开云体育

As a result of my recent purchase of a Tek scope with 1 ghz. bandwidth, I have begun looking for probes. I have looked at the specs for a P6243 and a TPP 1000, both rated at 1 ghz bw. Both instruction manuals show a chart of impedance vs frequency. In either case the impedance decreases with frequency (makes sense because of capacitance) to a few hundred ohms at 1 ghz. Am I missing something here? Doesn't that load down the observed ckt in the DUT to the point where the probe (and attached scope) are essentially useless at their rated high end? Thanks in advance for taking the time.
Jack Reynolds

Hi Jack,
The only thing you're missing is a recent edition of Tektronix' obligatory document "ABC of Probes", downloadable from e..g. <>.
Yes, a probe presents quite a load at high frequencies and it depends on the signal source whether that is a problem.
Where source impedance can be controlled, a low-Z probe (usually 50 Ohm impedance) is preferable.
It's all in ABC of Probes.

Raymond

Hi Raymond,

Thanks for your ever so prompt response.?? I am not missing the ABC's of probes and in fact have previously read it a few times. Rather than look for my paper copy, I just downloaded it and skimmed it again and I am afraid that I am still at a loss as to the usefulness of either of the 1 ghz probes I mentioned at a 1ghz frequency under test.? If you could describe what I am looking for in the primer, I will certainly go take another look.? I also understand the convenience and utility of using 50 ohm cable and loads at higher frequencies.? Have I happened on a couple of "dogs" for the probes I have investigated and mentioned?? The 6243 active probe has only 1 pf of capacitance at the tip and looks like a 50 ohm input to the scope.? Are virtually all 1 ghz circuits of such? low impedance that throwing say 500 ohms capacitive across them in parallel doesn't perturb them?? Is there a mysterious phase angle thing here that I don't fathom?? Should I be looking at other probes?? Tektronix was or is selling these things as 1 ghz capable probes.? Of what use are they?

On Fri, 26 Apr 2019 16:36:38 -0700, you wrote:

As a result of my recent purchase of a Tek scope with 1 ghz. bandwidth, I have begun looking for probes. I have looked at the specs for a P6243 and a TPP 1000, both rated at 1 ghz bw. Both instruction manuals show a chart of impedance vs frequency. In either case the impedance decreases with frequency (makes sense because of capacitance) to a few hundred ohms at 1 ghz. Am I missing something here? Doesn't that load down the observed ckt in the DUT to the point where the probe (and attached scope) are essentially useless at their rated high end? Thanks in advance for taking the time.

I have found a similar problem, not with probes, but with circuits.
The higher the impedance, the more that the circuit capacitance limits
the bandwidth.

Most of the 200 Mhz or higher circuits look like a low impedance
driver, I've put a 10:1 50 ohm probe on a TTL circuit, and the probe
worked well enough (driving 500 ohms).

Naturally, an active probe has less capacitance and lower loading at
the expense of more limited voltage range and more fragile
electronics.

Harvey

Jack Reynolds

On Fri, 26 Apr 2019 21:17:09 -0400, you wrote:

Hi Raymond,

Thanks for your ever so prompt response.?? I am not missing the ABC's of
probes and in fact have previously read it a few times. Rather than look
for my paper copy, I just downloaded it and skimmed it again and I am
afraid that I am still at a loss as to the usefulness of either of the 1
ghz probes I mentioned at a 1ghz frequency under test.? If you could
describe what I am looking for in the primer, I will certainly go take
another look.? I also understand the convenience and utility of using 50
ohm cable and loads at higher frequencies.?

that depends on the circuit. Lots of RF circuits are 50 ohms, some
75.

However, high speed TTL, ECL, and the like circuits fake being RF (and
quite well, since the boards often have to be designed like RF
circuits). So the scope loads the circuit, as all things do.

Have I happened on a couple
of "dogs" for the probes I have investigated and mentioned?? The 6243
active probe has only 1 pf of capacitance at the tip and looks like a 50
ohm input to the scope.?

what it looks like to the scope is meant to match the scope. What
does it look like to the circuit? The P6201 probes look like few pf
to the circuit, 100k ohms, but present the plugin with it's required
50 ohm input.

Are virtually all 1 ghz circuits of such? low
impedance that throwing say 500 ohms capacitive across them in parallel
doesn't perturb them??

for 50 ohms, that's about 10%.

At 1 GHz, what's the reactance of 20 pf of a standard high bandwidth
probe?

Is there a mysterious phase angle thing here that
I don't fathom?? Should I be looking at other probes?? Tektronix was or
is selling these things as 1 ghz capable probes.? Of what use are they?

how would you connect a scope to a circuit that is running at 500 Mhz
to 1 Ghz, or has equivalent bandwidth?

take the best bandwidth 10x probe you have, put it on a relatively
high impedance high frequency circuit (not a signal generator with a
low output impedance), then throw a 1x probe across that and see what
happens.

