There is a good overview video here
https://pinside.com/pinball/forum/topic/youtube-my-vintage-pinball-pinball-restoration-

Talking Roms can be quite controversial as a subject, and you seem quite excited over the prospect...Burners are all different, that is to say they are only as good as the chips they support, as result you are looking at a substantial investment in time money and most likely multiple machines (many on obsolete platforms)...The Q4 does work, and would consider it an excellent entry level machine.

Marco is a pretty reliable place for chips, probably about 90 percent good, and they'll replace them if not.

Most roms are on IPDB.

With a power supply you can check that the game boots, which at least means your roms are valid. I wouldn't trust that to mean your board is good though, but if you put the board in one of your bally/Stern games and it works there you're probably good.

Roms with windows are UV erasable. You'll need a eraser for them. Eeproms are erasable by the burner, which is nice. Often there are eeproms (that start with the number 28) that are compatible with games they take older eeproms (that often start with the number 27). There are also proms, which have no window and are one time use.

Those old Bally/Stern boards use 2k byte and 4k byte ROMs.
2kB capacity are 9316 (PROM) and 2716 EPROM
4kB capacity are 9332 (PROM), 2532 and 2732 (these two EPROMs have minor differences in pinouts)
The 2532 is pin compatible to the 9332 PROM found on Bally -35 boards.
2716 is NOT pin compatible with 9316 and requires board jumper mods for the different pinouts.
2732 is NOT pin compatible with 9332 and requires board jumper mods for the different pinouts.
2732 are generally cheaper/easier to get than 2532.
The black plastic (non-window) PROMs found on those old Bally/Stern boards are either 9316 or 9332 PROMs.

You can use 2532/2732 in place of 2716 by simply doubling the data.

Boards with a ROM at U1 and U2 can be combined into a single 2532/2732 installed at U2 (with board jumper changes).

EEPROMs = Electrically Erasable PROMs
There are electronically erasable versions of 2716 which are 2816 and 28C16.
There are NO electronically erasable versions of 2532/2732. You will have to use EPROMs.

The EPROMs you're programming for those old boards require 21 - 25 volts. USB ports are limited to how much current they can provide when being boosted to those voltages so makes programming some chips unreliable. I don't use a GQ-4X programmer but in demo mode it advises that if write failures occur when programming these old EPROMs to use an external power pack. Their help guide specifications for the power pack is:
9V DC, 200ma or above current
I'd recommend getting a power pack.

China.. They're $1 a piece. You can probably find local sources on ebay.
Jameco are selling ST branded 2732. Guess what you're getting when buying from China.. Another pinball part supplier was selling counterfeit SCRs that had crossed over legs a year or two ago, and there's a thread today of another pinball part supplier potentially selling fake LM323 regulators. Didn't you also get a fake LM323 from a pinball part supplier a year ago?
Other than GPE, I think you see where I'm going here.
The EPROMs required for these old boards are long obsolete. You cannot buy brand new, you'll be getting used parts.

The older EPROMs required a high voltage to program the cells and is partly based on the technology of the silicon die.
2716/2532/2732 require 25 volts for programming
2716A/2532A/2732A require 21 volts for programming
You can get some 27C16B and 27C32B that will program at 12.5 volts but they are quite hard to come by now.
Here are links showing pictures of many types of 2716 and 2732 EPROMs - some are marked with their programming voltage:
http://cpucollection.ca/2716.htm
http://cpucollection.ca/2732.htm

I guy called Leon (RIP) wrote some diagnostic ROMs for a number of boards - they're available for download on PinWiki. There are others out there I believe, but I don't use any of them.

I have some professional programmers (LabTools-48 and BK Precision 866C) but only use them for more specialised chips. It won't surprise you but I just use a cheap Chinese TL866A programmer for this stuff. It's small and portable.
There's some caveats when programming 2716 chips where they need to be programmed multiple times but I'm used to it. The TL866A was made obsolete after too many counterfeits were being made. The latest counterfeit is this one which I bought recently as a cheap backup.

ebay.com link: XP8710 USB Programmer Nand Flash 24 93 25 MCU Bios EPROM AVR IC Chip

The TL886 was replaced with the TL886II+ but maximum programming voltage is 18 volts which isn't suitable for the EPROMs discussed here.

Most programmers you buy these days include the ability to test 7400 series and 4000 series logic chips which can come in handy.

