Oscilloscope buy --- Dilemma

[Gp!]

HI,
total newbie here. Just getting back into electronics (thanks to Dave's videos) and immediately stuck on a little dilemma.
So many scopes are available and at such good prices that's it's almost embarrassing to choose.
My dilemma however is slightly different, would it make more sense to buy a unit that perhaps is a little bit more expensive and offers more features so as to buy it just once so as to cover all the future necessities....or get the cheap realiable model and then upgrade to something more powerful in the future ?
Obviously I am not talking about 16000 euro version but in the range from your basic 350 euro model to your 2000 euro model (if that). Some of the new models that have just come out seem amazing but I am thinking that they might be overkill, but again could come in handy in the future...
Is it worth planing that far ahead ?
Thanks for any pearls of wisdom !
Gp

[rstofer]

It's always possible to start at a higher level and there's nothing wrong with that.  Bandwidth is everything!  But Dave has videos where he explains that other features are right near the top.  Channels, decoding, measurements, decent UI, all these are part of the equation.

OTOH, you can go a long way with electronics with a lower level scope which means you have money left over for other things:  Signal generator, soldering station, power supply, etc.  Few people can afford top-of-the-line models for everything.

Probably the fastest selling scope around is the Rigol DS1054Z which, when unlocked, provides a decent bandwidth of 100 MHz, decoding of various serial protocols and more measurements than I know what to do with.  It's a great scope to start with.

There are better scopes!  The thing is, anything that has much better features costs 2 to 3 times as much.  There is a new 200 MHz Siglent (dual channel) that may make some headway against the Rigol for the entry level kind.

There's a lot more information over on the Test Equipment forum.  Lots of opinions on beginning scopes and features.

BTW, you buy the Rigol DS1054Z and decide a few years down the road that you want a $2000 scope so you sell the 1054 for 1/2 price and you have gotten several years worth of education for a couple of hundred $.  By then you will know exactly what you want.

Or, you will come to the conclusion that the 1054 will do everything you need.

Still, there is no substitute for bandwidth and 200 MHz is more than 100 MHz.  Remember, for square waves, the fundamental frequency is only part of the waveform.  You need to be able to get to at least the 5th harmonic and the display is better if you can get to the 7th (and so on out to infinity).

A 200 Mhz scope may realistically display only a 40 MHz square wave.  A 100 MHz scope can only do a 20 MHz square wave.  Bandwidth is king!

I still have the 350 MHz Tek 485 that I bought about 13 years ago for about $200 on eBay.  It has no features whatsoever but it has bandwidth.  I added the 1054 to get features I never knew I needed - especially single shot.  There is no comparison of features between an analog scope and a modern DSO.

I use them both!

[stj]

the 200MHz Siglent actually samples at the same rate as the 100MHz Rigol.
i noticed that while deciding which to buy.

[tggzzz]

It sounds like you don't yet know what you will do with a scope.
Define what you want to test, and from that the minimum necessary specification.
Find the cheapest working scope that satisfies the specification. Cheapest means you will still have money for all the other things you need, e.g. probes, PSUs, sig gens etc.
If/when you find something you cannot do with that scope, you will know exactly what you need to find in your next scope, and why.

The single key parameter in a scope is bandwidth; for modern jellybean logic 100MHz is the minimum.

The single key advantage of a digital scope is the ability to capture single-shot events.

More complex and capable => longer learning curve, and (arguably!) more ways to mislead yourself.

Use a scope to measure analogue signals, e.g. those being generated by and received by digital logic gates. Once you know the signals are valid digital signals, switch to debugging in the digital domain, using logic analysers, protocol analysers, and printf statements.

[james_s]

the 200MHz Siglent actually samples at the same rate as the 100MHz Rigol.
i noticed that while deciding which to buy.

That's not really an issue except for single shot captures of high speed signals. Most signals are repetitive so you can get a far higher effective sample rate. You will notice the much higher bandwidth whether the sample rate is the same or not. Whether you *need* that bandwidth depends on what you're doing. I think the Rigol is a good starting point unless you have a specific need in mind but it sounds like you don't at this point. You will probably go a long time before you really need something better and at that point you might start looking at older 500MHz scopes if you really need more than 100MHz.

