This is the post I had planned on making for awhile and just never really found the time to do until now; I previously posted some information for this project in mac622's post on starting with a partially assembled TwoBits board. https://pinside.com/pinball/forum/topic/partially-assembled-twobits-wpc-mpu In this topic, I provided a link to a Google sites spreadsheet I created which contains the complete BOM for the WPC CPU Board (A-12742-XXXXX) and includes links to Mouser Electronics for every part used in the assembly.

The primary reason I like this board enough to build one is that the component layout and schematic are a near perfect match to the original Williams board. The only thing I don't care for is the component values for all the resistors are listed under the part making it impossible to read them on an assembled board, this will make future troubleshooting of the board a little more difficult.

For this build I ordered the complete list of materials from my Google sheets page https://docs.google.com/spreadsheets/d/1C4znEqZd4K4eywhi__qDwI-3AsX6Y5qdw9Pzb0p178A/edit#gid=1430569459 substituting the 5mm LEDs for the option 3mm LEDs which are a better fit on this board; additionally I found that the 7 pin header from TE Connectivity is better fit with the IDC connector originally used by Williams than the Molex one I had used in my previous build, and based on feedback from the previous post I decided to go ahead and socket U20 even though it goes against my personal preferences.

As of this writing the Partially Assembled TwoBits board is no longer available on ebay, and I was previously informed that that it will no longer be available once supplies are exhausted. Note: The partially assembled version of this board contained an incorrect R91 resistor (2.2M ohm intstead of 22M ohms) as identified by user A12742; this caused the oscillator used to control the RTC to operate in an unstable harmonic. Please keep in mind that in order for the RTC on this board to operate properly, you must use some kind of uninterruptible power source (typically batteries), any board using NVRAM will cause the RTC to cease operating whenever the game is off, which makes this incorrect resistor value irrelevant.

The blank TwoBits board is still available at the following link for anyone interested in building this board ebay.com link: Williams WPC MPU brand new old stock bare circuit board build it yourself

Just for the record building this board from scratch is not for the faint of heart, while its a fun project (if you enjoy doing this type of work), it took me twelve hours to get the blank board to the same state as the partially assembled board. The majority of time is spent inventorying, sorting, bending and trimming all the components, and finding their board locations; the time you will spend actually soldering components to the board pales by comparison. Full disclosure, I'm very detail oriented and I'm certain it takes me a greater time than most to build this board as it's important to me for the component part numbers to be visible on the board after soldering, as well as for the component to be centered between through holes and all facing the same direction; it really bothers me when resistor color bands don't line up.

Close up of the finished board, seems like no matter how careful I am I always end up with one capacitor with the writing orientated the wrong direction...this is likely due to my refusal to acknowledge my need for bifocals
Closeup (resized).jpg

The parts:
IMG_1815 (resized).JPG

The BOM calls for a pair of 72 pin headers, these are used to make the 4 ribbon cable headers (J201, J202, J204, J211).
IMG_1822a (resized).jpg

I like to build the board starting with the lowest profile parts and building upwards, on this boards the glass diodes are the shortest part and are installed first and working towards taller components, followed by the resistors, capacitors, and larger diodes, I do a quick flux cleaning as needed to help control how messy the board is, not to actually remove it all at this point.
IMG_1848 (resized).JPG

continued...

At this point, the only difference between the blank board and partially assembled board are some chip sockets, before I install any sockets though, I like to solder in any IC's that are not going to be in sockets.
IMG_1849 (resized).JPG

Then the sockets for the remaining IC chips, and the rest, I don't have the DIP switch installed yet as I plan on washing the board before installing this part, it would likely be fine to do so, but since the switch isn't sealed I don't want to risk having water trapped inside it.
IMG_1855 (resized).JPG

To wash the board, I like to just toss it in the sink and spray it down with scrubbing bubbles, after is soaks a minute or two, I go over it with a stiff cleaning brush to get the flux out from between components and repeat as required until the board is clean. Afterwards I go over the board with my air compressor to blow any water out of the sockets and off the board.
IMG_1856 (resized).JPG

Finished board (except for battery holder and plug in chips)
IMG_1865a (resized).jpg

All that's left is testing the board and installing the battery holder, this is where things did not go quite as planned and the board failed to complete the boot (not what you want to see).
IMG_1867 (resized).JPG

Since all solder connections were inspected under a magnifier as the board was being assembled, I immediately ruled that out as a possible cause. The power to for the board is coming from my workbench power supply and is dead on +5 and +12 leaving me with one of 4 possible chips as suspect for this, two of them are the ASIC and ROM but since they were both pulled from a known working board, I ruled them out, leaving only the Alliance Memory and Motorola Processor as unknowns; swapping these one at a time with parts from a working processor proved that the Motorola MC6809 was DOA; looks like I'll need to get a replacement for it from Marco.

