EEVblog #1367 - 5 Types of Oscilloscope Passive Probes COMPARED

[EEVblog]

Dave looks at the pros and cons of 5 different types of oscilloscope passive probes.
Switchable x1/x10, Fixed x10, High voltage single ended, DIY Transmission line Z0 probe, and BNC to banana/croc leads.
PART 1 of 2

[Tom45]

The Siglent 2000X Plus can be set to 21X probe attenuation. It also has 50 ohm termination and probe readout pin support.

Thanks for the Zo probe information. I look forward to the rest of the story in part 2.

[Anthocyanina]

why is "high capacitance for high frequency work" listed as a con? isn't it a pro of x10 probes that they offer a wider bandwidth and are better for high frequency than x1 probes?

[HyperSpectral]

You may have implied that a 350MHz probe on a 350Mhz scope will give 350MHz bandwidth. The actual bandwidth of this system will be slightly less that 250MHz. The calculation of this is done using rise time and the system rise time is the square root of the sum of the squares of the individual rise times. The result of this is that the bandwidth of a scope and probe of identical values is 70.7% of that value.

So, in order to approach the scope bandwidth, a much higher bandwidth probe must be used. Note that in theory, the scope bandwidth can not be reached if a probe of finite bandwidth (any actual probe) is used. A 500MHz or higher probe should be used to get realistic results near the 350MHz scope bandwidth.

All that said, the quicksand that many people fall into is forgetting that any waveform other than a Sine wave has harmonics starting at 2 or 3 times the fundamental. So, a 350MHz scope will display a 350MHz Sine wave at 3dB loss, but any other waveform will be seriously distorted.

[madires]

Would a BNC plug with an IR photo-diode be a passive or active probe?

[Tom45]

why is "high capacitance for high frequency work" listed as a con? isn't it a pro of x10 probes that they offer a wider bandwidth and are better for high frequency than x1 probes?

Consider this simplified circuit of a 10:1 scope probe:



At DC, the reactance of the tip capacitor is infinite and the probe's input resistance (impedance) is 10 Meg.

For AC the capacitor impedance is no longer infinite. As the frequency goes up, the capacitor's impedance drops.

Dave calculates the capacitor's impedance and shows the formula.

At 500 MHz, a 10 pf tip capacitance has an impedance of only 32 ohms. So the 10 Meg probe impedance is only 32 ohms at 500 MHz.

[Anthocyanina]

oh, i see, and that lower impedance is a con since it may affect how the circuit being measured behaves, and i guess that also limits what voltages you can measure at those high frequencies?

thank you!

[Bud]

@19:30 -20:30   talking about the 3ft piece of coax having 100pF capacitance which at 1GHz represents 1.6 Ohm shunt impedance   

 - NO.  At RF frequencies coax is a transmission line and when terminated to a matched load (50 Ohm in this case on the scope input) will only have resistive losses in the wire , which were specified right there in the datasheet that you showed , in the next table right  below the capacitance/ft data. That datasheet specified only 1 db of loss at 1GHz for 1m length of the coax used..

If coax cables had THAT low shunt capacitance at RF, they would be useless  at high frequency, which is not the case and is an exact opposite.

[tautech]

 
Basic errors....a few minutes in, the compensation trimmer on a passive probe is LF not HF although some passive probes have both and then they are in the BNC end compensation box.

13.30 that 100x probe is specially designed for isolated channel probes in that all connections on it provide isolation from voltages that might bite whereas your standard 100x probe is little different to any other passive probe.

[Bud]

Here is an excerpt on a home made 21:1 probe from "High Speed Digital Design, A Handbook of Black Magic" by Johnson/Graham.

The 1kOhm series resistor is not to "isolate high capacitance", but  to reduce  the rise time degradation L/R due to inductance of the sense loop, see the attached picture. With a 1kOhm resistor the ratio L/R is much smaller.

[tchicago]

@19:30 -20:30   talking about the 3ft piece of coax having 100pF capacitance which at 1GHz represents 1.6 Ohm shunt impedance   

 - NO.  At RF frequencies coax is a transmission line and when terminated to a matched load (50 Ohm in this case on the scope input) will only have resistive losses in the wire , which were specified right there in the datasheet that you showed , in the next table right  below the capacitance/ft data. That datasheet specified only 1 db of loss at 1GHz for 1m length of the coax used..

If coax cables had THAT low shunt capacitance at RF, they would be useless  at high frequency, which is not the case and is an exact opposite.