[Electro Fan]

HI,
total newbie here. Just getting back into electronics (thanks to Dave's videos) and immediately stuck on a little dilemma.
So many scopes are available and at such good prices that's it's almost embarrassing to choose.
My dilemma however is slightly different, would it make more sense to buy a unit that perhaps is a little bit more expensive and offers more features so as to buy it just once so as to cover all the future necessities....or get the cheap realiable model and then upgrade to something more powerful in the future ?
Obviously I am not talking about 16000 euro version but in the range from your basic 350 euro model to your 2000 euro model (if that). Some of the new models that have just come out seem amazing but I am thinking that they might be overkill, but again could come in handy in the future...
Is it worth planing that far ahead ?
Thanks for any pearls of wisdom !
Gp

Hi, and Welcome!  As of today, The Scope Du Jour is likely the Rigol DS1054Z but that's open to some debate and in any event it's different than your question.  Here is a chart to help think through the question(s).

[tautech]

GP
If you've had zero DSO experience prepare yourself for a shock. There is much more they can do than a CRO, even quite basic ones.
One can never have enough scopes so grab something cheap with specs you think you want/need and grab something more upmarket later on. Flick the first one you buy if you must, the Buy/Sell/Wanted needs more DSO's for beginners.
By then you will have some idea of what you really need and what you really don't want in a DSO and furthermore if you think you're spoilt for choice now watch the next year or so for things to get even more crazy.

[Andreas]

Hello,

no, bandwidth is not the only key feature.

especially when you want to zoom in, want to do long single shot measurements
 or find short glitches the sample memory size matters.

Most scopes also appear very noisy because of the 8 bit ADC. (and sometimes of a poor analog stage).
If you are working in the analog world you might want a higer resolution.
E.g. 12-16 bit ADC with specified noise floor level and defined filters for a reduced (20kHz) bandwidth.

For documentation purposes it might also play a role if you can do only screen shots
or transfer the whole sample memory to a PC for further evaluation.

With best regards

Andreas

[bitseeker]

Today's entry-level scopes are quite feature-filled and are likely to last you a while. Especially when you're a beginner, it's likely unnecessary to plan for significant future growth since you may not know where you'll be going in the field and, hence, what your requirements will be. You may even find that your interests change before you get deep enough into things to warrant higher-grade test equipment.

Thus, get what will fit you today. It's highly likely to work for you for a while. Then, should you outgrow it, you'll know much better what you really need and can buy it then. Since scopes these days are fancy computers, you'll also have more power, features, etc. at that future date.

Finally, if the need to upgrade never comes, then you've saved yourself $1000 or more that you can apply to a DMM, soldering station, components, or something else for your lab.

[rstofer]

the 200MHz Siglent actually samples at the same rate as the 100MHz Rigol.
i noticed that while deciding which to buy.

The samples per second is the same: 1GS/s
 
All Channels In Use

The 200 MHz Siglent splits the 1GS/s over 2 channels so 500 MS/s per channel divided by 200 MHz/channel -> 2.5 samples per cycle

The 100 MHz Rigol splits the 1 GS/s over 4 channels so 250 MS/s per channel divided by 100 MHz/channel -> 2.5 samples per cycle

The samples per cycle are identical, we are trading bandwidth for channels.  This 2.5 number has been a standard in the industry for a very long time.  Somewhat above the Nyquist rate.

One Channel In Use

The samples per channel change when only 1 channel is in use*.  I'm not sure about the Siglent but the Rigol will use the entire 1 GS/s on a single channel.  I assume the Siglent is identical.

The Siglent would have 1 GS/s divided by 200 MHz or 5 samples per cycle.  The Rigol would have 1 GS/s divided by 100 MHz or 10 samples per cycle.  So, we trade off bandwidth for samples per cycle.

I like my DS1054Z and it will be sufficient for a long time.  Not everything is way high in frequency and the features, now that they have been debugged for a couple of years, are worth the time to learn.  And there is a lot of learning to do.

