Is Bandwidth/memory depth a waste of money in oscilloscopes?

[rf-loop]

The time acolytes have great interest in measuring jitter of the 1 pulse per second timing output available from GPS receivers.  This signal is almost always synchronous to the GPS receiver clock which is asynchronous to GPS time itself producing jitter equal to the clock period.  This jitter limits precision of a GPS disciplined oscillator over medium timescales unless it is corrected.

[/quote]

I have done it several times. Never even thinked to use any kind if oscilloscope trigger delay for it. When I have old times done these things I have many old Tektronix 7000 series analog Rolls-Royces with couple of plug'ins including different time bases with normal simple delay up to trig after delay after delayed trig. One Tek have even real independent beams and both beams have dual timebases with delayed trigs. Of course after then there was also more low grade Tek's starting from some older and many 2000 series analog scopes. With this and that delayded time bases. Also some more or less boat anchor DSO's was under and over tables.
Dual independent timebases A and B and  Trig B after delay after trig A is rare in cheap DSO's

But then...

When I come to 1pps jitter measurements it was never even in mind even one half of second  that I use oscilloscope  delay after trigger or trig and delayed sub trig and so on  for this purpose. 

Even simplest cheap DSO can do it without any kind of delays after trig or dual time base with this and that delayed things.
In any case result quality depends trigger delay time jitter + trigger itself jitter. 
Result quality is depending scope itself timing quality.
But, it is many times more easy to trust some known external reference what have some known jitter (all have) or just so much better that no need think.

Here in very simple example image is 1pps (yellow) and 10MHz very good reference with quite low jitter and drift in observation time.  Just do simple things using simple methods. Here I do not need think so much how is possible scope own jitter. THis result can easy trust (enough).
Of course I can use 1s delay after trig but then I need know how much jitter this 1s delay may have. With simple xtal clocked scope... this jitter is so huge that for this kind of measurements using scope delay alone is just garbage collections, waste of time and even more crap if we go to old analog sweep timing scopes (only reliable method is compare SUT and known reference signal.
What scope have reference clock quality what is needed here if use 1s delay time. If then need use external reference for accuracy it can also use with cheapest DSO. Even with scope what have external reference input I will not use it for this purpose but connect reference to other input channel and 1pps to other channel and simply compare and I feel I can trust  result much better. 

Why make simple things complex.

And as can see it also do not need long memory. 14 sample points memory is enough for this.
And in this case, in history buffer I can look also every single acquisition if I'm interested to look ecah one separately.

So its is how we do things. Some need complex things for simplest things and many times result may be then "reliable looks like" complex garbage.

[GeorgeOfTheJungle]

Actually, the sound of that Delco is quite good!

Should I be using half-inch copper pipe to ground my loudspeakers or is 8-gauge wire good enough?

Only oxygen free copper-silver alloy.

[GeorgeOfTheJungle]

Yeah, 3 channels is better than 2, 4 is better yet, etc. But the point is you don't need them as in: two is enough to do it perfectly and comfortably. Perhaps the things 4 channels advocates are saying they want to look at are not what you really need to look at to fix these things. For example, you don't need to peek at the two outputs at once. Another one: you can look first at speaker + then at speaker -, there's no need to look at both at once. Etc.

[bd139]

If it's a differential output, you might not be able to see the differential signal due to the common mode signal being large. So no that isn't the case.

Even my shitty little £20 car head unit has differential outputs.

[Pete F]

I highlighted the relevant bit you forgot you said.

"That's a point worth mentioning: To look at amplifier outputs, etc. you need to connect a pair of probes and look at the difference between them. For stereo that means a four-channel oscilloscope."

If it's a differential output, you might not be able to see the differential signal due to the common mode signal being large. So no that isn't the case.

Even my shitty little £20 car head unit has differential outputs.

 

I'm fairly certain I said "might", not "need" when referring to four channels.

