8
\$\begingroup\$

This is a tangential thought from this answer, specifically the line "There are 10 and 12 pin DIPs".

After some searching around, I finally located only one 10-pin DIP part: NTE1450, 5 Watt audio power amplifier. The part is evidently obsolete, and the one datasheet I found, I cannot locate again.

I found many other components in 10-pin DIP though: LED displays, DIP switches, resistor packs and of course DIP sockets.

There are many 8 pin PDIPs 14 pin PDIPs flood the logic gate listings, but 10 and 12 are apparently pariahs. I haven't actually searched for a 12-pin DIP IC yet, but I suspect that too will be a rarity.

Why are 10-pin DIPs so uncommon?

\$\endgroup\$
4
  • 17
    \$\begingroup\$ The half-life is short enough so that most have decayed into two 4-pin packages and a puff of smoke. \$\endgroup\$ Commented Jan 14, 2013 at 19:01
  • \$\begingroup\$ Nice question..! :D There appears to be a strange 12-lead DIP thing out there. \$\endgroup\$
    – user17592
    Commented Jan 14, 2013 at 21:05
  • 1
    \$\begingroup\$ Even stranger there are no 13 pin ones. ;) \$\endgroup\$
    – kenny
    Commented Jan 14, 2013 at 23:48
  • 1
    \$\begingroup\$ When viewing this title on the questions page it shows as "Why are 1-pin DIP ic's so uncommon"... \$\endgroup\$
    – Grant
    Commented Jan 19, 2013 at 20:06

4 Answers 4

3
\$\begingroup\$

While today's computer-assisted manufacturing would make it relatively inexpensive to create and manufacture new styles of lead frames, tooling used to be much more complicated and expensive, and there were relatively few discrete sizes of chips. As chip production increased, manufacturers may have had to produce new tools simply to keep up with demand (if one has a four punches to produce DIP14 lead frames, one can only produce four such lead frames at a time), but they would have had more inclination to add tools for sizes that were slightly larger than existing ones (so they could offer chips that could do things previously not possible) than for sizes that were smaller. The only real use for a 10-12 pin DIP would have been for an application where an 8-pin chip would have been insufficient, but a 14-pin part would have been too big. Even before the invention of surface-mount technology, there weren't many such applications, and once surface-mount technology came on scene there really weren't any: someone who needed a chip with 9-12 useful connections but couldn't afford the size of a 14-pin DIP wouldn't use a 10-12 pin DIP--they'd use a surface-mount part instead.

\$\endgroup\$
1
\$\begingroup\$

You don't see very many 10-pin DIPs for historical cost reasons. Extra pins are expensive. Especially back when these were extremely common.

Back in the 90's, DIP packages were mostly for digital logic (TTL back then). I don't recall a lot of analog parts in DIP packages. Transistors would come in 3-leaded packages. Anyway, most logic gates (OR, AND, NAND, etc) are going to have two input and one output pin. Plus you need to add a power and ground pin. So, if you have two gates in a package, then it ends up being 8 pins. If you have 4 gates in a package, you end up with 14 pins.

Any other combination give you extra pins, and extra pins are expensive, especially for large volume commodity parts like they were back then. That was the reason back then, anyway. The cost of the extra pins isn't as big of a deal now as it was back then, but now these parts have been in catalogs for decades, so it doesn't make sense to change.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Ref. I don't recall a lot of analog parts in DIP packages ... Op amps, diff amps, in-amps, VGAs, VCOs, filters, I could go on all day. :-) \$\endgroup\$ Commented Jan 24, 2013 at 2:21
  • \$\begingroup\$ Hah...that's what being a digital guy gets you. You never turn to the analog section of the catalog. :-) \$\endgroup\$
    – SDGator
    Commented Jan 24, 2013 at 15:09
0
\$\begingroup\$

You need power and ground. That leaves 8 pins. Two standard two-input gates requires 6 pins. That leaves 2 pins extra. On the other hand an 8 pin part fits that perfectly. Similarly 12-pins would be 3 gates with 1 pin extra.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ If one does not limit oneself to logic gates, which I didn't... What about the rest of the universe of ICs? Audio, general analog, and everything else? Those don't necessarily fit into this profiling. \$\endgroup\$ Commented Jan 14, 2013 at 18:37
  • 2
    \$\begingroup\$ 10 pin DFN packages don't seem to be so rare. \$\endgroup\$
    – Phil Frost
    Commented Jan 14, 2013 at 18:55
  • 1
    \$\begingroup\$ @PhilFrost Without doubt... but in SMT form factors, there are enough examples of 10 and 12 contact parts, and even odd-numbered contacts! The whole question is around DIP form factors. \$\endgroup\$ Commented Jan 15, 2013 at 6:16
  • \$\begingroup\$ @AnindoGhosh: right. I'm saying, if there's something intrinsically underisable about 10 pin packages, then we shouldn't see them in any format, yet it seems only DIP that doesn't have 10 pin packages. Must be something special about DIP. \$\endgroup\$
    – Phil Frost
    Commented Jan 15, 2013 at 12:12
  • \$\begingroup\$ @PhilFrost My point exactly :-) \$\endgroup\$ Commented Jan 15, 2013 at 12:33
0
\$\begingroup\$

You need power and ground, that leaves 8 or 10 'used' pins. This is just a wild guess, but I'd say packages with a 'nice' number of 'used' pins are most common. A number would be 'nice' when it has lots of divisors. These packages would be more common because there are multiple things of the same thing stored in one package, most of the time.

Look at the MAX232 device, which randomly pops into my head. It is a 16-leader, of which 12 'used', since there are two power pins and VS+ and VS-. All pins are in pairs: C1+ and C1-, C2+ and C2-, R1IN and R1OUT, R2IN and R2OUT, T1IN and T1OUT, T2IN and T2OUT. Additionally, these pairs are in pairs: C1 and C2, R1 and T1, R2 and T2. Since we have pairs of pairs, the number of 'used' pins is dividable by 4.

14-leaders are quite common. That gives 12 used pins, which can be divided by 1, 2, 3, 4, 6 and 12. 12-leaders with 10 used pins give only 1, 2, 5 and 10. 10-leaders would be more used, since 8 used pins is dividable by 1, 2, 4 and 8 - powers of 2.

Disclaimer: it's just a wild guess.

\$\endgroup\$
5
  • \$\begingroup\$ So... by your rationale, 10-pin DIPs should be popular. :-) \$\endgroup\$ Commented Jan 15, 2013 at 4:02
  • \$\begingroup\$ Not really, 14-pins should be more popular. But Phil stated in his comment on Brian's answer that 10 pin DFN isn't so rare. \$\endgroup\$
    – user17592
    Commented Jan 15, 2013 at 5:54
  • \$\begingroup\$ Following this reasoning, it appears to be that 3 is a useful number for the amount of pins: 14-pin leave 12 used pins, which is dividable by 3. 8-pin leaves 6, also dividable by 3. That explains why 10-pin isn't that common, although 8 has a lot of divisors. \$\endgroup\$
    – user17592
    Commented Jan 15, 2013 at 6:01
  • \$\begingroup\$ DFN isn't the issue, DIP is. Sure, I found enough 10-pin packages in various SMT form factors. \$\endgroup\$ Commented Jan 15, 2013 at 6:14
  • \$\begingroup\$ Okay, so that supports my guess, right? \$\endgroup\$
    – user17592
    Commented Jan 15, 2013 at 6:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.