Circuit question

[CaseyB]

I am making a simple optical isolated circuit and there are two ways to lay it out.



Is there any reason to do it the first way instead of the second?

[RayJones]

Definitely circuit #1. Each signal is fully independent.

The voltage across your resistor will vary with the number of inputs supplying volts (of whatever level).

Depending upon the input levels, this may make the circuit susceptible to false outputs with all inputs supplying volts with circuit number 2.

[RayJones]

Even better would be to add reverse biased diodes over each input LED.

That way -ve voltages will be handled nicely without possible damage to the LED's within the optocouplers.
Some optocouplers actually include this diode, especially the quad units (Used them at work ;-) )

[saturation]

X1-4 are floating, its good idea to ground them to insure zero input when unused.  Use a high ohm resistor between each input and ground.

[EEVblog]

Err, the grounds are tied together on both sides of the optocoupler, that's not very isolated!

Dave.

[armandas]

Err, the grounds are tied together on both sides of the optocoupler, that's not very isolated!

Dave.

I thought the same thing, but maybe that's just the diagram. The right hand side does not make much sense either (connecting Vcc to GND through some resistors?)

[KTP]

Well I assumed on the right side he would be pulling the signal from the transistor collector and sending it to some other part of the circuit, so the resistor to Vcc makes sense to me.

[RayJones]

The voltage across the resistor will vary with the number of inputs supplying volts (of whatever level).
Depending upon the input levels, this may make the circuit susceptible to false outputs with all inputs supplying volts with circuit number 2.

still dont quite understand

Simple Thevenin summing of currents, more inputs supplying voltage will cause more current in the resistor, so the voltage across it will rise.
If it rises high enough it may impinge on the LED's forward voltage, and hence the current through the LED reduces, and the output transistor may not turn on enough for a solid low level out.

Err, the grounds are tied together on both sides of the optocoupler, that's not very isolated!

just noticed after you posted this. yea! quite funny

Not a disaster, it all depends whether you have a solid earth either side.

I thought the same thing, but maybe that's just the diagram. The right hand side does not make much sense either (connecting Vcc to GND through some resistors?)

maybe the "squared R" symbol indicating a load or this is some sort of variable/adjustable resistive circuit? (where V+ actually goes to a load?

I assumed for the benefit of doubt the takeoff would be from the collector/resistor junction.

[NiHaoMike]

The second circuit will work fine as long as only one output is active at a time.

Resistors are cheap so there's rarely a need to save there.

[CaseyB]

Ok, I guess I should give a bit more detail.  Firstly, I am definitely a software guy, so please forgive my noobiness!  Secondly, what I am trying to do is build a Nixie Clock and I already have all of the software written, but I am struggling with the circuits.

Here is the complete circuit for a single Nixie module.

The chip guarantees that only one output will be high at a time, can I used that ti simplify the circuit?

And here is a power circuit that I found online.  The Nixies need 160VDC to turn on and this circuit is supposed to take 9-12VDC and boost it to 100-250VDC.  Does this look right?

[KTP]

As the other guy said, resistors are cheap, so why try to save $0.05 there?

The power supply you have drawn might be ok, but you could also make a simple transformer if it is ok for the unit to run from wall current.  For 115VAC like here in the states, you could use a small 120VAC to 9VAC step down transformer and then a 6VAC to 120VAC step down transformer hooked in reverse which after rectifying and filtering would give you an isolated 180VDC or thereabouts.  That might be easier than the circuit you drew...at least it would be easy for me to wire up in about 5 min vs locating the parts for the switcher design.

[tecman]

The power supply is a bit crude, but will work.  The only item in question is the diode, that looks like a 1n4148.  You will need something else for the voltage desired.

Paul

[CaseyB]

I am open to suggestions!  What would be an improved power circuit that operates from ~9VDC?

[mjz0010]

I used a UF4004 diode in a similar setup.