The only time I said "need" was when referring to two-channel, differential measurements. Loudspeakers, etc., are inherently differential devices. A differential measurement is the only way to measure the signal correctly when you get to that point in the circuit.

If you're doing it differently then I'm going to be blunt: You can do it better.

While you two guys were doing this "occasionally" and pontificating about how the pros could "do it better", I was earning my living from it and can assure you, you are talking out your clacker! All due (blunt) respect of course

Like I said, if you're designing or developing audio components, then sure, you'll probably want as many 'scope channels as you can get as you'll be tweaking things. However if you're repairing and fault finding faulty systems you barely need 2 channels, and talk about faults causing common-mode faults that can't be otherwise easily seen with a standard 2 channel oscilloscope is frankly ridiculous.

[Pete F]

To get this thread back from theoretical Fantasy-Land to the topic. I do like the Siglent SDS1202X-E however it's not available for the moment. I found the fault with the 2215 but will need to order the part, no idea how long that will take. It seems to me the SDS1202X-E has some features that aren't found on other oscilloscopes at the same price range (eg serial decoding standard, deep memory, etc). Is that an accurate conclusion or are there others that have exactly the same features? In relation to the thread topic, it's 200 MHz and 14 M memory from memory.

Cheers!

[GeorgeOfTheJungle]

They know very well that to see a differential signal you only need one channel not four. And if you can float the DUT thingy you don't even need to float the scope's GND.

[GeorgeOfTheJungle]

To get this thread back from theoretical Fantasy-Land to the topic. I do like the Siglent SDS1202X-E however it's not available for the moment. I found the fault with the 2215 but will need to order the part, no idea how long that will take. It seems to me the SDS1202X-E has some features that aren't found on other oscilloscopes at the same price range (eg serial decoding standard, deep memory, etc). Is that an accurate conclusion or are there others that have exactly the same features? In relation to the thread topic, it's 200 MHz and 14 M memory from memory.

Admittedly, those look good on paper, so I bought one but had to return it because it couldn't decode properly at low sweep speeds. Siglent came with a FW update that "fixed" the problem I was having: their solution: disable decoding at low sweep speeds. Huh.

Of the el cheapo brands I've tried, I only keep the hantek 5072p, it's a wonderful cheap scope, IMHO!

[rf-loop]

To get this thread back from theoretical Fantasy-Land to the topic. I do like the Siglent SDS1202X-E however it's not available for the moment. I found the fault with the 2215 but will need to order the part, no idea how long that will take. It seems to me the SDS1202X-E has some features that aren't found on other oscilloscopes at the same price range (eg serial decoding standard, deep memory, etc). Is that an accurate conclusion or are there others that have exactly the same features? In relation to the thread topic, it's 200 MHz and 14 M memory from memory.

Admittedly, those look good on paper, so I bought one of these but had to return it because it couldn't decode properly at low sweep speeds. Siglent came with a FW update that "fixed" the problem I was having: their solution: disable decoding at low sweep speeds. Huh.

Of the el cheapo brands I've tried, I only keep the hantek 5072p, it's a wonderful cheap scope, IMHO!

I have not available 1202XE now but using 1104XE
just running 10s/div UART decode. 300baud, bit, 1 stop, no parity
But, limiting factor is decode buffer length.
In this case with continuous data stream record acquistion length is 140s. But it can decode "only" 100s what include 3000 bytes (Full duplex both Tx and Rx have 3000). But I think it can say that 10s/div is somehow "slow sweep speed". 

like example here: (note, decode table time is related to trigger time position)

[David Hess]

The time acolytes have great interest in measuring jitter of the 1 pulse per second timing output available from GPS receivers.  This signal is almost always synchronous to the GPS receiver clock which is asynchronous to GPS time itself producing jitter equal to the clock period.  This jitter limits precision of a GPS disciplined oscillator over medium timescales unless it is corrected.