But the scope's input impedance is not 50 Ohm. It is 1 MOhm. Or, you are saying that the RF impedance is 50 Ohm, while DC impedance is 1 MOhm? I don't think this is the case here.

[Bud]

The 21:1 probe is made using 1kOhm external resistor at the test point and 50 Ohm terminator either switched on on the scope input or using an external termination resistor at the scope input connector if the scope does not have 50 Ohm input feature . Otherwise it will not have 21:1  voltage divider ratio. (1000+50)/50=21.  You may be confused because cheap and some mid-range scopes do not have built-in 50 Ohm input impedance feature and only have 1M input. For such scopes you have to use an external 50 Ohm terminator at the scope input.

[tggzzz]

The 21:1 probe is made using 1kOhm external resistor at the test point and 50 Ohm terminator either switched on on the scope input or using an external termination resistor at the scope input connector if the scope does not have 50 Ohm input feature . Otherwise it will not have 21:1  voltage divider ratio. (1000+50)/50=21.  You may be confused because cheap and some mid-range scopes do not have built-in 50 Ohm input impedance feature and only have 1M input. For such scopes you have to use an external 50 Ohm terminator at the scope input.

And then it isn't 50ohms, but 50ohms//20pF. That capacitance will change the attenuation as a function of requency, depending on the cable length.

That's why I like my 1970s Tek 485: it has a proper 50ohm input attenuator.

[nctnico]

@19:30 -20:30   talking about the 3ft piece of coax having 100pF capacitance which at 1GHz represents 1.6 Ohm shunt impedance   

 - NO.  At RF frequencies coax is a transmission line and when terminated to a matched load (50 Ohm in this case on the scope input) will only have resistive losses in the wire , which were specified right there in the datasheet that you showed , in the next table right  below the capacitance/ft data. That datasheet specified only 1 db of loss at 1GHz for 1m length of the coax used..

If coax cables had THAT low shunt capacitance at RF, they would be useless  at high frequency, which is not the case and is an exact opposite.

Added to that is that both sides of the coax need to see an impedance of 50 Ohms. Just putting 1k Ohm in series is not going to cut it. Terminating the end with 50 Ohm is only helping to get rid of reflections.

[tggzzz]

@19:30 -20:30   talking about the 3ft piece of coax having 100pF capacitance which at 1GHz represents 1.6 Ohm shunt impedance   

 - NO.  At RF frequencies coax is a transmission line and when terminated to a matched load (50 Ohm in this case on the scope input) will only have resistive losses in the wire , which were specified right there in the datasheet that you showed , in the next table right  below the capacitance/ft data. That datasheet specified only 1 db of loss at 1GHz for 1m length of the coax used..

If coax cables had THAT low shunt capacitance at RF, they would be useless  at high frequency, which is not the case and is an exact opposite.

Added to that is that both sides of the coax need to see an impedance of 50 Ohms. Just putting 1k Ohm in series is not going to cut it. Terminating the end with 50 Ohm is only helping to get rid of reflections.

Er, no.

I suggest you understand section 3 of
http://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf

[nctnico]

@19:30 -20:30   talking about the 3ft piece of coax having 100pF capacitance which at 1GHz represents 1.6 Ohm shunt impedance   

 - NO.  At RF frequencies coax is a transmission line and when terminated to a matched load (50 Ohm in this case on the scope input) will only have resistive losses in the wire , which were specified right there in the datasheet that you showed , in the next table right  below the capacitance/ft data. That datasheet specified only 1 db of loss at 1GHz for 1m length of the coax used..

If coax cables had THAT low shunt capacitance at RF, they would be useless  at high frequency, which is not the case and is an exact opposite.

Added to that is that both sides of the coax need to see an impedance of 50 Ohms. Just putting 1k Ohm in series is not going to cut it. Terminating the end with 50 Ohm is only helping to get rid of reflections.

Er, no.

I suggest you understand section 3 of
http://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf

Just look at these screendumps. The first is without 50 Ohm at the beginning of the coax cable, the second is with.





And for kicks using a Tektronix P6156 with a 20x attenuator and a probe-to-BNC adapter:


The 2nd image looks identical to the actual signal when connected directly. The others don't.

Direct connection:


In the end you can't beat transmission line theory. It is clear that not putting a 50 Ohm termination at the beginning of the coax cable is a a trade-off to get a reasonable result with minimal attenuation.

Edit: fixed inline images.

[tggzzz]

What is the scope, the scope's input impedance[1], the source, the source waveform, the cable length, the probing technique?

At these speeds the details matter.

[1] many are 50ohms//20pF, which causes problems.

