From: "John Jardine"
Subject: Re: AN: New Logic Analyzer, 500MHz, $166
Date: Sun, 17 Nov 2002 19:00:50 -0000
References: <3DD4890B.email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com>
NNTP-Posting-Date: 17 Nov 2002 18:49:28 GMT
X-Newsreader: Microsoft Outlook Express 5.00.2615.200
Jeff Liebermann wrote in message
> On Sun, 17 Nov 2002 10:44:18 +0100, "Frank Bemelman"
> >"chris" schreef in bericht
> >> ">
> >> > Input circuit is 100K in parallel with 10pF. As you would
> >> > expect, the sampling logic reads this signal every 2nS (when
> >> > sampling at 500MHz).
> >> >
> >> > /Tim
> >> Hi, your input is only 3ohms a 500M, not much will drive that! At
> >> 500kHz its still 3kohms so it looks like my initial guess was about
> >> right.
> >Or is it 30 ohms at 500Mhz? Of course it's not to analyse a 500Mhz bus,
> >it's just that it captures at 500Mhz, so you can expect some detail
> >in the results.
> 30 ohms is correct if we ignore the 30cm probe lead length. However,
> at 500MHz, two 30cm leads (one for signal, one for ground), yields
> about 3000 ohms of reactance. That makes a nice -37dBv attenuator.
> L = 2W [ ln (2W/r) - 1.00]
> L = inductance in nHy
> W = Wire length in cm
> r = wire radius in cm
> I'll assume #26awg = 0.41mm dia
> L = 2 * 60 [ ln ( 2 * 60 / 0.021 ) -1 ]
> L = 120 [ ln 5713 ] = 120 [ 8.65 ] = 1038 nHy
> L = 1.038 uHy
> The inductive reactance at 500MHz is:
> XL = 2PifL = 6.28 * 500 * 10^6 * 1.038 * 10^-6
> XL = 3125 ohms
> So, now we have a real network. Assuming negligible probe wire
> resistance, and ignoring the 100K input resistance, at 500Mhz, we get:
> >---- XL= 2135 ohms------------------>
> probes Xc = 30 ohms to analyzer
> Since the vectors are 180 degrees apart, we can assume that they
> cancel resulting in an input impedance at the end of the probes of
> about 2165 ohms, and a loss of about:
> 30/(2135 + 30) = 0.014
> dBv = 20 log 0.014 = -37dBv
> In other words, a 3.3v p-p input signal at 500MHz will appear as 46mv
> on the input to the A/D converter.
> [Note: I'm half asleep and doing some of this in my head. please
> check my calcs, and ignore my sloppy abuse of significant figures,
> before assuming they're correct].
> Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
> (831)421-6491 pgr (831)336-2558 home
> http://www.LearnByDestroying.com WB6SSY
> firstname.lastname@example.org email@example.com
Offhand, as there is stray capacitance between the connecting wires, I would
have thought it better if the wires are looked at as a distributed
transmission line. In which case at 500Meg the wires may have a simple
'resistive' impedance of a few hundred ohms. But of course ... the
100kohms//10p meter input impedance is in no way going to terminate that
line, which still makes the makers claim of a 500Mbit/S measurement,