Advice on MV1-FZ stuck in Watchdog.

[jew90]

As per the title this board is stuck in watchdog, the bios was also fried so i've replaced it with a bios that i know works and it is still in watchdog.

Checked continuity from SRAMS to CPU and from bios to CPU and everything seems to check out.

Replaced the two main work SRAMS with two from a known good board and its still in watchdog.

On the SRAMS

[20] CE: Is low on both RAM1 / RAM 2
[22] OE: Is High / Pulse on both RAM1 / RAM2
[27] WE: High on both RAM1 / RAM 2.

WE is active low I believe so it should not be high?

Any suggestions would be great, im coming back to it in the morning, having a break for now.

 

[Xian Xi]

How did the bios fry? Is the CPU working properly?

 

[jew90]

How did the bios fry? Is the CPU working properly?

No idea on the bios.

Not sure if the bios in it was one used for testing, or a dead one i'd got.

The replacement i've put in still works.

I'll check the controlling lines to the bios in a second and see what state of play they are in, was just wondering where they connect to the work sram.

I can always change the 68000 too, but if it was dead would it keep resetting the board?

I figurered theres probably a missing trace from cpu/bios/sram?

 

[channelmaniac]

Watchdog is separate hardware so yes, it would hammer a dead CPU too.

Stuck in Watchdog can be cause by ANYTHING that affects the main CPU address/data busses or address decoding.

Check ALL the pins on the address/data busses on the Work RAM, Backup RAM, BIOS, and CPU for continuity. Next check the BIOS ROM for activity on the enable lines. If that's not working then there's an issue with the CPU or with the address decoding for the enabling of the chips at the proper block of address space.

Next check the enable lines on the work RAM and the backup RAM. If those are floating or stuck high then those lines will never come active and the system will freak out. If stuck high it should give you a backup RAM error, but not always. If floating then the RAM could be outputting spurious junk and cause the CPU to be stuck in Watchdog.

It's common for the system to be dead when the backup battery leaks and corrodes traces around the 74HC32 and other chips/components that are powered by that battery. The enable lines for the backup RAM go through that 74HC32 as those lines have to be kept in certain states for the RAM to go into low power standby mode to save the system/game settings.

Raymond

 

[jew90]

Thanks Xian, Raymond.

I didn't realise that the watchdog was a seperate circuit, if the bios was killed with extra power then the cpu might be dead too, fortutuantly i have a parts board i can take one off.

The backup ram error is interesting, I got this on an identical 1FZ, here is what I recorded to fix it.

Ram test produces an immediate Backup Ram ERROR Read: 5555 Write: 55FF

Replaced backup ram IC at RAM3 with known good spare. Still error persists.
Repaired trace from Backup Ram 3 W/E Line (Pin 27) to (Pin 8)74HC32 at U17

I shall test the bios / cpu / backup ram and see what the state of play is

 

[jew90]

OK this is interesting.

All of the Data/Address lines have continuity with both the work and backup SRAMS.

All of the Address lines on the BIOS are Hi/Pulse on the logic probe except A10 which is just Pulse.

All of the Data Ouput lines on the BIOS are HI without pulse.

The BIOS chip works i've tried it in another board.

Any suggestions?

 

[Xian Xi]

All of the Address lines on the BIOS are Hi/Pulse on the logic probe except A10 which is just Pulse.

Is it connected to CPU(39) and RAM(21)?

 

[jew90]

Is it connected to CPU(39) and RAM(21)?

It's connected to Ram (21) and CPU(42)

CPU(39) connects to BIOS(28) [A7] and RAM(3)

 

[jew90]

Interestingly theres an LS245 Octal Transciever at U9 that is getting slightly warm... miles away from the coin-cell battery that isn't leaking.

All of the Bus A lines are pulsing.

All of the Bus B lines are High.

I don't have another SOIC 245 to swap it out with yet but I have a feeling this could be part of the problem!

