Fullscreen HTML5Med Res (??MB)HTML4

This is the top model in the Agilent 54600 series. It has two analog input channels and 16 digital input channels so it can perform as a combination analog scope and logic analyser. Agilent 54622D Mixed Signal Oscilloscope Back in October 1991 we reviewed the first of the “new generation” digital scopes from Hewlett-Packard, the HP 54601. Now we review one of the successors of that unit, the Agilent 54622D, a 200 megasample/second “mixed signal” machine. But while the two machines might bear the same 3-digit prefix, the perfor­mance of the new series is radically better. REVIEWED BY LEO SIMPSON 26  Silicon iliconCChip hip This is the rear panel of the 54621A 60MHz 2-channel oscillo­scope. This is identical on the 54622D mixed signal scope except that the latter unit also has the trigger input socket on the back panel. I N NOVEMBER 1999, the instrumentation side of Hewlett-Packard was spun off into a new company called Agilent Technologies Inc and this company has finally released its new series of digital scopes. There are five models in the series, two 60MHz models and three 100MHz units: the 54621A 2-channel 60MHz; the 54621D 18-channel 60MHz mixed signal scope; the 54622A 2-channel 100MHz; the 54624A 4-channel 100MHz; and the mixed-signal 54622D which has two analog chan­ nels and 16 digital channels. Over a period of several weeks we had the chance to try out the 2-channel and 4-channel 100MHz models and the 54622D mixed signal oscilloscopes. Apart from the more comprehensive trigger­ ing facilities and 16-channel input capability of the 54622D, all the scopes have the same operating features so virtually every­thing in this review will apply to all models in the series. First up, we will state that the ba­ sic “spec” for the 54600 series seems pretty standard – 100MHz or 60MHz bandwidth and 200 megasamples/ second (for analog channels). How­ ever, this tells only a small part of the story because this series of scopes does not have the sampling rate tied to the timebase speed, giving low sample rates at low timebase speeds. Instead, the 54600 series run at a high sampling speed all the time and in fact, it runs at the maximum 200MSa/ sec speed for all timebase speeds above 1ms/div. This is a huge step up in performance compared to most digital scopes. And it has a large memory, with 2MB of RAM per channel or 4MB in single channel mode. These two factors make it a potent scope indeed. We’ll come back to the performance aspects later. For the moment, let’s look at the basic operating features. Compared with other brands of digital scopes with CRT screens, the Agilent 54600 series is quite compact, measuring 352mm wide (including handle), 172mm high (including feet but not accessory bag) and 317mm deep. It is quite light at 6.3kg. The scope screen is a nominal 7-inch CRT with a live screen area measuring 127 x 99mm while the screen graticule has the standard arrangement with 8 vertical and 10 horizontal divisions. Along the bottom of the screen are six “soft keys” and their functions all change, depending on what button has been pushed on the control panel. Below these soft keys is the 3.5-inch 1.44Mb disk slot, the power on-off switch and the intensity control. This latter knob only controls the intensity of the channel traces, not the graticule or the various screen labels. The screen labels are fixed in intensity but you can change the brightness of the graticule (the ruled grid on the screen) by pushing the Display button, the Grid soft-key and then rotating the multi-purpose knob to the right of the screen. The control panel has a number of control groupings which make it easier to use. For example, there are groupings for the analog (vertical) and digital inputs, the horizontal timebase, triggering and so on. In the analog group, there are two BNC sockets, one for each channel, a sen­ sitivity control and position control for each and three buttons, one to select each channel and one to select MATHematic operations. Each of these three buttons is brightly illuminated when it is press­ ed. Digital inputs Similarly, for the selection of digital (logic) channels, there are two buttons plus a “label” button and these are also illuminated when pressed. The 16 digital inputs are connected via a special 16-way connector and two 8-way woven ribbon cables each of which termi­nate in a “pod” which has nine leads – ground and eight probe leads which can be connected to header pins on PC boards, or when probe tips are attached, connected to individual IC pins. You can select either or both of the 8-channel banks which are labelled D0-D8 and D9-D15. Individual chan­ nels can be turned off and any or all channels can be separately labelled by using the “label” button. You can (laboriously) feed in your own labels or you can select from the scope’s own fairly comprehensive logic library. The digital channels can be set for TTL, CMOS, ECL or user-defined threshold levels