Ok, let's say i need a 470 ohm 1/2 watt resistor, but all i have are 1/4 watters. I connect them in series parallel so i can have a 470 ohm 1/2 watt resistor bank. should i connect them in parallel and then series, or series then parallel (parallel then series, each resistor will connect at one point, looking like a figure 8, while series then parallel will look like 2 lines... i dunno how else to describe it)
I know in theory it probably won't matter, but I'd like to consider real world things such as ground loops, hum, etc (i mostly do old tube radios/amps)
Thanks,
-Josh

You can connect 2 1/4 watt 470 ohm resistors
in parallel to get 1 235 ohm 1/2 watt.
Connect the 2 parallel pair in series =
235 + 235 = 470.

yes, but one can argue that you could connect 2 470ohm resistors in series, making a 940 ohm resistor, then connect 2 series pairs in parallel and get 470 ohms. the only difference is with Bob's method, all 4 resistors connect at one point. the question is, is one better (real world things, such as hum if used in audio, etc.)
Also, I'm now thinking, would the max. dissipation of this arrangement be 1 watt (assuming each is 1/4 watt)?
Bob - (apologies if this sounds abusive, I'm not intending to be abusive) That was one of the things I mentioned in my post. I mentioned the other b/c i think it would be easier to differentiate between the 2 methods, but you didnt answer the actual question (probably my fault since i often dont ask the right questions), but thanks for helping :-)
::edited for spelling

[This message has been edited by PEdoubleNIZZLE (edited 07-31-2005).]

Interesting question, I would think that neither method would be prone to introducing any hum, after all the entire resistor bank is going to be small, and only have 2 'rest of the world' connections. For what it's worth, I always have used the parallel first then connect the parallel groups in series method (ending up looking like a figure 8), simply for mechanical rigidity.

There are some high-reliability applications (aerospace and military) in which these arrays are used intentionally, and the decision is based on an analysis of the effects of a failure of any one resistor. For the parallel-then-series arrangement, the total resistance change is less (50% increase) than for the series-then-parallel arrangement (100% increase), but the stress on the resistor paired with the failed one is higher.

For audio work, there's no electrical difference between series-parallel and parallel-series connections. Either way, the power rating will be 1 watt.

In a theoretical case where all resistances are exactly the same, or where R1 = R3 and R2 = R4, then the instantaneous voltage at points X and Y will always be identical.

Thus, by Kirchoff's Law, if you change this from a series-parallel to parallel-series arrangement by adding a link between X and Y, nothing will change. As the voltages at X and Y are identical, no current would flow in that link.

I would add the "link" to minimize slight differences and to improve current sharing.

And John is correct: power capacity adds with either series or parallel resistors.

Believe it or not resistors do add noise to a circuit. I worked with a fellow that designed anti-submarine warefare equipment for the defense department. He was explaining to me how poor quality resistors created noise in the listening cicuits.

If you look deeply into the specs of metal film resistors you'll find noise specs. In the case of a carbon film resistor I understand it to be the way the carbon reacts to the current traveling through it.

For the rest of us we'd never notice it or hear it or probably be able to measure it with the off the shelf equipment available.

There we all learned something new today...

RSlater,
RSmike