Harvey

On 4/26/2019 7:57 PM, Raymond Domp Frank wrote:

Hi Jack,
The only thing you're missing is a recent edition of Tektronix' obligatory document "ABC of Probes", downloadable from e..g. <>.
Yes, a probe presents quite a load at high frequencies and it depends on the signal source whether that is a problem.
Where source impedance can be controlled, a low-Z probe (usually 50 Ohm impedance) is preferable.
It's all in ABC of Probes.

Raymond

Hi Jack,
I would recommend thoroughly reading the downloaded document again. Just having a quick look at it just now revealed relevant content at least on page 10 ("Usually, the loading of greatest concern..."), 13 ("keep ground leads short"), 16-17, 29 ("Signal Source Impedance"), 29-30 ("Input Resistance and Capacitance"), 32, 33, 34, 35, 37 ("Bandwidth to the probe tip"), 40, 43.

The capacitance of *all* Hi-Z probes has a large influence at high frequencies, because the decreasing input impedance becomes an ever increasing short on the resistance. FET probes with their relatively small input capacitance present a higher high frequency impedance than most passive High-Z probes but their input resistance (as observed at DC and low frequencies) will still drown (be shorted) in the capacitive reactance at higher frequencies, just at a lower rate than higher-capacitive (passive) probes.
Low-Z (50 Ohm) probes have the advantage of a constant impedance from a relatively low frequency until their highest useful frequencies, albeit that that impedance is rather low.
Active (FET) probes isolate their output from their input, making a constant 50 Ohm output across their bandwidth impedance possible.

The bandwidth of probes is specified assuming a source impedance of 25 Ohm, see "Bandwidth to the Probe Tip" on page 47.

All this is a matter of physics, not a matter of choosing the right probe to completely get rid of the problem. You can only optimize by making the right choices. You cannot get 10 Mohm or even 1 Mohm input impedance at high frequencies with any regular (FET) probe. Also, transmission line effects (signal reflection, standing waves) would destroy the "improvement".

If you want to do measurements on high-frequency signals or signals with high-frequency content, always consider using Lo-Z probes like 1:1 50 Ohm, 1:10 500 Ohm or even 1:100 5 kOhm probes. These have relatively constant impedance and many can be made at home. Search for that on the Internet.

Among High-Z probes, FET probes with their relatively small input capacitance have the relative advantage of making the setup less sensitive to ground lead inductance, resulting in less "ringing". You can read about that in the ABC as well.

Raymond

On Sat, Apr 27, 2019 at 05:59 AM, Raymond Domp Frank wrote:

If you want to do measurements on high-frequency signals or signals with
high-frequency content, always consider using Lo-Z probes like 1:1 50 Ohm,
1:10 500 Ohm or even 1:100 5 kOhm probes. These have relatively constant
impedance and many can be made at home. Search for that on the Internet.

You may want to look for an old-fashioned Lo-Z probe like a P6056 (500 Ohm) or P6057 (5000 Ohm). Even with their standard BNC connection, they work well with your more modern digital 'scope and can be bought for little money.

Raymond

You may want to look for an old-fashioned Lo-Z probe like a P6056 (500 Ohm) or P6057 (5000 Ohm).
Even with their standard BNC connection, they work well with your more modern digital 'scope and
can be bought for little money.

Raymond

I'd second that.

Also the ABC of Probes was published at intervals, and the content changed as time went by.

Craig

I am not missing the ABC's of probes and in fact have
previously read it a few times. Rather than look for my paper copy, I just downloaded it and skimmed
it again

Raymond

The other things to download and read cover to cover are the Concepts series books "Oscilloscope Probe
Circuits", "Probe Measurements" and "Sampling Oscilloscope Circuits"

Craig

On Sat, Apr 27, 2019 at 05:59 AM, Raymond Domp Frank wrote:

All this is a matter of physics, not a matter of choosing the right probe to
completely get rid of the problem. You can only optimize by making the right
choices. You cannot get 10 Mohm or even 1 Mohm input impedance at high
frequencies with any regular (FET) probe.

Have a look here for something special, a bit more extreme: <>

Raymond

Hello Everyone,

Thank all of you who responded to this query and Raymond in particular.? I hope that there are a few other people in the group who benefited from this discussion.? I certainly did.? My aha moment came when I saw Raymond's mention of the 25 ohm output impedance for the DUT (confirmed by the ABC's).? To summarize what I think I have learned:? Obviously capacitive loading of higher z circuits at high frequencies would be a problem with any sort of mechanical probe.? In order to be able to sell and have people be able to compare the performance of higher frequency probes, the industry(Tek) had to draw a line in the sand for comparison purposes and that was done at 25 ohms.? The typical probe impedances at rated frequency of about 500 ohms then make perfect sense as that represents somewhere around an additional 5% load on the circuit being measured.? If the user is bent on trying to measure higher impedance circuits at high frequencies, then as the younger generation says, It is what it is!

Thank you for your time and bandwidth,

Jack Reynolds