(Topic ID: 222797)

NEW! Williams System 3-7 In-Game Test ROMs


By pincoder

1 year ago



Topic Stats

  • 236 posts
  • 34 Pinsiders participating
  • Latest reply 28 days ago by pincoder
  • Topic is favorited by 61 Pinsiders

You

Linked Games

Topic Gallery

There have been 30 images uploaded to this topic. (View topic image gallery).

image (resized).jpg
LTSpice sim of BLANKING keep-alive in test ROM (resized).png
02-blanking (resized).png
0C200399-BFEC-4C5C-9381-C10E3CD5EC73 (resized).jpeg
IMG_8222 (resized).JPG
IMG_8221 (resized).JPG
9FCCB059-2325-4DE3-9C6B-3A04E7B2B8D3 (resized).jpeg
215E54F7-9746-4248-B427-DDB3492CA329 (resized).jpeg
7C15D61E-AA9B-4F81-A8B8-2D2704E2C616 (resized).jpeg
SST39SF0x0 (resized).png
CERDIP-PINS (resized).png
512 (resized).PNG
28c512 (resized).PNG
wang (resized).jpg
25616 (resized).jpg
2e3f85951413d05e56ae1fe9aee7d8d99408b7a7-1-1-1-1-1-1-1.jpg

There are 236 posts in this topic. You are on page 3 of 5.
#101 1 year ago

The other thing I need to do is rebuild switch matrix for firepower, which I figured. A bunch of switches out, but no full rows or columns. Chips are ordered from eBay.

#102 1 year ago

I've used a GQ-4X for years and it has never cared about the file extension name. I choose show all files, pick the one I want to use, then program the EPROM. It has never made a difference if it is named .bin, .rom, .L1, .L2, etc.

#103 1 year ago

Nice. Glad you figured it out. Have you gone through all the ROMs yet? It would be great if you could send me some feedback on your experience with the whole process of using them, and your thoughts on the documentation files.. do they make sense etc. Its always good to get a fresh set of eyes on things. Any comments you or anyone else has on using them is certainly appreciated and will help make them easier to use.

#104 1 year ago
Quoted from pincoder:

Nice. Glad you figured it out. Have you gone through all the ROMs yet? It would be great if you could send me some feedback on your experience with the whole process of using them, and your thoughts on the documentation files.. do they make sense etc. Its always good to get a fresh set of eyes on things. Any comments you or anyone else has on using them is certainly appreciated and will help make them easier to use.

Will pm you with deets, I feel like I'm cluttering up your thread too much. TLDR is that these ROMs are awesome though, thanks for putting these together!

#105 1 year ago

Sure, a PM would be great! Glad you find the ROMs useful. I had fun writing them!

#106 1 year ago

Just ran a small update.. edits to documentation.. new version is 2019.04.07.1941

#107 1 year ago

Did anyone happen to order some of these x16 adapter boards and have extras to spare that they'd like to sell? Ideally in Canada...

#108 1 year ago

so i orderd the cs but they sent me a iiplus. any one know about this error. do i need to add external power to it

Captsaure (resized).PNG

Ok so this is a verification error. just states that at the address. The wrong value was verified

#109 1 year ago

Don't know about the error but the II plus won't do over 16 volts so you won't be able to do older eproms like 2732,2716. When they announced the changes to the product that it won't go over 16vdc I ordered a couple extra of the CS's.

If you keep using the eeproms you should be fine but I have 100's of the old eproms that I erase and reuse constantly so it's a non-starter for me if it won't do them. I have an ancient data i/o that I keep for the 25v ones I run across but the software is terrible and a 2732 takes almost 5 minutes to program, and I have to keep a dos based machine with a parallel port to use it. Strictly for emergencies.

#110 1 year ago

So programing them is as simple as selecting ic, opening the file, and hitting program.
Maybe they were fakes, they did come from china. They look decent, atmel, look legit.
Can't seem to find any info on error codes.
Any way i have ordered an atmel and a catalyst 28c256, was going to make up one on a perfboard.