Hope you enjoyed the read, and should you decide to tackle building one of these, please post your thoughts and experiences on it.

Follow up:

When I found out that the processor was DOA, I contacted Marco and left a voice mail message as they are closed on weekends. This morning I was contacted back to verify it is the external clock version of this processor that I require and they are sending out a replacement processor for the DOA one. I felt this is worth mentioning as Marco is a company that consistently provides outstanding customer service; unfortunately, this has become a rare thing to find anymore.

Chris, good catch on this. R95 and R99 are in fact 1Mohm resistors. There was a slight error on my parts list that showed the 1Mohm resistors as R95 and R96, but also showed R96 as 1.5K. Special thanks to the j_m_ for sending me a PM about this correction; I have corrected the Google sheet to reflect the proper part information. The best thing about using Goggle sheets for this is that anyone following the document link will get an immediate correction to the parts list

Assembly update and notes:

Final functional assembly picture with replacement MPU. Please note, the MPU, Williams Battery Holder, and ASIC were all sourced from Marco. The Battery holder required a slight modification (with my crimp tool) to make the leads fit into the TwoBits CCA.

IMG_1911 (resized).JPG

The EPROM crowd are going to cringe when they see I install a battery holder on this board

Afterthought: I do have a 50023 PN label (to cover the TwoBits info) and Warranty labels printed and ready to be placed on this board

For those that are really looking close, you will notice that Marco replaced the DOA MC6809E with a MC68B09E, the only difference between the teo is that the 68B09E is capable of running at 2.0 MHz instead of 1.5MHz but since it's an externally clocked processor, it will still run at the clock frequency we provide it with.

This is a valid question and just comes down to the board having a RTC, since the RTC clock circuit must have power in order for the clock to work you really have only one option. For me, having the clock there and not working is just not acceptable. I did socket the memory chip so that if someone obtains this board in the future, then they can easily add NVRAM.

Another valid question, I wanted to source all the parts from one merchant and use only parts that are stocked, this is why I chose Alliance memory for the project. The 256K memory chip has per unit cost of $2.10 where as the 64K version has a unit cost of $3.04; this is likely the same reason Williams used the 256K chip in late generation WPC89 boards.

Please note, some of the resistor quantities are higher than what you need to take advantage of Mousers bulk pricing; for example, this assembly only uses one 470K ohm resistor but ordering a single resistor is $0.21, if you order ten of them they are only $0.02 each making it cheaper to buy ten.

This board is a superior to the Two-Bits board as the voltage rail on top of the board is far enough from the mounting screws that you will not need to use insulating washers to mount the board; if you are building this board, you will want to use the 5mm LEDS.
If building the Two-bits board get the 3mm LEDs.

Essentially, any others will be cosmetic. If you build the twobits board, don't forget the insulating washers or you will short your 5V power to ground.

No they are the same length, this isn't the only difference, there are many minor changes that have no operational impact. For instance, the Twobits board uses square pads for the .1" headers and larger component pads for the .156" headers, rather than the more commonly used external connector pads commonly used during this time frame; if this were a playfield board I would be a little concerned, in the backbox...not so much.

It's my opinion based that the TwoBits board is a re-engineered version of the original, there are numerous subtle differences in the the track layout, component locations, and the silk screen is obviously completely different.

pasted_image (resized).png

pasted_image (resized).png

Yes, the list has been verified, you will want to buy the 3mm LEDS for D19-D21 for the TwoBits board

VERY NICE! It looks almost exactly like mine I do like the attention to detail in orientating all of the resistor color codes in the proper direction, top-bottom and left-right

SIDE NOTE: I'm not sure if you did this or not, but I found it's easier and cleaner to remove all of the header key pins prior to installing them in the board as it's super easy to just pull the pins out of the new header with a pair of pliers.

Very nice, I love the ZIF sockets you have your CPU and ROM in! What are you using for your input power transformer?

Correct.

Neither one is.

3 for 3