Amardeep-AC9MF
Experienced Member
It was a real and lingering disservice to use the same numeric portion of the part number for an incompatible part. Unless the TMS2716 was the first one on the market. Anyone know which came first?
- VCF South West - June 14 - 16, Davidson-Gundy Alumni Center at University of Texas at Dallas
- VCF West - Aug 2 - 3, Computer History Museum, Mountain View, CA
- VCF Midwest - Sept 7 - 8 2024, Schaumburg, IL
- VCF SoCal - Mid February 2025, Location TBD, Southern CA
- VCF East - April 2025, Infoage Museum, Wall NJ
-
Please review our updated Terms and Rules here
EEPROM availability 2532 & 2732 replacements.
- Thread starter Hugo Holden
- Start date
Chuck(G)
25k Member
No, I recall it was a bit of a competition between Intel and TI. Intel was first (at least in official announcement) with their 2716, but TI didn't yet have the single-supply fab worked out and so announced the TMS2716, compatible with the Intel/TI 2708 (which the Intel 2716 was not) TI eventually came out with the TMS2516 which was Intel-compatible.
The madness doesn't end there. The TMS2532, one would expect would be compatible with the Intel 2732. No such luck--different pinout.
TI didn't straighten out until the 2764 in the 28 pin DIP. Motorola didn't quite get the message and introduced their 64Kb EPROM in the old 24-pin package before toeing the line with the 28 pin 2764.
The madness doesn't end there. The TMS2532, one would expect would be compatible with the Intel 2732. No such luck--different pinout.
TI didn't straighten out until the 2764 in the 28 pin DIP. Motorola didn't quite get the message and introduced their 64Kb EPROM in the old 24-pin package before toeing the line with the 28 pin 2764.
Last edited:
Chuck(G)
25k Member
Which brings up the evergreen question "Is it the job of the standards organization to certify existing practice or to be forward-looking and set practice for things not yet in existence?"
Amardeep-AC9MF
Experienced Member
Indeed an interesting thought experiment. There are altogether too many variables in the development cycle of products that fall under the need for standardization at some point of their existence to arrive at One True Answer (tm).
BTW, are you the same fellow who founded Sydex? If so I have you to thank for an enormous amount of knowledge gained over the years regarding floppy disk storage.
Chuck(G)
25k Member
That's me.
The standards thing has long been a thorny issue. I recall (served as an alternate rep) with Fortran X3J3 (eventually became F90) that the vector extensions working group came up with a syntax that was at odds with IBM's VECTRAN. Both IBM and DEC threatened to leave the committee if we didn't certify the IBM vector scheme.
Moral: Don't blame the standards group--they're doing the best they can.
The standards thing has long been a thorny issue. I recall (served as an alternate rep) with Fortran X3J3 (eventually became F90) that the vector extensions working group came up with a syntax that was at odds with IBM's VECTRAN. Both IBM and DEC threatened to leave the committee if we didn't certify the IBM vector scheme.
Moral: Don't blame the standards group--they're doing the best they can.
Hugo Holden
Veteran Member
When I was browsing a Rockwell data book, to look up the pin connections of the 2332 mask ROM (these were used in the Rockwell RM65 video board) they had created two types, the 2332A and 2332B. One replaced the type with a basic 2732 pin out, the other type with a 2532 pin out otherwise they were the same. It must have been a headache for people making these mask style or OTP style ROMs too.
I would like to get the 2532 and 2732 types in a OTP version with fusible links (a pipe dream no doubt) because I think these styles of ROMs are super reliable and don't get ROM rot. But normally they seemed to have much lower capacities, presumably the die could only be miniaturized to a certain extent.
Back onto the problem of the defective ROMs though, I could not emphasize enough how sinister this problem is.
The reason being that the byte file in the ROM looked normal and verified in the programmer. However, in use on the video board, as a character generator, the ROMs malfunctioned. Presumably because they were being tasked by higher frequency operation and somehow that causes a problem where some data bits seemed on the border between 0 and 1 (but cured with the higher program voltage than the ROM was designed for). At least in the application I could "see" the defects in the video and the data corruption.
