24 May
2000
24 May
'00
12:49 p.m.
> Yes, but ... the 5-volt parts were costing on the
order of $100 per
piece.
> I bought a bunch for a project and still remember
the "sticker-shock,"
> though it wasn't any better a year later when I had to buy 2732's. In
the
> latter case, I sold them for $25 each to a
west-coast surplus vendor who
> turned around and sold them for $80 each.
I paid $11 each for a handful of 2508s in late 1976. A year later 2716s
were 12$. 2732s were much later and more costly, in 1979(late) I paid
a whopping 15$ each.
I find it interesting that after the 2716's were
out for a while,
and the price had dropped to someplace in the $5-$10 range someplace,
The halfgood part 2758 was used on the IMSAI IMP-48 in 1977!
the intel 2508's were still in the high $30's.
This always
amused me because there was no difference at all between the
Intel 2716 and the 2508 other than the 2508 was a half bad
2716. I call that real good sales to take bad parts and get
BZZT! 2708 was three voltage and the 2716 was 5V huge
difference in pin out too.
The problem was many vendors were gouging for them and STILL!
Allison
24 May
24 May
2:14 p.m.
"allisonp" <allisonp(a)world.std.com> wrote:
Hi
I was talking about Intel's 5 volt '08. They may have been
called 2758 or 2508, I don't recall. As far as I know,
Intel never specifically made a 5 volt 1Kx8 EPROM die, only
the 2Kx8. TI was first out with a single rail '08 and was
doing in the 2708 sales from Intel. That is when Intel started
relabling the half bad 2716's( Yes, I know that the pinout are
not compatable between the 2716 and 2708, what does that have
to do with what we were talking about? ). Early on these half
bad parts were cheaper than 2716's but later in their life,
they were list as being more expensive.
Dwight
the intel 2508's were still in the high
$30's. This always
amused me because there was no difference at all between the
Intel 2716 and the 2508 other than the 2508 was a half bad
2716. I call that real good sales to take bad parts and get
BZZT! 2708 was three voltage and the 2716 was 5V huge
difference in pin out too. 
3:42 p.m.
Yeah . . . I remember those, too. Intel made essentially the same parts
with different numbers, apparently for marketing purposes. They had the
8708 and the 2758, both of which were 24-pin 8 k-bit EPROMs, similar pinout,
etc.
Several vendors were competing to see who could foul up the numbering scheme
the worst. I don't know who won.
Dick
----- Original Message -----
From: Dwight Elvey <elvey(a)hal.com>
To: <classiccmp(a)classiccmp.org>
Sent: Wednesday, May 24, 2000 5:14 PM
Subject: Re: Altair parts substitutions
"allisonp" <allisonp(a)world.std.com>
wrote:
Hi
I was talking about Intel's 5 volt '08. They may have been
called 2758 or 2508, I don't recall. As far as I know,
Intel never specifically made a 5 volt 1Kx8 EPROM die, only
the 2Kx8. TI was first out with a single rail '08 and was
doing in the 2708 sales from Intel. That is when Intel started
relabling the half bad 2716's( Yes, I know that the pinout are
not compatable between the 2716 and 2708, what does that have
to do with what we were talking about? ). Early on these half
bad parts were cheaper than 2716's but later in their life,
they were list as being more expensive.
Dwight
the intel 2508's were still in the high
$30's. This always
amused me because there was no difference at all between the
Intel 2716 and the 2508 other than the 2508 was a half bad
2716. I call that real good sales to take bad parts and get
BZZT! 2708 was three voltage and the 2716 was 5V huge
difference in pin out too.
3:39 p.m.
I didn't realize that Intel made any 2508's. My sense is that they called
their 5-volt parts 8708's. They're the parts that were commonly shipped on
their iSBC-8024's if you ordered a monitor PROM with them.
TI was the one famous for the numbering screwup made by assigning an
industry-standard number to parts that didn't meed the industry standard.
They seem to have reversed output enable and chip select on their parts, not
to mention that their 2716's were multi-voltage parts. Their 2516's were
comparable with the Intel 2716's, however, and MOT made them as well as the
2508's, which I think were half-good 2716's. Since MOT didn't develop
either part, I don't know just offhand which one they pursued. It didn't
matter much, since only the really large ROM arrays used the different
select/enable timings to advantage.
