Favourite DSO user interface

[Howardlong]

Folks

What's you favourite DSO UI?

Some DSOs are real pigs to operate, and yet others are a piece of cake. I've probably used a dozen different DSOs at any length, and this is about ease of use of the UI, not really about features.

My favourite is the Agilent 54642D/54641D/54622D series from the early 2000's. They boot up even more quickly than a CRO (<5s), the UI is lightning fast to respond, high resolution (horizontally) screen, everything about it is intuitive. The only minor irritant is the splash screen that comes up at boot but you can clear that easily enough.

The least favourite is the Tek TDS460A, the UI is very slow to respond, slow boot time, shared vertical channel controls, finding things in the menu system often ends up being a case of random button pushing. Everything feels like a fight.

Another "pig" I've fought with was the HP 54500 series (single rotary knob), I was on a customer site about ten years ago and this was what I was offered, I was under time pressure and having to learn this scope didn't help. I'd have preferred an honest old CRO to be honest. Looked the part on the bench though.

I have a 10yo 54831D Agilent Infiniium here too which is great from the point of view of features and performance, but is less than stellar for making quick measurements. Boot time is horrendous, 2mins 15secs to the scope screen.

Here are some (non-exhaustive) measurement criteria:

o Ease of access to functions (eg, not buried in menus, separate channel controls)
o Intuitive operation
o Speed of UI in terms of response to user input and update rate
o Reasonable screen resolution
o Boot time

[Wuerstchenhund]

What's you favourite DSO UI?

These days, LeCroy's MAUI without doubt. IMO the best UI of any advanced scope.

Another "pig" I've fought with was the HP 54500 series (single rotary knob), I was on a customer site about ten years ago and this was what I was offered, I was under time pressure and having to learn this scope didn't help. I'd have preferred an honest old CRO to be honest. Looked the part on the bench though.

Strange because I loved these scopes. I used a 54510A at that time, and I loved the single rotary knob, as it was easy to use and left enough space for a full numerical keypad (something I really miss on modern scopes). And being able to type in a setting via the numerical keypad was great and much quicker than fiddling with separate rotary knobs as on other scopes.

When we later got 54542As I was quite disappointed that HP moved to the conventional layout with independent rotary knobs. They still were great scopes, though.

I have a 10yo 54831D Agilent Infiniium here too which is great from the point of view of features and performance, but is less than stellar for making quick measurements.

I never liked the 54800 Series. The first ones (running Win95) were terrible scopes and so unreliable that HP had to take lots of them back (although most issues have been fixed in subsequent models running Windows 98 and later).

The way the 54800 Series was designed isn't great, either (the scope app can't even run in windowed mode on the Windows desktop as it has to run on a separate and very slow graphics adapter), and the waveform update rates are really abysmal poor (around 2100 wfms/s on the faster models). It also has a small (8.4") low res (640x480) display which doesn't help for displaying lots of measurement data.

The worst thing however is that Agilent has used the poor 54800 UI as base for the UI of their later Windows scopes, instead of designing a proper one with touch support from ground up.

Here are some (non-exhaustive) measurement criteria:

o Ease of access to functions (eg, not buried in menus, separate channel controls)
o Intuitive operation
o Speed of UI in terms of response to user input and update rate
o Reasonable screen resolution

You should add

o Reasonable screen size

to the list. A high resolution is pretty useless if the screen is so tiny that data is barely readable.

o Boot time

Really, who cares about boot time? Especially when the specs of almost every scope out there tell you to wait a certain amount of time (usually between 10 to 30 minutes) after cold boot so that the scope is within its specs?

If a few minutes of boot time make that much of a difference then the real question is if a Windows scope is the right tool here. Windows scopes are useful for complex measurements, which usually take some effort to set up and can't (and shouldn't) be rushed. If it has to go fast then a scope with an embedded OS like the DSOX2k/3k or the Rigol DS1kz or DS2kA are probably a better choice as they boot in a few seconds.

[Howardlong]

I've never used a LeCroy ever, but I wouldn't mind having a go to see what they're like.

The problem I had with the 54500 was mainly that I was under the gun to get something fixed, I was working in Noordwijk at ESA's ESTEC R&D facility and was working on some BER measurements on a satellite downlink that was being integrated in the clean room. Being presented with this scope wasn't great, I had to spend valuable time figuring out how to drive it. I am sure if I had had time things would have been better, but we were very time constrained due to project pressures and others wanting to test their bits and pieces too. I have a dear old HP 1980b that is a "one-arm" scope here in the lab for a bit of retro, I used to yearn after this scope back in the day when it was a couple of years' salary. I quite like it now, but it did take quite a while to get used to that UI.

Boot time to me is pretty important, even though it might not be on top of your list. For almost everything I do (mixed signal embedded with RF) I just don't need better than 1 or 2% accuracy on Vp-p and whatnot, it's seeing if a signal's there or not. Timebase-wise, it's not like we're on free-running oscillators anymore, they're all quartz based (at least). If I need an accurate frequency measurement, any better than about 1000ppm, I generally wouldn't be using a scope to do it anyway. I can't imagine having to start work in the morning and have to wait half an hour to do anything. Seriously speaking, do you really regularly and studiously wait 10 minutes to half an hour before using a scope for it to warm up? In edge cases possibly, but in general, no. Happy to be corrected, but this is largely a hangover from the analogue days.

The 54622/41/42 series boot up in less time than it takes you to hook up the probe, so it's pretty convenient for your workflow. Having to sit around while your mind inevitably wonders onto other things simply isn't the best way of working for me.

I agree that Windows isn't the best UI for a scope a lot of the time, but there are some benefits, mostly that extracting data directly, or even fiddling with the data directly on the scope, can be pretty convenient.