I have done it several times. Never even thinked to use any kind if oscilloscope trigger delay for it. When I have old times done these things I have many old Tektronix 7000 series analog Rolls-Royces with couple of plug'ins including different time bases with normal simple delay up to trig after delay after delayed trig.

A time interval counter would be a better choice for this measurement but they are less common than oscilloscopes and older ones do not provide any statistical information.

Tektronix made a special plug-in, the 7D11 digital delay, to allow the 7000 series mainframes to make exactly this sort of measurement.

I used to use an oscilloscope to make this measurement all the time when doing embedded programming to measure the latency and execution time of interrupt service routines and for profiling.  The histogram display produced by an oscilloscope can be very enlightening.

Even simplest cheap DSO can do it without any kind of delays after trig or dual time base with this and that delayed things.

The problem as I outlined is that some DSOs cannot do it even though they should be able to.  The cheap Rigols cannot delay past the acquisition record so a slower time/div has to be used to capture the second pulse limiting the sample resolution.

In any case result quality depends trigger delay time jitter + trigger itself jitter. 
Result quality is depending scope itself timing quality.
But, it is many times more easy to trust some known external reference what have some known jitter (all have) or just so much better that no need think.

In this case it is not the jitter of an oscillator which is being measured so it is not one oscillator versus another oscillator.  It is the asynchronous products of two oscillators, the GPS receiver clock and GPS time, being measured so even a mediocre oscilloscope sample clock is sufficient.  An analog oscilloscope would need some help like with the 7D11 digital delay mentioned above or a DD501.

This is the "hanging bridges" issue with GPSDOs.  When I was using my best time interval counter to make this measurement (1 nanosecond), I could read the outside temperature which affected the GPS receiver clock from the jitter.

I'm taking a look at the 2215 in between about a billion other things and have to laugh at the irony. The fault would be a perfect opportunity for a DSO as the PSU is starting but then blowing the fuse. I'd love to capture that start up. I'd like a DSO to repair an oscilloscope that I'm replacing with a DSO so I can then throw out the repaired oscilloscope. Something is definitely wrong with that picture 

That is a common issue with repairing analog oscilloscopes which power supply problems.  Digital storage oscilloscopes, even low performance ones, are particularly useful in diagnosing power supply startup problems.  Below is a photograph of my 2232 and 2440 DSOs measuring a couple of signals from my working 7904 switching power supply during startup.  Two things of note are the relatively low noise in the 2440 display despite peak detection and full bandwidth and even more so, the absence of noise in the second trace of the 2232 display despite peak detection and full bandwidth.  I did not notice the 2232 issue at the time or I would have made a second capture showing exactly what witchcraft is going on there.

[Fungus]

The cheap Rigols cannot delay past the acquisition record so a slower time/div has to be used to capture the second pulse limiting the sample resolution.

Yes they can. As shown earlier I can set a 1s delay at any sample rate and up to 500s at lower sample rates.

One second is 40x the length of the buffer at the maximum 1GSa/sec.

[wpwrak]

One warning about the SDS1202X-E: there are some features you may find interesting Siglent only put in their 4-channel models, e.g., remote control from a Web browser. (And, judging from videos posted here, that Web interface seems to be very well implemented, with satisfyingly fast waveform updates.)

Whether having four channels would be a real benefit depends also on what sort of digital work you expect. You can debug I2C with two channels, but already SPI is painful. And many "analog" circuits have digital components these days. I have an ancient Rigol DS1102CD and every once in a while I found myself forced to use the - abysmal (in this particular model) - "logic analyzer", just to be able to see what's happening on some bus. With four analog channels, that may not have been necessary.

I'm also planning on buying a new scope and based on specs and research people in this forum have been doing, I'd also consider the Siglent SDS1000X-E series today's best overall performers. Go much lower and you may encounter unpleasant surprises. Go higher and prices skyrocket. I also appreciate that Siglent include protocol decoders, instead of either making them an - IMHO - grossly overpriced option or tempting customers to do something they ought to find unethical. Besides, having a bit of headroom never hurts.