Still, I'm looking at the Siglent SDS1202E-X and watching the comments.  If Siglent is diligent about upgrading the firmware, I might buy one just because...  Bandwidth is still king!

* The entire samples per channel thing is a little more complicated.  For 2 channels the Rigol drops to 500 MS/s and for 3 or 4 channels, it drops to 250 MS/s.  I don't know how the Siglent works but I would bet it works as 1 channel -> 1 GS/s, 2 channel 500 MS/s.

[rstofer]

Here are a couple of videos re: the new Siglent SDS1202X-E.  The second video is a detailed 'show and tell' which really demonstrates the modern DSO.




It is worth watching reviews re: any scope you consider buying.  There are a LOT of DS1054Z videos, just Google for them.

[David Hess]

It sounds like you don't yet know what you will do with a scope.
Define what you want to test, and from that the minimum necessary specification.
Find the cheapest working scope that satisfies the specification. Cheapest means you will still have money for all the other things you need, e.g. probes, PSUs, sig gens etc.
If/when you find something you cannot do with that scope, you will know exactly what you need to find in your next scope, and why.

I agree with tggzzz here.  If you lack experience to know what you want in an oscilloscope, then buy an inexpensive low end oscilloscope to learn on.  I would even recommend a used analog oscilloscope or DSO if it means cutting the price to one half or less assuming that you are in a position to maintain it.

The single key parameter in a scope is bandwidth; for modern jellybean logic 100MHz is the minimum.

The single key advantage of a digital scope is the ability to capture single-shot events.

I went the way rstofer describes; my fastest oscilloscopes are analog simply because a modern high bandwidth oscilloscope is an order of magnitude more expensive and no set of features can replace bandwidth.

But my "go to" oscilloscope is a Tektronix 2232 combination analog and DSO with 2 channels and 100 MHz of bandwidth.

[tggzzz]

no, bandwidth is not the only key feature.

I suspect something is being lost in the translation.

Bandwidth is the key feature. Bandwidth is not the only important feature.

especially when you want to zoom in, want to do long single shot measurements
 or find short glitches the sample memory size matters.

You are missing the fact that many very useful scopes manage to allow you to zoom in and see short glitches - without having long memories. (Key point: look for a "dual timebase", plus good triggering )

If you don't have sufficient bandwidth for your task, you are stuffed. If you don't have a long enough memory then there are ways around it; they are awkward and annoying but they are possible.

Most scopes also appear very noisy because of the 8 bit ADC. (and sometimes of a poor analog stage).
If you are working in the analog world you might want a higer resolution.
E.g. 12-16 bit ADC with specified noise floor level and defined filters for a reduced (20kHz) bandwidth.

True for digitising scopes. If you are doing audio work then many other aspects may become important, such as linearity and overload recovery.

For documentation purposes it might also play a role if you can do only screen shots
or transfer the whole sample memory to a PC for further evaluation.

You can always take a photo! (The aphorism in my .sig has analogies in the electronics world!)

The OP might care to note that analogue scopes are very very cheap (<£1/MHz) nowadays, to the point where they are often given away free (my local Hackspace has so many we'll be junking some soon). They would probably be sufficient for the OP to play around with, to find out what they really need.

[David Hess]

especially when you want to zoom in, want to do long single shot measurements
 or find short glitches the sample memory size matters.

You are missing the fact that many very useful scopes manage to allow you to zoom in and see short glitches - without having long memories. (Key point: look for a "dual timebase", plus good triggering )

And my favorite, peak detection during decimation.

Oddly enough dual timebases were not the only way to achieve magnification; they are rare but some oscilloscopes had single delayed timebases with the delay function being controlled by what is effectively a very limited and non-displayable timebase.  What distinguished these from true dual timebase oscilloscopes is that the later could almost always display both timebases simultaneously in one form or another.