I am using two burners right now. I had a GQ-4X for like 8 years now. Great device, does the 25Vpp chips like 2532, 2732, 2716 that some of the lower end burners will not do.

I bought a TOP3000 to be able to program the DIP42 EEPROM MX29F1615 which the GQ4X does not support. The TOP3000 software is a little more quirky, but it is faster to program and overall has larger range of compatibility GQ4X. Downfall for TOP3000 is it does not program 21Vpp and 25Vpp chips. They are listed in the device list but they won't program. I am using TOP3000 for everything but 2732, 2532, and 2716 which I burn on GQ-4X.

I think the GQ-4X would work well for you needs if you want to focus on the old bally/stern games. Yes it likely needs an external power supply depending on the USB port. I reused an old wall wart DC power supply.

How long does a chip take to erase?.... depends.... 5min to 60min. The M2732A bootlegs commonly available with a tiny die erase very fast. Some of the late 70s and early 80s chips with very large die inside the window can take much longer to erase. Best I can tell is generally erase time mostly based on how big the die is inside the chip.

How many are bad? ... depends again. The M2732A bootleg batches range from 100% all work to 50% are dead on arrival. Pretty much if it burns OK and verifies OK the chip is fine and will run forever. If you get an immediate write error or the chip refuses to erase it is bad. W27C512, M27C040, M27C160 and MX29F1615 are chips I commonly use and those ones have a much high success rate than the DIP24 chips like 2732

Can you erase non windowed chips?.... depends on the device. If the device is OTP (one time programmable) than you get one shot at burning it, no erasing. If it is a UV EPROM you bake it in UV light to erase. If it is an EEPROM the burner will erase it pretty much instantly. AT28C16 is an EEPROM version of 2716. M27C512 is an EEPROM version of 27515.

Where do you get chips? ... All the chips used on old bally games are long obsolete so you are reusing chips pulled off of board or at the mercy of the grey market sellers mostly from China. They are typically less than a dollar each and expect some to be dead on arrival. Most of the time, most of the chips will all work. Occasionally i've had bad lots. Usually they come from China erased, cleaned up, and legs refinished.... so ready to use.

Typically you pick what EPROM to use based on the file size of the ROM AND how the board is jumpered. A smaller memory size can be replaced by a larger memory size when the pin count and pin out matches. So a 2732 can replace a 2716. A 27512 and replace a 2764, 27128, or 27256 because they are all DIP28 ROMs following a standard pin out. You just fill up the larger EPROM with repeated copies of the data so no matter where the CPU is set to look for the program it will find it.

This chart from wikipedia helps with EPROM sizes.
Untitled (resized).png

If you're dealing with ballys I'd definitely get the gq4x. The minipro can't do 2532s, and the new version can't do 25v chips.

Doing this cheap is not really an option, I thought that at one time...But the chips won't work on a bad board, so many years later, you will also have a couple of oscilloscopes, a solder station, bins for storing parts,processors, flip flops, rectifiers, power supplies, frequency counters, UV eraser and bla, bla, bla..I got about 10k into my simple cheap chip blowing project...and that was on the relativity cheap end (IMO) since I calibrate everything myself with a 10MHZ frequency standard.

I used the GQ 4x4 and its a solid programmer, $99 from Amazon https://www.amazon.com/gp/product/B00OD7O9G2/ref=ox_sc_act_title_1

This programmer is very easy to use and a beginner can start using it fairly easily; I've never needed an external power adapter but this is going to depend solely on your PC's mother board.

I prefer to use the windowed EPROMS as I also have a commercial grade UV eraser; however, if I'm going to burn EPROMs that are going out the door, I typically use new burn once EPROMs from Mouser as the person getting these will likely not be able to reprogram them, and they will usually have no reason to ever replace it.

The old version can't do 25v chips either but at least it does 21v which is good enough for most.

I'd ignore even trying to buy 2716's unless you can get the 28 eeprom version which zacaj refers to - 2716's are just a huge PITA to deal with. Use 2732's and double up the 2716 image to burn. Easier to stock one type.

2532's can be dealt with, http://www.therealbobroberts.net/mspacrom.html covers how to build a very simple adapter to burn 2532's as 2732. Works great, I built a couple of them on protoboards with ZIF sockets so I can deal with them (there's no easy way on williams system 7 board to get 3x2732 AFAIK... just 3x2532, for maximum romspace.... I don't like the idea of a bunch of extra jumpers and cuts when I can just burn the proper eprom in the first place....)

