Lower bandwidth probes for a higher bandwidth oscilloscope - bad?

[fazeka]

Hi all,

I have an older Tektronix 2445A 'scope that needs a new pair of probes. The only work I do is in the audio realm. Highest frequency I would test would be the bias frequency of an analog tape recorder (432 kHz).

My friend has a set of Pomona 6493 probes he would consider selling me. I did some research and found that these are 60MHz. The Tek 2445A is a 150MHz scope.

I've researched online, seems everyone suggests going with probes that are higher bandwidth.

What's issues would I encounter (for audio work, if it matters) with using 60MHz probes with a 150MHz scope?

Thanks!

 

[JohnRoberts]

The scope should not care, the image will just be LPF.

About the only thing I can imagine is a 10x probe not being accurate.

Will your friend let you borrow them to check...? Most scopes have a front panel square wave output to tweak overshoot with. If the square wave looks square you should be cool...

JR

 

[gyraf]

..AND x10 probes all have a small trimmer somewhere to adjust for optimum square-wave performance..

/Jakob E.

edit: that's what you say already, isn't it

 

[Monte McGuire]

The bandwidth numbers specified for probes are not all that useful or reliable, since they usually are defined when the probe is driven by a super low impedance, such as 25Ω or so, using a really compact ground connection to the probe body at the tip. It's usually defined as 0.35/(risetime), so if you're measuring square waves straight from a generator, then this is something that will relate to what you see on the scope.

A more important factor is the capacitance of the probe at the tip. This capacitance gets added to your circuit and will affect not only the bandwidth of what you see, but the operation of the circuit itself. A lower value will always be better. With a typical fast 10x probe, values around 8pF to 11pF will be about as good as you can find. Incidentally, this is why 1x probes are a really bad idea - there is no way to isolate the probe cable capacitance from your circuit - it's all just sitting there in parallel with your circuit. Values like 60-100pF are not uncommon for typical 1x probes, and that's an enormous capacitance, sometimes enough to destabilize an op amp in an already marginal circuit.

If you actually need to  do fast things, then an active probe can be a good idea, but they're pricey. Some FET probes have capacitances around 0.5pF to 2pF, and can give you real bandwidths of 1GHz or more. Also remember that up there, a couple of pF turns into a low impedance, so capacitance is really important to minimize.

For audio work, 50MHz is probably 'enough', and a capacitance around 8-11pF will be suitable for everything but super fast logic or possibly aligning a helical scan machine like a DAT.

The Pomona 6493 has a capacitance spec of 12.5pF at the tip, which is a little on the high side, but not a problem for most work. I think it'll work well even though your overall system bandwidth will be lower than what it could be. In reality, most people end up using their scope input channels with the bandwidth limit switch on to reduce the bandwidth to 20MHz or thereabouts to reduce trace noise and remove useless interference. Sure, high bandwidth is useful sometimes, but this probe will be very useful.

It's also not a bad thing to have several probes, so buying this now and getting something else later is not bad either

 

[JohnRoberts]

The bandwidth numbers specified for probes are not all that useful or reliable, since they usually are defined when the probe is driven by a super low impedance, such as 25Ω or so, using a really compact ground connection to the probe body at the tip. It's usually defined as 0.35/(risetime), so if you're measuring square waves straight from a generator, then this is something that will relate to what you see on the scope.

A more important factor is the capacitance of the probe at the tip. This capacitance gets added to your circuit and will affect not only the bandwidth of what you see, but the operation of the circuit itself. A lower value will always be better. With a typical fast 10x probe, values around 8pF to 11pF will be about as good as you can find. Incidentally, this is why 1x probes are a really bad idea - there is no way to isolate the probe cable capacitance from your circuit - it's all just sitting there in parallel with your circuit. Values like 60-100pF are not uncommon for typical 1x probes, and that's an enormous capacitance, sometimes enough to destabilize an op amp in an already marginal circuit.

Alternately there have been cases where the probe capacitance can stabilize the circuit. I heard one story about a government/military system prototype where they shipped the first unit with a scope probe still attached hidden inside the back of the chassis.  :

JR

If you actually need to  do fast things, then an active probe can be a good idea, but they're pricey. Some FET probes have capacitances around 0.5pF to 2pF, and can give you real bandwidths of 1GHz or more. Also remember that up there, a couple of pF turns into a low impedance, so capacitance is really important to minimize.

For audio work, 50MHz is probably 'enough', and a capacitance around 8-11pF will be suitable for everything but super fast logic or possibly aligning a helical scan machine like a DAT.

The Pomona 6493 has a capacitance spec of 12.5pF at the tip, which is a little on the high side, but not a problem for most work. I think it'll work well even though your overall system bandwidth will be lower than what it could be. In reality, most people end up using their scope input channels with the bandwidth limit switch on to reduce the bandwidth to 20MHz or thereabouts to reduce trace noise and remove useless interference. Sure, high bandwidth is useful sometimes, but this probe will be very useful.

It's also not a bad thing to have several probes, so buying this now and getting something else later is not bad either