hi all,

a few questions on impedances...

1. i know load/source terminations is necessary to prevent signal distortions at either end due to reflections for high speed signals... the best way will be to have both source and load terminations for impedance matching. How about DC signals? is impedance matching necessary?

2. How do we easily measure our load impedance or source impedance? by their typical operating voltage and their typical current drawn? what if we do not have these info, what is a quick way?

3. for a load, is the input impedance refering to the load impedance? similarly, output impedance = source impedance?

4. assume a scenario where we have 2 ICs (IC_1 and IC_2). Pin 1 of IC_1 is output and connected to pin 2 of IC_2 (input) via a discrete circuitry with parts like resistors and caps... therefore, its like we have IC_1 connected to a discrete circuitry (of resistors, caps) and in turn this discrete circuitry connected to IC_2.

Do we match IC_1_pin1 impedance to IC_2_pin2 impedance straight?
or
We match IC_1_pin1 to discrete circuitry input and then match discrete circuitry output to IC_2_pin2?
or
The discrete circuitry can actually act as a impedance matching network for IC_1 and IC_2?

Thanks for your patience!!!!

 

A VNA(Vector Network Analyzer) is the tool for making impedance measurements. Last time I looked their prices were headed toward the $100,000.00 range. You got that much in your budget?

 

A VNA can be rented for only a few thousand $ per month.

 

The other (cheaper?) option is to model the circuit to calculate the values. Depends if you have access to suitable software.

Dave

 

Hmmm, last I checked you could get SWR meters pretty cheaply. Might not be applicable, but then again...

Whenever you are dealing with high speed signals you are dealing with a form of RF (even if it is digital). All the same theories apply. If you are headed for the microwave region then there are gadgets called circulators that do an excellent job of matching impedance, by rerouting the reflection and then absorbing it. There are lower frequency versions using transformers, but they are a lot more complicated and less reliable.

Might help a little to give a clue as to what kind of signals you are trying to match, the impedance, and the upper frequency limits.

Matching impedance at DC isn't usually an issue, but if it doesn't match at DC then what are the odd it matches higher up? DC is easy. The problem is primarily reflections, they really screw up anything digital, and don't do analog signals much good either.

 

How about DC signals?

Are you aware of the maximum power transfer theorem?

To paraphrase what Bill said, where are you coming from? Your subject matter is too wide.

On the subject of measurements see this recent question.

http://forum.allaboutcircuits.com/showthread.php?t=13723

Also time domain reflectrometry meters (AKA cable length meters) are now available quite cheaply. My recent Farnell catalogue starts listing at under £100.

 

I've used TDR's. Neat gadgets. They graph the impedance of the line vs. distance. They were orginally invented to locate breaks in long distance telephone lines under the ocean (where it was impractical to inspect the cable). They are good for long runs between points too, for the same reason.

 

Hi msoff.
1.- For DC, it is resistance matching, not impedance matching. Like a battery made with eight AA cells won't start your car even being also 12Volts ---> as resistances don't match.

2.- Are you back to impedance? Check data sheets.

3.- ??? Impedances can not match and still work. Matching is to transfer maximum power. Listen to your stereo with headphones; sounds perfectly fine. But no full transfer of power, ---> impedances don't match.

4.- Depends on how the discrete circuit affects/loads the devices/signals;
- Do you know for sure you need to match impedances from one IC to the next ?
- Yes, a circuit can act as a impedance matcher. The speaker output of your stereo may have a ~200 ohm resistor inserted in the connection for the headphone jack.

Miguel