Mask ROM ..... EPROM equivalent?

Solder

Member
Joined
Apr 29, 2007
Messages
862
Reaction score
7
Location
UK, United Kingdom
What is the EPROM equivalent of a 40-pin '4M-16Bit' mask ROM?

I ask as I'm currently working on a Konami 'Bells & Whistles' PCB - the sprite colors are wrong. All other chips (custom and otherwise) are fine, so the fault must be the mask ROM at 3K or 7K.

Now somebody has already socketed the mask ROM at 3K (all connections are good) ..... and there's no mask ROM for the socket so I programmed up a replacement, namely a 27C4000 EPROM (the datasheet indicates that the internal layout and pinouts match). That passes the built-in mask ROM test just fine, but still the colors are wrong.

As an experiment I socketed the remaining mask ROM at location 7K (this passes the self-test okay too) but on putting it into my programmer and trying to read it as a 27C4000 there are errors.

So am I using the wrong EPROM at location 3K ? Perhaps the Konami self-test results are erroneous? Yet maybe not, as if I use a blank 27C4000 (or one incorrectly programmed) or no ROM/EPROM at all then the self test reports 'BAD' for the relevant EPROM.

Either way, I need to be sure that I'm using the correct EPROM type.

Any ideas please? I also can't understand why I can't read the good mask ROM as a 27C4000 - perhaps indicating that I'm using the wrong EPROM type. Yet if this is the case, why does a suitably programmed 27C4000 pass the Konami mask ROM self test okay?

Thanks
 
Last edited:
how do you know it should be a 27c4000?
if all test is okay but color is still wrong, the problem should be somewhere else. might be a color ram?
btw, you can refer to guru's xref here.
 
From verifying the pinouts, etc I'm pretty sure that a 27C4000 should work ........ but of course it doesn't appear to (maybe!).

Yes, checked the color RAM, etc - however, the color RAM handles ALL the colors, while the only ones affected with this color fault are the sprites - the background tiles are fine.
 
Look for something along the lines of a Fujitsu 74LS273 at the edge of the video circuitry close to the actual R/G/B out. Check it for bad/stuck inputs or outputs. Konami boards of this era are notorious for very high failure-rate TTL logic.
 
Thanks, but reading it as a 4096 or 4002 won't work as the pinout is very different to that of the mask ROM (for example, on a 4002 or 4096 the voltage pin is pin 40, while on the mask ROM it's pin 21).
 
27C400 is another one that will sub out...

If you have a known good EPROM but the board says it's bad then check the signals on the pins of the EPROM to see what is missing or is screwing looking. You might have a bad trace (those Konami boards have VERY thin traces) or a problem with a custom IC.

RJ
 
An update on this:

I've now tried a mask ROM from a working Bells & Whistles (from location K3) and it works perfectly on the board that I've been having problems with, ie the board is 100% definitely NOT at fault.

It's all down to finding a suitable replacement for the mask ROM, yet a 27C4000 won't work.

I haven't tried an MX4100 as:

a) I don't have one
b) The internal layout and pinout exactly match the 27C4000, so they are the same?

Any ideas please on an EPROM equivalent for the 40-pin '4M-16Bit' mask ROM please? The file size of the binary is 512K.

Note: there is provision for a 42-pin chip in that location, so maybe I need a 42-pin EPROM equivalent?

Thanks!
 
Last edited:
Addendum: Reading the original 40-pin mask ROM as a 27C4000 and then programming up a 27C4000 helped a bit - the sprite colours were much improved (compared to using a 27C4000 programmed with the MAME binary which produced totally wrong sprite colours), the only problem now being some horizontal lines in the sprites (about quarter of an inch between them).

So we're getting somewhere I guess ................ not quite there yet though! I guess I should be reading the mask ROM as something else, but what?
 
Last edited:
You could double the ROM .bin and put it on a 42pin M27C800, but I doubt that would help if it won't read properly from the 4Mb chip.

I always read 4Mb mask Rom as AM27C400...you might try programming one as that too and
see if it helps.
 
Thanks for the sugestions, but have tried both and no luck.

This is so damn weird, especially considering that reading from the mask ROM and programming a 27C4000 (it's the same as a 27C400) produces a much better result (albeit with horizontal lines) than using the MAME binaries.

Have also socketed the 100% known good mask ROM at location K7 and repeated the process, but the results are the same.

I can't understand this - why is a 27C4000 not working?
 
Update: Have resolved the problem!

A 27C4000 IS the correct EPROM type to use.

Causes of the problem were as follows:

a) My Galep-4 programmer wasn't correctly reading/writing 27C4000 EPROMs - once I tried my Wellon VP-280 programmer all was well (but due to point b (below) I had to read the original mask ROMs - I could then correctly program a 27C4000 from the new binary with the Wellon).

b) Most importantly, the Bells and Whistles binaries that I was using (and which I downloaded some time ago) were incorrectly dumped/corrupt. I would like to ensure that I am using the latest binaries - have they been re-dumped? If so, where is the best place to download the latest dumps? I would naturally like to verify the binaries that are being made available. Obviously they work in MAME ......... so why not when programmed onto an EPROM and used on a PCB?
 
Back
Top Bottom