While on-paper specifications are easy to compare, they only tell part of the story. While a device might be excellent on-paper, it could be difficult to use or have a number of important limitations. In this section, I look at the usability aspects of the RTM3004 – just how enjoyable is it to use on a daily basis? Are there any annoyances which get in the way of getting things done?
As with all complex test equipment, I always recommend prospective users read the manual and other supplied documentation to make sure they are able to make the most of their equipment. The RTM3004 doesn’t come with a printed manual, only with a quick start guide. The full manual is available for download online and is a hefty 787 pages as of the fourth revision.
I spent the better part of the first month of the RoadTest period with my printed copy of the third revision of the manual, reading through the first half of the manual which details operations of the base unit and all of the options. The manual was very well written with clear explanatory diagrams and copious amounts of full-colour screenshots to illustrate the procedures. The manual also comprehensively detailed the available features and data formats used by the oscilloscope, although due to the way the manual is arranged, some information might require reading parts of several chapters. Only a few very minor typographical and grammatical errors were found in the manual and reported for correction.
The latter half of the manual is a SCPI command reference which appears to be extremely comprehensive but is not required reading until you need to construct an automated testing solution.
On the whole, I enjoyed reading the manual and I feel that I have benefited because I was able to learn about features which I wouldn’t otherwise be aware of. The best thing is that manuals of a similarly high standard are also available for the software.
Front-Panel and Touch-Screen LCD Operation
The first thing that strikes you about the RTM3004 is the way the 10.1” capacitive touch-screen occupies a significant majority of the front panel area. The LCD is large, bright and glossy, which is slightly unfortunate as this means that glare from the room lights does show up as well as fingerprints. This is understandable because of the fine pixel pitch of the display – a matte screen may look grainy due to the interaction of the grain and the pixels themselves, whereas this screen is sharp like an Android tablet is.
As promised, pressing the LED-backlit power soft-button starts up the oscilloscope in under ten seconds. Compared to some other units I’ve used, this is like night and day. By the time I’ve pushed the button and got the probes into the right position, it’s already ready to display its first trace. While this is a convenient feature to have, its usefulness is somewhat moot in a lab setting where you might power-up the instrument at the beginning of the day and leave it running, being sure to wait the allotted 30 minutes to ensure it has warmed up and is providing accurate readings.
Almost of the oscilloscope’s operations can be achieved using the touch screen. The main menu is the rhomb menu in the bottom right corner of the screen, which brings up a sliding menu which can be swiped to select the function, calling up a submenu where options can be selected, toggled and values manipulated. For example, in this case the display settings are being configured, with clear toggles and values which can be manipulated by selection, direct entry or by turning the multi-function knob. In my experience, the menus were quite responsive, although swiping the main menu up and down did result in some jerkiness.
On-screen keyboards are available for entering numeric values and filenames. On-screen help is also available, making the operation rather intuitive. Aside from the rhomb menu, there is a quick toolbar menu across the top where you can “pin” up to eight quick buttons from a selection.
Something which wasn’t immediately obvious to me was that each of the channel labels in the toolbars and the sample mode/trigger icons in the top right area are actually shortcuts to a short menu for configuring the most commonly toggled settings that can also be used to quickly invoke the full menu for more advanced configuration.
Other features are available through the Apps selection which can be invoked from the rhomb menu or via the dedicated front panel button, allowing access to installed options.
To the right of the screen, a more traditional compliment of buttons and knobs are separated into controls for various subsystems. While the number of buttons is somewhat reduced to a non-touch-screen, the remaining controls are well-thought-out and allow for more traditional manipulation of the oscilloscope parameters which helps ease the transition. However, even with the realisation that most of the functions can easily be accessed by touch instead, having the physical interface can still be a faster way to access important menus (e.g. channels, trigger, apps) without needing to go through a multi-step process. It also allows a more “analog” feel when it comes to adjusting values sequentially, although I found the knobs to have an “acceleration” behaviour that meant that I often overshot the value I was intending to reach. Instead, I have changed my habits to favour direct entry of the intended value to be even more efficient.
Some of the front panel buttons are translucent for LED backlighting, which is used intelligently to provide information “at a glance”. For example, the LEDs for each active channel display a colour code – Yellow for C1, Green for C2, Orange for C3, White for C4, Blue for Logic and Teal for Math which corresponds to the colours used on the LCD display. The Trigger button lights up with this colour to allow you to identify the trigger source and toggle it with a single push. The LEDs are a little bright by default, but the brightness is adjustable, which helps immensely. There is a slight annoyance that the supplied probe identifying collars aren’t exactly these colours, but it’s still close enough.
