R&S to launch MXO5

[Bernd_2]

Hello,
I have some questions regarding the maximum input voltage for the 1 MOhm input of the MXO5.

The spec sheet states that the maximum allowable voltage of 300 Vrms with 20 db/decade derates to 5 Vrms greater than or equal to 250khz.

At 1 MOhm with RT-ZP11 passive Probe for derating see RT-Zxx Standard Probes Specifications (PD 36.07.3851.22). The ZP11 is not (yet?) included in the version available to me, so I consulted the user manual for the ZP11.

As best I could from the small voltage derating diagram in the manual, I transferred the values for the ZP11 (orange color) and for the MXO5 (red color), as I understood the spec sheet, into a diagram on double logarithmic paper.

If it is true that only a maximum of 5 Vrms may be applied to the input of the MXO5 above 250 kHz, a measurement up to 300 Vrms would only be possible up to a maximum of 50 kHz with the ZP11 with a divider ratio of 10:1 and only a maximum of 50 Vrms would be permissible above 250 KHz, even if the probe itself can withstand more than 50 Vrms up to 25 MHz.

Compared to other oscilloscopes, this seems very low to me.

For comparison, I have also drawn the derating curves of the Tektronix 5 B Series (blue color) and the Batronix Magnova (green color) according to the specifications in the spec sheets.

For the Keysight HD3 (gray dashed line), the spec sheet only states that a maximum of 135 Vrms may be applied to the 1 MOhm input, without providing any information on the derating curve with increasing frequency. Is it really possible that this device can handle the full 135 Vrms over the entire frequency range?

I would be delighted if someone here could help me to understand this, because of course I don't want to blow the front end of my MXO5 under any circumstances.

Best Regards
Bernd

[egonotto]

Hello,

I don't understand the “If it is true that only a maximum of 5 Vrms may be applied to the input of the MXO5 above 250 kHz, a measurement up to 300 Vrms would only be possible up to a maximum of 50 kHz with the ZP11 with a divider ratio of 10:1 and only a maximum of 50 Vrms would be permissible above 250 KHz, even if the probe itself can withstand more than 50 Vrms up to 25 MHz.”

According to the data sheet of the R&S ® RT-ZP11 that I have, you can measure up to 400 V rms sine wave if the signal is below 500 kHz.
With a 10 MHz sine wave, 80 V rms is still permitted.
With a 100 MHz sine wave, 20 V rms is still permitted.
With a 500 MHz sine wave, 10 V rms is still permitted.

It must be remembered that the impedance falls with the frequency.
At 1 MHz sine wave you have about 10 kOhm.
At a 10 MHz sine wave you have about 1 kOhm.
At 100 MHz sine you have about 100 ohms.
At 1 GHz sine you have about 10 ohms.
 So with a sine wave with 5 V rms and 1 GHz you have a current of about 0.5 A rms

Best regards
egonotto

[Bernd_2]

Hello egonotto,
thank you for your quick reply.

I have added the output voltage curve of the ZP11 (black line) in the diagram. In each case 1/10 of the input voltage (upper orange line).

What I meant was that between about 32 kHz and 25 MHz the permissible input voltage for the ZP11 cannot be used, because otherwise the output voltage of the ZP11 would be higher than the permissible input voltage on the MXO5.

For example, at 1 MHz, approx. 330 Vrms may be applied to the input of the ZP11.
As I understand it, that would be 33 Vrms at the output. According to the data sheet, however, the MXO5 can only handle 5 Vrms at 1 MHz. So at 1 MHz you should not apply more than 50 Vrms to the input of the ZP11 when it ist connected to the MXO5.

Unfortunately, there is no impedance-frequency diagram for the oscilloscope, so I have compared the voltage derating curves in my diagram.

What surprises me is that the derating curve goes down so steeply, although the input capacitance of the MXO5 is lower than that of the other models:

MXO5      12 pF
Tek 5B     13 pF
Magnova  19 pF
HD3         24 pF

The picture below in post #150 shows the original derating curve from the ZP11 manual.

Best Regards
Bernd

[egonotto]

Hello,

from datasheet of MXO5: "300 V (RMS), 400 V (V p ),
derates at 20 dB/decade to 5 V (RMS) above 250 kHz"

So you can estimate:
300 V rms for sine below 250 kHz.
30 V rms for sine below 2.50 MHz.
15 V rms for sine below 5 MHz.
7.5 V rms for sine below 10 MHz?
5 V rms for sine over 25 MHz.

At 10 MHz it could be a bit tight. However, since the 20 dB/decade between 2.5 MHz and 25 MHz does not quite apply (at 25 MHz only 3 V rms would be allowed for a sine wave, but they allow 5 V rms).
So I wouldn't worry about that. As 5 V rms is always permitted.
 
The reason for derating is as follows. As the impedance decreases with increasing frequency, the current increases.
For example, when using the R&S ® RT-ZP11, the current is lower at a sine wave of 10 MHz and 90 V rms than at 100 MHz and 20 V rms.