Since the original bios doesn't appear to work either, could this board have suffered power overload / short circuit damage I wonder.

 

[channelmaniac]

Check the control lines for that 245 before you go too crazy...

Also, if the BIOS was socketed it's very likely that it was plugged in backwards by someone not paying attention.

 

[jew90]

Check the control lines for that 245 before you go too crazy...

Also, if the BIOS was socketed it's very likely that it was plugged in backwards by someone not paying attention.

Ah yes the bios could have been, I don't know the history of the board or if anyone has tried anything with it.

I think everything on that 245 was pulsing bar the B Bus but i'll double check it before replacing it.

I tihnk the fact it gets warm enough to the touch to notice is a sign it isn't quite well but will double check.

Thanks for the suggestions / input so far.

Hopefully it will be booting before long. :-)

 

[jew90]

Check the control lines for that 245 before you go too crazy...

Also, if the BIOS was socketed it's very likely that it was plugged in backwards by someone not paying attention.

Ok.

[1] (DIR) - Pulse
[2]-[9] A1-A8 - Pulse
[11]-[18] - High
[19] E*- High

This chip is noticably warmer than the other TTL's too.

Wont have one for a few days to swap it out with. So onto the other board (making new thread)

 

[channelmaniac]

Ok.

[1] (DIR) - Pulse
[2]-[9] A1-A8 - Pulse
[11]-[18] - High
[19] E*- High

This chip is noticably warmer than the other TTL's too.

Wont have one for a few days to swap it out with. So onto the other board (making new thread)

GO READ THE DATASHEET.

http://focus.ti.com/lit/ds/symlink/sn74ls245.pdf

OE being high means the busses are isolated. There is no data flowing across that 245. So unless those data pins (11-18) are shorted to Vcc (+5v, a.k.a pin 20) there's probably no reason to replace that chip.

It's normal for TTL logic to run warm. If it generates enough heat that it's taking your skin off then you need to replace it.

 

[jew90]

GO READ THE DATASHEET.

http://focus.ti.com/lit/ds/symlink/sn74ls245.pdf

OE being high means the busses are isolated. There is no data flowing across that 245. So unless those data pins (11-18) are shorted to Vcc (+5v, a.k.a pin 20) there's probably no reason to replace that chip.

It's normal for TTL logic to run warm. If it generates enough heat that it's taking your skin off then you need to replace it.

Um I did read the data-sheet, I have it printed out.

Although I did miss the bar over the OE... :/

 

[Kyuusaku]

Some computer theory would really help here, this is a crazy way to troubleshoot.

 

[jew90]

Some computer theory would really help here, this is a crazy way to troubleshoot.

Can you elaborate further? I don't see how it is 'crazy'

Symptoms:

Watchdog.

Troublshooting Path:

1. Check continuity between CPU / BIOS / Work & Backup Ram. [No problems]

2. Check and replace faulty bios. Still problem.

3. SRAM WE is held high.

Likelyhood that problem is between BIOS / CPU / RAM.

While testing traces I notice that LS245 near bios is quite hot (much warmer than normal ttl). Outputs are high but so is output disable (mis-read datasheet)

Still don't trust this chip, even if it might not be causing watchdog. Afterall original bios chip was dead. CPU could be dead too but appears normal.

 

[Kyuusaku]

I mean read up on digital logic and computer architecture (knowing what active low means is really important). I've fixed many things and not once have I had to replace an IC. OK I can think of one or two times, but it's almost never.

The only way the '245 would be relevent is if it buffers the 68K data bus to the carts P ROM, otherwise it's probably not pulsing because it leads to controller switches, the palette or other things not being accessed.

There are two critical chips you haven't mentioned checking continuity to, the NEO-C# which decodes the address and control signals, including your non-pulsing /WE, and the P ROM (perhaps by way of '245 buffers). The P ROM is just as critical as the BIOS, and moreso than the backup RAM since the CPU shouldn't execute code there. The SRAM's /WE is probably nothing to worry about because the CPU will immediately crash and there's little chance it will be writing to that memory.