But, here is the thing:
If these ROMs were not used in a character generator application, and they say were holding an O/S or data, this fault could cause random malfunctions and intermittent behavior of the computer. It could be a total nightmare to track down, because a ROM reader would keep telling you the ROM file was normal.
Probably after this episode, before I will be 100% convinced a UVeprom is good I will have to program it to be a character generator and check that it works properly when tasked with higher frequency operation than it gets in the GQ-4x.
It reminds me of another episode where I had bought a 3.5" external disk drive with a USB link. (from guess where, ridiculously low price at $15 or something) I was using it to get files onto those disks so I could transfer them to my 5155 computer which I had fitted with a dual 5.25/3.5 drive unit.
The programs I tried occasionally worked , most crashed the computer and it threw up CRC errors. Initially I could not figure out what was wrong.
Mercifully I had transferred some photo images to a disk and I noticed them starting to change.
The longer the disk was in the USB drive "doing things", the more and more the image files were corrupted.(I have a copy of the corrupted images somewhere) There was a fault in the drive that kept randomly altering and partially erasing the files. But it depended on how long the disk was in there. Initially a file freshly loaded to the disk could look ok. This was also corrupting the program files too, and explained why the programs were malfunctioning.
I opened the drive up, it had some sort of IC interface added to mate the USB cable and the type number on the IC had been ground off the top of it.
That was an interesting lesson. So I bought some new old stock drives in their original boxes from the early 2000's era Imation USB external drives and they work perfectly.
I would like to get the 2532 and 2732 types in a OTP version with fusible links (a pipe dream no doubt) because I think these styles of ROMs are super reliable and don't get ROM rot. But normally they seemed to have much lower capacities, presumably the die could only be miniaturized to a certain extent.
Back onto the problem of the defective ROMs though, I could not emphasize enough how sinister this problem is.
The reason being that the byte file in the ROM looked normal and verified in the programmer. However, in use on the video board, as a character generator, the ROMs malfunctioned. Presumably because they were being tasked by higher frequency operation and somehow that causes a problem where some data bits seemed on the border between 0 and 1 (but cured with the higher program voltage than the ROM was designed for). At least in the application I could "see" the defects in the video and the data corruption.
But, here is the thing:
If these ROMs were not used in a character generator application, and they say were holding an O/S or data, this fault could cause random malfunctions and intermittent behavior of the computer. It could be a total nightmare to track down, because a ROM reader would keep telling you the ROM file was normal.
Probably after this episode, before I will be 100% convinced a UVeprom is good I will have to program it to be a character generator and check that it works properly when tasked with higher frequency operation than it gets in the GQ-4x.
It reminds me of another episode where I had bought a 3.5" external disk drive with a USB link. (from guess where, ridiculously low price at $15 or something) I was using it to get files onto those disks so I could transfer them to my 5155 computer which I had fitted with a dual 5.25/3.5 drive unit.
The programs I tried occasionally worked , most crashed the computer and it threw up CRC errors. Initially I could not figure out what was wrong.
Mercifully I had transferred some photo images to a disk and I noticed them starting to change.
The longer the disk was in the USB drive "doing things", the more and more the image files were corrupted.(I have a copy of the corrupted images somewhere) There was a fault in the drive that kept randomly altering and partially erasing the files. But it depended on how long the disk was in there. Initially a file freshly loaded to the disk could look ok. This was also corrupting the program files too, and explained why the programs were malfunctioning.
I opened the drive up, it had some sort of IC interface added to mate the USB cable and the type number on the IC had been ground off the top of it.
That was an interesting lesson. So I bought some new old stock drives in their original boxes from the early 2000's era Imation USB external drives and they work perfectly.
Last edited:
Amardeep-AC9MF
Experienced Member
Data corruption is an insidious thing. I dealt with two major instances in my life that defined how I look skeptically at every storage and transport medium. Literally every single file I handle and store gets a hash sum immediately upon reception or creation and preferable it had one from its origin.