[nctnico]

What is the scope, the scope's input impedance[1], the source, the source waveform, the cable length, the probing technique?

At these speeds the details matter.

[1] many are 50ohms//20pF, which causes problems.

This is measured using the calibrator output on a Lecroy Wavepro 7300A set to 50 Ohm with full bandwidth. The coax cable (about 50cm) is soldered onto an SMA (with 1k series resistor) which is connected to the calibrator output (BNC). Lead lenghts about 2cm. The test with the Tektronix P6156 probe (with 20x attenuator) is done with a BNC adapter on the probe in order to connect to the calibrator output. No ground leads or ground spring:


The P6156 is basically a 1k Ohm resistor in series with a coax cable so it is no surprise the result is the same. Edit: I see something went wrong with the inline images which I fixed now.

[PartialDischarge]

Here is an excerpt on a home made 21:1 probe from "High Speed Digital Design, A Handbook of Black Magic" by Johnson/Graham.

The 1kOhm series resistor is not to "isolate high capacitance", but  to reduce  the rise time degradation L/R due to inductance of the sense loop, see the attached picture. With a 1kOhm resistor the ratio L/R is much smaller.

I don’t have access to the rest of that book, but it is amazing how misleading the treatment of the risetime in that probe is. The discussion is centered around L/R and it is mentioned that this probe is going to be very fast, nope, the limiting factor is the input C to the scope and the dominating effect is RC with now R=1k. With a 1MOhm scope and external 50Ohm termination, that probe is going to be slow as molasses, with scopes with internal 50 Ohm better but not spectacular.
I would not waste time in building that probe when a FET probe is what is needed there

[tggzzz]

What is the scope, the scope's input impedance[1], the source, the source waveform, the cable length, the probing technique?

At these speeds the details matter.

[1] many are 50ohms//20pF, which causes problems.

This is measured using the calibrator output on a Lecroy Wavepro 7300A set to 50 Ohm with full bandwidth. The coax cable (about 50cm) is soldered onto an SMA (with 1k series resistor) which is connected to the calibrator output (BNC). Lead lenghts about 2cm. The test with the Tektronix P6156 probe (with 20x attenuator) is done with a BNC adapter on the probe in order to connect to the calibrator output. No ground leads or ground spring:


The P6156 is basically a 1k Ohm resistor in series with a coax cable so it is no surprise the result is the same. Edit: I see something went wrong with the inline images which I fixed now.

It sounds like the 50ohm at the source is connected calout-50ohmToGnd-1kohm-cable-scope. In that case the effect you see without the source 50ohm might be due to the cal out behaving properly when correctly loaded.

Alternatively, I suspect the scope isn't 50ohm input but is really 50ohm//15pF. In that case you will see the effect stated in the P6156 manual, viz:

Use with 1 Megohm Input Oscilloscopes

High resistance inputs require external 50 W terminations (Tektronix part number
011-0049-01). Introduction of a termination will result in a slight signal
reflection in the system (appearing at twice the cable delay time) due to the shunt
capacitance of the oscilloscope in parallel with the 50 W termination. To reduce
the effects of this reflection, add a 2X attenuator between the probe connector
and the 50 W termination. This will decrease the reflection by a factor of four
while increasing the attenuation by a factor of two. For sine wave measurements,
the 2X attenuator is recommended to minimize standing waves.

That one fooled me once in the past!

As you say, you can't beat transmission line theory

[nctnico]

What is the scope, the scope's input impedance[1], the source, the source waveform, the cable length, the probing technique?

At these speeds the details matter.

[1] many are 50ohms//20pF, which causes problems.

This is measured using the calibrator output on a Lecroy Wavepro 7300A set to 50 Ohm with full bandwidth. The coax cable (about 50cm) is soldered onto an SMA (with 1k series resistor) which is connected to the calibrator output (BNC). Lead lenghts about 2cm. The test with the Tektronix P6156 probe (with 20x attenuator) is done with a BNC adapter on the probe in order to connect to the calibrator output. No ground leads or ground spring:


The P6156 is basically a 1k Ohm resistor in series with a coax cable so it is no surprise the result is the same. Edit: I see something went wrong with the inline images which I fixed now.

It sounds like the 50ohm at the source is connected calout-50ohmToGnd-1kohm-cable-scope. In that case the effect you see without the source 50ohm might be due to the cal out behaving properly when correctly loaded.

No because with the second test with the coax, the cal output still sees a 1k Ohm resistor at it's output and yet the resulting signal looks fine.