I used to use an ancient data i/o programmer but switched to the TL866A people are referencing... unfortunately, it's discontinued. Right when this happened I bought a couple spares just because I like the programmer, a lot. It's nice and cheap and works well enough for what I use it for.

You can stick your refurbished MPU boards into any game and test them out at least for bootup functionality. There is a test rom on pinwiki that will test all the solenoid and lamp circuits if you want to make sure everything is working (until they fix pinwiki the latest version won't be there that also tests the aux lamp driver board.... there isn't one that tests solenoid expander ones, and probably never will be.)

For bally Stern you only need 2716s, 2732s, and/or 2532s. None of the bigger ones.

You tell the software what part number chip you're burning, it knows the voltage level needed

That will be reality.
$125.50 for the GQ 4x4 and an eraser:
ebay.com link: GQ PRG 109 GQ 4X V4 GQ 4X4 Universal EPROM Programmer EPROM UV Eraser

You can get 10pcs of 2716 or 2732 out of China for less than $8
US wise these 27C32Q-200 look like a good deal - these only require 12.5V for programming.
ebay.com link: PROM EPROM EEPROM EEROM UVPROM Modules

Technically speaking you can skip getting 2716 EPROMs and just use 2732 instead (with the data doubled). They're the same price. I'm pedantic and still use 2716 where appropriate.

The 4x4 just means version 4 of the 4x

Load the lower half and upper half of the buffer with the same data then burn.
Alternatively you can double the file before loading in the programmer software by copying it twice - in a Windows Command Prompt type:
COPY /B ROM + ROM ROM2

ROM refers to your source file. ROM2 will become your double sized ROM file.

For the most part, all 2716 have the same pinouts (except for Texas Instrument TMS2716 which has different pinouts - they were first to market at 2kB capacity but required 3 different programming voltages. Intel then came out with a single programming voltage so their solution won out). You won't come across any TMS2716 so don't worry.
All 2732 have the same pinouts, likewise with all 2532.

That's what the board jumpers are for. PROMs versus EPROMs have different pinouts (except in the case of 9332B PROMs where 2532 EPROMs are a direct plugin replacement). Bally switched to 9332B around '81.

Bally-17 and Stern MPU-100 require mods to use 2716 or 2732 EPROMs. Below I documented a simpler method to convert to 2732 that's less convoluted than the long time approach as listed on PinWiki. It also has the benefit of directly supporting *some* 2716 chips (listed on the last post).
https://pinside.com/pinball/forum/topic/converting-a-bally-17-to-35-mpu-with-fo597-service-bulletin#post-5270218

The PPS website lists the jumper settings for different ROMs on Bally boards here:
http://www.planetarypinball.com/mm5/Williams/tech/bally_read1st.txt

For Stern MPU-200 boards refer to the ROM jumper settings on the MPU-200 board schematics.

I use a Needhams EMP-21 for programming chores. Stand alone power supply and USB interface to the PC so I can keep it on a shelf when not in use. I know Needhams went under years ago but you still see them for sale on eBay practically all the time. Same for Craigslist. I just did a quick eBay completed auction search and all EMP-20's went for less than $100. EMP-21's seem to hover in the 200-300 range and less of them.

If I'm reading OP's message correctly you're primary concern is to program EPROMs used in 6800 microprocessor Bally and Stern games. I made a deliberate decision early on (20 years ago) that all my games would be strapped for 2732 EPROMs creating as broad of a universal platform as I could. There are a couple of exceptions, but the vast majority of Bally and Stern boards will happily run 2732's (or as Quench advised can be modified for 2732 use) The Needhams has never let me down and does pretty much any device available during the 80's and 90's with the exception of 3 voltage EPROMS again as Quench mentions in the post above.

