One of the things I never even thought of to check when buying an oscilloscope was to see if the probes and/or scope were limiting each other (as a result of their combined rise time from what I understand). I was of the opinion that an X Mhz scope would have X Mhz probes because that's what every scope I looked at had; and I had looked at more than 5 different models across 4 different manufacturers.
Well, I recently read that you have to do 1/sqrt(1/(O_BW**2)+1/(P_BW**2)) in order to see if the combination of both scope and probes can really achieve the advertised level of BW. For example, if you have a 350Mhz scope with 350Mhz probes, then you'd have a BW of ~247Mhz to probe with. Likewise, if you connected up a 250Mhz probe, you wouldn't have 250Mhz BW, instead you'd have ~203Mhz of BW. In fact, in order to get within 5% of the scope's 350Mhz of BW, you'd have to separately purchase 1000Mhz probes!

So, am I making a mistake in understanding how to calculate total system BW here? Is the above paragraph about BW a correct understanding?
If "yes", then why do manufacturers include probes that result in much slower than advertised performance?
Does this also affect sinusoidal waveforms or just waveforms with sharp edges?

Thanks!

EDIT: Tried to clarify what I'm asking.

 

I'm not sure but I think using a probe with the same bandwidth rating as the scope means you will see the bandwidth of the scope.

 

I don’t know the exact answer you are seeking.
I do know that you have to use a x10 attenuation probe.

 

If "yes", then why do manufacturers include probes that result in slower than advertised performance?

It depends on the manufacturer. The last time I checked, Tektronix probes didn't limit system bandwidth when used with their scopes. For example, if you had a 100MHz scope and used a 100MHz probe, the system bandwidth would be 100MHz (3db down). If you mix and match, the scope and probe would each be 3dB down at the maximum rated frequency, so overall system response would be 6dB down.

 

It depends on the manufacturer. The last time I checked, Tektronix probes didn't limit system bandwidth when used with their scopes. For example, if you had a 100MHz scope and used a 100MHz probe, the system bandwidth would be 100MHz (3db down). If you mix and match, the scope and probe would each be 3dB down at the maximum rated frequency, so overall system response would be 6dB down.

Rigol scope and probes, both 350Mhz. Also a 250Mhz probe from Pintek.

 

It occurred to me that it might be helpful to cite where I got this info from: https://edadocs.software.keysight.c...probe-with-the-wrong-bandwidth-587208673.html

 

The bandwidth is spec'd as the "3db down" point. As mentioned, if the probe is -3dB and then the scope is -3dB, then the combined loss is -6dB and you are no longer at the "spec point of -3dB for bandwidth" Hence the fancy formula that gives you the combined -3dB point taking into account both probe and scope loss.

 

The bandwidth is spec'd as the "3db down" point. As mentioned, if the probe is -3dB and then the scope is -3dB, then the combined loss is -6dB and you are no longer at the "spec point of -3dB for bandwidth" Hence the fancy formula that gives you the combined -3dB point taking into account both probe and scope loss.

I got that part.
I also made further replies towards the end of helping those who said they were confused.

"It depends on the manufacturer..."
Is what I'm now interested in.

 

"It depends on the manufacturer..."
Is what I'm now interested in.

I recall reading in some Tektronix documentation that they spec'ed overall system bandwidth when using their scope and recommended probes. I'll see if I can track down the reference again.

 

For example, if you have a 350Mhz scope with 350Mhz probes, then you'd have a BW of ~247Mhz to probe with

I see your point. If you have two similar amplifiers in series, 350 & 350 = 247.
I think the scope & probe are rated at 350 when used together.
I think the probe is rated to be used with at 350 scope.

 

It depends on the manufacturer, read some Tek and Keysight probe datasheets and you will see Tek makes no reference to "matching o-scope" while Keysight adds footnote that BW only achievable with certain model of their scope. Neither provide any details in test setup or how BW was measured. So its not a real spec if they don't detail how the measurement is performed. Just salesmanship. Probes may come with calibration sheet that would provide real details.

 

Keep in mind, if you're going to try to do some higher frequency stuff, you can't always just touch the circuit with the probe and expect to get accurate results. Sometimes you need to take very specific steps to get accurate readings. For example, I recently had to so some low level USB debugging and needed to use active and differential probes to get good info. If you will be trying to take readings from SMPS type stuff, you might need to use coax cable to carry the signal away from high EMF areas to where you can take a good reading, etc.. What exactly are you expecting to take accurate readings of?

 

What exactly are you expecting to take accurate readings of?

Well, I could tell you, but I was kinda saving it as a surprise to post here... I think you'll enjoy it. (I'll give you a hint though, it's a measurement of noise and then how to eliminate it.)

But bear in mind that in order to measure a sinusoidal waveform, you need at least 3x the BW of the waveform you want to measure; at least based on what I've read.

 

But bear in mind that in order to measure a sinusoidal waveform, you need at least 3x the BW of the waveform you want to measure; at least based on what I've read.

3-5x higher bandwidth is used for square waves. You don't need to worry about harmonics with sine waves.

 

There is a big difference between analog and digital scopes.
With analog scopes is common for people to plot the bandwidth of the scope.
-1db at xxxmhz, -2db @ xxxmhz, .......-10bd at xxxmhz.
When you run off the top end of the scope, you can look at the high end graph and correct for the loss of gain at very high frequencies. Not a great idea but extends the frequency range (with some errors).

Have not tried it with digital scopes. I am working with an 8 channel, 6ghz, 16GS/sec scope as I type this.