So, 50 Ohm is not a E12 or E24 resistor value. As a result, if you want a 50 Ohm resistor, you tend to get quite expensive resistors, at least for through-hole.
So what are the options for getting 50 Ohm resistance? Ofcourse, since this is through-hole we are only talking about "low" frequencies, up to perhaps a few hundred megahertz. Options are:
51 Ohm resistors, these can often be found reasonably priced in hobby-user-volume (ie, where the difference between 1% and 5% is negligible)
Two 100 Ohm resistors in parallel.
47 Ohm resistors.
Any other tricks you can use to get cheap, simple and ok performing 50 Ohm resistance? I'm modding a video distribution amplifier to have 50 Ohm output impedance (even though I don't think it will make a difference for 10 MHz clock distribution).
A 49.9 ohm 1% resistor (E96 series) is cheap and probably easier to find than a resistor from the E12 series.
For most applications i would think that 51ohms is fine.
If your building calibration components or measurement equipment you probably want something better and you probably want to look at temperature stability and parasitics etc.
If you still need it cheap and accurate you could buy a batch and test them to see which is closest.
47 (or 51) ohm standard value should work fine for output R for 10 MHz distribution amp. It will have a very low VSWR.
If you really want to get closer try 2 paralleled 100 ohm.
Hiya
I've been using an extron matrix video switcher for a while now fed off a z3801a for clock distribution to a few items of test equipment with no problems.
However my measurements only go to 8 decimal places.
Cheers
A 49.9 ohm 1% resistor (E96 series) is cheap and probably easier to find than a resistor from the E12 series.
This, plus some resistance from connections, will give the nearest typical value. But is it needed?
Reading the datasheets of "50 Ohm" cables reveals that the real impedance is somewhere between 47 and 54 Ohms.
Also, most important is the termination of the line, except if it is for signal amplitude measurements.
If you are talking about through hole resistors then, in general, two 100s in parallel will work better than a single 51 as the parasitic inductance will be lower.
....or...instead of replacing the 75 ohm resistors add a 150 Ohm in parallel to end up with ....50 Ohms
Depending on your requirements, they do exist, and don't cost that much (~0.5 EUR/pc http://ie.farnell.com/vishay/cmf6550r000fkeb/resistor-metal-film-50ohm-1w-1/dp/2419001 ). Or, if you're in a cheap mood, you can go with 2x 25 Ohm in series - it might look better than 2x 100 Ohm in parallel, but with slightly worse inductance.
Sure, those are still not very expensive, but I meant expensive in respect to other resistors. When I get my standard 1/4 W resistors the price is in the 1 cent per resistor range - half a euro per resistor means they are 50 times more expensive.
....or...instead of replacing the 75 ohm resistors add a 150 Ohm in parallel to end up with ....50 Ohms
That would be an option. I prefer to stick to single resistors though because the existing footprints are for a single resitor and are quite tight and I don't think I can cleanly fit two resistors in place.
Why not calculate the voltage reflection coefficient with a 51R in a 50R system and from there work out the VSWR.
The return loss for 47 ohm resistor in a 50 ohm circuit is around -30 dB which is pretty darn good.
Anything better than -15 dB should be good enough unless you are working on a calibration standard.
Coaxial cables and connectors are often 51 or 52 ohms anyway. If you are using BNCs, they are not generally 50 ohms either; Tektronix ended up making their own for their 1 GHz 7104 oscilloscope.