need help with txmission line xformer

For starters, the data sheet is confusing... It seems that the manufacturer doesn't know which is the primary and which is the secondary, or which is the start end of the windings - one of them has two start ends????. They seem to be having two-bob-each-way (betting slang for each way even money bet for those who might ask).

Are you using a network analyzer to evaluate every step of your 2GHz evaluation? If so, do you know how to use it? Do you have a high-Z low-capacitance probe to go with it? It's very easy to be fooled when taking measurements above 0.5GHz or so. I have about $20k invested in my old HP's 3GHz measurement capabilities. I've read of cheaper alternatives, but these may well compromise your ability to make the exact measurement you need at some point.

In my experience, transmission-line transformer designs can be relatively foolproof, provided you are in an appropriate "sweet spot." But for ordinary ferrite materials, the sweet spot runs out by 50 to 100MHz. Go read Jerry Sevick on the subject. To go higher you have to dramatically reduce the transmission-line transformer _and_ coax size, and find ferrite materials that will work to high frequencies. And you have to fabricate them properly. I have found that each of these four issues is a serious challenge.

--
 Thanks,
    - Win

I think salesman write them.

We call it boxing a bet, but it's risky!

Have a look at the MiniCircuits TC1-1-13M, the picture of the part is clearer and you can see why the schematic is drawn like that with the 2 dots where they are. The schematic itself also gives a hint.

For the OP, what are you using for a balanced port?

Which pins are you driving?

You will often see in the literature baluns tested as pairs, so the mid point (the junction of the 2 identical devices) becomes properly balanced even with single ended drive and load. Insertion loss per balun is then inferred as half the observed dB loss.

Regards Ian

I'm using as shown here, in fig 1:

The signal gen is driving the primary with a power meter on the secondary. I don't think this is right, shouldn't I have a balanced load or something? How would I connect it to check it's insertion loss? I need to do it, the burn them in for days, then check them again.

As I stated: "using an HP8753D in CW mode as a source and an HP power meter across the outputs as RL."

I'm not probing anywhere in a circuit. I used the network analyzer in CW mode out of port 1. I also tried insertion loss by connecting it between port 1 and port 2 ( like you'ld do with a filter). I see no need for a HI Z probe here. I have used them tshooting receivers.

I didn't design these, have no control, just trying to use them.

Back to my question: How would I connect them to check insertion loss?

The left hand end of fig 1 in the url you give should be driven by the

8753 source. If you connect a second balun to the right hand port with no ground, you can hook up a grounded measuring instrument to the remaining pair of terminals. ______ _______ ______ ---^^^^----¦ ¦----¦ ¦-----¦ ¦ ¦ ¦ Bal1 ¦ ¦ Bal2 ¦ ¦ Det ¦ O ¦ ¦ ¦ ¦ ¦ ¦ ¦___________¦ ¦____¦ ¦_____¦ ¦ Gnd ¦______¦ ¦_______¦ Gnd ¦______¦ ^ ¦ This middle section is not grounded, and provides the balanced port. As the 2 baluns are identical, you assume the losses are equally split between them. (View with fixed width font)

Do you have the 8753 s-parameter set? Just wondering why the power meter.

Regards Ian

I was using the power meter because I didn't wanted to ground on output, I wanted to 'float' the output.

The problem is that the power meter sensor head is not a balanced load at high frequencies - the centre pin has a nice defined 50 ohms to the screen/N type shell, but the screen has a poorly defined extra impedance to ground.

Using the 2-balun trick lets you build a symmetric (middle) sense point, so it is properly balanced.

Regards Ian

Thanks, I hope to re-visit it tommorrow. I give the two balun trick a try. Also, how does this sound: I drive input with a network analyzer port1, load one arm of output with 50ohm, and look at other arm of output with port2 of N.A.?

That would be measuring the device from a 50 ohm source into a 100 ohm "balanced" load. There might be other effects due to cable lengths and how good the NA input match is.

Regards Ian