Oscilloscope - used vs new - revival

[Fungus]

DS1052E has 1 Mpts which in their day was a revelation for a cheap DSO, then Siglent came out with the 2 Mpts SDS1102CM ....and so on until today.

The OP has already said he's looking for something cheap to keep him going until he saves up enough to get a real oscilloscope (moving straight past the SDS1xxx). 

[bestel]

Another option is a used Rigol DS1052E/1102E if you can get one cheap and in good shape. This also does not have deep memory (25kpts IIRC) but it has higher bandwidth (50 or 100MHz).

DS1052E has 1 Mpts which in their day was a revelation for a cheap DSO, then Siglent came out with the 2 Mpts SDS1102CM ....and so on until today.

Sorry, my memory keeps failing me. I'll shut up now. 

Don't as we each have something to offer those that are trying to get their heads around the complexities of the current DSO offerings let alone understand how a DSO works and the relationship between GSa/s and mem depth.

I found that I took "bad habits", I mean using DSO like analog scope : never zoom in / out, and retake a single shot for inspection.
But after I've done some research, I will totally consider memory depth for my next "real/pro" purchase.

For now I still search for a not overpriced old Tektronix... I also search for agilent/keysight, ... Maybe Rigol and Siglent ?
I saw a video on youtube titled "Oscilloscope performance vs. specifications", and this is also my feeling :
I see really good specs with Siglent/Rigol, even Owon and Unit-t, but how do they compare in real life?


In his video, we see that specs aren't the only parameter, because we just know the "maximum" rating, not the rating we'll have for our measure.
I will continue my research and post my thought here...

[Fungus]

I see really good specs with Siglent/Rigol, even Owon and Unit-t, but how do they compare in real life?

Owon and Uni-T aren't well regarded because the raw numbers on the sales brochure don't tell the whole story. Firmware with a decent amount of features that that isn't constantly locking up counts too.

[tautech]

I found that I took "bad habits", I mean using DSO like analog scope : never zoom in / out, and retake a single shot for inspection.
But after I've done some research, I will totally consider memory depth for my next "real/pro" purchase.

Good, understanding the features of what a good DSO offers is key to making good purchase choices.

I saw a video on youtube titled "Oscilloscope performance vs. specifications", and this is also my feeling

Well yes it's an old video however it does offer some good core DSO understanding but 7 years later today there are now offerings that were barely dreamed of in 2013.
Pricing has mostly remained unchanged but feature sets have improved immensely.

[Fungus]

But after I've done some research, I will totally consider memory depth for my next "real/pro" purchase.

Lots of memory is good but some of the best oscilloscopes really don't have much memory (Keysight), they're just better at being able to capture the interesting parts of the signal.

[bestel]

I think we need to make some measurements become a standard when reviewing a scope.
Testing max bandwidth with frequency folding in mind, testing the peak detect, ...

Specs are good (when they are true) but not enough, it would be fun to make a board that can "auto test" a scope and give a score at the end...

[Berni]

Yep you will always get a better deal on a used piece of gear.

For the lower performance scopes it still makes sense to buy new since you get a new scope with warranty and all for around 400 bucks(if you can afford it that is). But when it comes to higher performance scopes of >300 MHz it can get pretty hard to justify a brand new scope as the prices start go up exponentially.

But when it comes to high performance scopes, i find that you often don't need the power of a high end scope in day to day work. So the clunky interface of an old high performance scope can be fine too as long as you also have a good modern low performance "daily driver scope". So instead of buying a stopgap scope towards a nice scope it might make sense to buy this daily driver scope now and then later on pick up an old high performance scope on the used market when you need it.

My own daily driver scope is a Agilent MSO6034A. Pretty old scope that is still modern enough to have most of the X3000 series functionality.

[tautech]

I think we need to make some measurements become a standard when reviewing a scope.
Testing max bandwidth with frequency folding in mind, testing the peak detect, ...

We each have different ways of evaluating the basic performance of a scope however the modern scope is more than just a scope so where must we begin and stop. For the online reviewer I challenge any to know a modern scope well within a week which is why Dave limits his reviews to basic functionality and a browse over the feature set.
There are some quite good reviews online but in most cases reviewers have owned and used that particular instrument for some time before attempting to make a video. Those that haven't well demonstrate their unfamiliarity with the scope. 

Specs are good (when they are true) but not enough,

Trust A and B rand specs to be a very accurate representation of the scopes capability, why wouldn't you when their brand reputation is at stake ?
Datasheets and manuals generally tell most of the full story yet omissions of some specs can be revealing so proper comparisons between competing models really need datasheets printed and specs ticked off one by one to reveal what's missing in one or the other. Beware of terminology too as 2 datasheets might not describe the same spec with the same words.

it would be fun to make a board that can "auto test" a scope and give a score at the end...

The subject is so vast such a board would be so complex and most likely not represent how I .....or you might use a scope.
We each have our own workflow and approach measurement problems from different paths and to add to this not all scopes work in the same manner.

