6
\$\begingroup\$

I always thought that number before G in the part number is a number of identical gates in the chip. 1G08 is one AND gate, 2G08 is two AND gates in one package. Simple and understandable.

I am going to use LVC74 in my design, and was astonished to find out that 2G74 has one D-type flip flop in it. Fortunately, I did not buy it yet being totally ensured that it must have two DFFs in SSOP-8 package - 2 power pins, one data pin, one clock pin, and one output pin per trigger (no set, no clear, no inverted output).

I see this question on TI forum, and in my opinion answer is not clear, and seems to be wrong. I have compared 1G74 and 2G74 datasheets from Nexperia, and they appear the same with difference in document formatting.

WTF?

\$\endgroup\$
8
  • \$\begingroup\$ ti.com/lit/ug/scbd152b/scbd152b.pdf - page 1-12. \$\endgroup\$
    – Eugene Sh.
    Jul 2, 2019 at 17:45
  • 3
    \$\begingroup\$ .. but it looks like it doesn't help. According to it 2G should be double gate... \$\endgroup\$
    – Eugene Sh.
    Jul 2, 2019 at 17:51
  • \$\begingroup\$ That's pretty weird, looking at those datasheets side by side. All the values appear the same, and there's just minor formatting differences. The 1G version existed for about a year before the 2G version, and it seems it existed like it is now for the last 15 years. I'm pretty sure you'd find the same chip in each of them. \$\endgroup\$
    – W5VO
    Jul 2, 2019 at 17:58
  • 1
    \$\begingroup\$ The short, cynical, and opinion-based answer is "because people are involved". Ultimately, part numbers mean what the manufacturer wants them to, and violations of some "system" are going to happen. \$\endgroup\$
    – TimWescott
    Jul 2, 2019 at 18:01
  • \$\begingroup\$ 1-2 = ? -1 ?... \$\endgroup\$ Jul 2, 2019 at 18:04

3 Answers 3

3
\$\begingroup\$

According to the Nexperia Logic Application Handbook, page 266, they are one and the same:

Why do the 74LVC2G74/1G74 have different part numbers but
same function? Both are single gate but the part number
indicates one has two gates included

Nexperia (then Philips/NXP) created the original device 74LVC1G74 (single D-type
flip-flop). A competitor released a functional equivalent later but named it the
74LVC2G74. The exact reason is unknown: either a simple mistake or an ingenious
method to create an apparent sole-source part number. To clear this confusion,
Nexperia now provides the same silicon under either part number, 74LVC1G74 and
74LVC2G74: one to match the original name and one to match the competition
name. There is no electrical difference between these two devices and they are in
fact the same silicon, package and top marking. We apologize for any confusion this
caused but we didn’t start it!

(At the time of writing TI is the only other manufacturer of the 74LVC[12]G74; I have no information whether TI is also selling identical flip-flops under different part numbers)

\$\endgroup\$
2
\$\begingroup\$

It might have something to do with the lithography differences for identical advanced high-speed low impedance low-voltage CMOS specs. ( e.g. cross-licensing of lithographic files with TI)

Yet in 2016 it was only updated for a name change when NXP (now Nexperia) was being considered with all its patents for purchase by Qualcomm for $47 billion bills.

74LVC1G74 v.1 2004 02 02
74LVC2G74 v.1 2005 11 03

My guess is that although using different processes with same results, there may be some customers ( aerospace ) that were notified of the differences and for reasons not evident in the datasheet only matter to those customers.

Although in 2016 a TI employee explained erroneously that 1G had one output and 2G had two.

\$\endgroup\$
0
\$\begingroup\$

My answer is a conjecture, I do not have any inside knowledge. I have been puzzled by this quite early on.

I believe that originally, the idea was to follow the number of outputs, which for the first models in the series was the same as the number of gates. When the first dual gate devices appeared in the portfolio, they needed 8 pins, and somehow the 8-pin package became associated with the concept of having two gates.

Next, devices appeared that needed 8 pins for a single function, such as the '74 and the '157. Both had two outputs, but this time they were complementary. Somehow product managers decided to keep the 2G code - and there is some "logic" in that. With the '53, however, things became murkier. This is an analog SPDT switch, and depending on which way around you operate it, it can be a multiplexer with one output, or a demultiplexer with two. It still was called the '2G53 according to my 2003 Logic Pocket Data Book from TI.

After that, someone must have realized that it was getting confusing, and when a device like the '1G139 appeared in 2004, it was considered to have one function, despite the 8 pins and the 4 outputs.

Now, of course, they needed to fix the past, since it was hard to explain why there was a '1G139, but a '2G74. They must have resolved to give the '2G74 another name, the '1G74, even though both are the same. I can't see an actual difference between them, except for the name. Maybe it was easier for them to have a component that has two names, than trying to discontinue the '2G74 and have people move over to the '1G74.

But I'd sure like to hear from an insider how that whole confusion actually came about.

\$\endgroup\$

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.