But if you still have concerns, you can ask Rohde & Schwarz to be sure.

Best regards
egonotto

[Bernd_2]

Hello,
Thanks again for the explanations.

Now I know where my misunderstanding was. I had understood the information in the data sheet to mean that the 5 Vrms limit was already reached at 250 kHz.

You are right. The derating only starts at 250 kHz and then goes down at 20 db/decade until it reaches the 5 Vrms limit at approx. 15 MHz.

I have drawn a new line with the corrected values in the attached diagram.

Best Regards
Bernd

[Bernd_2]

Hello,
here are a few pictures of the Leo Bodnar fast pulser LBE-1322 on MXO5 at 10 MHZ.

First at 2 GHz bandwidth at 50 Ohm. With bandwidth = 0.45/risetime, the measured risetime of 214 ps results in a bandwidth of approx. 2.1 GHz.

Then at 1 MOhm at 700 MHz bandwidth. With a measured risetime of approx. 1.5 ns, the calculated bandwidth is now only 300 MHz.

Best Regards
Bernd

[egonotto]

Hello,

Bernd wrote: "Then at 1 MOhm at 700 MHz bandwidth. With a measured risetime of approx. 1.5 ns, the calculated bandwidth is now only 300 MHz."

If the source has 50 ohms and the input of the oscilloscope has 1 Mohm || 10 pF then the input impedance at 320 MHz is approximately 50 ohms. This corresponds to a low pass of less than 300 MHz. This is therefore compatible with the 700Mz bandwidth of the oscilloscope.

Best regards
egonotto

[Bernd_2]

Hello egonotto,

thank you for your explanation of the connection.

With another oscilloscope with 17 pF input capacitance, the result was accordingly a risetime of approx. 2.3 ns.

I have attached a screenshot where the pulser is connected to the input at 1 MOhm via a 50 Ohm feed-through terminator. The measured risetime is now 507 ps, which corresponds to a bandwidth of 887 MHz according to the above formula.

Best Regards
Bernd

[Bernd_2]

Hello,

here are a few pictures with the probes R&S ZP11 (700 MHz) and the Testec TT-HF 612RA (500 MHz) on the Bodnar Pulser LBE-1322 at 10 MHz, 1 Vpp.

The connection to the LBE-1322
is established with a Telegärtner SMA-BNC adapter and the 2.5 mm Testec BNC adapter.

As can be seen in Fig. 1, the factory compensation of the ZP11 can still be improved somewhat.

Figure 2 shows the comparison of the ZP11 with the TT-HF 612RA. The compensation on the TT-HF is a compromise between the straightest possible roof on the rectangle and overshooting.

Is it possible that the dents in the roof at 12 or 13 ns are caused by reflections at the input of the oscilloscope? With a caliper length of 1.3 m and an assumed velocity factor of 0.65-0.67, this would be about right mathematically.

13 ns x 0.29972m/ns x 0.67 / 2 = 1.305 m

Best Regards
Bernd

[Martin72]

Tell me where you live and when no one is usually at home....

[Bernd_2]

Even if you knew, you wouldn't stand a chance against the hungry cats guarding my test equipment. 

[2N3055]

You need to terminate Pulser with 50 Ω pass through terminator to have 25Ω source impedance..

[Bernd_2]

Hello 2N3055,

I had done this a few days ago. At the Moment I only have a cheap Chinese feed through. So I don't know if the massive Overshoot is caused by the Terminator.

Therfore I didn't post these pictures until now.

I'm going to buy a better one. Not sure if I should buy the R&S HZ22 or spend the extra money for a Huber&Suhner one.

But already with the cheap terminator an improvement in risetime with the ZP11 from 1.086 ns to 0.5268 ns visible.

When I have the better terminator I'll repeat the measurement and post the results.

Best Regards
Bernd

[Bernd_2]

Tomorrow I'll try it with a BNC-T with a 50 Ohms termination resistor at one End and the Probes at the other End. Let's see what the results are then.

[Bernd_2]

Hello,
Here are the results of the measurements with the two probes on the LBE-1322 with a BNC-Y and a 50 OHM terminating resistor on one leg.

Figure 1 shows the diagram of the ZP11 with little overshoot and significant improvement in rise time compared to the direct connection.

Figure 2 shows the diagram of the TT-HF 612RA with considerably more overshoot and also a smaller improvement in rise time compared to the ZP11 compared to the direct connection.

Figure 3 shows the measurement curves of both probes in one diagram.

Best Regards
Bernd

[Bernd_2]

Hello,

the Digikey package with the new Cal Test CT2944C-50 Feed Through Terminator (FTT) and an Amphenol RF 000-46650-51RFX 50 Ohm BNC Terminator Plug arrived today.

I couldn't bring myself to buy the almost 100 euro more expensive Huber+Suhner 6701.01.B FTT, especially as the data sheets show no noticeable difference.