The touch-screen lends itself to a more intuitive note-taking process as well, with the inbuilt annotations tool allowing you to draw over the screen in a number of colours and add a single line of text to label your discoveries. This is potentially quite useful, although the single text box with a short line of text is somewhat limiting.
In case you want to disable the touch-screen, there is a touch-lock button on the front panel, which can be quite useful in educational circumstances in case you might be pointing out certain aspects of a signal and don’t want it to be interpreted as a screen touch or accidentally manipulate any settings. Even then, it is still possible to interact with the screen as the front USB host port can be used with a USB hub and USB HID-class keyboards and mice to operate the oscilloscope and key in values, in addition to using it with USB Mass Storage Devices formatted in FAT/FAT32 for data storage. A limitation in that the keyboard is assumed to be in QWERTZ layout – this has been reported for improvement in the future. The unit also features an education mode which locks out password protects certain menus and settings.
So, whether you like to touch the screen, twiddle knobs, push buttons or use a keyboard and mouse, the RTM3004 has you covered. In fact, if you check out Chapter 8, you can do it remotely too.
Experiment: Probing Around a Raspberry Pi
To give the RTM3004 a bit of a test, I decided to grab a Raspberry Pi and give it a probe, since it should have some decently challenging signals to look at.
For example, the above is the 4-level signal from the magnetics of the Ethernet jack when a 100Mbit/s connection has been established. Unfortunately, the RTM3004 doesn’t have Ethernet decoding.
The above shows the two oscillators on the Raspberry Pi – one at 25Mhz and the other at 20Mhz. The hardware trigger counter gives us a good idea as to its frequency without using the measure tool.
The above shows an intermittent clock – the one used on the SD Card which is only active when the card is being accessed. A nice 25Mhz, as expected.
Finally, two traces – one showing a 1.5Mbit/s low-speed USB device (keyboard) and its D+/D- lines. The other shows the Wi-Fi adapter connected in 480Mbit/s high-speed mode, although the signal is somewhat noticeably corrupted by the influence of the long ground wire. Unfortunately, the RTM3004 doesn’t feature USB decoding.
Experiment: Looking at Video Signals
To test the external triggering facility, I used a VGA to RGBHV lead with the terminations set to 50 ohm (close enough to the 75 ohm expected) to plot the signals.
The RGB lines are connected to C1, C2 and C3 respectively. C4 has the horizontal sync signal, whereas the vertical sync signal is connected to Ext Trig. In this case, I have triggered on the horizontal sync, but it was also possible to change to vertical sync simply by changing the trigger source.
I also decided it was worth testing with a composite video source, as there is a dedicated Video Trigger mode. In this case it is synchronised at the vertical synchronisation pulse.
Selecting a particular line number, it was possible to zoom in and observe the horizontal sync, colour-burst signal and the image data for that line.
Data and Setup Save and Recall
A number of data saving options are available. It is possible to save screenshots, instrument settings, displayed waveform data, buffered waveform data, history waveform data, reference waveform data and power analysis reports. A number of save formats are offered, some of which are native to the RTM3004 (e.g. TRF, SET), whereas others are interchangeable (e.g. CSV, TXT, PNG, BIN).
The oscilloscope features internal memory which is used by default for reference waveforms and settings and the ability to use external USB storage in FAT/FAT32 format. A OneTouch save feature is also provided, which allows for “one button” save of multiple types of data into a single ZIP file for easier data management. By default, filenames feature an automated number increment on each save.
Manual filenames can be specified on saving, however, due to (what I assume to be) patent restrictions in implementing long file-name support on FAT/FAT32 filesystems, the filenames are limited to just eight characters. For settings, this is a significant limitation, so a description field and screenshot are saved as well to ease identification.
During testing, an issue was identified where if waveform saving was used with more than one channel (i.e. Vis. Channels selected as source), the oscilloscope would lock-up once the save button was selected. A workaround is to save each channel separately, which works correctly. This issue has been reported to Rohde & Schwarz for rectification.
Measurement, Statistics, Cursors and Math
Eight measurement slots are available which support measurements such as burst width, pulse count, edge count, mean value, RMS value, peak value, frequency, period, amplitude, base level, overshoot, pulse width, duty cycle, rise/fall time, delay, phase, crest factor and slew rate. Manual cursor measurements are also available.
The active measurements can have the statistics mode turned on, allowing for means and standard deviations to be recorded across acquisitions.
An automeasure feature is provided which only operates on one channel but automatically chooses relevant measurement parameters and illustrates it visually on the screen.
Math channels are available, up to five, with two oprands each and configurable operation and units.