and by setting dif­ ferent thresholds for the two-channel banks you can work on mixed logic circuits. This whole 16-channel digital fa­ cility means that you have the best of both worlds in the Agilent 54622D – a high perfor­mance two channel analog (well, digital storage) scope and a 16-channel logic analyser. Well, not quite but you get the picture. In the “trigger” panel there are five buttons and a level control. Again, four of these buttons (edge, pulse December 2000  27 Fig.1: this is the “help” screen you get when you press and hold down the channel 1 selector button. Similar help screens are available when you press any button on the Agilent scope, regard­less of whether it is a front panel button or one of the softkeys below the screen. These help screens would make it much easier for any novice to become competent in using an oscilloscope. width, pattern & more) are illuminat­ ed when pressed and they also bring up their associated softkeys below the scope screen. Four other buttons can be illuminated: run/stop, single (for single shot mode), cursors and “quick meas”. Finally, the knob below the main time­ b ase control has an adjacent anticlockwise arrow and when some function can be varied by this knob, the arrow is illuminated. This last fea­ ture might be simple but it indicates just how much thought has gone into making this scope easy to use. Overall then, the control panel of the new Agilent 54600 series is dis­ armingly simple. You immediately feel as though you can sit down and use it straight away without having to read through a manual. And just in case you do need help with any feature, all you have to do it hold down the relevant button and some text will immediately appear on the screen to give the information. For example, say you are looking at the channel 1 menu which has soft keys for coupling (GND, DC, AC), BW limit, vernier and probe. You want to know what BW limit is. Press the softkey and text will come up to tell you that this key sets the scope bandwidth to 20MHz (instead of 100MHz) and this can be used to remove unwanted high frequency noise from the signal. Bandwidth limiting has been around for a long time on scopes but if you were a novice user you may 28  Silicon Chip Fig.2: here is a 100kHz sinewave with a small amount of noise superimposed, as shown by the small amount of frizzle on the waveform. Note the vertical cursors which are brought into play when the frequency measurement is called for. not have an inkling of what it meant. By the way, while all the oscillo­ scope labelling and operat­ing features are in English, the help screens can also be in Chinese, French, German, Italian, Japanese, Korean or Spanish. Fig.1 shows the help screen you ob­ tain when you hold down the channel 1 selector button. Before we leave the channel 1 menu we should mention the probe soft­key. The 54600 comes standard with 10:1 passive probes and these will auto­ matically be sensed by the scope as soon as they are plugged in. The scope then changes the input sen­sitivity by a factor of 10. So if the sensitivity was 1V/div with no probe connected, it will automatically go to 10V/div as soon as the 10:1 probe is connected. On the other hand, if you are using a probe without the auto-sensing fea­ ture, the probe softkey will let you set the attenuation to suit, anywhere from a factor of 0.1:1 to 100:1, in a 1,2,5 sequence. Apart from using probes without auto-sensing, this feature could be useful when you are moni­ toring a circuit via a voltage divider and you want the appropriate factor to be used and displayed by the scope. Screen saver We mentioned the controls for screen brightness (intensity) previ­ ously and that brings up another good feature of these Agilent scopes – a screen saver. Many a time I have come across a scope with a bright display (with no-one watching it) and I in­ stinctively lean across and turn down the intensity. But these scopes have a screen saver which moves around the screen and thereby avoids burning any pattern into the screen, as well as increasing the life of the tube, as the beam current is reduced. There is a choice of three screen saver patterns (Agilent logo, etc) and you can set the time after which it cuts in, from one minute to 435 min­ utes (default time is 360 minutes or 6 hours). Anyway, this is another good idea from Agilent. Triggering If nothing else, the triggering facili­ ties on the 54600 series are outstand­ ing. If you can’t get a stable display with this scope, you’re not likely to do so on any other machine. We’ve already mentioned the illu­ minated buttons which bring up the triggering options: Edge, Pulse Width, Pattern & More. “Edge” triggering al­ lows you to trigger the scope timebase on the positive or negative edge of pulse waveforms. The “Pulse Width” button allows you to select triggering on positive or negative pulses and to select the pulse width, greater or less than a variable value or between two variable