Thank you for your work.

Some of these look very handy. Great work. I am going to play with the sound test to explore the noises in each sound ROM. The CMOS clear is useful for a system 7 board running older software and NVRAM as switching between games leaves you no easy way to clear the RAM on the board.

Since there is sixteen 2716 images one could build a 27256 EPROM board with a 4 position dip and pullup resistors to bank switch between each test rom.

..why would that require ram?

Yes but why does reading the dips and doing a jump require subroutines?

Are you planning on writing anything similar for Bally/Stern MPU's?

Nice work, I have some weird problems on a System 6 board. I'll get some more 2716 ROMS and do some testing. My board is modified to run the Firepower Combo ROM. will these tests still work?

Where's a good place to buy ROMs for this?

I have a new rom burner and this looks like a good project to learn more about this type of stuff.

Hobbyroms.com

Did anyone test/try any of this yet? Curious what people are seeing as far as results go.

I'd be interested in something like this. I want to try this out to play with it but having to locate and program 16 ROMs has kept me from doing so. I'm sure I have 16 laying around, but a lot of them have other test programs on them or are for specific games that I use to test. I really don't want to erase them for this and then later have to do it all over again to use them for the same purpose they're being used for now.

Seems like using a single 28c16 is the best option right now

The blank PCBs shouldnt be expensive to have made. Perhaps I can upload the gerber file to a site people can order from or just have some made the next time I order blank PCBs.
top (resized).pngbottom (resized).pngtop_copper (resized).pngbottom_copper (resized).png

The four upper most address bits are attached to a dip switch bank. When the dip switch bank is closed it is a zero, grounded. When dip switch bank is open it is pulled up by 10K sip resistor to make a one. That should split the 256K EPROM into sixteen 16K chunks.

You're not even reading the dips, you're simply using them to present a different chunk of rom to the board, which is reading it as a 2716. Same concept that the Alltek uses to select its games, and lots of old computers from the 80s did when you had different OS chips in them for various "purposes".

Now, if you mean the dips on the mpu board, yes, that would be a bit different, but you are correct in that performing the JMP doesn't require any ram.

That was the original discussion, I think? Most of the roms probably don't need the full 2kb or anything even. Should be possible to read the mpu dip switches, jump to one of the tests, repeat.

27512 is the bggest in a dip28 package. With a five pin dip switch bank would give you 32 selectable ranges of 16K ROM.

I can create a shareable link on the place I normally order PCBs from but the minimum order is 10pcs for $40usd something shipped slow boat from China. Message me for the link. I could upload the gerber files somewhere or use another board house I can upload the files that other people can order from.

Hey - just downloaded your images - I have a FLASH game that isn't booting I fail the first test - leds don't flash, so it's time to start removing ICs. Bummer! All the signals (looking with a scope) look good - the uP is running. Some of the databus lines look questionable so it could be the RAM or the PIA. Neither is in a socket... It would be nice if there was some alternate signal you could toggle besides the LEDs so I could isolate it between the PIA or the databus... Anyway - glad you are doing this - it's been hard to find the previous testroms (Andre/Leon) and their instructions on the web.

looking at Mouser and Digikey - the mainstream EEPROM are now the 1Mb parts. the 28C256 parts are ~$6 while the 39SF010 (1Mb) is ~$1.37. However, the 1Mb part is a 32 pin dip. While $5 isn't much, it might be better to change the layout to support the 32pin footprint to future proof the design? The UV 27x256 parts are, however, all over ebay between $2 and $10 bucks, but since I don't have a UV eraser, I like the EE parts better... just a thought.

You can look for other non UV DIP28 EEPROMs on ebay. 27SF256, 27SF512, 27E512, W27C512, etc.

Good point.- forgot to search for those.

Andre:

https://pinside.com/pinball/forum/topic/wms-sys-3-7-andre-test-rom-usage-video-v2
https://pinside.com/pinball/forum/topic/andre-test-rom-v2-in-action
https://code.google.com/archive/p/lcd-pinball-display/downloads

Leon's stuff is archived at Pinwiki. There is also info on using the ROMs in the respective game sections.

https://www.pinwiki.com/wiki/index.php?title=Leon_Borre%27s_Repair_Articles

thanks - I got the NVRAM pulled off and still not running. I'll head back to the garage (I mean my lab!) and check those other signals. I don't have a programmer here (it's at the office), so I'll try that bus test tomorrow night.

