Probes for 1GHz scopes?

[alanambrose]

I've been looking at the used MSOX3104T 1GHz scopes from Keysight and have a question re probes. If you buy this scope new it comes with 4 off 500MHz probes just like a 200MHz scope. Why only 500MHz - what's the point of supplying those probes with that scope? Also, the datasheet lists only the N2795A 1GHz and N2750A 1.5GHz high frequency probes - but don't you need say a 3GHz probe to use the scope's 1GHz bandwidth successfully?

Alan

[AndyC_772]

Bear in mind that the probe is just as much a part of the measurement system as the scope, and it's essential to consider how it will affect the device being probed.

A passive probe is robust and versatile, but has inherently limited bandwidth, and its capacitance can place a significant loading on the circuit under test. A "500 MHz" probe may indeed have a reasonably flat response out to 500M when connected to a low impedance source, but it's not high impedance at 500M. The N2890A probe is rated 10M Ohms // 11pF, and that 11pF has a reactance at 500M of just 29 Ohms.

To the best of my knowledge there are no passive probes out there which are significantly better.

Active probes have much lower capacitance, so they give a much more accurate and representative reading at high frequencies. The N2795A is 1M Ohm // 1.0pF. I also have an older 1156A probe which is rated to 1.5 GHz and is 100k // 0.8pF. In both cases the frequency response is still going to depend much more on the source and how it reacts to having 1pF connected to it.

I use the passive probes on my MSO-X3104A 95% of the time. The other 5% I use an active probe, normally when I'm verifying timing and signal integrity on a digital interface.

[edpalmer42]

How often do you expect to be working at frequencies above 500 MHz?  Probes for those frequencies are expensive.  That's why most vendors don't supply them with a 1 GHz scope.  If most of your work is at lower frequencies, hide the 500 MHz probes and use cheap 100 MHz probes for daily use.

As Andy stated, even the best 'high impedance' passive probe can load down a circuit at high frequencies.  However, an alternative to the high priced active probes is a 'low impedance' passive probe.  There are a few message threads around here on that topic.  Basically, the probe connects to a 50 ohm input instead of a 1 Mohm input.  A 10:1 probe has a resistance of 500 ohms which doesn't sound good, but the shunt capacitance of 1 pf or less makes it compatible with many (most?) uses.  With a little care and luck, these probes can be a good DIY project.

Ed

[tggzzz]

How often do you expect to be working at frequencies above 500 MHz?  Probes for those frequencies are expensive.  That's why most vendors don't supply them with a 1 GHz scope. 

Frequently, whenever assuring signal integrity! 500MHz is only a rise time of 700ps, and jellybean logic is faster than that.

Besides, if you are only looking at 10MHz , you wouldn't need a 1GHz scope.

[tggzzz]

To the best of my knowledge there are no passive probes out there which are significantly better.

Passive resistive divider Z0 probes have much higher bandwidth e.g. keysight  n2874a, hp10020a.

[David Hess]

Also, the datasheet lists only the N2795A 1GHz and N2750A 1.5GHz high frequency probes - but don't you need say a 3GHz probe to use the scope's 1GHz bandwidth successfully?

Oscilloscope bandwidth is usually specified including a suitable probe so a 1 GHz oscilloscope with a 1 GHz probe results in a 1 GHz bandwidth at the probe tip.  But be warned that this assumes a 25 ohm source impedance to drive the capacitance at the probe's input.  Higher source impedances will result in lower bandwidth.

[BravoV]

Decent active probes, even used are still quite expensive, took me almost 3 years of lurking to score this one at "decent" price. 

HP 54701A, 2.5GHz 100kOhm // 0.6pF @input.

[OwO]

Resistive divider + MMIC amplifier (e.g. BGA616) can give you reasonably flat gain to 1GHz easily and have capacitance in the range of 100fF, while costing less than a few dollars in parts. Of course you must have a 50ohm input on the scope, but that is a given at >50MHz. Of course, expecting 1MOhm input impedance or even 100kOhms is completely unreasonable with this kind of design.