75R to 50R converter

"Joe G (Home)"

** No need to do that.

The source Z of a wide a band TV antenna is all over the place anyhow.

** Just use 50 ohm co-ax for the feeder.

No useful 75 /50 transformation is possible or needed.

...... Phil

I'll give it go.

Joe

"Joe G (Home)" schrieb im Newsbeitrag news:45ff929a$0$15781$ snipped-for-privacy@news.optusnet.com.au...

Hello Joe,

you could do it with a resistive divider. Using the right values will give you the corerct termination resistance from both sides.

R1=43.3, R2=86.6

This R2 is simply twice the value of R1. So just use 3 resistors with 43 Ohm. It's important to have a very low lead inductance. SMT-resistors would be fine.

43 Ohm ____ 75 Ohm side ----|____|-----o-------------- 50 Ohm side | ____ ----|____|--- GND 86 Ohm (or 2*43 Ohm)

Such a passive divider has a gain of 0.44 compared to 0.8 without this Z-match circuit. So you will loose half the signal, but you have the advantage of the cable impedance matching.

Best regards, Helmut

"Helmut Sennewald"

** Like chucking out the baby with the bath water.

When are these demented ham radio FOOLS going to learn that cable matching is something you do at LOAD end.

....... Phil

Then why didn't you suggest a balun? Or do you think it wouldn't make a difference?

"Anthony Fremont"

"Anthony Fremont = Psycho CUNT from Hell "

The most asinine, most ASD FUCKED pile of sub human

YANK SHIT on all usenet !!!!!!!!!!!!

** Nope .....

Satan would be soo jealous of losing your undivided attention.

Sucking his tiny, cold dick is your full time occupation.

** Shame about all the UTTERLY ASININE FUCKWIT assumptions

brain dead, ASD f***ed shitheads who post on SEB make !!!!!!!!

Oh - now I see my tiny oversight .... all those brain dead scumbags

must be your six toed, pointy headed, dribbling, farting, puking relatives.

Yet to come down from the trees.

Still hanging by their tails & eating their own droppings.

Just like YOU !!!!!

You MOTHER FUCKING ASSHOLE

....... Phil

"Helmut Sennewald" schrieb im Newsbeitrag news:eto8hv$2to$00$ snipped-for-privacy@news.t-online.com...

Hello,

Here is a picture from an adapter. Most probably using resistors as shown above. The specified insertion loss of 5.7dB less is exactly the same as calculated from the resistors above.

IL = 20*log10(0.422/0,8)

Best regards, Helmut

You wan't the world to killfile me, but you won't do it yourself? Why not?

Joe should find a broadband balun (unless one is built into the antenna) so that the coax shield is less of an antenna and more of shield. Odd that a log periodic would be 75 Ohms, IIRC they are usually in the vicinity of

200 - 300 Ohms UNBALANCED.

Cheers! A bit of searching and I also found Mini Circuits have these with

5.7dB loss...

At this stage I am tinkering... but will get serious later...

Simple to build....

Regards Joe

Oh... I see you are active on LTSpice... too.

Regards Joe

You need something called a "Minimum Loss Pad" Lots of people make them. From cheap $ to very expensive $$$$. I know HP makes them. But that's the expensive variety. Or you can make one. Ebay, etc.. For TV (US Terrestrial anyway) focus on 50 MHz to 700 MHz. Anything more that that would just be a waste of money. After the conversion to DTV, I believe the highest UHF channel will be

51, with probably very few exceptions if any.

-mpm

"Helmut Sennewald" snipped-for-privacy@t-online.de

= Some Asinine Kraut FUCKWIT !!

** Only an UTTER ASS uses resistors for any RF matching job.

FUCK the HELL OFF !!!!!!!!!!!!!!!!!!

you DEMENTED KRAUT IDIOT !!

...... Phil

Or just place a 25ohms resistor in series with the coax. Then the antenna sees a 75ohms impedance. If the coax is properly terminated at the other end (by the spectrum analyzer) it should be fine.

However, a TV antenna will show a real 75ohms only at one or two small frequency bands, if that.

--
Regards, Joerg

http://www.analogconsultants.com

There's probably not much real need to do it, but a reasonably low- loss way is to use a transformer. MiniCircuits ADT1.5-122T or ADT1.5-17 will do a good job over the whole frequency range you mention. A log-periodic has at least a chance to maintain a reasonably constant impedance over a range of frequencies, but you'd do well to check what the impedance really is, if you have access to a network analyzer, before messing with matching it to your analyzer.

Before you do any of that, though, consider that there is only 0.18dB loss of power incurred by connecting a 50 ohm load to a 75 ohm source, as compared with connecting a 75 ohm load to the 75 ohm source. That's clearly not worth worrying about when the component you add introduces more loss than that, and where the gain of the transducer (antenna) is a function of frequency, and isn't known to even ten times that inaccuracy. Where it could be worth considering is in a system with rapid variations over time, where echos at mismatches cause reflections, and only then in some specific instances. That doesn't seem to match what you've described.

Cheers, Tom

Huh?!??

Impedance Power Loss (dB) Voltage Loss (dB)

75 to 50 -5.72 -7.48 50 to 75 -5.72 -3.96 Round Trip -11.44 -11.44

Note that power losses are the same regardless of which way you're converting. Voltage losses differ depending on the direction of the impedance transformation. As you would expect.

-mpm

Huh indeed. 75 ohm generator, 2Vrms open circuit. Connect a 75 ohm load. Power delivered to the load is 1V at 75 ohms, or 13.33 milliwatts. Do you accept that that's the most power you can get from that source? Connect to a 50 ohm load instead; the load voltage is

2V*50/(75+50) = 0.800 volts. 0.8V at 50 ohms is 12.80 milliwatts. 12.8 milliwatts is 0.96 times 13.333 milliwatts. 10*log10(0.96) is

-0.177dB.

The power loss attributable to such a modest impedance mismatch is generally not enough to worry about.

I really don't know where you got your numbers, but they don't seem to stand up to analysis. Show us the math and maybe we'll understand where they came from.

Cheers, Tom

Oh, I see--your numbers are for a minimum loss pad. But what I posted was that you connect the 50 ohm load directly to the 75 ohm generator, NOT through a pad! Unless you can convince me that the antenna performance suffers in some way because it's terminated in 50 ohms instead of 75 ohms, there simply is no need for the pad. In any event, if you do have some hangup about terminating the antenna into

75 ohms, don't use a pad--use a transformer. The loss will be much lower. Pads are nice for DC-to-daylight performance, but for the modest one decade or so of frequency range the OP wanted, the transformer will be fine.

Cheers, Tom

How do you know the Z of the antenna is 75R? I've just done a web search, and all of the designs that have turned up so far are 50. In this case, a

1:1 balun would be all you need.

Good Luck! Rich

Sorry about that. In any event, I'm glad you deciphered that post because I was definitely not looking forward to typing in those equations using plain text format. :)

I don't know? The OP asked for 1 GHz. That's quite a lot. I think a simple transformer would introduce unacceptable bandwidth limitations. Certainly there would be group delay and response issues.

Also as I'm sure you're aware, most useful "answers" are not displayed on the spectrum analyzer screen itself, but are instead calculated from the displayed results. I'm not sure I would (or could?) trust calculations where I was not controlling for RF Impedances.

Now of course, for un-amplified off-air measurements (if that's what the OP intends to do), the loss of an L-Pad might be too high, in which case, he needs a network analyzer too so he can properly characterize the impedance of his setup. Ditto if he ever adds a preamp.

But if he's just gonna check center-freq and maybe some simple CW deviations, I see your point.

-mpm