[Amps] Ouch - 23 Ohms ?

[Dr. David Kirkby]

Dan Sawyer wrote:
> Thank you. The devices are Mini-Circuits ZFDC-20-5-23. Normally these 
> devices either have no third numeric indicating 50 Ohms, or a 75 
> indicating 75 Ohms.

That is clearly a special. You might find Minicircuits will tell you - 
perhaps if you phone rather than email. I have sometimes got companies 
to tell me about specials, but on other occasions they will refuse to 
give any information at all.

I have a few minicicuit couplers around here. Here are the measurements 
at DC.

ZFDC-10-2 (10 dB, 50 Ohm)

In = close to 0 Ohms @ DC
Out = Close to 0 Ohms @ DC
Coupled = close to 0 Ohms @ DC

So that bears no relation to the 50 Ohm of the device. Yet the part 
number is not too dissimilar from yours.

ZEDC-15-2B (I assume 15 dB, but look it up if you want)

In = close to 0 Ohms @ DC
Out = Close to 0 Ohms @ DC
Coupled = close to 0 Ohms @ DC

OK, part number not so similar (this is the newer smaller device)

> I am trying to determine both the coupler impedance and the device pass 
> through impedance. I believe the coupler impedance is 23 Ohms as that is 
> what is measured with a DVM. I will try to measure reflection with a 
> variable resistor.

I would tend to forget a DVM. The devices are not likely to be rated for 
use to DC so their characteristics at DC will not necessarily bear any 
simple relationship to that at RF.

In practice, I what I suspect you will want to do with these is use them 
in a 50 Ohm system. I would tend to put

1) 10 dB or more 50 Ohm attenuator on the output put, to give a decent 
return loss there.
2) Power meter or spectrum analyser on the attenuator.
3) Put power in the input port
4) Adjust the load resistance at the output port until you find that 
which gives minimum coupling. That is then where its isolation is best, 
which is what you want to know. Depending on the design, you may be able 
to adjust that.

I don't think you will gain anything useful from connecting all 6 in 
series which is what you are doing.

If I can find a ZFDC-20-whatever  I'll take a photo of one - it is quite 
possible I have one around here.

Take a look inside - do the tracks look very wide compared to other RF 
tracks you have seen? To get to 23 Ohms, the track would need to be 
quite wide - even if on 0.8 mm PCB.

> Dave Bowker wrote: