On 20 Jul 2017 17:50:21 +0000, you wrote:
I am confused regarding the stated bandwidth of an oscilloscope probe.
I thought that the bandwidth was computed from the input impedance, almost purely capacitive at high frequencies. as the frequency where the reactance drops to 50 Ohms.
50 ohms is not the magic point. It is on 50 ohm probes, though.
The oscilloscope input is standardized at 1 meg (for most scopes) and
somewhere between 15 and 25 pf (tek = 1/20, HP I think goes 1/17 on
some scopes).
The actual probe is not just a resistive voltage divider feeding a
resistive/capacitive load, it has some peaking networks and
compensation networks to optimize the equivalent (for Tek) pulse
response to the frequency response.
This idea seems to work most of the time but I am seeing some probes with stated bandwidth somewhat higher. For instance I see some Chinese probes advertised as 100 MHz, 200 MHz, even higher, yet their input capacitance is still around 20 pF more or less.
I have Tek probes generally to 200 Mhz, although the cable length has
an effect. 1X probes have very low bandwidth, 10x have higher, and
some 100X probes have higher than that. Past about 200 Mhz, the
inputs go to 50 ohms and I forget the capacitance. You're dealing
with low impedance here, while a 10x 1 meg probe has a loading of
perhaps 4-5 pf in parallel with 10 megs, a 10x 50 ohm probe has a
loading effect of 500 ohms, not sure about the capacitance.
So my opinion is that the bandwidth is a combination of the peaking
networks, the cable capacitance, the scope capacitance, the probe's
stray capacitance and the effective capacitance of the actual dropping
resistor.
The 20 pf is just really the capacitance that the probe can be
adjusted to match.
NB: some HP probes can adjust to 20pf TEK inputs, and vice versa, but
don't depend on it. Wonder why they did that?
Before you jump down my throat about how terrible Chinese stuff is, let's keep it at a technical level and explain just what bandwidth means so I can decide what to buy.
Equivalent pulse response. While TEK and HP probes are held to higher
standards, I'm not so sure about the few Chinese probes I have. You'd
probably do well to see if what you can get is available from multiple
vendors so you might be able to find spare parts as/if needed.
I suspect that if you found a Rigol brand probe, you might get a
higher quality than an unnamed brand. That said, the Chinese probes
will work, but I'd be tempted to look across the spectrum of the same
probe (if you can ID them) and pick the lowest bandwidth as reality.
I see probes very cheap (not really rugged but still a good value) for various prices from around $4 to $13 but they appear to be the same, except for the bandwidth rating. How can this be?
They may be playing specmanship. Tektronix probes of different cable
lengths have different bandwidths. The bandwith is also limited by
the scope you're using it with.
Since my use is very casual and I don't need to pay for high quality, this is a somewhat important question. And of course, after a few years they fall apart and I need to purchase again, but over my lifetime I don't usually get to where I have spent the same money as for a Tek or HP probe.
Had you considered the Tektronix (or perhaps HP if they did this)
modular probes? P6108 is a 10x, modular (can replace parts) probe, 2
meters, 100 Mhz, 13 pf loading, 10 megs loading. That is, the probe
looks like a 13 pf capacitor to ground in parallel with a 10 meg
resistor.
Ebay used is quite possible.
Just went through about 10 probes, all of the P6105/6/8 variety,
identical specs at 2 meters length. A similar switchable probe 1X/10X
has a load of 1 meg and 104 pf in the X1 position and 10 megs/13 pf in
the 10 x position.
Harvey
Bob