Alternatively, I suspect the scope isn't 50ohm input but is really 50ohm//15pF. In that case you will see the effect stated in the P6156 manual, viz:

No; I have used a 3GHz scope to do the test. I'm damn sure it has proper 50 Ohm inputs! Please check the specs before making these kind of statements.

I suggest you to do a simulation with transmission lines in Spice.

[tggzzz]

What is the scope, the scope's input impedance[1], the source, the source waveform, the cable length, the probing technique?

At these speeds the details matter.

[1] many are 50ohms//20pF, which causes problems.

This is measured using the calibrator output on a Lecroy Wavepro 7300A set to 50 Ohm with full bandwidth. The coax cable (about 50cm) is soldered onto an SMA (with 1k series resistor) which is connected to the calibrator output (BNC). Lead lenghts about 2cm. The test with the Tektronix P6156 probe (with 20x attenuator) is done with a BNC adapter on the probe in order to connect to the calibrator output. No ground leads or ground spring:


The P6156 is basically a 1k Ohm resistor in series with a coax cable so it is no surprise the result is the same. Edit: I see something went wrong with the inline images which I fixed now.

It sounds like the 50ohm at the source is connected calout-50ohmToGnd-1kohm-cable-scope. In that case the effect you see without the source 50ohm might be due to the cal out behaving properly when correctly loaded.

No because with the second test with the coax, the cal output still sees a 1k Ohm resistor at it's output and yet the resulting signal looks fine.

In that case I don't understand your circuit - please supply a photo or schematic.

Alternatively, I suspect the scope isn't 50ohm input but is really 50ohm//15pF. In that case you will see the effect stated in the P6156 manual, viz:

No; I have used a 3GHz scope to do the test. I'm damn sure it has proper 50 Ohm inputs! Please check the specs before making these kind of statements.

I suggest you to do a simulation with transmission lines in Spice.

I've seen "50ohm" scopes that simply put 50ohm in parallel with the 1Mohm//15pF.

Looking at the pictures of the scope, there are many connections that can be used to communicate calibration info from the LeCroy probe to the scope. The scope might then be using DSP to adjust the display accordingly. If the Tek probe doesn't supply that info, then obviously the correction cannot be done.

I'll do a simulation once I have your schematic.

[nctnico]

No. Again, you are making assumptions which aren't valid. Just take the schematic (1k in series with the source at the beginning of the coax cable and a 50 Ohm terminator at the end) from the Tektronix paper you linked to and simulate it using a lossy transmission line model. You'll see that you get steps indicating an impedance mismatch (just as transmission line theory predicts).

[tggzzz]

Alternatively, I suspect the scope isn't 50ohm input but is really 50ohm//15pF. In that case you will see the effect stated in the P6156 manual, viz:

No; I have used a 3GHz scope to do the test. I'm damn sure it has proper 50 Ohm inputs! Please check the specs before making these kind of statements.

Here is evidence that you don't have a 50ohm input.

The pictures below are with my Tek 485 and HP10020a 1.5GHz Z0 probe. The key point about the 485 is that it has two input attenuators: one is the traditional 1Mohm scope attenuator, the other is a traditional RF 50ohm attenuator. An RF relay connects the input to either one or the other attenuators.

The first picture shows the source: the Tek485 cal out (<1ns risetime) correctly loaded with a 50ohm through terminator that ought to be good to a GHz or so. The probe (500ohm input) is directly connected to the source, without leads. Timebase is 5ns/div.

The second picture shows the trace when the scope's (real) 50ohm attenuator is used. I think you will agree that is a nice enough waveform.

The third picture shows the trace when using the scope's 1Mohm input with an inline 50ohm through terminator. Hence the input "seen" by the cable is really 50ohm//20pF. You will note the dip you see in your traces. Where do you think that has come from?

The fourth picture is used to highlight the probe cable length, by switching to the 50ohm attenuator so the input impedance is 25ohms. You will note the position of the step coincides with the dip in the third picture, thus demonstrating that the dip is due to the 20pF input capacitance.






Incidentally, those familiar with using TDRs will realise that a dip in a trace is due to capactance, and a peak is due to inductance.

BTW, a Z0 probe does not use a lossy transmission line; it uses standard coax.

[nctnico]

Just do the simulation and you'll see I'm right. You will get a step in the signal if the beginning of the coax cable isn't terminated.

I'm 100% sure the Lecroy 7300A has proper 50 Ohm inputs. Proof of that is that the 'dip' doesn't show with a direct cable connection. A 3GHz oscilloscope would be utterly useless without proper 50 Ohm inputs. People would have come to Lecroy's headquarters with pitchforks, tar and feathers if they screwed it up.