As for a UV eraser, I'm a cheapskate and went to the specialty lighting store and asked for a UV germicidal bulb in a F15 T8 configuration. This is most often used in food areas or aquariums. The bulb cost me something like $3-$4 about 20 years ago but like everything else it's gone up quite a bit.

https://www.bulbs.com/product/TUV15T8-G15T8

The upside is you can erase a LOT of EPROMs and use probably the most common small fluorescent tube/ballast out there so finding a fixture to put the bulb into is quite easy. I use a old desk lamp and put the bulb into it when erasing EPROMs. I also put a black plastic pail over the bulb before lighting it. Some say the light is harmful to the eye, others say not. I'm not chancing it. I just set a timer for 5-10 minutes and turn it off. Take the EPROMs you just UV exposed back to the Needhams and verify they have been erased. If not they go back in for another 5-10 minutes. If they still do not erase after a second trip through the eraser I figure the EPROM is bad and move on.

I know there are others much more skilled than I who have contributed to this thread so I'll put out this question. Can you "over" erase a EPROM? Is there any risk of damaging an EPROM by exposing it to too much UV (erasing) radiation?

F15 T8 Germicidal bulb.jpg

It's recommended to place an opaque sticker over the EPROM window after programming to ensure any stray UV light doesn't affect the programmed cells.

The spec on your linked germicidal lamp says "CAUTION: can be very harmful to eyes and skin. DO NOT expose eyes or unprotected skin to these lamps. SERIOUS injury and blindness could result."
My eraser warns against looking at the light.

Apparently yes. Over exposure to UV can render the EPROM un-programmable. How long is over exposure?, not sure.
Is your lamp connected to a timer?
Did you install a reflector behind the lamp?
How far from the chips are you placing the lamp? Specs on all the EPROMs I've seen recommend 1" (2.5cm).

I erase my EPROMs for about 15 minutes.

The typical erasure info I see in EPROM spec sheets say:

The recommended erasure procedure for the M2732A is exposure to shortwave ultraviolet light which has a wavelength of 2537 Å (254nm). The integrated dose (i.e. UV intensity x exposure time) for erasure should be a minimum of 15W-sec/cm2. The erasure time with this dosage is approximately 15 to 20 minutes using an ultraviolet lamp with 12000 µW/cm2 power rating. The M2732A should be placed within 2.5cm of the lamp tubes during erasure.

I have a stand alone 4 chip erasing unit with timer setting, I always use the minimum setting of 5 minutes due to warnings about excessively exposing chips. Worse case one or two chips need to get another five minute pass when I do another batch of four.

The erasure time depends on the light intensity of the UV lamp. I must admit my eraser is about 25 years old.
If yours only does 4 chips, is it one of the cheap Chinese erasers commonly listed on ebay?

UV-Eraser_Grey2.jpg
UV-Eraser_White2 (resized).jpg

No, one I picked up about 20-25 years ago, Dataerase II. Same time I picked up my parallel port Pocket Programmer.

One pictured in middle of this Ebay auction listing.
s-l1600 (resized).jpg

Seems like the over exposure is pretty extreme and not something to worry about for the most part. Sometimes I just set it to the max time length and let it go. I've had one or two stubborn chips end up baking for multiple hours before it finally erased all the way. Then it burned and worked fine.

The little Blue/grey Chinese eraser Quench posted is the one I use. I've had it nearly 10 years now I figure and it is still going on fine. Lamp has never burned out or anything. It erases a M2732A with a small die in about 10 min.

You could get into burning Eproms for well under $100. The minipro is about $50 - although you'd have to find the older one, and the cheapie erasers are about $20. 2732's run about $1-$2 depending on where you get them.

Probably the reason no one has posted a single supplier for eproms is there isn't a reliable, repeatable source for them. (At least not old ones).

WOW after looking a bit at the pages you are completely right. So glad I put the pail over the fixture to avoid direct exposure. Here in the US you see these clam-shell looking fixtures on the wall in just about every kitchen. You can't see the bulb directly but the purple-ish glow washes the wall facing up to the ceiling. Maybe the reflected light does not contain the harmful rays. Nonetheless I'm not taking any chances!

I don't use a directly connected electrical timer. I just set my smartphone and when it goes off I remove power.

Yes the desk lamp has a white reflector basically giving the light 180 degrees. I place the chips on the floor of my garage and direct the unreflected light towards the chips / floor.

I place the chips approximately 1" as you indicate. I read this somewhere when I was building my homebrew eraser.

As barakandl mentioned there have been stubborn ones. I have tried leaving them exposed for 30 to 60 minutes. A few do reprogram without incident, but most seem to fail at reprogramming even if they pass the erase test. For whatever reason I have had the most trouble with Mitsubishi and Intel EPROMs demonstrating this resistance to erasing.

