XOR gate: Discrete parts or SMT? (ans'd)

Thread Starter

metermannd

Joined Oct 25, 2020
211
If one needed a single XOR gate, what would be the best approach?

SOT-23-5 (or similar) single-gate device
SOIC-14 package, using a single gate
Discrete components (bridge rectifier, transistor, resistors)

All 3 have a disadvantage that I can see. First option is the tiny package (I can do 3-pin SOT-23s, but the 5-pin package taxes my soldering skills). The second option I don't like because of the real estate it takes up. The third option - well, additional parts when options 1 and 2 exist.

Thoughts?
 

Papabravo

Joined Feb 24, 2006
16,832
If one needed a single XOR gate, what would be the best approach?

SOT-23-5 (or similar) single-gate device
SOIC-14 package, using a single gate
Discrete components (bridge rectifier, transistor, resistors)

All 3 have a disadvantage that I can see. First option is the tiny package (I can do 3-pin SOT-23s, but the 5-pin package taxes my soldering skills). The second option I don't like because of the real estate it takes up. The third option - well, additional parts when options 1 and 2 exist.

Thoughts?
What does you discrete XOR consist of?
I'll bet large sums it's more than any single part solution.
 

schmitt trigger

Joined Jul 12, 2010
464
Personally, I like single gates in SOT23 packages a lot. That is what I choose every time.

But if you can’t solder it, I would go for the SOIC 14. It is probably as cheap as the single gate device, and definitely a lot less components than a discrete solution.
 

Thread Starter

metermannd

Joined Oct 25, 2020
211
I figured the SOT-23-5 or similar would be the way to go.
I'll just try for the biggest package I can find and/or use the hand-soldering pad layout.

Here's the discrete XOR / XNOR diagram I found on the 'net:

discrete xnor and xor gate.jpg

What prompted this question was a setback I had with my ongoing project: The (unobtainium) CPU at its core may be dead in the water, and there's no easy replacement (HC11 is closest and that's already just past NRND status).
 

Papabravo

Joined Feb 24, 2006
16,832
GND is missing from diagram (a), and Vee is missing from diagram (b)
It is also not clear what the reference point is for logic signals A and B in either diagram.
The resistor values seem a bit high for an RTL like circuit.
 

AnalogKid

Joined Aug 1, 2013
9,362
Diagram (a) is interesting. A, B, and Y all are referenced to the same GND, which is the same as the return for the positive rail, but that GND is not a part of the circuit. Cute.

(b) is even cuter - there is no explicit power supply connection because none is needed.

ak
 
Last edited:

Thread Starter

metermannd

Joined Oct 25, 2020
211
(grin) I thought those diagrams were a bit weird as well, especially the the fact they call for a bridge rectifier of all things.

I'll stay the course and carry the existing design (single 7486 gate equivalent) over to the next iteration, switching up to the SOIC-14 only if I find I need a second gate. Just wanted to make sure I wasn't passing up a hidden benefit by changing designs.

Thanks for the input, all. :)
 

Bordodynov

Joined May 20, 2015
2,909
And look at the parameters of the 2SD2704K transistor. It has an allowable base-emitter voltage of at least 25 volts.
In the circuit I showed, the value of the one is larger and closer to the power.
 
Top