3 weeks later
#111 1 year ago

I'm working on a project pin and these roms are helping out a lot. I think the blanking.txt calls out the wrong pins. Instead of "IC7 pin 6" it should be "IC7 pin 8". Instead of "IC12 Pin12" it should be "IC23 pin 12".

What frequency signal should I be getting at IC7 pin 8 with the blanking rom installed? My wavelength is 0.118ms (~847hz). It is on 50% of the time. When I look at IC23 pin 8 the signal is high for 73% of the wave. So it is only draining Q5 for 26% of the cycle. I have two head scratchers as well...when I put my logic probe on the Q5 trigger (IC23 pinthe blanking signal goes high(TP4 goes high)...also if I replace C32 with a smaller capacitor the blanking signal works (TP4 goes high).

I can't get the blanking signal to go high (TP4) with this rom. I am able to pass the leds, displays, cmos, and ram tests no problem. I have a meter, logic probe, and a logic analyzer to work with and from what I can tell pin 8 on IC7 (the Q5 trigger) is pulsing but IC23 pin 12 (the blanking threshold) stays high around 3.6V. I've replaced Q5, C31, C32, and IC23. Basically all the components that relate to the blanking signal.

#112 1 year ago

Thanks for the detailed information and positive feedback!

Before I can give you any detailed answers I'll need to know which game and system board are you running this on, and which schematics you are looking at.

To get you started in the right direction, the blanking signal should be showing up on PIA1 (PIA at address 2800) pin PA2 and waits for 3 clock cycles before toggling the pin.

#113 1 year ago

I don't want to highjack your topic so I created a new topic with a response to your questions here:

https://pinside.com/pinball/forum/topic/repairing-a-lucky-seven-pin#post-4981848

I appreciate the help.

#114 1 year ago

Just started making a diy selectable rom board, hopefully will finish tomorrow, off to bed for now.
Them 28c16 were never able to be programed on my 866ii plus. Received a 28c256 and worked first go

20190509_011625.jpg
#115 1 year ago
Quoted from Nusilor:

Just started making a diy selectable rom board, hopefully will finish tomorrow, off to bed for now.
Them 28c16 were never able to be programed on my 866ii plus. Received a 28c256 and worked first go
[quoted image]

Sweet! Let us know how it turns out!

#116 1 year ago

will test on the weekend, hopefully i didn't over look anything.

top (resized).jpgbot (resized).jpgsip (resized).jpgs1 (resized).jpg

new (resized).jpg
#117 1 year ago

sorry not that great at photos. not sure how to make them smaller either

#118 1 year ago

just realize you broke it down in to 8k chunks. i just made the pcb from the first page, that does 16k chunks. hopefully i can figure this out, so if i just copy and paste every 16k is that how its done.

#119 1 year ago

Actually the 2K-adapter.bin requires a 64K chip (32 chunks of 2K) and the 4K-adapter.bin requires a 128K chip (32 chunks of 4K).

Since you are using a 32K chip you will only get the first half of the 2K-adapter.bin image loaded onto it, or the first 1/4 of the 4K-adapter.bin image.

Of course, you can selectively pack your own image file using a concatenation of the single .bin files you want to use. Then Just write that file to your chip. If you don't end up filling all of the chunks just fill them with FFs.

Let me know if you need help creating your own file

#120 1 year ago

No i only have a 4 pos dip. 16 combinatons. I have a 28c256 that will get broken in to 16k chunks 256k÷16=16k

Would need a 5 pos dip for 32, but i only have a 4 pin.

Also whats the method to this madness
1000
1100
1110
1111
0111
0011
0001
1010
0101
1011
1101
0100
0010
0110
1001
0000
How do i figure out what selects what chunk

I think i should be fine making the file. Will i be fine putting in about 1-2k in, cant imagine it would select bang on every 16k. What im saying is will it read if there is a blank at the start.

#121 1 year ago

Since you have gotten these programs to work on a system 7 board, is there any reason why they won’t work on a System 9?

#122 1 year ago
Quoted from Nusilor:

No i only have a 4 pos dip. 16 combinatons. I have a 28c256 that will get broken in to 16k chunks 256k÷16=16k
Would need a 5 pos dip for 32, but i only have a 4 pin.
Also whats the method to this madness
1000
1100
1110
1111
0111
0011
0001
1010
0101
1011
1101
0100
0010
0110
1001
0000
How do i figure out what selects what chunk
I think i should be fine making the file. Will i be fine putting in about 1-2k in, cant imagine it would select bang on every 16k. What im saying is will it read if there is a blank at the start.