I have found TI TMS2732A are the best at erasing / reprogramming cycles but are painfully slow to program. However since I personally only program EPROMs a few a month, it's a non-issue to me.

Frankly any 2732 I can get my hands on I use. I prefer pulls as other posters have said the Chinese knock-offs are not very good quality wise. I can say the same for 6821's too, but that's a completely different topic.

My eraser has a semicircle chrome metal reflector above the lamp.

The slowness could be an issue with the algorithm of your programmer?? TMS2732A spec is pulse program at 11ms max. Rough calculation is about 40 seconds max to program?

The MiniPro TL866 programmer has an incorrect algorithm for 2716 and 2732 EPROMs with programming pulse way too short. I believe it's the reason 2716 don't program properly first time, rather than the fact the programmer doesn't have 25V support. My LabTools-48 programmer takes nearly 2 minutes to program 2716 chips and has no issue. The MiniPro takes less than 10 seconds and produces an unreliable result.

I haven't received any bad 2732 from China. The last two recent batches of 2716 I bought from China were used pulls with original markings and some label glue still on them with different date codes from the early '80s (i.e. not remarks).

If you decide to use 2716's personally I think you'll quickly figure out that only the earliest titles used xx16 sized memory chips (ROM or EPROM). Your choice but I think if you put the question out to the group the consensus would suggest 2732's. Almost every replacement MPU for Bally made with dedicate U2 and U6 sockets are expecting you to burn the code on 2732's. 2732's are also about the easiest to source making them attractive from that perspective also. As also mentioned in this thread China is manufacturing repros. That in itself says a lot for a IC that is approximately 40 years old.

There is a movement now to expand memory space of the Bally MPU to take advantage of the extra memory the processor chip is able to use. This can be used for additional game features and tests. Again, back at the time these games were made, the cost of the memory chips were a BIG deal! barakandl (nvram.weebly.com) makes a replacement Bally MPU that takes advantage of the full amount of EPROM space the 6800 can utilize)

2716 and 2732 are EPROMs. The pictures of the boards you sent are all masked ROM's. ROM's (9316, 9332) are one-time burnable and in the late 70's and early 80's were significantly cheaper than EPROMs (sometimes called masked or one-time programmable ROMs). Bally would do all their prototyping or early engineering on EPROM's but when it came time for production software they uploaded the code via modem (the old telephone type modem) to have bulk quantities of ROM's made. I was told who and where one time and if I remember I'll edit this post.

Correct. There is no direct drop in replacements of 9316 or 9332 for use in Bally -17 and Stern MPU-100 boards. This is again why we mod these boards to use the 2732 EPROM. Imagine never having to fiddle with jumpers again. Once your collection is all 2732 jumpered or modded, every MPU works in every Bally game. Yes the Stern MPU-200 is a different animal but for this discussion I'm speaking of both Bally MPU's (-17 & -35) or Stern MPU-100 games with a 6800 processor chip.

I'm not certain about the 2532 / 2732 differences but I think for the most part 2532's were Texas Instruments patent / pinout, and 2732 were Intel's. Intel was the Microsoft of the day and more designers chose to use the 2732. Just my shot in the dark. I'm probably wrong and again I'll edit the post if need be.

Bally ROM details

Interesting question

Note China is just coming out of New Nears vacation. There's going to be a huge backlog on purchases and shipping so you might be waiting longer than normal to get your parts.

Many AliExpress vendors source their parts from the same suppliers. You're not (rarely are) buying direct from the source. If it doesn't cost you anything (much) choose a vendor with faster shipping and see if they have any customer reviews maybe with pictures included.
For the past few years, ST branded 2716 were being supplied but it looks like they've just about dried up and are now supplying different manufacturer parts as used pulls and no longer remarking them.

You must modify these boards, there is no direct plugin replacement for the original 9316 PROMs

Three pinouts are different between 2532 and 2732 chips - see pins 18, 20 and 21 in the diagram below.
/CE = Chip Enable (active low)
/OE = Output Enable (active low)
Vpp = Voltage Program pin
Axx = Address line xx
Ox = Data Output line x

EPROMs1.gif

When you download ROMsets from IPDB, look at the file size of each ROM.

