Roughly measure PCB and short cable trace impedance

[kudjung]

Dear All,

I just wonder whether we can have any method that roughly measure PCB impedance value or look at impedance discontinuty if we only have a high speed oscilloscope but not the TDR.

I use to try emulating the TDR using a 25 MHz signal generator and a scope but that work for a very long cable.

As i understand we will need a very short pulse to look at PCB impedance.

I just read through some of the very fast pulse generation method and found some topic about SRD(step recovery diode) which can generate pulse in ps range.

Anyone has any experience with this SRD or any thinking whether this method will work in trying to look at impedance discontinuty or measuring PCB impedance value. If we just generate the pulse using SRD and then look at reflected signal using a high speed oscilloscope.

Thanks for any suggestion.
Kudjung

 

[volker_muehlhaus]

A pulse generator is not difficult, but you would need a really fast scope.

As an alternative, maybe you can measure the series inductance and shunt capacitance of the line. Line impedance ZL=SQRT(L/C)

 

[FvM]

A pulse generator is not difficult, but you would need a really fast scope.

Respectively, it doesn't make sense to have the pulse generator much faster than the oscilloscope. The required bandwidth or risetime depends of course on the electrical length of your impedance test pattern. A fast step will be basically sufficient for the pulse generator, standard ECL logic (about 200 ps risetime), said step recovery diodes (< 100 ps), avalanche transistor circuits (about 150 ps) or mercury wetted reed relays (200 ps) are promising simple solutions.

 

[kudjung]

**broken link removed**

This particular circuit cellar magazie has a very well information on the TDR and some circuit for experiment base on avalanche transistor. I'll give this a try. Look like this might work.

I posted it here in case it may helpful to some folks that looking for the same information.

 

[kudjung]

Guys,

This is the first time I'm using transistor in avalanche mode. I'm doing some experiment with the transistor that I have first, 2N3904. I doesn't seem to be able to get the transistor in avalanche mode.

I'm doing a simple test first by connecting 100 ohm resistor between B and E. Feed the Collector and Emitter with current-limited power supply and gradually increase the supply voltage. I've maxed out the power suply at 120 volts without damaging 2N3904. The circuit was not drawing any current. 2N3904 only has Vceo about 40 volts. I was thinking that when the power supply voltage reach above 40 volts, 2N3904 should reach the avalanche region and trip the power supply or otherwise damage but nothing happen. But when I turn off the power supply then the 2N3904 is damaged.

I think I might have mis some of the concept behind this avalanche mode.

Anyone can help give some explanation about this or suggest some simple setup to study the avalanche mode in action.

TIA,
KJ

 

[FvM]

I previously used a 2N708 as avalanche transistor. But I had an original Motorola type, it's not said, that presently available devices from other manufacturers show the same behaviour https://www.microsemi.com/datasheets/2n708.pdf

I generally experienced, that many transistors, particularly small planar-epitaxial types, are suited for avalanche operation.

P.S.: You didn't tell about applying a trigger pulse.