Neo Diagnostics ROMs

schmake

n00b
Joined
Aug 6, 2012
Posts
3
On your good MV1F you'll get a blank screen if don't have the Z80 test cart inserted and you don't hold down button D. Try again with D in P1 port held and it should get through to the end. Was this tried on both of your bad 4 slots? If the BIOS works fine in the 1F but gives CRC error in your 4 slots then the board must have an issue. My first guess would be a bad burn for CRC errors but the working 1F rules that out. If it's the same CRC value on screen everytime I think it's a lower address line from 68k to BIOS. That can easily stop the system from working completely but it's possible to get far enough to display an error like this.

Silly me, I held the D button and all tests pass on the 1F. I also tested it on another good 4 slot board I have here a tests pass, so the BIOS is good.

On the 4 slot which gives the CRC error it gives the same value on screen each time. I will investigate lower address line from 68k to BIOS

Thanks again mate.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
I don't hear any beeps or tones.

LOL, I forgot that when I made an AES cart that I ran out of 27C1001/010 so I had to use the one from the MVS one I made and forgot to put it in.

So I put it in and it passes all the tests but still trying to figure out why it's resetting randomly. Could it be a failing Neo-B1 or CPU?

Edit: The reset is coming from the Neo-B1. All connections are good halt, reset, AS, LDS and R/W. Could work ram be going bad spitting out things?
 
Last edited:

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
I probed it today and found that R/W on the 68k(9) pulses at random rates and when the soft reboot happens it goes low. Is that normal?

Edit: Tried testing again with the Diagnostics and if I don't press anything on boot I get the normal backup ram message and a 001100. If I press the normal B and C on the AES it passes all tests. If I hold B and D then it passes all tests but I get the 001100 as well.

I checked all data and address lines on the 68k and all are fine. I notice that it doesn't randomly reboot with the Diagnostics, is this due to no 68k program running?
 
Last edited:

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
I checked all data and address lines on the 68k and all are fine. I notice that it doesn't randomly reboot with the Diagnostics, is this due to no 68k program running?

68k program always runs even if the only thing it has to do is stop watchdog reset from running. It doesn't depend on RAM as much as a game would so if you get intermittent RAM problems causing weird resets like this, that could be the reason.

I posted 016 with a RAM test loop option like the MVS BIOS has when you flip all the DIPs. Try that on the AES. If it detects it's running AES it only does WRAM so you don't have to do anything special. Power on with ABCD held to get the menu and select the test. Try letting it run for a while to see if it's a RAM issue.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
Went through 2200 WRAM passes. What devices can trigger a soft reboot?
 

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
2200 passes must have been around 15 minutes. If this AES doesn't have the reset problem with this BIOS as games do, you can scratch the B1 and WRAM from the list. B1 must be good or else it would've reset since preventing unwanted watchdog resets is done the same way as any game. If the 68k doesn't get proper interrupts from the LSPC like it should, that can lead to a reset since the watchdog will time out. That's just a guess considering the BIOS doesn't depend on interrupts to work which could be the reason it only resets in gameplay. It's hard to blame any one device with this problem, but I don't think it's the B1 based on what you're getting and it's definitely not the RAM.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
2200 passes must have been around 15 minutes. If this AES doesn't have the reset problem with this BIOS as games do, you can scratch the B1 and WRAM from the list. B1 must be good or else it would've reset since preventing unwanted watchdog resets is done the same way as any game. If the 68k doesn't get proper interrupts from the LSPC like it should, that can lead to a reset since the watchdog will time out. That's just a guess considering the BIOS doesn't depend on interrupts to work which could be the reason it only resets in gameplay. It's hard to blame any one device with this problem, but I don't think it's the B1 based on what you're getting and it's definitely not the RAM.

How does the LSPC interrupt work? Could it be the fast video RAM? Is there a way to loop a test for video RAM?
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
How does the LSPC interrupt work? Could it be the fast video RAM? Is there a way to loop a test for video RAM?

Well, I pulled the fast video RAM aka 5814s and tested them and the upper was fine but the lower failed once out of 20 tests. So I pulled one from a parts board and so far it isn't rebooting anymore, need to do more testing.

