Hi l'm building a circuit and l need an 8 input OR gate(does anyone know of one ?) this is not available so l'm using 74LS32 - Dual input Quad or gate. the problem is can l connect all the outputs together as one just like that? thanks
If you take 8 small switching PNP transistors and join all the Emitters and collectors together, you can create a 8 input OR gate. the collector cluster can connect to one R to common to provide a low side signal on the output. Each Base of each transistor can be driven with a 10k R from a TTL level signal, this will provide you 8 inputs. The Emitter cluster connects to your 5V+ supply..
The output will be at your Collector cluster..
I made an error with the last post.
Make the same circuit but use NPN instead of PNP's and just do the reverse of what I posted before.
Emitter cluster to common, Collector cluster to a Load R from the
5 Volt supply.. 10k driving each NPN for your 8 inputs. On the Load R, have that go to a NPN stage to invert the output, other wise, you have a Nor gate instead..
Nope. :) But you could use a pair of 74LS32s.
Or any of a zillion other ways of combining gates.
Also search for "open collector".
-- sam | Sci.Electronics.Repair FAQ:
Important: Anything sent to the email address in the message header above is ignored unless my full name AND either lasers or electronics is included in the subject line. Or, you can contact me via the Feedback Form in the FAQs.
Devices that have "open collector" outputs can, generally, be tied together. Normal TTL outputs, however, are active (see ) and should not be tied together in this manner.
You can get the results you are looking for just by using a second
74LS32. Take the four outputs of the first one to two gates (four inputs) on the second. Take the resulting two outputs to one of the remaining OR gates and you're done, modulo the cumulative effects of the gate delays.A "real" 8-to-1 OR-gate could be realized in a small programmable logic device. A GAL could do this easily but programming one could be a challenge if you don't have access to a proper device programmer.
I'd go with the two 74LS32s.
-- Rich Webb Norfolk, VA
No. Connect pairs of dual OR-gate outputs to two dual OR-gates, then those to a single OR gate. What could be simpler?
How about 4 diodes and a resistor? __ -\ \ | |-->|-| -/__/ | __ | -\ \ | | |-->|-| -/__/ | __ | -\ \ | | |-->|-+----- out -/__/ | __ | -\ \ | | |-->|-| -/__/ | | .-. | | 10k | | '-' | | === GND (created by AACircuit v1.28.6 beta 04/19/05 www.tech- chat.de)
yone
sose
How about simply 8 diodes and 1 resistor with no gates? Depending on the load connected to the output, the resistor may not be necessary.
Very true. The OP said he was using 7432's, but he could get by without them...
Some advantages of the diode-only approach may include:
1) It's inexpensive 2) It's easily expanded if another input is desired 3) Small-signal diodes rarely fail, so it's likely to be very reliable 4) Assuming through-hole technology, it can occupy about the same space as one 16-pin DIP 5) If PCB real estate is limited, leaded-diodes can be incorporated into wiring harnesses
Presumably you (only) need LS speed?
The clumsy way to do this is to feed two signals to each of four gates. This consumes a DIP14.
Take the outputs from the four gates and run them, in pairs, to the inputs of two more '32 gates (half of a second DIP14).
Finally, connect the outputs of these two gates to a third gate in that second package.
32's are slow so this approach is *very* slow.Another approach is to replace the quad OR with a quad *NOR*. This uses a DIP14.
Feed these four outputs into a quad *NAND* -- the output of which will be your desired logic function. Should be faster end-to-end at the expense of having to use two different devices.
Another approach is to pass the signals through open collector inverters. Six per DIP14 so you will need 1 and a third packages.
Tie *all* of the outputs together -- and two a pullup resistor. Finally, pass this signal through one of the remaining inverters being sure to add a pullup on its output, as well.
Other variations are possible: e.g., a pair of quad NORs feeding a dual NAND; three triple NORs feeding a triple NAND; etc.
Usually, one checks to see what "spare gates" one has available elsewhere in the circuit in order to select the most opportune implementation.
There are good reasons why Diode Logic & DTL died in the '60s.
-- Politicians should only get paid if the budget is balanced, and there is enough left over to pay them.
That is true Michael, but depending on the exact application, implementing an OR function with diodes can still be the simplest way to get a 'wide' array I tend to use schottky types these days for the speed and lower forward voltage drop. Otherwise, as others have said in the thread either use two common TTL gates, or better yet, just use a CMOS 4078. This device will work happily at TTL supply and drive levels and is an 8 i/p OR function in a single 14 pin package.
Arfa
The disadvantages are slow speeds and higher operating currents. Both can cause problems in some designs. If it is a very slow device, and has no power restraints it is acceptable.
-- Politicians should only get paid if the budget is balanced, and there is enough left over to pay them.
[start with a bundle of 2-in NOR gates]
Just to expand on this a bit... The basic one-stage TTL circuit can do AND/OR/INVERT, so NAND and NOR and INVERT are all one-stage logic functions. AND, OR, and BUFFER(noninverting) gates all have to have two stages, so are all slower.
It's different for CMOS logic.
How about an STM 74HC4078? Mouser has stock. Part #511-M74HC4078.
54/74 4078
I already suggested that a coupla days back ... :-)
Arfa
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.