values. “Pattern” allows for triggering on a logical AND condition of selected channels and each channel may be high (H), low (L), don’t care (X) or a rising or falling edge. Thus you can Fig.3: this low level 90kHz burst signal is the source of the frizzle on the waveform of Fig.2. Expanding out the waveform and using the frequency measurement facility showed that it was ringing at frequencies up to around 100MHz. set up triggering on a whole range of pulse/channel combinations. By the way, you also have the option of inverting the channel 1 & 2 inputs, giving a further range of conditions. The “More” button brings up an advanced trigger menu encom­passing TV, Sequence, I2C (for I2C bus signals) and Duration. TV triggering is really impressive as you can set triggering on NTSC, PAL, PAL-M, SECAM or generic TV waveforms, using positive or negative sync polarity. You can trigger on field 1 or 2, all fields, all lines or you can select a particular line number in field 1 or 2. Inter­estingly, the maximum line count catered for in a generic TV waveform is 1024. Agilent have certainly catered for every pos­sibility. Sequence triggering is pretty fancy too. In this mode the scope looks for a specified pattern and then triggers on another specified pattern. You can even specify a reset event. Duration triggering lets you trigger on an AND logic combi­nation of in­ puts with a duration greater or less than a specified value, with or without a timeout and so on. If you think about it, the specified triggering possibilities are enormous. Measurement options As you might expect, the measure­ ment options on the Agilent 54600 series are pretty wide as well. You can initiate measure­ments by push­ ing the “quick meas” or “cursor”. In Fig.4: monitoring video waveforms is dead easy. Here we’ve fed in a signal from an NTSC colour camera and with line 238 selected from field one. The line frequency comes up as 15.65kHz but we assume that the error (15.65kHz versus a correct 15.75kHz) is due to noise in the measurement. cursor mode you can have vertical (X) or horizontal (Y) cursors. For vertical cursors, the scope displays the time difference between the two (ie, along the X (time) axis) and also displays the reciprocal value which is the frequency. When horizontal cursors are set, the scope displays the voltage difference between the two. By the way, you can also have binary and hex cursors – the mind boggles. Pushing the “quick meas” button allows you to select time (frequency, period, rise & fall times, etc) or am­ plitude measure­ments (RMS, average, peak-peak, etc) from five menu pages. Only three measurements can be dis­ played at a time though, which is a bit of a downer. Mathematics While it may not be thought of as part of the measurement features, the mathematics of the 54600 series are particularly strong. Not only can you do standard operations like multiply­ ing channel 1 by channel 2 or taking the sum or difference, you can also get a plot of the derivative (ie, dV/dt) or integral (∫ V.dt) of the analog channels. The integral is calculated using the “trapezoidal rule” (now there’s a blast from my distant past – I was surprised that I could even remember it!). And there’s FFT (Fast Fourier Transform) analysis which can be done with Hanning, Flat Top or Rec­ tangular window settings. The number of points in the FFT is fixed at 2048 while the noise floor ranges from -70dB to -100dB depend­ ing on the level of averaging. High resolution display On the face of it, the display of the new Agilent scope is not anything out of the ordinary; the screen resolution is quoted as 255 vertical by 1000 hori­ zontal points (for the graticule area) but, and it is a big BUT, there are 32 levels of grey scale. However, there are several factors which lead to a scope display which is far better than you would expect, even given that there are 32 levels of grey scale. First, there is the very high sampling rate, even at low timebase speeds, and second, there is the deep memory. Third, the acquisition system has a vertical res­ olution (linearity) of 12 bits. Combine all of this with Agilent’s so-called MegaZoom tech­ n ology and the result is a display which is far better than most digital storage scopes. Not only that but it is better on most signals then even the best analog scopes. No longer do you see jitter on “smooth” signals such as sinewaves, due to the sampling system being tricked by noise. Instead, if there is noise superimposed on a signal, you see a smooth trace with lower intensity “frizzle”. The waveform of Fig.2 is a classic example of this. Here is a 100kHz sinewave with a small amount of noise superimposed. Look December 2000  29 Fig.5: using the same signal as in Fig.4, we wound up the time­base to 1µs/div, and set the horizontal position knob to bring up the colour burst. A rock steady waveform is the result. closely and you will see the frizzle. We were a little puzzled when we first saw this frizzle