I've enbedded a comment or two below.
Dick
----- Original Message -----
From: allisonp <allisonp(a)world.std.com>
To: <classiccmp(a)classiccmp.org>
Sent: Wednesday, May 24, 2000 3:49 PM
Subject: Re: Altair parts substitutions
who
TI's 2716 was a 3 voltage part, too, IIRC. They had a 5Volt-only 2516,
though.
> Yes, but
... the 5-volt parts were costing on the order of $100 per
piece.
> I bought a bunch for a project and still
remember the "sticker-shock,"
> though it wasn't any better a year later when I had to buy 2732's. In
the
>> latter case, I sold them for $25 each to a west-coast surplus vendor > turned
around and sold them for $80 each.
I paid $11 each for a handful of 2508s in late 1976. A year later 2716s
were 12$. 2732s were much later and more costly, in 1979(late) I paid
a whopping 15$ each.
I find it interesting that after the 2716's
were out for a while,
and the price had dropped to someplace in the $5-$10 range someplace,
The halfgood part 2758 was used on the IMSAI IMP-48 in 1977!
the intel 2508's were still in the high
$30's. This always
amused me because there was no difference at all between the
Intel 2716 and the 2508 other than the 2508 was a half bad
2716. I call that real good sales to take bad parts and get
BZZT! 2708 was three voltage and the 2716 was 5V huge
difference in pin out too.
The problem was many vendors were gouging for them and STILL!
Allison
4:01 p.m.
"Richard Erlacher" <richard(a)idcomm.com> wrote:
I didn't realize that Intel made any 2508's.
I don't think they did. That was TI's part number for the 5V 1K*8 EPROM.
Intel's 5V 1K*8 part was the 2758. Other posts have claimed that an Intel
2758 is a half-bad 2716, but my fuzzy recollection from that era was that
the ones I looked at had substantially different die sizes. For half-bad
parts Intel usually added a suffix to the part number to denote which half
was bad, and I didn't see such a designation on any of the 2758s I used.
My sense is that they called their 5-volt parts
8708's.
Nope, that required the same supplies as the 2708, since it was
really the same part.
TI was the one famous for the numbering screwup made
by assigning an
industry-standard number to parts that didn't meed the industry standard.
Because there *wasn't* an industry standard on the 2K*8 EPROM at the time
when TI and Intel both announced their parts. I think Intel's move to
the single 5V supply caught TI completely by surprise.
4:47 p.m.
Eric, I hope I'm not crazy . . . My recollection of the Intel 8748 board,
for which series I recall Intel cooked up the 87xx part numbers, had
silkscreened on it "It's a 5-Volt world" though they didn't use
5-volt-to
whatever they wanted to use for RS232 converters. The memories didn't see
any power supply other than +5. I do believe you're right about the 2508's.
Needless to say, all this stuff is really difficult to sort out 25 years
hence, since it wasn't too clear what was the case even then.
Dick
----- Original Message -----
From: Eric Smith <eric(a)brouhaha.com>
To: <classiccmp(a)classiccmp.org>
Sent: Wednesday, May 24, 2000 7:01 PM
Subject: Re: Altair parts substitutions
"Richard Erlacher"
<richard(a)idcomm.com> wrote:
standard.
I didn't realize that Intel made any
2508's.
I don't think they did. That was TI's part number for the 5V 1K*8 EPROM.
Intel's 5V 1K*8 part was the 2758. Other posts have claimed that an Intel
2758 is a half-bad 2716, but my fuzzy recollection from that era was that
the ones I looked at had substantially different die sizes. For half-bad
parts Intel usually added a suffix to the part number to denote which half
was bad, and I didn't see such a designation on any of the 2758s I used.
My sense is that they called their 5-volt parts
8708's.
Nope, that required the same supplies as the 2708, since it was
really the same part.
> TI was the one famous for the numbering screwup made by assigning an
> industry-standard number to parts that didn't meed the industry
Because there *wasn't* an industry standard on the 2K*8 EPROM at the time
when TI and Intel both announced their parts. I think Intel's move to
the single 5V supply caught TI completely by surprise.
5:08 p.m.
