Page 1 of 1

SN74HC139N vs. 74HCT139

Posted: Mon Jul 05, 2010 10:03 pm
by zodiac1968
Hi all. At my local electronics supplier, I managed to find a couple of SN74HC139N 16-pin address decoders. Does anyone know if these would be comparable to the 74HCT139 spoken about in the snes cart-mod guides online?

Posted: Mon Jul 05, 2010 11:30 pm
by Bregalad
If I'm not mistaking, the first 2 letters are the maker's code. It doesn't matter who makes the chip really. I'm not sure about N after the code - but it likely have no importance.

However, I guess it's important you get a LS or HCT chip (not just HC). This is because the input voltage range for high and low levels is different - LS technology using bipolar transistors instead of MOS transistors.
HCT chips are using MOS transistor but inputs are compatible with LS voltage.

Posted: Tue Jul 06, 2010 7:00 am
by tepples
Bregalad wrote:I guess it's important you get a LS or HCT chip (not just HC).
I thought the discrete mappers in NES games were using HC. For example, see 74HC161 and 74HC02 in an UNROM game. What is different about the Super NES?

Posted: Tue Jul 06, 2010 7:26 am
by lidnariq
The N at the end means it's in a through-hole DIP package (as opposed to some kind of surface mount one)

Posted: Tue Jul 06, 2010 12:33 pm
by Bregalad
What is different about the Super NES?
I have no idea, but SNES carts I have just DO have HCT or LS chips, I never saw any with HC so I guess there's a reason to it.

Posted: Tue Jul 06, 2010 5:13 pm
by whicker
Even though the SNES is a 5V CMOS system, the TTL signal level versions were more common and thus were pennies cheaper. Especially LS.

At SNES speeds, the family doesn't matter. So it was a cost and availability issue.

However to give an example, in a recent hobby experiment, I've been interfacing a 3.3V output from a microcontroller out of laziness to a 5 Volt HC serial-to-parallel shift register, clocking at 10 MHz, and let me tell you that it doesn't really work consistently. But at least I know why and it only wastes my own time.

HC is "High speed CMOS"
HCT is "High speed CMOS with TTL levels"
LS is "Low-power Schottky" and TTL levels


TTL input "Low" is 0.0 to 0.8V
TTL input "High" is 2.2 to 5.0V

5V CMOS input "Low" is 0.0 to 1.0 V
5V CMOS input "High" is 3.5 to 5.0 V (worst case)

Posted: Tue Jul 06, 2010 7:33 pm
by zodiac1968
Thanks for the responses everybody, great stuff.

whicker, just to clarify: are you saying that (as far as snes goes) the HC and HCT will both work, its just that HCT is always seen for some reason ie cost/availability? Your example of your hobby experiment went right over my head, was it just to illustrate that there are non-snes cases when the two wouldn't be comparable? thanks again.

Posted: Tue Jul 06, 2010 9:22 pm
by whicker
zodiac1968 wrote:Thanks for the responses everybody, great stuff.

whicker, just to clarify: are you saying that (as far as snes goes) the HC and HCT will both work, its just that HCT is always seen for some reason ie cost/availability? Your example of your hobby experiment went right over my head, was it just to illustrate that there are non-snes cases when the two wouldn't be comparable? thanks again.
yes, there's plenty of noise margin and settling time that any 5 volt 74xxx series will work. and yeah, my contrived example is when things start to work unreliably.

just watch out for cold solder joints, meter long wire lengths acting like antennas, and remembering to actually power the logic chips (VCC, GND).

Posted: Wed Jul 07, 2010 6:15 am
by blargg
and remembering to actually power the logic chips (VCC, GND).
And remember that with CMOS, a chip with either or both of these unconnected might still appear to work, since it gets powered by the inputs through the protection diodes. With CMOS, you must have power and all inputs connected, not floating, or it'll seem to work when you're looking, but fail when you get it all soldered together and done.