- Werner

[Pete F]

Thanks to the recent few contributors, that's interesting stuff. The Tek 2215 was pretty easy to diagnose and the part should be here, well, probably not until Monday now. So hopefully I'll get that back. It still has some other issues that make it more difficult to use, but hopefully it will at least provide something.

The comment about 4 channels and diagnosing digital is interesting, and not something I would have considered. While I studied digital in college it's not something I really enjoyed (ironically I was much later offered a job as a Field Service Engineer with Digital DEC at one stage of my career though, go figure!). However with digital of one form or another now being embedded in even the most mundane of applications, it's certainly a good point. However having pushed the budget from an entry level Hantek DSO5072P (which I'm quite sure would do the job for me perfectly well) to a Siglent SDS1202X-E (at twice the price), I'm not going to push it any further. I can't imagine myself ever wanting to control this by web browser etc.

[wpwrak]

However having pushed the budget from an entry level Hantek DSO5072P (which I'm quite sure would do the job for me perfectly well) to a Siglent SDS1202X-E (at twice the price), I'm not going to push it any further. I can't imagine myself ever wanting to control this by web browser etc.

I was thinking of the 1104X-E, which should be "only" about 30% more expensive than the 1202X-E. The 1204X-E would cost significantly more, almost twice the price of the 1202X-E, but it seems that you should be fine with 100 MHz bandwidth anyway. Besides, the 1104X-E can apparently be "hacked" to support 200 MHz (which may or may not yield the same performance as a model sold as supporting 200 MHz, may or may not work in the future, and so on.)

Access from the PC should be convenient if the PC controls other parts of the system and isn't right on your bench. (E.g., when examining a configuration change.) Maybe someone with hands-on experience with the two scopes could explain better whether it's a worthwhile feature or just a gimmick. (Like WiFi dongle support, which is also limited to 4-channel models. Signal generator and LA options, and Bode plots (requires a signal generator) are also limited to 4-channel models, but I don't know if you'd care about any of that.)

- Werner

[Pete F]

Sorry I didn't really understand what you were getting at there Werner. Like I said, the price of the 1202X-E is about as much money as I want to spend on this. No more, as it's already much more than I originally anticipated. So I'm not sure quite what you're suggesting.

There is absolutely zero chance I will be examining a piece of equipment that isn't sitting right in front of me on my bench. Zero! I'm sure that feature would be handy for some people, but it's completely irrelevant for my application.

The parts for the 2215 arrived already, absolutely incredible service from RS Components I have to say. I used to use them years ago and they were always good, however I order the mosfets after business hours closing last night, and here they are sitting on my desk by lunchtime the next day, free postage too. Unbelievable!!!!

[tautech]

Access from the PC should be convenient if the PC controls other parts of the system and isn't right on your bench. (E.g., when examining a configuration change.) Maybe someone with hands-on experience with the two scopes could explain better whether it's a worthwhile feature or just a gimmick.

- Werner

For some it's no gimmick but a very useful feature, your requirements may vary.
The webserver offers not just remote control but a way to export the imagery onto other devices easily whereas in many more expensive DSO's that might have HDMI or VGA that only offer the display imagery to another visual device.
But of course it's all about 'need'.
Using the WiFi dongle you can access the webserver from another WiFi device such as a tablet, laptop or even a smart phone where if you wanted you can grab a screenshot and email it or text it to another.....all wireless !
We keep finding ways to use it. 

Here's a crappy vid I did on projecting the display wireless to a laptop connected to an 18" monitor but it could've easily been a big screen projector.

https://youtu.be/W30AqcDQL18

[Specmaster]

Your link seems to be dead?