Experiment: Oscilloscope Channel Input Noise
Given the ability to do basic measurements and statistics, I thought it’d be nice to just run a quick check of the channel input noise. I disconnected all inputs, set C1 to 1:1 at the highest sensitivity and shortest timebase and let it average for a while.
The channel input noise measured an average RMS value of 108.11µV which is rather unexpectedly low. Keep in mind, this was inside my room with RF interference, unshielded LAN cables, USB cables etc. The peak-to-peak voltage was averaging 312.17µV which isn’t bad either. I suspect this is why I found the instrument to give pretty “quiet” traces compared to other units I used in the past. In fact, it’s that quiet that sometimes I forget I’m measuring small signals at the end of a 10:1 probe!
The oscilloscope offers four main sampling modes and other modes which are a combination of them. The default mode is Sampling, which takes a single sample to produce a reading on the screen. The High Resolution mode reduces the effective sampling rate for longer timebases by averaging a number of samples together to produce a more representative value. Unlike some of the other oscilloscopes I have used, there is no configuration options, so the trade-off between sample rate and resolution increase cannot be tailored. There is also Envelope mode which shows the highest and lowest values of the samples representing the signal on the screen and a Peak Detect mode which only shows the highest values. Examples of them and their influence on the signal from my Mega 1280 LFSR are shown below.
The RTM3004 is capable of triggering on a variety of conditions, including Edge, Edge A/B, Width, Video, Pattern, Runt, Risetime, Line, Timeout and Serial Bus. Most types also include a hysteresis setting. The serial bus trigger feature depends on the serial bus decode options installed and activated at the time. The trigger level can be changed by the dedicated knob which can be clicked to automatically set the 50% level, working quite well for most signals. Further details can be found in the manual.
The Mask Test functionality is included as standard with the RTM3004 which is a nice feature to have. Masks can be defined from measuring a “golden template” waveform and scaling it in the X and Y axes, or through importing reference waveforms. The mask test can be configured to save failures to the segmented memory buffer, beep on failure, save a screenshot, output a pulse, etc.
Experiment: AC Power Line Waveform
A waveform that can have occasional “violations” is the mains power. Direct measurement of mains power is dangerous and the oscilloscope inputs are not rated for such operation, so I measured the output from an isolated AC step-down transformer. A mask was developed by scaling the “regular” waveform generously – in this case, a number of violations were caught over 12 hours which were stored in a memory buffer and output to screenshot. This was a case where a power transient was caught.
Such an interruption can occur due to a switching transient – imagine if this happened at the peak of a waveform, the voltage can be almost 30% higher than the nominal value.
Digital Volt-Meter, Trigger Counter and FFT Display
Other features included as standard include a digital voltmeter, trigger counter and FFT analysis.
Experiment: 433Mhz ASK Module Signal
The RF domain is one space where FFT displays are commonplace, so I measured the signal from a “cheap” 433.92Mhz ASK transmitter module. This would be “close” to the 500Mhz bandwidth limit of the passive probes.
Using the more-accurate hardware counter, the frequency was 433.876Mhz (compared with the measurement report of 433.369Mhz), the signal showing a good 820mV peak to peak roughly. Because of the frequency limit of the hardware DVM, the reading was nearly zero.
Using the FFT mode, it is possible to see the signal in the frequency domain. The settings for RBW/Span/etc. are connected with the recorded signal length, so adjustment of the time-domain timebase to longer values is necessary to achieve narrower RBWs. In this case, we can see stray emissions about 16 and 32Mhz either side of the intended frequency.
Zooming out on the FFT also reveals a second harmonic emission at twice the intended frequency.
For more details on how the FFT display works in practice, see Chapter 7: Arbitrary Waveform, Function and Pattern Generator where it is used to graph the gain transfer function of an opamp.
Memory Segmenting and History Mode (RTM-K15)
The RTM-K15 History and Segmented Memory option increases the oscilloscope memory from 40Mpts per channel to 400Mpts per channel while adding fast segmenting, allowing for under 1.5us blind-time by capturing without visualisation. This feature has been invaluable, as many events happen extremely quickly and trying to obtain an accurate trigger on an infrequent event with just one shot is something which can be extremely frustrating. With the segmentation and history, it is possible to capture a number of events first and then go back to find a particular trace of interest. Depending on the record length, you can have between 5 and 34,952 history segments.
Experiment: Infrared Remote Signals
To illustrate the advantages of segmented memory, I set-up an experiment to look at raw infrared remote signals as received by an IR phototransistor. In this case, the oscilloscope triggered at the beginning of each signal and recorded the signal up to the configured sample length.