The 28C256 is a 256k *bit* chip so you have to divide by 8 which gives you 32768 addresses of 8 bits each. Dividing 32K into 16 chunks means each chunk is 2K. You will have room on that chip for 16 images.

The numbers you listed (although not sorted) are binary 0-15 and would correspond to which 2K chunk you want to address in the chip. Depending on how you wire the DIP switches to the chip of course.

The purpose of the board you are making is to trick the MPU into thinking there is a single 2716/2816 chip in the socket you are plugging into. Setting the DIP switches means it will show a different 2K portion of the bigger chip to the MPU board. The MPU board will only ever be able to see one portion/chunk at a time.

So, any of the 2K chunks can contain nothing. But, if you select that chunk and turn on the game, it won't have anything to boot from. Selecting a chunk that contains a valid test ROM image will cause the MPU board to run the code contained in that chunk.

#123 1 year ago
Quoted from uncivil_engineer:

Since you have gotten these programs to work on a system 7 board, is there any reason why they won’t work on a System 9?

They might work as-is, but I doubt it. I am currently in the market for a system 9 game so that I can create a set for system 9. How quickly I'm able to adapt the existing code to system 9 depends on how much they actually changed on the board (mainly addresses, sizes, and quantities of PIAs, RAM, CMOS, etc). I believe they went into alphanumeric displays too, so I'd have to write a driver for that.

Is it okay to assume you have a sys9 game? If so, what do you think of the onboard diagnostics? If they're better than these ROMs there might not be any point to keep expanding this project.. Thoughts anyone?

#124 1 year ago
Quoted from pincoder:

They might work as-is, but I doubt it. I am currently in the market for a system 9 game so that I can create a set for system 9. How quickly I'm able to adapt the existing code to system 9 depends on how much they actually changed on the board (mainly addresses, sizes, and quantities of PIAs, RAM, CMOS, etc). I believe they went into alphanumeric displays too, so I'd have to write a driver for that.
Is it okay to assume you have a sys9 game? If so, what do you think of the onboard diagnostics? If they're better than these ROMs there might not be any point to keep expanding this project.. Thoughts anyone?

Right now I am dealing with a dead system 9 Sorcerer. I think I may have a reset issue on the board. The onboard diagnostics work ok, unless you are like me and have a locked up board.

I don't think there was a whole lot of difference between the system 7 setup and the system 9 boards in terms of the PIA interface (both use 3). I think the biggest difference is in the memory size, CMOS, and obviously the sound section. However, if you get it to work with system 9, it will also work on System 11. You can actually use a System 11 board in a System 9 game in a pinch.

#125 1 year ago

sorry, i figured it was binary. didnt think at the time to add 0s infront of the first 7 numbers, and that threw me off.
0 0
1 1
2 10
3 11
4 100
5 101
6 110
7 111
8 1000
9 1001
10 1010
11 1011
12 1100
13 1101
14 1110
15 1111
ok, so i need 28c512 chip for 32 slots.

#126 1 year ago
Quoted from uncivil_engineer:

Right now I am dealing with a dead system 9 Sorcerer. I think I may have a reset issue on the board. The onboard diagnostics work ok, unless you are like me and have a locked up board.
I don't think there was a whole lot of difference between the system 7 setup and the system 9 boards in terms of the PIA interface (both use 3). I think the biggest difference is in the memory size, CMOS, and obviously the sound section. However, if you get it to work with system 9, it will also work on System 11. You can actually use a System 11 board in a System 9 game in a pinch.

That's good to know thank you. Does an 11 board require any jumper changes to run a 9 game?

#127 1 year ago
Quoted from pincoder:

That's good to know thank you. Does an 11 board require any jumper changes to run a 9 game?

I'll take a look at the schematics for 9 and see about making a set of ROMs.