2kB capacity are 9316 (PROM) and 2716 EPROM (16 kilo bits = 2 kilo Bytes)
4kB capacity are 9332 (PROM), 2532 and 2732 EPROMs (32 kilo bits = 4 kilo Bytes)

2532 are a drop in replacement for 9332. Bally used 9332 chips from around Flash Gordon onwards. Before this they used 9316 PROMs though on some games before Flash Gordon, 2716 EPROMs was used at U1 and U2.
Stern mostly stayed with 9316 PROMs but did use 2716 EPROMs on some MPU-200 games.

oldschoolbob can you please post a single picture of your Eightball MPU board clearly showing both ROM chips for my reference? Thanks.

You can get a siegecraft adapter that was originally made for williams system 6 games and mod if to use on Bally:
https://sites.google.com/site/allentownpinball/romblaster

Then you can put all bally/stern code on 2764's, or larger.

Just to stir the pot a little.....

I bought a Needhams EMP20 years ago and got tired of farting around with old computers to make that thing work since you need a parallel port to use it. Its a good programmer it just seemed odd to have this thing on the off chance I was going to need to program some oddball ancient eprom. I ended up buying a Batronix BX32 and its worked with everything I've needed so far.
For 2532 eproms I just use the old Bob Roberts 99 cent home made adapter just put the eprom in the adapter and set it for 2732 to program. http://www.therealbobroberts.net/mspacrom.html

I've modified two boards for test. I just used any wire handy. I don't actually run the short wire shown in the mod - instead I just use a lead clipping very close to the leg and cut trace to bridge them. I'll post a picture later. Somebody in this thread sent me a picture of how they did the mod - I'll let them decide if they want to post their picture here

For Bally, the ROM part numbers are listed in their parts catalogs which you can download from PinWiki:
https://pinwiki.com/wiki/index.php?title=Bally/Stern#Parts_Catalogs

Thanks, those Eightball MPU board pictures have the original version ROMs. The Eightball ROM pictures you attached in post #27 were the later version Eightball ROMs that aren't documented (not even in their parts catalogs) as being used together. That's the reason I'd like a picture of that board to keep for reference as evidence they co-existed. I'm a part of a separate documentation project.

2516 is not pin compatible with 9316. I want to be clear about one aspect of the 93xx series PROMs.
The 9316 has three CS (Chip Select) pins. The active logic levels per CS pin can be programmed at factory. If Bally had the foresight they could have ordered 9316 chips that were compatible to the 2716 pinout. Instead they ordered 9316 that were pin compatible with previous gen smaller size PROMs.
The 9332 has two CS (Chip Select) pins. Bally chose to order them to be compatible with 2532 pinouts. Don't quote me on this but from memory, the 9332 PROMs used on Williams boards are pin compatible with 2732 chips.

The attached PDF datasheet contains both the 9316 and 9332 specs and details about why the CS pins were programmable to simplify product design.

93xx_PROMs.pdf

-17 modded Front.jpg

-17 modded Back.jpg

Here is my handiwork on a -17 I modded for a customer last night. I used the through holes wherever possible as anchor points for the jumper wires. As @quench stated the white jumper wire on this board could be eliminated completely by just taking a lead between the trace and pin 5. I will do that going forward. BTW this is by far the least invasive way to mod a -17 board I have ever used and I've been modding them for 20 years. Quench, you hit a home run with this one!

As for the type of wire I have found that Kynar wire works best because the insulation does not melt when hit with the soldering iron like most other insulation does. The only downside is that you need a special tool to strip the insulation off Kynar wire but well worth it in my book.

Kynar Wire.jpg

Kynar Tool.jpg

I just burned (programmed) a set of Mata Hari code on two TI2732A chips. Took exactly 3:30 each. This is on a Needhams EMP-21.

I wanted to program them on AMD 2732DC EPROMs I have, but the Needhams doesn't show the "D" as an available device type. Google seems to think it's the same as the "B" version, but when I select that on the Needhams, it doesn't like it and I get errors. I can select the "A" version and it verifies OK (blank) without error. However I seen to remember a programming voltage difference between the two. Anyone reading this know the differences in suffix A, B, and D? Didn't want to risk programming the "D" version at the wrong voltage so I reverted to the standby TI2732A

https://www.datasheetarchive.com/pdf/download.php?id=e8467a019888cd98787e9bfd7582e8a236d07d&type=O&term=AM2732DC

looks like AM2732DC is a +25v Vpp chip.

