You have a new MPU so it would not be the old MPU board reset problems that used to occur. Maybe the power driver board needs repair.

Would post to the Whitewater Club... As well as here !
Should check out the club though as there is a lot of game specific information there.

Sounds like a typical WPC reset issue:

https://www.pinwiki.com/wiki/index.php?title=Williams_WPC#Game_resets

I'm going to offer a different opinion based on what was written above and then pose another question to the OP. I am not disputing that it cannot be a WPC reset but opining that it MAY not be.

This does sound like a typical reset but the speaker popping is a sign that it is not typical. The speaker pop happens due to a voltage / current surge at initial power on in the analog side. Once the circuits have reached an equilibrium any resets of the digital logic side don't change the analog circuit.

OP: When the "hard crash" happens does the machine automatically return to "game over" or "attract" mode or does it just sit there and do nothing until you physically power cycle the machine?

A slam tilt is equivalent to a "WPC reset" (but not identical) so you can trigger a slam tilt and see if the result is the same.

I would think the NEW MPU would have the WPC reset fix already on it

The CPU board is rarely (if ever) the source of the cause for a WPC reset. It's often the connectors between the transformer or the boards or components on the power board.

A new board is also not a guarantee that any issue will be fixed without root causing the actual problem first.

The power supply is typically the source of the famed WPC reset issue, since that's the board supplying the 5v. When the 5v drops too low, the MPU resets. Typically, problems on the MPU are caused by battery damage which result in various other symptoms (such as issues with the switch matrix), but don't generally cause resets.

There are of course a couple other problem points before you get into the mess of components on the driver board, but the pinwiki troubleshooting guide helps guide you through tracking down the issue.

[edit]: Nevermind, not a WPC reset, as per new info below.

Try reseating the 4 connector ribbon cables from the CPU to fliptronix to sound board to the DMD controller board. I bought a Wh20 last weekend and the machine crashed on me as well and went completely black. I went and reseated those cables and fixed it.

I'd also check all of your stand up targets. Be sure they are wired right. The base colors ( not the stripes ) should be the same on everyone. White wire and black end of diode to a lug. Green wire to a lug. Silver band end of diode to a lug.

White Water is sensitive to a switch wired wrong.

LTG : )

This is NOT a WPC reset. This is either a hardware instability (causing incorrect software execution) or a software error (bug) that results in either an infinite loop or the processor possibly executing some form of illegal instruction and then trapping to an exception vector that is not equipped to recover from that situation. I don't know enough about the 6809 to know for sure but it's probably not complicated enough to have exception handlers - it only has IRQ, FIRQ and NMI. It likely executed an undefined (in the hardware instruction decode) instruction and went off to la-la land.

This is what the watchdog is supposed to prevent. When the watchdog detects a low enough voltage so that TTL voltage levels become indeterminate it's supposed to reset the system to restore reliable execution. It's almost certain that your digital logic voltage level was fine - otherwise the watchdog would have reset the system.

Honestly ... it's unlikely (but not impossible) that it's a software bug. It's more likely some hardware issue causing incorrect software execution.

This is what a typical WPC reset does. The momentary drop in voltage below the threshold of the watchdog causes the watchdog to pull the ~RESET line low (that causes the system to reset) and when the ~RESET line is released (to high logic level) the system restarts (by itself).

Agree with the hang / crash in attract mode. In my experience it's indicative of hardware failure / instability.

If you want to exacerbate the classic WPC reset then "chimp flip" during multiball.

If the game is completely blacking out, no GI or anything, I would unplug the game and open the power box. The thermistor would be the first place I would look. When they're bad, you can usually see it. Can be bypassed to confirm.

Would also check the main fuse and main fuse holder.

Multiball has a greater potential to trigger more solenoids more frequently. Three balls in the playfield = three times the chance a solenoid will fire.

I believe you mean the LH-5 rom. That's the one with the reset issue. LH-6 fixed this.

Then I suspect that you have an issue somewhere with your machine. I've been running the LH-6 rom since it was first released, without ever having a reset.

When I had the LH-5 installed, I would get resets quite often.

And I have yet to hear of anyone else having such an issue with LH-6. Of course, I'm running original hardware, so it could easily be that one of your replacement boards is part of the issue.

Asking everyone else, has anyone else had a reset issue due to using LH-6 roms? And if so, what board set are you running?

Repin J114. I have seen many WPC games with loose tension on these pins and plugs in too easy. There's your sign.