Seeking wisdom on long slotted line construction

[mazurov]

Hello all,

Inspired by http://users.monash.edu.au/~ralphk/slotted-line.html I'm about to build a 1.5 m slotted line to do rx-only antenna measurements at about 150 MHz.  So far it seems that one can easily be built from (US) standard copper water pipes by placing them coaxially, diameters here -> http://en.wikipedia.org/wiki/Copper_tubing . If I pick 3/8" pipe for the inner conductor and type L wall thickness then pipe diameters for the outer conductor will be 1" and 1.25".  I calculated the impedance of the resulting line (air-filled) to be 43.04 ohm for the former and 55.65 ohm for the latter. The question(s)  I have to the knowledgeable folks are:

1. Is this impedance too far off for measurement purposes? If no, which one of the two is better? The other question only makes sense if the answer is yes.
2. What is the good impedance range to aim for? I can play with other available diameters mixing metric and imperial stock to get close to 50 ohm if necessary. This, however, will make the apparatus more expensive.

I can make parts accurately - to ten thou with no difficulty and to one thou with so if there are other ways to arrive to fifty ohms like milling a slot with a specific width or offsetting the inner conductor I can do this too.

Thank you for your time!

Regards,
Oleg.

[rs20]

I'm not an expert at all, but option number 3 could be to have a simple lossy resistive impedance matcher in there. Since the impedances are so close to begin with, the losses would be small.

EDIT: Sorry, disregard probably, I misread and thought you were making an antenna.

[XFDDesign]

Your interest is going to be on the Reflection Coefficient. What is 43 ohm? Gamma = (Zload - Zsource)/(Zload + Zsource) so

(43-50)/(43+50) = -0.075

What does this mean for your VSWR?

VSWR = (1+|Gamma|)/(1-|Gamma|)

So 0.86:1, which is pretty damned good. You don't start to really worry until your VSWR gets to more than 1.2:1. This is RF you're playing at, experiment and find out! Don't expect the guys at 220GHz are getting perfect 50Ohm matches everywhere. It's all about managing what you can!

[mazurov]

This is what I'm really trying to understand - whether any particular Z0 is important here as long as it is known before measurement. To me, making a line and then measuring its Z0 is much easier than making a 50 ohm line or 75 ohm line.

[VK5RC]

Either calculated impedance value looks close enough to me,  I suppose if you are getting really fanatical what you build will never exactly match the calculation,  the slot and the probe would have some effect on impedance. 
Also does the slot need to be so long 1.5m, IIRC you only need a bit over 1/4 wavelength ie 50cm.
By the way,  I hope you have seen that great 1960s youtube video on impedance matching!
 

[mazurov]

You need to be able to measure distances half wave apart plus you'll need some slack. General Radio 874-LBA manual describes a trick to see slightly more than half wave.  Stephen F. Adam's "Microwave Theory and Applications" book contains a number of example measurements using a slotted line. Both texts are available online for free from multiple sources. I like old electronics (and chemistry) books - the available tools were crude and what they were capable to measure with these tools is quite amazing.

From looking at the pictures of slotted lines I can see that the slot takes about middle 1/3 of the overall length of the transmission line, the other two one thirds being presumably left uncut for the current to flow. There is likely some magic ratio and I'm going to experiment with it too. The current standing waves must also be present in the line and the discontinuity presented to them by the slot must actually be beneficial to the experimenters, otherwise the designers would have simply close the slot in the vicinity of the probe.