[rsjsouza]

Based on your requirements, your price point and on the feedback others have been providing on modern entry-level oscilloscopes, I think the Siglent SDS1202X-E is your best option. It is a modern architecture, it works reasonably well (I played with it just for a brief time) and quite compact. One caveat: it may not be as durable (no track record yet) and certainly not as repairable as your Tek, but that is par for the course in such highly integrated test equipment.

I couldn't help but notice you used a Kenwood oscilloscope; I still have a CS4025 that is incredibly basic but it is so easy to operate and such workhorse that I keep coming back to it (I confess I tried to pass it along a number of times, but never had the courage to actually do it).

[Pete F]

I can't recall if the Kenwood is the same model as I had, but yes it was a no-frills but extremely well designed oscilloscope that always did precisely what it said on the packet.

To close the book on this whole affair, I finished up buying the Siglent SDS 1202X-E and am quite happy with it. The build quality as as good as one would expect and it seems to operate well. It's certainly not as intuitive to use as a basic analogue oscilloscope and I often find myself staring at it trying to recall basic functions that would be immediately obvious on any analogue 'scope. Obviously with more experience with it I'll remember the functions more easily, however they still require far more keystrokes compared to simply flicking a switch in most cases on an analogue machine. The light weight of the more modern oscilloscopes also makes them a bit of a PIA for constant use and selecting keys. I think if I was still doing this professionally I would probably prefer an analogue oscilloscope for general signal tracing, fault finding, etc. I've seen others refer to maintaining and using the two types of oscilloscope and I couldn't understand why, but now see that the two types really compliment each other rather than being a substitute. Having found the fault with the Tek 2215 I will definitely fix it and keep it as an alternative. Nevertheless for somebody just starting out and only having one oscilloscope, something like the SDS 1202-E would be ideal from what I've seen.

[Specmaster]

I can't recall if the Kenwood is the same model as I had, but yes it was a no-frills but extremely well designed oscilloscope that always did precisely what it said on the packet.

To close the book on this whole affair, I finished up buying the Siglent SDS 1202X-E and am quite happy with it. The build quality as as good as one would expect and it seems to operate well. It's certainly not as intuitive to use as a basic analogue oscilloscope and I often find myself staring at it trying to recall basic functions that would be immediately obvious on any analogue 'scope. Obviously with more experience with it I'll remember the functions more easily, however they still require far more keystrokes compared to simply flicking a switch in most cases on an analogue machine. The light weight of the more modern oscilloscopes also makes them a bit of a PIA for constant use and selecting keys. I think if I was still doing this professionally I would probably prefer an analogue oscilloscope for general signal tracing, fault finding, etc. I've seen others refer to maintaining and using the two types of oscilloscope and I couldn't understand why, but now see that the two types really compliment each other rather than being a substitute. Having found the fault with the Tek 2215 I will definitely fix it and keep it as an alternative. Nevertheless for somebody just starting out and only having one oscilloscope, something like the SDS 1202-E would be ideal from what I've seen.

Yes, analogue scopes are easier to use because of their rotary switches, toggle and stay put pushbuttons, you can instantly see what the scope is set at make adjustments. Digital ones controls cannot give you the same visual feedback because it's all hidden behind many layers of on screen menus. This is one reason why so many people advise noobs to get a basic analogue one first and then add an digital later after the basic functions are learnt.

They both have their place on a modern electronics bench IMO I personally have 2 analogue and 1 CombiScope on mine.

[nctnico]

I can't recall if the Kenwood is the same model as I had, but yes it was a no-frills but extremely well designed oscilloscope that always did precisely what it said on the packet.

To close the book on this whole affair, I finished up buying the Siglent SDS 1202X-E and am quite happy with it. The build quality as as good as one would expect and it seems to operate well. It's certainly not as intuitive to use as a basic analogue oscilloscope and I often find myself staring at it trying to recall basic functions that would be immediately obvious on any analogue 'scope. Obviously with more experience with it I'll remember the functions more easily, however they still require far more keystrokes compared to simply flicking a switch in most cases on an analogue machine. The light

This can be a limitation of the user interface on the Siglent. There are many other DSOs with complicated or simple user interfaces.