Thanks! After digging around a bit more on the datasheetarchive site I came up with the 1980 AMD master selection catalog. It appears the "DC" moniker indicates no suffex in terms of device selection. The "A" and "B" versions came later. My mind was stuck that "D" came after "B" so I thought it was a later version, when in fact it was earlier.

Thankyou Bob, those pics are great!

Here's the issue. Your programmer probably has a generic 2732 algorithm where they used the standard timing spec from Intel 2732 of 45-55ms programming pulses. 2732 has 4096 bytes. 4096 times 50ms programming pulse for each byte equals 3:24 minutes.
TI TMS2732 program pulse spec is 11ms max so it should be taking 45 seconds to program. Ring up Needham tech support and tell them to fix it

BTW since you're the early Bally enthusiast, it doesn't seem widely known that MPU boards for 7 digit games have a different resistor in the display interrupt generator circuit compared to the 6 digit game MPU boards. See here:

https://pinside.com/pinball/forum/topic/two-bits-mpu-with-flickering-strobing-displays-#post-5408121

and here a few posts down for pictures.

https://pinside.com/pinball/forum/topic/two-bits-mpu-with-flickering-strobing-displays-#post-5408417

I think you're probably right. I've used other Mfr's 2732 and when selected on the EMP-21 program much faster. AMD for example program at what my memory recalls to be about 45 seconds.

I've definitely run into that issue in the past and have changed the resistor to clear up the flickering issue. Your research gives a lot more clout to the reasons why.

I've also fiddled with the CPU clock speed. As @barakandl says there seems to be a "sweet spot" for the 6800 as well. I had an 8-ball one time that just played dead. No matter what you changed physically on the playfield, it just seemed anemic. Changed the MPU and it was night and day. Play improved and was more lively. After some research on the two MPU's I found the clock speed on the 6800 was much different between the two. This was like 25 years ago, but I recall changing something to get the clock speed increased to the 2nd MPU's and then the original MPU produced game play that was comparable to the second MPU. If that makes sense

resistor or cap in the clock circuit can adjust the speed. The original caps probably have big tolerance ranges.

Changing these caps to 220pF gives you pretty close to the MPU200 clock rate.

I am learning a lot by reading this thread and I'm sure many others have too - thanks to all that have contributed! Another thread I also read is https://pinside.com/pinball/forum/topic/help-burning-a-stern-rom#post-4592642.

I have some follow-up questions, mainly surrounding terminology in stating size of a chip, size of a ROM image file, and corresponding pinball manufacturer documentation.

When someone refers to an EPROM or PROM only by size, say "a 512K EPROM", does the number represent Kbits or Kbytes? One example would be Stern SB157 that discusses using a "blank 512K EPROM" to update the sound OS on a Whitestar board. I looked at the biosv8.u8 file for that sound update, and Windows shows the file as 128KB (bytes) in size on my hard drive, so it would seem that Stern must have meant a chip of 512Kbytes in size, not 512Kbits. So, a 27C010 chip. But that doesn't make sense as the chip at U7 is physically smaller than a 27C010 chip. U7 (the location where the sound update chip would temporarily be installed) is described in the LOTR manual as "27256 / 27512, CPU Sound, 512K". Both of these chips would be the correct physical size for the socket at U7, so Stern must have been talking about a 512kbit EPROM after all. But neither a 27256 or 27512 chip can hold a 128Kbyte file, so I'm still missing something here. Perhaps the 128KB file size I see for the biosv8.u8 file on my hard drive is not the correct way to identify the actual size of the image in the file?

Here is another example, just in case there was something uniquely odd about the example above. If I look at the directory entry for a ROM file extracted from a zip file on my computer, and that file is shown as 128KB (KB means bytes in this case) on disk, then how does that relate to the size chip needed to burn that file to an EPROM? For example, the CPU file in the LOTR 10.0 or 10.02 zip files both result in an extracted file that is 128KB as stored on disk. The readme file says "Game ROM 1M". The LOTR manual shows U210 as "27C040 CPU Game, 1MB. The PinWiki chart shows the 27C040 as a 4Mbit / 512Kbytes chip. Why would Stern say 1MB in the manual? Would one use a 27C040 chip and just multiply the software file as many times as necessary (quadruple it) to fill the 27C040? Could one instead use a 27C010 and burn the image without multiplying it to fill the 27C010?. What about a 27C020, could that chip be used with a doubled software file? Was Stern thinking to the future in specifying a 27C040 chip, so future games could have significantly bigger code and still run on this board? Or by specifying a 27C040 chip do they for some reason really want that size chip to be used and not a smaller chip? Or was there a typo in the Stern manual and they should have said "27C010 CPU Game, 1MB"?