[tggzzz]

... For the online reviewer I challenge any to know a modern scope well within a week which is why Dave limits his reviews to basic functionality and a browse over the feature set.
There are some quite good reviews online but in most cases reviewers have owned and used that particular instrument for some time before attempting to make a video. Those that haven't well demonstrate their unfamiliarity with the scope.

That's pretty much the case, and brings some corollaries:

• the complexity is an impediment for a beginner that doesn't understand the fundamentals of scopes
• the complexity can lull those with some experience into various more subtle errors[1]. Ditto those that lack understanding of fundamental theory
• the complexity represents a learning curve even for experienced users, and doubly so for beginners
• it is difficult to get a decent comparative review of scopes
• it is likely that those who make very definite statements are only familiar with one brand and/or type of scope

[1] gliding instructors (and probably more) know that dealing with inexperienced Bloggs is easier that dealing with Bloggs that are nearing going solo. The former make simple obvious mistakes; the latter make more surprising (and dangerous) mistakes.

[Howardlong]

My own daily driver scope is a Agilent MSO6034A. Pretty old scope that is still modern enough to have most of the X3000 series functionality.

I run an MSO7104B as my daily driver, so the same scope as yours, but with a bigger telly, and that's despite having a reasonable selection of other more modern and better specified scopes (including an MSOX3104T that's also out on the bench). It's the 7104B that used before anything else because its UI is so straightforward and responsive, and it boots faster.

[Berni]

I run an MSO7104B as my daily driver, so the same scope as yours, but with a bigger telly, and that's despite having a reasonable selection of other more modern and better specified scopes (including an MSOX3104T that's also out on the bench). It's the 7104B that used before anything else because its UI is so straightforward and responsive, and it boots faster.

Yep i would ideally want a MSO7104 as a daily driver, but that 6000 was the best 2nd hand scope i could find for a good price back then. But it has stuck with me because that boot time and fast UI makes it the fastest at getting the job done. Hence these old 5000/6000/7000 series scopes still fetch a fairly high price on ebay these days.

[Howardlong]

But after I've done some research, I will totally consider memory depth for my next "real/pro" purchase.

Lots of memory is good but some of the best oscilloscopes really don't have much memory (Keysight), they're just better at being able to capture the interesting parts of the signal.

It's the law of diminishing returns IMHO. Being able to stop the scope and still see the trace is the killer feature of a DSO over a CRO. Then having 10k points over just a screen full. Then having 1M points over 10k points. But the benefit of having 100M points over 1M points isn't really such a step change, more of a nice to have.

The other thing about the Keysights is the automatic management of deep memory, having the option of manual configuration taken away from you actually improves your workflow, and the scope always uses all of its memory anyway. With other scopes, particularly Tek, I find myself continually having to adjust the memory depth for each situation, not least because it's so darned slow. It's difficult to fall in love with another scope once you've experienced a Keysight.

[mawyatt]

I've always enjoyed working in the lab with scopes and such, but ~20 years ago executive management moved me out to teach others and develop new chips

They said I was worth more outside the lab and the bottom line is all that mattered, so in front of a computer screen I went 

I retired last year and got a couple old HP34401s to repair, then a couple Tek 2465 scopes to fix. Never had a chance to play with any of the new DSOs, although I did use an old LeCroy (rebranded Siglent I found out) when I need to "capture" a waveform now and then (wasn't impressed tho).

Figuring I need a new DSO, I've studied all the posts here on the various scopes and decided on the Siglent SDS2104X Plus only to find out it was backordered, so got the SDS2102X Plus which was in stock. I needed to do some work for someone working a university research project right away, so needed the scope.

After a week with this new DSO I must say I'm impressed. Will it replace my Tek 2465....probably not, but already has shown so much capability it will likely take much more of my "scope time" away from the Tek 2465 than I would have thought

Can't say much about the other DSOs (suspect they are all really good), but this SDS2102X Plus is much better than I thought, and that's coming from an old diehard analog type

Best,

[Berni]

Yep this "always using max memory and always running max speed" is also one of the things that makes the Agilent/Keysight scopes feel a lot like an analog scope.

Not only you never need to tell it how much memory to use, but it doesn't even let you use less than max memory because there is no menu to even set it. The point is that all of the sample memory is inside the acquisition ASIC, this means that there is such massive memory bandwidth available so the ASIC can render the entire memory to screen within pretty much a single LCD refresh cycle, yet as a downside this limits the amount of memory before the ASIC die would start becoming too big. But with the 8Mpts of memory on my MSO6000 i can't say i ever had a situation where i really needed more memory than that. Its still plenty enough memory to single shot capture a long busy waveform and then zoom into parts of it for a closer look. I gladly take the "small" sample memory in return for the responsive UI. Pretty much every touch of a knob responds on the screen instantaneously (Like literally the next LCD refresh, faster than the blink of an eye) this is what makes it feel so much like its an analog scope, yet you can still freeze the waveform whenever you want.