[Specmaster]

Personally, I'd go with the advice that Dave gives, get your self a cheap analogue CRO scope to start with, especially while you're building up your scope knowledge. You'd be surprised just what you can do with a 20Mhz scope anyway. As your experience grows and you decide just what you want to do in electronics, then you can slowly upgrade in the scope dept to something that matches your requirements and then you can dispose of or keep your old scope, personally I keep them.

Do you really need to measure frequency, voltage etc accurately from your scope, when a reasonable auto ranging DMM can be purchased for reasonable money and secondhand but perfectly functioning frequency meters are also going on Ebay for reasonable prices?

I do mainly repairs to radios and audio equipment etc so accuracy of measurements on a scope are of less importance to me then the ability to see what the waveform looks like and also the ability to see if a signal is present at various points on a circuit.

[Brumby]

There are some good points being made here about various features and performance - but if you are a total newbie, these might be more confusing than helpful.

It would seem you have the financial means to choose from a more advanced range of scopes than the usual high school aspirants - but without the knowledge to know what to look for.

There is something to be said for grabbing an old, cheap CRO.  Yes, it will be limited in what it can do - but that can sometimes lead to creative approaches and problem solving that will teach you a lot more about the fundamentals than walking in with a box that has all the answers.  It's also a good start if you ever have the misfortune to do something which might destroy it.  The loss is bearable (even if the embarrassment is not).  On that note, make sure you watch this:



If you want to step out into something with a better feature set, I would suggest the Rigol DS1054Z as a good start.  Even if you move on to something bigger and better, you could still hang on to it as a backup - or assigned to less demanding work.

THE MOST IMPORTANT thing is to realise you won't know what you want/need in a scope until you actually start using one.  Something like the Rigol DS1054Z will set you up quite nicely to start with without breaking the bank.  It may adequately serve your needs for many years - or it may not.  You need to get your hands dirty to find out.

In time, you will get to know what features/specifications will be important to you and to know what questions to ask when you feel you need to step things up a notch.

[pitagoras]

IMHO Rigol DS1054Z is the best value for money, 4ch 100Mhz (hacked) at EUR 400. If you discover quickly you need something else, it will sell quickly (even here in the forum) as it is kind of the standard entry model now.

[FlyingHacker]

If you plan on moving up later, or using a scope professionally, you might want to consider the Keysight 1000X series. You would be learning on the same interface as the real, professional scopes. Thus, you would not have to relearn much to use high end scopes at work or school. Almost no additional training required from a prospective employer, etc.

They are also very nice scopes, with a very fast UI compared to the low end Chinese scopes.

As others have said, you can never have too many scopes. So buy more... I like to cascade the trigger from one scope to one or more additional scopes to get a buttload of channels all in sync. Stack the scopes vertically (works with old Tek gear for instance) and you can see all those channels lined up vertically. Makes you feel like a scope god to have three scopes and 8-12 channels.

[james_s]

If you can get it really cheap, an analog scope is a good way to get started and learn the basics of using a scope. I would not suggest paying much for one though, usually they can be found for around $50 with some patience, you might even find a free one.

[Gp!]

WOW ! Exactly what I was looking for !


 The fog has cleared !

Thank you very very much everybody for all the advice !

[Muxr]

If you can get it really cheap, an analog scope is a good way to get started and learn the basics of using a scope. I would not suggest paying much for one though, usually they can be found for around $50 with some patience, you might even find a free one.

Weird how prices on these fluctuate.. a year ago, you could barely find a decent analog scope for under $200-$300. Now I am seeing all sorts of deals for $100 or less.

Of course it's going to depend on the local market where the person lives.

[james_s]

That's just the nature of things, enough older DSOs have trickled down and low cost modern compact DSOs have become common enough that bulky old analog scopes and CRT based DSOs are out of fashion. I've had at least 5 TDS300 and TDS400 series DSOs and never paid more than $100 as-is for one of them. The key is to find auctions that are actual auctions rather than BIN prices from fantasy land. I've had success with the Best Offer thing too, toss in a lowball offer a few hours before the listing expires and more often than not it has been accepted. For a while I even saw several 500MHz 4 channel TDS500 series scopes sell for <$150 in working condition, that's a steal for a very capable instrument if you have the space for one.