The biggest screen I have used for a scope is 8.5", although obviously these days there are bigger, particularly in the five figure scope department.

[tautech]

Seriously speaking, do you really regularly and studiously wait 10 minutes to half an hour before using a scope for it to warm up? In edge cases possibly, but in general, no. Happy to be corrected, but this is largely a hangover from the analogue days.

No, full warmup is still today required for a DSO to be sure it is within spec.

Just look at any SM, calibration is to be done after a minimum warmup of 20 minutes.

However you are quite right, few, if any of us would wait for full warmup before use and as you say mostly for simple measurements, precision is not always really necessary.

[Hydrawerk]

Agilent 54600 series was nice, but I prefer DSOX2000 nowadays.

[Wuerstchenhund]

The problem I had with the 54500 was mainly that I was under the gun to get something fixed, I was working in Noordwijk at ESA's ESTEC R&D facility and was working on some BER measurements on a satellite downlink that was being integrated in the clean room. Being presented with this scope wasn't great, I had to spend valuable time figuring out how to drive it.

Understandable, being confronted with it's somewhat different UI when under stress probably isn't the best thing. It really took a while to get used to it.

Boot time to me is pretty important, even though it might not be on top of your list. For almost everything I do (mixed signal embedded with RF) I just don't need better than 1 or 2% accuracy on Vp-p and whatnot, it's seeing if a signal's there or not. Timebase-wise, it's not like we're on free-running oscillators anymore, they're all quartz based (at least).

If boot time is important, then you should seriously consider getting a DSO running some embedded platform. Windows scopes are notorious for their extended boot times, although SSDs and faster disk interfaces have brought this down quite a bit.

If I need an accurate frequency measurement, any better than about 1000ppm, I generally wouldn't be using a scope to do it anyway. I can't imagine having to start work in the morning and have to wait half an hour to do anything. Seriously speaking, do you really regularly and studiously wait 10 minutes to half an hour before using a scope for it to warm up? In edge cases possibly, but in general, no. Happy to be corrected, but this is largely a hangover from the analogue days.

Yes, I do, and that's pretty common in all the labs I work with. Of course people sometimes take a scope and use it instantly for stuff similar as you describe (i.e. just to see if there's any signal), but they're well aware that if you want your test kit to be within specs you have to wait until its warmed up, no matter if it's an old analog scope or the latest high-end DSO. The only difference is that nowadays you're waiting 10 to 15 minutes only while in the old days it was more 30+min (and sometimes even more than an hour).

The only scopes that I'm aware off that are within specs right from the start are LeCroy scopes (aside from the low-end LiteRunner/WaveAce/WaveJet) as they regularly do thermal self calibration of their analog front end, which compensates for temperature differences. These are also the only scopes that fully remain within specs if used outside the standard test environment (i.e. lab with stable 21deg C temperature), which for other scopes is the only environment where spec compliance is guaranteed by their manufacturers.

The biggest screen I have used for a scope is 8.5", although obviously these days there are bigger, particularly in the five figure scope department.

I found 8.4" VGA just fine for scopes with a purely menu driven UI like the old LeCroy WaveRunner LT Series, but it's pretty small for a Windows scope. LeCroy used 10.2" SVGA displays when they came out with their first Windows scopes (WavePro 7k, WaveMaster 8k) in around 2001, and aside from the WaveRunner 6k(A) which uses an 8.4" SVGA display (which for me is also too small) has stayed with larger displays for all their Windows scopes since.

[T3sl4co1l]

My favourite is the Agilent 54642D/54641D/54622D series from the early 2000's. They boot up even more quickly than a CRO (<5s), the UI is lightning fast to respond, high resolution (horizontally) screen, everything about it is intuitive. The only minor irritant is the splash screen that comes up at boot but you can clear that easily enough.

The least favourite is the Tek TDS460A, the UI is very slow to respond, slow boot time, shared vertical channel controls, finding things in the menu system often ends up being a case of random button pushing. Everything feels like a fight.

Ding, same here.  I used a 54622D for a few years.  Everything is just, there, and works.  They don't shy away from the display being a digital data display, because that's what it is.  Yes, it responds somewhat differently than a CRO; it should!  Trying to fake CRO behavior bothers the hell out of me.  It's only been in recent years that DPO approaches have finally become reasonable enough to do it.

Tek DSOs from beginning until recently have been just pigs.  It literally takes me three seconds just to change the trigger polarity on my TDS460.  My 475 does the same in a single switch-flip! (Well, when said switch isn't feeling flaky and I need to cycle it a bunch.  Blah, bad contacts.  Lots of bits I need to replace in that -- okay another pet peeve, CROs in bad condition -- anyone?)

Or are the recent Tek offerings any better?  I know the DPO2000s and MSO4000s sucked.  But I don't keep up on what's new in those lines or others.  And frankly, don't care...

My next purchase is probably going to be a moderately aged, 4 channel 1GHz, color LCD, Agilent.  When I have the need, and the money, of course.

Tim

[nctnico]

I have used about 10 to 12 different DSOs over the years (never a real Lecroy though). I can't say I prefer one over the other. Every DSO has it's pros and cons. It depends whether the designed and you are on the same wavelength. For example: what I like about the old Tektronix TDS500/600/700 series is that you can select between the left/right or top/bottom cursor with a dedicated select key on the front panel when in cursor mode. On other oscilloscopes you have to cycle through 3 options (or more) to get back to changing a cursor.

IMHO the more features a DSO has the more of those are hidden somewhere inside a menu so it doesn't make much sense to say a simple to operate DSO with less features is better than a DSO with a more features.