"Richard Erlacher" <richard(a)idcomm.com> wrote:
Eric, I hope I'm not crazy . . . My recollection
of the Intel 8748 board,
for which series I recall Intel cooked up the 87xx part numbers, had
silkscreened on it "It's a 5-Volt world" though they didn't use
5-volt-to
whatever they wanted to use for RS232 converters. The memories didn't see
any power supply other than +5.
True of things like the 8755, which was introduced at about the same time.
But the 8708 was from a few years earlier, and was definitely marketed as
an EPROM for use with the 8080. I don't have any pre-1980 Data Catalogs
handy, but I'm 99.99% certain that the 8708 was just a different designation
for the 2708, just as the 8102 is a 2102, and the 8205 is a 3205.
25 May
25 May
7:10 a.m.
Eric Smith <eric(a)brouhaha.com> wrote:
"Richard Erlacher"
<richard(a)idcomm.com> wrote:
Some times there were slight differences between family, similar
names. There was a 4702 that had a 1 volt different rating on
the negative rail as compared to the 1702.
If one looks in many of the older Intel development systems,
they may find what look like 2716's but were labeled 2617 ( as
I recall ). These were 2716's that had a few bits bad that
were used as system ROM's. If the bad bits didn't interfere
with the data needed, they used them.
Intel did a lot of strange numbering. Not everything they did
was consistent.
Dwight
Eric, I hope I'm not crazy . . . My
recollection of the Intel 8748 board,
for which series I recall Intel cooked up the 87xx part numbers, had
silkscreened on it "It's a 5-Volt world" though they didn't use
5-volt-to
whatever they wanted to use for RS232 converters. The memories didn't see
any power supply other than +5.
True of things like the 8755, which was introduced at about the same time.
But the 8708 was from a few years earlier, and was definitely marketed as
an EPROM for use with the 8080. I don't have any pre-1980 Data Catalogs
handy, but I'm 99.99% certain that the 8708 was just a different designation
for the 2708, just as the 8102 is a 2102, and the 8205 is a 3205.
24 May
24 May
5:03 p.m.
I wrote:
Intel's 5V 1K*8 part was the 2758. Other posts
have claimed that an Intel
2758 is a half-bad 2716, but my fuzzy recollection from that era was that
the ones I looked at had substantially different die sizes. For half-bad
parts Intel usually added a suffix to the part number to denote which half
was bad, and I didn't see such a designation on any of the 2758s I used.
Checking the 1980 Component Data Catalog reveals that those who claimed
that 2758 was a half-good 2716 are almost certainly correct. The electrical
specifications match the standard speed-grade (no suffix) 2716. Pin 19,
which is A10 on a 2716, is called "Ar" on a 2758, and is specified in the
DC and Operating Characteristics table as an input which should be in the
range of -0.1 to 0.8 V (i.e., logic 0). However, footnote 4 of that table
states:
Ar is a reference voltage level which requires an input current
of only 10 uA. The 2758 S1865 is also available which has a
reference voltage level of Vih instead of Vil.
Essentially the same information is restated in the paragraph "Device
Operation".
It's very interesting how carefully they avoided making any statements
that pin 19 is an address line, instead calling it a reference voltage.
6:05 p.m.
Eric Smith <eric(a)brouhaha.com> wrote:
"Richard Erlacher"
<richard(a)idcomm.com> wrote:
Hi
I don't think that die size meant much. I was working for
Intel at the time these came out ( not in memory department )
and they were making die size changes on regular basis.
It seemed like every 3 or 4 months, we'd look into the windows
and see a smaller die for the same labled part. They may
have made a 1Kx8 part but I never saw one. The ones I saw
were always 2716 chips. I think later on, they were not
even separating out the half bad parts, they were just
using good die from the 2716 line because of the extra
cost of handling and the higher yield. I think you'll still
find that the extra address line was to be tied to one level
even though they didn't have the Hi/Lo select that the
earlier ones did.
Dwight
I didn't realize that Intel made any
2508's.
I don't think they did. That was TI's part number for the 5V 1K*8 EPROM.
Intel's 5V 1K*8 part was the 2758. Other posts have claimed that an Intel
2758 is a half-bad 2716, but my fuzzy recollection from that era was that
the ones I looked at had substantially different die sizes.
8819
days inactive
8820
days old
9 comments
4 participants
participants (4)
-
allisonp＠world.std.com -
elvey＠hal.com -
eric＠brouhaha.com -
richard＠idcomm.com