because it came from our low distortion sine/square oscillator described in the January & February 1990 issues. And since Fig.6: to measure the colour burst frequency on the signal of Fig.5 we wound up the timebase setting on the Agilent to 100ns/div, took a single-shot trace and the result is at the bottom of the screen: 3.584MHz. we thought our oscillator was above reproach, we initially thought that the scope was wrong. Looking further, we turned the oscillator’s output down to zero and the signal was still there. Even when we turned the oscillator off, the noise was still there. Not only that but we were able to obtain a stable triggered display with all sorts of spikey noise on it. Fig.3 shows the result; a low level 90kHz burst signal which is apparently radiated by our 100Mbit/s Cat.5 network (and it has a CE mark too – so much for EMC compliance!). The point is that the Agilent scope had no trouble seeing this noise. By the way, we got that scope trace with an averaging setting of one; the Mega­ Zoom cleans up the signal while still leaving the noise detail in there. We also measured the waveforms from a 555 circuit recently featured in the magazine and it was interesting that the 54622D showed up the spikes at the crests of the sawtooth; these are not visible on other digital scopes. TV waveforms The 54622D mixed signal oscilloscope has a special input socket for the 16 digital channels and these are brought out to two pods, each of which has nine leads – ground and eight probe leads. These can be connected to header pins on PC boards, or when probe tips are attached, connected to individual IC pins. 30  Silicon Chip The 54600 is particularly good on TV waveforms, because of its com­ prehensive sync facilities, including line number selec­tor. Fig.4 gives an indication of this. Here we’ve fed in a signal from an NTSC colour camera and selected line 238 from field one. As you can see, on the scope picture, we’ve measured the line frequency and video amplitude. The line fre­ quency comes up as 15.65kHz but we assume that the error (15.65kHz versus a correct 15.75kHz) is due to noise in the measurement. Want to measure the colour burst frequency? Easy. Just increase the timebase to 1µs/div and wind the horizontal position knob to bring up the colour burst. Fig.5 shows the Fig.7: using a 100MHz analog scope on the signal of Fig.5, this was the best result we could get. Since the analog scope did not come with a TV line selector there was no way to get the rock steady and bright waveform which is so easy to obtain with the 54600 series. result. This ability to display TV wave­ forms was so good that I thought I might crank up our workhorse analog scope, a Kenwood 100MHz model which I’ve always regarded as being fairly good in the sync department (for an analog scope). Anyway I fed in the same waveform and by judicious use of the main and delayed timebases I was able to obtain the waveform shown in the photo of Fig.7. It’s not bad but since the old faithful Kenwood analog scope did not come with anything as fancy as a TV line selector there was no way I was going to get the rock steady and bright waveform which is so easy to obtain with the 54600 series. Oh, we wanted to measure the col­ our burst frequency, didn’t we? Wind up the timebase setting on the Agi­ lent to 100ns/div, take a single-shot trace and the result is Fig.6. There is the measurement at the bottom of the screen: 3.584MHz. It should be 3.579MHz (alright, 3.579545MHz, to be precise but who is count­ing?). We could go on but by now you should have the overall picture. The Agilent 54600 scopes are delightfully easy to drive and the overall features in the package are very comprehen­ sive. Sure, it doesn’t feature runt triggering but you could play with sync settings to achieve a given result if you want. We think the 54600 series will be a real winner. Agilent have had plenty of time to think through all the features on this scope and the competition will be scrambling to match them. For further information on the 54600 series you can check the web­ site at www.agilent.com or phone 1 800 629 485. Pricing on the 54600 series starts at $4300 plus GST for the 54621A 60MHz 2-channel model, ranging up to $8514 plus GST for the full-featured 54622D 100MHz mixed signal oscilloscope complete with two analog probes and the digital SC probing pods. Own an EFI car? Want to get the best from it? You’ll find all you need to know in this publication  Making Your EFI Car Go Harder  Building A Mixture Meter  D-I-Y Head Jobs  Fault Finding EFI Systems  $70 Boost Control For 23% More Grunt  All About Engine Management  Modifying Engine Management Systems  Water/Air Intercooling  How To Use A Multimeter  Wiring An Engine Transplant  And Much More Including Some Awesome Engines! AVAILABLE DIRECT FROM SILICON CHIP PUBLICATIONS PO BOX 139, COLLAROY NSW 2097 - $8.95 Inc P&P Call (02) 9979 5644 9-5 Mon-Fri with your credit card details! December 2000  31