This allowed me to hit the “Single” button on the oscilloscope, mash every single button on the remote in sequence and have all the button data in the memory without having to do it one-button-at-a-time. This data can then be saved in a batch export as well.
The history mode has a segment table which shows the time for each recorded segment in a number of different time formats. The segments can be explored one at a time, directly accessed by number or via the slider. Otherwise, they can be replayed either singly or in a loop.
Three modes of basic analysis are provided – overlay, average and envelope. Using this, it is easy to spot “constant” parts of the signal between captures amongst other commonalities.
As I was formerly awarded a Tektronix PA1000 Power Analyzer, I thought I’d run an experiment just to check what the standby and operational power requirements for the RTM3004 are.
Using my test rig consisting of a “pure sine wave” synthesised mains source, Variac and PA1000, the standby power as measured to the IEC 62301 standard at 230V was just below the required 1W level, measuring 0.9745W. This meets the requirements but is still a little high for just powering a power LED – the USB and Ethernet interfaces are shut down. It suggests that maybe there is something else going on – maybe there is a processor running or some RAM being refreshed to enable faster boot-up times?
Turning the unit into active mode, the power consumption was measured using the same standby test to show power trends over time. The power consumption stabilised about 72.555W in operation, which is similar to that of a high-end laptop computer. Despite this, the large fan kept the unit at its operating temperature without making a distracting level of noise. The fan was only noticeable when the room was silent but did not have any strong tonal qualities. My Rigol DS1102E had a fan which is quite significantly louder.
I’m glad to report that, on the whole, using the RTM3004 was an intuitive and highly satisfying experience. The documentation available was clear, well-written and fully illustrated, reducing the learning curve and effectively introducing newcomers to the Rohde & Schwarz touch interface.
The RTM3004 is fast to start, taking less than 10 seconds and has an intuitive and responsive touch-screen interface that is as simple to use as most mobile applications. The 10.1” screen is large, sharp, clear and glossy. The rhomb menu, quick toolbar, short menu and dedicated front panel buttons provide a number of ways to accomplish actions. Having the knobs available preserves the “analog” interface and eases the transition from non-touch instruments, although the knob acceleration makes it feel less “precise” in my opinion. The RTM3004 is versatile enough that you can directly key-in most values using an on-screen keyboard. The panel is also smart, using colour-coded LEDs with adjustable brightness to provide feedback about channel status. External USB input devices are supported, so whether you like to touch the screen, twiddle knobs, push buttons or use a keyboard and mouse, the RTM3004 has you covered. The unit is also whisper-quiet, with its fan barely noticeable in ordinary use.
The RTM3004 offers eight measurement slots and a wide array of automatic measurements. Cursors are available for manual measurements. Statistics mode is supported to provide mean and standard deviation values across acquisitions. Automeasure is available on a single channel, allowing for automatic measurements and labelling of parameters. Five math channels are available with two oprands each and a wide array of mathematics function. This was very much as expected from an oscilloscope of this class.
During these tests, I (informally) measured the channel input noise at an average RMS value of 108.11µV which is quite unexpectedly low and explains why I find the traces from the RTM3004 to be “quiet” compared to some other oscilloscopes I have used before. Because of this, sometimes I’m not aware that I’m actually measuring a small signal at the end of a 10:1 probe, so the signal reaching the oscilloscope is miniscule.
It is possible to save screenshots, instrument settings, displayed waveform data, buffered waveform data, history waveform data, reference waveform data and power analysis reports. The oscilloscope features some internal storage but can also access USB Mass Storage devices formatted in FAT/FAT32. A OneTouch feature is also provided so multiple types of data can be saved at a single keypress into a single .ZIP file for easier data management. Filenames are limited to eight characters, which is unfortunate.
The only major detraction was an issue in exporting Visible Channels in the waveform save dialog which resulted in the oscilloscope locking up, where individual export of each channel worked just fine. The amount of time spent debugging the problem as a little irritating, but the issue has been reported and is expected to be rectified in the next firmware update.
The RTM3004 has included within its base offering a digital voltmeter, trigger counter, FFT display and mask test capabilities which is an additional sweetener and makes the instrument useful for automated testing applications and basic RF tasks even without the Spectrum Analysis option.
The RTM-B15 History and Segmented Memory option proved to be extremely useful and highly recommended, as it upgrades the 40Mpts memory into a segmented memory of 400Mpts per channel. This allows you to capture anywhere from 5 to 34,952 segments for replay and analysis later, with fast segmentation ensuring a minimum blind time to ensure infrequent events are captured. With an industry-leading amount of memory, there is also less need to compromise on sample rate or record length.