#128 1 year ago

ill do some reading on address lines. This seems to be the config,
dip 4 a11
dip 3 a12
dip 2 a13
dip 1 we
5v a14

#129 1 year ago
Quoted from Nusilor:

ill do some reading on address lines. This seems to be the config,
dip 4 a11
dip 3 a12
dip 2 a13
dip 1 we
5v a14

Not quite..

The image I've attached is from a system 6 schematic showing the pinout of the IC17 socket. Pin 21 (WE) is tied to ground and pin 20 (CS1) is used to select the chip.

So:

- Pin 20 from the IC17 socket should be connected to the output enable of whatever chip you decide to use.

- You then tie A0 through A10 to the new chip.

- A11 and up on the new chip should each be tied to a DIP switch. A11 to rightmost DIP, A12 to the left of that.. until you have no more address lines to connect on the new chip. If you have more address lines than DIP switches you CAN tie them to ground, but that would be a waste. Get more DIP switches.

- 5V pin on the new chip should be connected to pin 24 on the IC17 socket. Ground to connected to pin 12 on the IC17 socket.

- Data lines are straight forward and connect one for one.

Hope that clears things up
2816-MPU (resized).png

#130 1 year ago

Ok, cheers, got it. There were a few errors in that original board picture. I can make a new one if you want.
Seems like its ready to go, will test tomorrow

#131 1 year ago

sweet. it actually reads in my rom burner.

#132 1 year ago
Quoted from Nusilor:

sweet. it actually reads in my rom burner.

Nice going!

#133 1 year ago

Heres a rough drawing with the fixes, blue just indicates jumping over wires

2e3f85951413d05e56ae1fe9aee7d8d99408b7a7-1-1-1-1-1-1-1.jpg
#134 1 year ago
Quoted from Nusilor:

Heres a rough drawing with the fixes, all the rest was the same

Crap, I made a board based on the original picture. Haven't tried it yet. So you're saying I need to re-wire it based on your corrections?

#135 1 year ago

Yeah, that original one was wrong. I have a confirmed working board based upon my drawing.

dip 1 was tied to WE when i was ment to be a14
5v was tied to a14
and there was no connection between the WE from both chips

#136 1 year ago
Quoted from stangbat:

Crap, I made a board based on the original picture. Haven't tried it yet. So you're saying I need to re-wire it based on your corrections?

I have no idea myself, on either board as I didn't create them and not all of the information is available based on the pictures.

But, if you follow the design outline I gave in my previous message, you ought to be able to make a board with any chip size from 4k to 64k bytes. nusilor can you post any other connection details I missed? Ie the tie up/down resistors and what to do with the extra chip selects of the new chip, if any?

If I get some time down the road I'll write a how to and provide a 64k board sample. I may even build a batch and put them up for sale, preloaded with the adapter image..

Oh and thanks nusilor for providing the details in how you did your board!

#137 1 year ago

pincoder You explained it perfect
stangbat Heres a pinout of both chips from the respective data sheets so you can picture what we are talking about, and come to your own conclusion.
As you can see if you line the pins up from the bottom, and let the top 4 pins hang out on the 256:
WE(lines up with a11) and VCC(+5v(lines up with a13) don't line up, so jumpers need to be run.
Then the address lines 11,12,13,14 need to be put on the dips for selection. weather it is pulled up(+5v(1) or down(GND(0) makes the selection
Im sure that the 512Kbit(64k) chip would be as simple as adding a15(probably) to the pull up resistor network. on another dip switch

25616 (resized).jpg
#138 1 year ago
Quoted from Nusilor:

pincoder You explained it perfect
stangbat Heres a pinout of both chips from the respective data sheets so you can picture what we are talking about, and come to your own conclusion.
As you can see if you line the pins up from the bottom, and let the top 4 pins hang out on the 256:
WE(lines up with a11) and VCC(+5v(lines up with a13) don't line up, so jumpers need to be run.
Then the address lines 11,12,13,14 need to be put on the dips for selection. weather it is pulled up(+5v(1) or down(GND(0) makes the selection
Im sure that the 512Kbit(64k) chip would be as simple as adding a15(probably) to the pull up resistor network. on another dip switch[quoted image]

I'm not certain this is quite right either. Specifically, we, or, and ce should be connected to their respective pins on the new chip.