The build quality of the Cal Test looks good to me, all contacts are gold plated, the VSWR is the same and the average power is significantly higher than that of the H+S

Figure 1 shows the diagrams for the direct connection of the LBE-1322 to the input with 1 MOhm, as well as with an intermediate P57 FTT and CT2944C FTT.

With the CT2944C the overshoot and undershoot is a little less, but considering that the P57 only costs 1/6 I find this very acceptable.

Figure 3 shows the whole thing again with the ZP11.

Here the differences in over- and undershooting are even smaller.

My new Highflexx 7 cables are due to arrive tomorrow, so I'll do an FFT sweep with the two FTTs. Let's see if there are any significant differences.

Best Regards
Bernd

[Bernd_2]

Hello,
here is an example of the XY display with the MXO5, which is only possible if an oscilloscope supports a math waveform as  source for an axis in the XY diagram.

The pictures show the hysteresis curve of the iron core of a Phywe experimental transformer as used in physics lessons in schools. The magnetic field strength H is plotted on the X-axis and the magnetic flux density B on the Y-axis. However, I have made the following simplifications.

The magnetic field strength H outside the conductor results from the quotient of the current strength I and the field line circumference and has the SI unit A/m and is therefore proportional to the current in the primary coil of the transformer (channel 2 on the oscilloscope).

H ~ I_p

The magnetic flux density _B results from the magnetic flux φ divided by the area of the single core. However, because the flux in the closed core is not so easy to measure, I use the relationship that the voltage induced in the secondary coil is proportional to the temporal change of the flux in the core, more precisely the first derivative of the flux φ with respect to time

U ~ dφ/dt

It follows that the flux φ is proportional to the integral of the secondary voltage over time. SI unit Vs.

φ ~ ∫ U dt

In the XY diagram, the math waveform integral C1 (secondary voltage) + C4 is therefore displayed as the Y-axis. A DC offset is applied via C4 to center the diagram vertically. The diagram is therefore not the BH curve known from textbooks but a φI curve, which also shows the effect as derived above.

In order to drive the core into saturation even at low voltages and currents, I have lowered the frequency on the AC power supply to 12 Hz.

Figure 1 shows the hysteresis curve at 80 Vrms sine wave voltage at 12 Hz. The distance between the intersection point of the curve and the zero point on the X-axis is a measure of the coercive field strength H_c and the distance between the intersection point on the Y-axis and the zero point is a measure of the remanence B_r.

Figure 4 shows the effect with a rectangular voltage of 40 Vrms and 12 Hz on the primary side. As can be seen, the core saturates already at 40 Vrms.

Best Regards
Bernd

[Bernd_2]

Hello,

The new firmware version 2.6.2.4 includes a practical new feature: measurement result lines.

As an example, I have attached a screenshot from the Bodnar Pulser, where the result lines show exactly where the measurements are taken.

Best Regards
Bernd

[2N3055]

Hello,

The new firmware version 2.6.2.4 includes a practical new feature: measurement result lines.

As an example, I have attached a screenshot from the Bodnar Pulser, where the result lines show exactly where the measurements are taken.

Best Regards
Bernd

It was long overdue time. Those are called measurement cursors and many other scopes have it for a long time. Even some that cost less than 500€
They are very useful and are good usability improvement for the users.
It is nice to see R&S is listening.

[Bernd_2]

Yes, I am familiar with this from my Siglent SDS2000X Plus, where the source for a cursor can be a measurement.

On the MXO5, the result lines are located in the measurement setup and are completely independent of the four possible cursor sets.

Unlike the SDS2000X Plus, the cursors on the MXO5 do not have a measurement as source.

[2N3055]

Yes, I am familiar with this from my Siglent SDS2000X Plus, where the source for a cursor can be a measurement.

On the MXO5, the result lines are located in the measurement setup and are completely independent of the four possible cursor sets.

Unlike the SDS2000X Plus, the cursors on the MXO5 do not have a measurement as source.

Measuremnt cursors can also be set from measurements. Higher models also have multiple cursor sets.

But that is not important. My point is that, simple as it is, this is very useful thing and you will enjoy it.
Best,
Sinisa

[Bernd_2]

Thank you Siniša,

since it is implemented I use it for almost every measurement.

Best
Bernd

[pdenisowski]

It is nice to see R&S is listening.

Oh, we're listening!  There's just a limit as to how much new functionality we can fit in each release. If you saw our roadmap (and no, I can't share - sorry!), you would see that we have a LOT more coming

[2N3055]

It is nice to see R&S is listening.

Oh, we're listening!  There's just a limit as to how much new functionality we can fit in each release. If you saw our roadmap (and no, I can't share - sorry!), you would see that we have a LOT more coming

Shame, that roadmap would be interesting reading...   

But no surprise there. Other guys have their roadmaps too.. So you need to look into the future all the time.

[Martin72]

If you have too much on your plate at once, you won't get anything done.
Or you'll get it done very slowly.