[Pete F]

I can't recall if the Kenwood is the same model as I had, but yes it was a no-frills but extremely well designed oscilloscope that always did precisely what it said on the packet.

To close the book on this whole affair, I finished up buying the Siglent SDS 1202X-E and am quite happy with it. The build quality as as good as one would expect and it seems to operate well. It's certainly not as intuitive to use as a basic analogue oscilloscope and I often find myself staring at it trying to recall basic functions that would be immediately obvious on any analogue 'scope. Obviously with more experience with it I'll remember the functions more easily, however they still require far more keystrokes compared to simply flicking a switch in most cases on an analogue machine. The light

This can be a limitation of the user interface on the Siglent. There are many other DSOs with complicated or simple user interfaces.

No it has nothing to do with the user interface in this case. "Specmaster" nailed it; there are many layers of functionality on a modern DSO, and the more capable, the more layers there will be. Contrast that to a conventional analogue oscilloscope and all the functions have their own switch/knob, with at worst a requirement to possibly pull a knob to get another function. If all the functions had their own knob/button/switch the DSO would finish up looking like the mixing console at Abbey Road Studios!  That's just the nature of the beast. I'm sure some brands are also easier to use than others however.

[rsjsouza]

Exactly this. The soft button paradigm dissociates the mechanical from the functional, aggregating functions formerly performed by precious levers, switches and potentiometers of the user's interface.

If you are comparing different DSOs, then the judgment of easeness and intuitiveness is quite valid. If you are coming from a CRO, all modern DSO interfaces will seem unintuitive.

[GeorgeOfTheJungle]

Case in point: it takes a single push of a button to recall a waveform in a tek 2232: MEMORY#, and it takes two to save it: SAVE + MEMORY#. In a modern DSO you've got to go quite deep into submenus to achieve the same, it's so cumbersome that often you simply end up not using waveforms.

[TheUnnamedNewbie]

So to bring the topic bang to the original question I'm keen to know how a (relatively) high bandwidth oscilloscope (let's say 100 MHz) would help somebody who generally works in the audio range of frequencies. Maybe clock freq, but that's about it at this stage. I personally can't see how 100 MHz has even the vaguest assistance, but that's why I'm here to ask.

The "MHz" in the bandwidth rating of an oscilloscope usually isn't related to the frequency of the signals you're looking at.

Any signal other then a perfect sine wave has harmonics. These harmonics can be much higher frequency than the base frequency. eg. Square waves have infinite harmonics, to see a perfect square wave on an oscilloscope screen you need infinite bandwidth. This is true even if you're only looking at a 1Hz square wave!

It's also true of zero Hz square waves. A single transition from low to high voltage may have infinite harmonics. You need a lot of bandwidth to see the detailed characteristics of a single rising edge. You may see discussion of "rise time" in forums like this, that's what it refers to. "Rise time" is a better measure of oscilloscope performance than "MHz" but rise time would confuse sales people (is a 4ns rise time twice as good as a 2ns rise time...?  )

So .... it doesn't matter if your clock freq has a base frequency in the audible range, you need a lot more than audio range in your oscilloscope if you want to look at it.

This seems a bit of a backwards statement. The 'MHz' (or GHz) in the rating is the 3 dB bandwidth of the system, and thus has everything to do with the signal you are looking at. A 1 MHz clock is not a signal with a 1 MHz frequency. Perhaps you are trying to put this into terms for beginners, who might indeed mix up the frequency domain thinking in pure (co)sines, but saying the MHz on the badge is not related will only confuse them more, I think.

You say ''rise time'' is a better measurement, but I disagree that this is true. I could care less about risetime in my applications. I care about Bandwidth - up to what frequency can I display waveforms. Sure, in digital applications the rise time may be easier to relate to the signals you are looking at, but this is not universally the case.