.01, and ran this to attempt to duplicate your binary:If you are capable, you can modify the TEST5065 code to output addresses AAAAA and 55555
nasm -f bin TEST5065.asm -o TEST5065.BIN