Edit: Yup, played it for 30 minutes with no reboots. If possible, a fast VRAM loop test could be useful.
 
Last edited:

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
I missed that edit. Cool to see the problem fixed. Palette/video memory loops will get added to the menu in 017 probably.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
I missed that edit. Cool to see the problem fixed. Palette/video memory loops will get added to the menu in 017 probably.

Cool. I think they will be useful for weird ones like mine.
 

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
I posted 017 with the display RAM test loops added. Most of them make the display freak out so button A can be held to pause it and see current results, then resume from the point where it stopped if wanted. It's worth trying if you get strange issues like Xian did and want something extra to try, or if you see graphic glitches that aren't down to a dirty slot and regular RAM tests don't show helpful or consistent results.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
WRAM DATA (0000) 1001000
Address: 1002C4
Actual: 007F
Expected: 0000

I'm guessing this means the work ram is bad?

Edit: It changes address sometimes to 100640 (00FF), 105340 (007E), 10F342 (0035), 105B40 (00BC).
 
Last edited:

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
In all those cases the upper byte worked properly so something is up with the lower chip. I'd guess it was the RAM if the addresses weren't random but it could be a trace that needs some work. I don't think it's data or the /OE based on those results so if the rest looks good then try replacing it.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
One thing I noticed was that I couldn't get to the sub menu by pressing ABCD, it goes into watchdog. I'll check the /OE in the morning. I know the data lines were all connected to the 68k.
 

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
That's normal if you have WRAM problems since the tester you posted results from doesn't depend on it to function, but the submenu does. The SNK BIOS would have similar problems.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
Got it, it was the A8 interconnect that was broken.
 

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
I figured it was either address or /WE causing that. If address is unstable then you'll just get random results.
 

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
Bump with an update. 018 is posted now. It covers a bunch of extra problems that have less obvious fixes compared to stuff like RAM. Certain cut traces between certain custom chips lead to some known problems that are detectable. My battery damaged MV4F had a bad trace with a NEO-F0 control signal that made DIPs unreadable (DIPs themselves were fine) and that's one of the things that now gets caught. Pinout info for the problem is put on screen for repair but only the 2nd gen stuff is complete with all the lacking info on the older chips.
 
Last edited:

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
Bump with an update. 018 is posted now. It covers a bunch of extra problems that have less obvious fixes compared to stuff like RAM. Certain cut traces between certain custom chips lead to some known problems that are detectable. My battery damaged MV4F had a bad trace with a NEO-F0 control signal that made DIPs unreadable (DIPs themselves were fine) and that's one of the things that now gets caught. Pinout info for the problem is put on screen for repair but only the 2nd gen stuff is complete with all the lacking info on the older chips.

Awesome work.
 

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
Just posting to say Xian Xi is selling this with the optional sound cart for the full package. If anyone wanted to try it out but couldn't set it up on their own, visit his site and have a look.
 
Last edited:

Mehunglo

n00b
Joined
Aug 21, 2007
Posts
25
Questions on the CHA board. Do I have to remove the C roms? I'm assuming I have to replace the M1 on the board with the M1 bin purchased.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
Questions on the CHA board. Do I have to remove the C roms? I'm assuming I have to replace the M1 on the board with the M1 bin purchased.

You can leave the C roms in, just swap the M1 and make sure you have an S1 in as well.
 

Mehunglo

n00b
Joined
Aug 21, 2007
Posts
25
I'm having a problem with my M1 cart. I keep getting the Z80 dead error. I've tried it on a 2 and 4 slot which do not have audio yet the games work fine. I've already tried cleaning the carts and slots many times. Xian has been helping me troubleshoot the cart.


 
Last edited:

smkdan

Galford's Armourer
Joined
Dec 30, 2009
Posts
452
The error is probably real (and not just the cart) if the 2/4 slot work fine but without game audio. That error covers the Z80 crashing when trying to access the cart M1 which explains missing game audio. If you have another 100% working board to test on then you can rule it out easily.
 
Top