I have another scope with 1000Mpts of waveform memory and i have yet to use it for anything useful. Using full memory makes the screen update turn into more of a fast paced slideshow of a few frames per second, but when you turn on a software feature like serial decode or math channels or even some automates measurements the thing suddenly slows down to seconds per frame! Turning on a bunch of them and you can literally end up waiting for more than a minute for the screen to update! So even if you have all the memory you would ever want it still pays off to set up segmented acquisition and do it with less memory because its more practical.

[FriedLogic]

but I need to check the signal I'm decoding first ! 

It probably depends what exactly you want to look at. One of the problems that I found with some of the cheaper old scopes is that they did not show clock edges very accurately. I ended up getting a Hameg HZ 60 scope tester which generated a good low overshoot square wave to test them - yes this was a long time ago. The newest scope I have which is a Picoscope 3206B does not do too bad a job, but is not up to a 1980s Philips!

In practice, the probe setup is often the bigger issue.

[Howardlong]

Can't say much about the other DSOs (suspect they are all really good), but this SDS2102X Plus is much better than I thought, and that's coming from an old diehard analog type

I got an SDS2104X plus in nearly two weeks ago. I have the same feelings, it's certainly impressed me. A shame you couldn't get the 4 channel version :-(

The front end produces super crisp waveforms, when I first got it I thought the timebase must have stopped! It's reasonably responsive for most use cases too. Touch screen gesturing is pretty crap, although single finger prodding operations such as accessing menus are OK, which it needs to be because a lot of functionality isn't available form the front panel buttons.

[mawyatt]

Can't say much about the other DSOs (suspect they are all really good), but this SDS2102X Plus is much better than I thought, and that's coming from an old diehard analog type

I got an SDS2104X plus in nearly two weeks ago. I have the same feelings, it's certainly impressed me. A shame you couldn't get the 4 channel version :-(

The front end produces super crisp waveforms, when I first got it I thought the timebase must have stopped! It's reasonably responsive for most use cases too. Touch screen gesturing is pretty crap, although single finger prodding operations such as accessing menus are OK, which it needs to be because a lot of functionality isn't available form the front panel buttons.

Needed 3 traces right off to plot a 16 bit DAC output voltage vs. the position feedback sensor voltage and the high drive voltage for the piezo element stage. This is a nanometer level positioning controller based upon a piezo electric element for supporting a university research project (for someone), each DAC step represents 4 nanometers. With the scope traces you can see the feedback voltage perfectly follows the DAC command voltage and the piezo element non-linearity and hysteresis in the driver voltage, had to use 2 different plots to show this with the 2 channels. If I had waited for the 4 channel I'd still be waiting since they are still backordered.

I did look at the pulse response and compared to the Tek 2465 (after upgrading the BW to 350MHz), pretty much the same response.
Agree the gesturing is useless, but the prodding seems to work OK, and the Bode is very slow.

Best

[bestel]

A little list of all scope I'm looking for 

[Fungus]

Apparently you can get a new Rigol DS1102Z-E for 215 Euros.

(on Aliexpress)

[tautech]

A little list of all scope I'm looking for 

Good work.
Now you need examine specs a little more closely and the feature sets. 

[Dubbie]

Yep, you will always get a better deal on a used piece of gear.

This completely depends on where you live.
Here in NZ, the options for buying second-hand gear are very scarce. And when you do see it for sale it is usually overpriced.

I just searched our local buy/sell site for scopes and this "bargain" popped up.

For that money, you would get a much much nicer Siglent or similar with a warranty and it won't use 3 square feet of your bench and 500W to run.

[Doctorandus_P]

If you buy a 15+ year old scope you are likely to get a small memory depth and slow update rate, and you'll probably be saving for your new keysight right from the start.

If you spend the EUR400 for the siglent. there is a good chance it's "good enough" for multiple years, and you may never need a "better" scope.

And if you upgrade later, you can still sell it for a good 2nd hand price.

[bestel]

If you buy a 15+ year old scope you are likely to get a small memory depth and slow update rate, and you'll probably be saving for your new keysight right from the start.

If you spend the EUR400 for the siglent. there is a good chance it's "good enough" for multiple years, and you may never need a "better" scope.

And if you upgrade later, you can still sell it for a good 2nd hand price.

"The plan" is to buy for 1 year (or less) and then re-sell when I'm ready to buy the real one (or my business does not works).
The advantage of buying an used tektronix/keysight/agilent/...  is the "no money lost", because normally I will be able to sell it at the same price I've bought it... If I bought it at a good price of course...

In both cases I know my first scope will not be enough for all my use cases... this is why I ask for your help to found a "good offer", but it seems there is no "magic" 

[bestel]

Hi guys!

Just an update, I just picked up a dead free Metrix Scopix OX 7104-C... No charger, battery was dead... I dismount it, check some datasheet for not destroying it before putting 12V in place of the dead NiMh and it works well!

It's a CDD based osciloscope, but specs are good for a free one 
Bandwidth of 4 x 100MHz, 4 x 1 GS/s with 12 bits of resolution (thanks to the CDD design) and all input are ground isolated.


I think I will be OK 

[rsjsouza]

That seems to be an excellent acquisition! Congratulations!