@nusilor can you verify the board in your chip programmer as a 2716 and use the dips to select each image?

#139 1 year ago
Quoted from pincoder:

Specifically, we, or, and ce should be connected to their respective pins on the new chip

yes, this is correct and is why jumpers need to be run. If you connected WE to the adjacent pin you would connect it to a11(added photo for explanation). WE is wright enable, the williams pcb ties this to 5v. Maybe you could get away with this, I imagine 5v would be enabled(i might test this later). OE is output enable and CE is chip enable

Quoted from pincoder:

@nusilor can you verify the board in your chip programmer as a 2716 and use the dips to select each image?

yes, i went in and added a number in each rom, went through every one of them and it matched the binary number. Didn't miss a beat.
wang (resized).jpg

#140 1 year ago

pinouts for 28c512 and 27c512

28c512 (resized).PNG512 (resized).PNG

#141 1 year ago

Awesome. That's much clearer

You shouldnt have to add numbers to the ROM images though to test this part.. Each image has it's name written into the header.. Should be visible in an ascii viewer.

Thanks for clarifying and posting your work. I'm glad you got it working!

#142 1 year ago
Quoted from Nusilor:

stangbat Heres a pinout of both chips from the respective data sheets so you can picture what we are talking about, and come to your own conclusion.

Thanks, makes sense.

#144 1 year ago

what does q5 have to do with ic 12 pin 12

#145 1 year ago

also how do you make the socket divide in to 4k chunks

#146 1 year ago

ok not sure if i found a fault with the rom. im no expert. my explanation is at the end of my thread. i was thinking duty cycle was to large since the cap was within spec.

#147 1 year ago
Quoted from Nusilor:

also how do you make the socket divide in to 4k chunks

2^12=4096

Since address lines start at A0 (and not 1) you untie A11 from the DIP switches and connect it to a socket that has A11 (ie a socket for a 2732/2532 (4k) chip).

#148 1 year ago
Quoted from denoument:

I'm working on a project pin and these roms are helping out a lot. I think the blanking.txt calls out the wrong pins. Instead of "IC7 pin 6" it should be "IC7 pin 8". Instead of "IC12 Pin12" it should be "IC23 pin 12".

You are correct for system 4 boards. Thanks for pointing it out. There is a small difference in the blanking circuit between system 6 and system 4.

Here is a revised paragraph for 02-blanking.txt:

Check for pulsing signals on (in this order):

* PIA1 Pin4 - If not pulsing then replace PIA1 (oscilloscope users: pulses at 120hz)

* System 6: IC7 Pin6 - If not pulsing then replace IC7

* System 4: IC7 Pin10, Pin9, Pin8 - If not pulsing then replace IC7

* IC23 Pin12 - If not pulsing then replace Q5

Sorry about the confusion. The paragraph has been updated for the next release.

#149 1 year ago
Quoted from pincoder:

2^12=4096
Since address lines start at A0 (and not 1) you untie A11 from the DIP switches and connect it to a socket that has A11 (ie a socket for a 2732/2532 (4k) chip).

So if i put a 4k chip in a system 6. its only going to be using the last 2k since it runs a11 to to 5v. pin 23. Which is the same as fliping a dip 1

#150 1 year ago
Quoted from Nusilor:

So if i put a 4k chip in a system 6. its only going to be using
the last 2k since it runs a11 to to 5v. pin 23. Which is the same
as fliping a dip 1

Correct. But for a 2732 it's pin 21 not 23. Here's the pinouts for 2716 (2K) and up:
CERDIP-PINS (resized).png

Note that Pin numbers change for 2764 and up, since they have more pins..

Promoted items from the Pinside Marketplace
Trade
Machine - For Trade
Dunkirk, MD
From: $ 20.00
Cabinet - Other
Rock Custom Pinball
$ 17.00
There are 236 posts in this topic. You are on page 3 of 5.

Hey there! Got a moment?

Great to see you're enjoying Pinside! Did you know Pinside is able to run thanks to donations from our visitors? Please donate to Pinside, support the site and get anext to your username to show for it! Donate to Pinside