rom garbled graphic

[solidcore]

[RESOLVED]
Custom ROM pcb+flash 29f160



Effected, 2.0 / 3.1
Not tried 3.0
No problem with 3.1.4 , 1.3

Flash chip tested Fujitsu MBM29F160

Requirement : RP106 & RP107 must be populated pulling high address lines. No boot otherwise.
^ above statement was true a the time, but now it's incorrect. Now the r6 Boot without rp106/rp107. Rp106/rp107 was causing the glitches.

Symptom: Animation plays out as usual but a weird horiz/vertical band of garble across the sprites randomly show.

System : a500 r6

Any smart ideas? or even silly ones... at this point I am pulling hair out as to what it may be, I've been told presumably a timing violation issue.

[PR77]

Few things to check;

1. What timing is the Flash Chip? As you mentioned timing might be an issue but the original ROM was around 120ns (I think) and given other ROM images are OK I suspect this is not the problem.
2. What happens with the early boot menu?
3. Does the system boot OK?
4. Is the /WR line of the Flash Chip floating or pulled-high?
5. Are the ROM images you're programming OK?

[solidcore]

ok just tested on my r8a.1 with rp106/rp107 populated with some fresh 4.7kohm's which arrived today. ALL good. now going to replace them in the a500 r6....

[solidcore]

Originally Posted by PR77

Few things to check;

1. What timing is the Flash Chip? As you mentioned timing might be an issue but the original ROM was around 120ns (I think) and given other ROM images are OK I suspect this is not the problem.
2. What happens with the early boot menu?
3. Does the system boot OK?
4. Is the /WR line of the Flash Chip floating or pulled-high?
5. Are the ROM images you're programming OK?

1. 70.. but think im onto the problem, it may have been bad arrays that i had salvaged from a donor board previously.
2. early boot menu didnt try that yet... will try if cant resolve with new rp106/107's
3. boots fine games work fine etc.
4. some lines are not pulled ... gary > rom... but pulling high didnt make diff on those lines
5. Rom images are GOOD, uae / other amigas working great .

*update*

ok arrays were fine rp106/rp107 not faulty changed to different ones.. HMM problem persists.

early boot menu looks fine.

Changed denise - Nope
Changed gary - Nope

*update*

at this point im starting to think either the fuji chips are just out of spec for the a500 to cope , the r8 works great but the r6 isn't happy ... the r3/5 wouldnt work unless i put the pull ups on the rom pcb itself.

Out of curiosity, if the a500 r6 is not broken, why would the r8 be able to handle this better? same rp106/rp107 populated.

Could this be caps?

[solidcore]

ok update... isnt the flash rom fault ! so we can rule that out,i tried another rom that was working 100% before, so the miggy has had a fault somewhere.

swapped cpu.

ill try the cias. nope swapped both didnt do anything.

[PR77]

R3/5 have A17 pinned differently so that explains the pull ups (or pull downs- depending on where in the memory space you are addressing. R6 and up have the ROM socket routed for larger ROMs sizes already.

I would also check RP105 as this is pulling the higher address lines high. But to be honest, if there is an issue with the address lines I would expect the Amiga to simply not work.

Perhaps check if AGNUS is correctly socketed and if RP201, 202 and 203 are OK. Might be helpful to put some pressure on the RPs during measurement as they are thin film ceramic, a potential crack may not be obvious.

[solidcore]

Originally Posted by PR77

R3/5 have A17 pinned differently so that explains the pull ups (or pull downs- depending on where in the memory space you are addressing. R6 and up have the ROM socket routed for larger ROMs sizes already.

I would also check RP105 as this is pulling the higher address lines high. But to be honest, if there is an issue with the address lines I would expect the Amiga to simply not work.

Perhaps check if AGNUS is correctly socketed and if RP201, 202 and 203 are OK. Might be helpful to put some pressure on the RPs during measurement as they are thin film ceramic, a potential crack may not be obvious.

tried putting pressure everywhere, no change.. ill try reseat agnus.

*update* No change.

only thing not tried now is paula. maybe the a500+ paula go in there? if so ill try that.

Hmm this has stumped me .

[PR77]

I would then try DiagROM and run a memory test to exercise U34 and U35 for the ChipRAM ADDR and CTRL paths. If the issue is 100% occurring at the same point and the same part of the image is corrupted perhaps the issue is occurring at a specific address.

The other ROM versions I would suspect to have a different memory map with regards to image data.

[solidcore]

Originally Posted by PR77

I would then try DiagROM and run a memory test to exercise U34 and U35 for the ChipRAM ADDR and CTRL paths. If the issue is 100% occurring at the same point and the same part of the image is corrupted perhaps the issue is occurring at a specific address.

The other ROM versions I would suspect to have a different memory map with regards to image data.

2.0 has same issue as 3.1, i havnt got 3.0 handy atm , but will try soon as i do, 3.1.4 and 1.3 no issues.

Running diagrom on memory now... i will try the suggestion regards to u34/u35

Thoughts are only thing i did to this miggy was add rp106/rp107, could this cause such a problem on r6?

Ram appears fine, games working fine too... hmmm

---

things I have tried thus far:
- replaced gary
- replaced denise
- replaced cpu
- replaced roms (all roms 3.1/2.0 causing issues ,3.1.4/ 1.3 is fine)
- replaced CIA odd and even
- Reseated agnus
- replace paula

things I have not tried yet:
- Removing rp106/rp107 pull ups, but can't see how they're at fault really unless something is struggling to pull them down, any ideas?
- replace agnus

[lesta_smsc]

Had similar issue when my A500 had poor connection on Agnus. But these artefacts could potentially come from any of the display chips.

May be worth just testing the traces using Amiga PCB site of the Agnus and Paula pin connections. That way you know they are making adequate contact before having to re-seat or replace them.

[solidcore]

so i removed rp106, knowing for well that it probably means the rom wont boot anymore, as it wasn't working before... to my suprise the rom booted but glitches were fewer, i removed rp107 expecting no boot and now everything is working 100% glitch free.

So rp106/rp107 being installed on my r6 is bad. except even more weird is that it wouldn't boot them prior to installing them the first time around - What the hell!... and my r8 needed them. This is so confusing!


All I can think now is that one of the chips on the address bus is pulling high better than a chip I had in the first time round .... but not sure why rp106/rp107 causing glitches other than failure to pull low?

a moderator should Mark this thread RESOLVED.