 

[jew90]

I mean read up on digital logic and computer architecture (knowing what active low means is really important). I've fixed many things and not once have I had to replace an IC. OK I can think of one or two times, but it's almost never.

The only way the '245 would be relevent is if it buffers the 68K data bus to the carts P ROM, otherwise it's probably not pulsing because it leads to controller switches, the palette or other things not being accessed.

There are two critical chips you haven't mentioned checking continuity to, the NEO-C# which decodes the address and control signals, including your non-pulsing /WE, and the P ROM (perhaps by way of '245 buffers). The P ROM is just as critical as the BIOS, and moreso than the backup RAM since the CPU shouldn't execute code there. The SRAM's /WE is probably nothing to worry about because the CPU will immediately crash and there's little chance it will be writing to that memory.

Hey.

it was my bad.. I understood the explanations but I misread the chart (didn't see the bar).. and thought that active was high not low.

Anyhow.

Back to this board

The MASK1 chip is enabled and data lines are pulsing.

The Bios (mask2) chip output is disabled though.

I've got continuity between this pin and pin 26 or 36 (will check when I get home) of one of the NEO chips.

This matches with my Working 1FZ's.

However, on the working 1FZ's I get continuity from this same pin to the Work/Backup Ram chips.

I don't get this on this watchdog board.

Also, which is interesting.. IF i have the multimeter on continuity / diode test. I get a 'beep' when the red lead is on the bios Output Enable leg (pin 20) and the black is on the battery +ve on my working boards.

I don't get this either way around on my dead board.

I can't see any bad traces though.

Faulty chip(s)?

Been meaning to post but been busy the last week or so

 

[Kyuusaku]

Data lines will pulse because it's a shared bus and a chip will only be enabled while it's being accessed, so neither of those are suspicious and shouldn't be mentioned. If you want help with the troubleshooting you should say the signal names instead of pin numbers or "this pin" and "that pin", it's confusing. I can only assume you mean the read strobe because that's the only control signal that should be shared between the BIOS and RAM. If you look it up and that's what it is, it very well could be the problem.

Output enable and battery "+ve"?

Faulty chips are not likely, I don't think you've systematically checked everything.

 

[jew90]

Data lines will pulse because it's a shared bus and a chip will only be enabled while it's being accessed, so neither of those are suspicious and shouldn't be mentioned. If you want help with the troubleshooting you should say the signal names instead of pin numbers or "this pin" and "that pin", it's confusing. I can only assume you mean the read strobe because that's the only control signal that should be shared between the BIOS and RAM. If you look it up and that's what it is, it very well could be the problem.

Output enable and battery "+ve"?

Faulty chips are not likely, I don't think you've systematically checked everything.

It's definatly Output Enable (Pin 20) on the bios that has continutiy to the Neo Chip

It goes to the Neo Chip which I dont' have the pinouts to. This same pin has continutity with the RAM and battery on the working boards.

On my faulty board it only goes to the Neo Chip, the Neo Chip doesn't connect to the ram / battery.

I'll make a detailed post tonight.

I have 3 working 1FZ's to compare too and there all the same...

 

[MKL]

You're nuts if you think the output enable pin on the bios (pin 20) connects to anything other than pin 59 of the NEO-E0.

 

[jew90]

You're nuts if you think the output enable pin on the bios (pin 20) connects to anything other than pin 59 of the NEO-E0.

Hey.

it is strange.

Yes this is the pin it connects to, I was confused before.

This is what I don't understand.

If I connect the positive lead of my meter to Pin 20 on the bios chip and the -ve lead to the battery positive I get beep. (on working board, not if i switch leads around)

On the dead board I dont get this beep.

Where is the E0 pin connected to?