The Stern manual's documentation for U7 puzzles me for another reason. The PinWiki chart shows 27256 as a 256Kbit chip, and 27512 as a 512Kbit chip, so these are different size chips. Could Stern really have used either chip in U7 even though the diagram also says 512K, or is this really always a 27512 512Kbit chip?

And finally, the Stern manual's documentation for U37, U36, U21 and U17 also puzzles me for a different reason. These are described in the manual as: "M27C04000I-12FI, Voice ROM x, 8MB". PinWiki shows the 27C040 as 4Mbit chip, but Stern lists as 8MB on the diagram. What chip type is really used in these locations, a 27C040 4Mbit chip or a 27C080 8Mbit chip? I guess I could understand if the "8MB" had been a size smaller than the specified chip (could multiply the image to fill the chip), but 8MB is larger than the specified chip size! The actual image files seem to be only 1MB, so perhaps the 8MB shown in the Stern manual is a typo?

Thanks for adding some clarity!

And another quick follow-up question. Some of the labels on EPROMs seem to fall right off with just a touch. Some are stuck on there really well. What do you recommend to safely use to remove the labels that are stuck on quite well? Would GooGone work ok, or would that damage the chip? According to this thread http://www.vcfed.org/forum/archive/index.php/t-57315.html, not much would damage these chips so I'm thinking of using Goo Gone followed up by alcohol to remove any residue.

Playing with my new GQ-4x4, and UV EPROM eraser. Interesting how different EPROMs take different amount of times to erase. I have to say that the most difficult one to erase out of the 20 I have erased so far was a National NM27C040. No idea why, but this one made 3 trips thru the eraser with increasing times before it would test as blank, final trip in the eraser was for 1 hour!

Having problems getting my GQ-4x4 to work on my new pc with windows 10. Getting a message that the burner isn't connecting to the software. Was fine on the old pc. Tried to update driver in case that may have been the problem, but was unable to - not sure if this was the problem anyway. Have also tried the early software that came with the burner & the latest software from the mfg site.

Anyone know if the problem is driver related & what to do, or what else I could try to get it working?

Thanks.

I have an older model which you have to install the driver manually. it does not plug and play properly.

Download and extract
http://www.mcumall.com/downloads/download.asp?DownloadID=56

From memory... plug in usb plug the burner

open up device manager (search for device manager in the win10 search by the start button)

find the unknown device in other devices expanded part. If there is more than one unplug / plug in the burner noticing what disapears/comes back.

Update driver. Choose browse for driver manually. Point at the folder from the download link.

Sounds like there is some bit / btye confusion going on and some of that may be Stern's fault. Documentation may be wrong or the EPROM size got larger in later versions but they did not update documentation.

EPROMs follow a standard pin out so a larger EPROM can usually replace a smaller size in the same pin count package. IE 27512 can replace a 2764, 27128, and 27256

Use the file size of the image to decide what ROM to use. In windows it shows bytes so mutiply by eight. If your file is 128KB in windows explorer would go fit on a 1Mbit 27C010 but you could also use any of the other 32pin JEDEC standard pin out EPROMs as a replacement IE 27C040. Just fill up the EPROM with repeated copies of data until full.

Here is a chart on wikipedia that will help pick an EPROM based on the file size.
https://en.wikipedia.org/wiki/EPROM#EPROM_generations,_sizes_and_types

Great, thanks, will try that.

one day, I will be smart enough to comprehend the (to me) huge amount of knowledge contained in this thread. Today though, is not that day. Thanks for all of this great information!

That is also true of the GQ-4x4 that i just bought. You must install the USB 3.0 driver manually. Go to http://download.mcumall.com to get what you need. I think this YouTube video can also help you:

In case the link stops working, you can search for "How To: Burn, Read, and Erase Pinball EPROMS with GQ-4X". In this video, he takes a GQ-4X4 out of the package, gets it ready for use and burns an EPROM.