Here's the testing...

on PIA1, on pins 1 - 20, only 1 and 18 are low and the rest are high (steady state). on the other side of the IC, all the pins are toggling - 21-24 are fast toggles (looks correct), 25 clock is ok, 26-33 looks like the databus, 34 high (reset), 35-36 looks like the address bus and 37/38 high and 39/40 low

no, but.... if I press SW 1 (diag) I can get the those pins (and therefor the LED) to turn on or off. The only change when I press the button (not all the time ~20% of the time) and usually they sequence top on, both on, both off... Since that is just toggling the NMI, is there something in the code executing ???

yes VMA looks good at the processor and the address buffers, chipselect logic etc

no difference, however, depressing SW1 doesn't change the LEDS (both on) - PA2 stays high

I'll try test 12 next. The test0 looked as expected. One issue is that the system 4 board has a 6800 processor - not the 6802/6808, so your text file could be updated as the pinout of the 6800 is slightly different. (pin 3/7/34/35/36/37/38 have different functions) the 6800 doesn't have the internal RAM either - are the test roms using that RAM?

Here's a summary of the differences for anyone interested..

From a code standpoint the MC6800, MC6802 and MC6808 mpu’s are interchangeable.
MC6802 is a MC6800 with 128 bytes of RAM and an on-board oscillator.
MC6808 is a MC6800 with an on-board oscillator but no RAM.

From a code point of view, it shouldn't be any different. The internal RAM is mapped to the same address space. You can also jumper the board to use the external ram with a 6802

fun tip about the internal ram... i have found most motorola brand 6808s found in wms sys 6-7 gear will actually function as a 6802. If you have a board on a bench you can check it by flipping the internal ram jumper/pull resistor and a test rom. Occasionally you see one with a ram error or it acts up in game play, but 90%+ a MC6808 stuffed in WMS boards of the time functions as a 6802. Same thing shows up in Bally world too specially with the moto 6808s with bally's custom part number imprinted. Save a 6810 for something else if it is in a socket. AMI 6808s don't have the ram. Haven't really checked other brands.

I have even seen evidence of a previous mc6802 mark rubbed off and 6808 put down on top of it with a date code 3 weeks later.

a lot of the WMS and Bally custom part numbered chips had a similar remarked from the factory look. Obvious abrasion marks where the imprint is.

cool... I was just concerned as the test rom goal is to isolate components. If it was using internal RAM, but actually with the 6800 it was external ram, it would be unexpectedly accessing the external RAM.

@barakandl Thanks for the board info above. I built a board based on your drawings, just on a perf board not a PCB. It all tests OK with a continuity etc. I burnt the combined image onto a 27C256 and installed the chip. I then tested the board in my Rom burner, reading every option for the 4 dips back into the Rom burner. All work exactly as expected, except for when all 4 select dips are set to ON, this should select image 0 - Bus, but it selects image 4 - RAM1. It seems like dip3 is not turning ON, but it is, otherwise I couldn't select 11 lamps, 10 timer etc. Any idea what is going wrong?

you can check the suspect address pin at the eprom socket. When the dip is closed it should be low and when the dip switch is open the resistor pulls it up to 5v.

I checked all 4 lines that are connected to the dips, I get 4.98v when the dip is open and 0v when the dip is closed. I also checked the image in the 27C256 and it looks to me that ROM 0 is the first one in the list. I did this in case ROM 4 was accidentally written to the 27C256.

Is there any chance that when all 4 addresses are low it does not select the first chunk, as in should it be some other combination of address pins?

Unfortunately I can't report back on very much yet. I have 2 System 6 boards that will not pass the LED test, so I am stuck at the BUS (ROM 0) test. given my issue above with selecting ROM 0, I burnt a 2716 with ROM 0 for testing.

While I'm here, I should check some things:
I understand that these test ROMs are designed to be used with the boards in the game. Can I use the ROMS with the boards on the test bench instead?
Do the ROMs, including the BUS ROM, only work correctly if the driver board is connected?

On 1 of my boards Pin34 on the CPU is high but J1 is installed and R4 is removed and a 6808 is installed - so I need to figure out what is going wrong there.

On both System 6 boards I have some address lines that are not as per your expected conditions. I'll look to trace those lines through the board to see where the problem lies.

pincoder Thanks, I'll burn a new ROM in a couple of days and I'll let you know how it goes.

pincoder

I cannot find the gerber files on your site. Can you post a direct link?