Recommend mid-price oscilloscope

[Fungus]

Unbelievably useless in sub-MHz region

Can you elaborate on that? I just connected the scope to a 1KHz sine wave and it shows what I expected:
a 1KHz sine wave.

This should be good. 

[MrWolf]

Can you elaborate on that? I just connected the scope to a 1KHz sine wave and it shows what I expected:
a 1KHz sine wave.

At it's absolute best could be compared with 3-digit multimeter accuracy in horizontal. So it could be 998 or 1002Hz. You would never know for sure (unless relying on 1CH-at-the-time hardware counter-bandaid, with pretty limited vertical sensitivity). Some think its normal and scope is only for weird and wiggly lines. I have used two 200€ and 400€ scopes for years which deliver accuracy comparable to 4-digit multimeter  in horizontal. And not only in "just right" timebase but in almost any timebase up to 3 orders on magnitude apart.
So for me it is absurd that something like Z box is produced and most weirdly, recommended...

[David Hess]

Like the old saying, buy a old scope but make sure you have another to keep it in working order.

My first answer when someone asks me what oscilloscopes are good for is that they are good for fixing other oscilloscopes.

For the record: I'm suggesting to use only ONE scope to measure itself using the working channel. I've done that before...

I have done it also.  It works great *if* you tie the probe's ground lead down preventing it from shorting out on something.

[Karel]

At it's absolute best could be compared with 3-digit multimeter accuracy in horizontal. So it could be 990 or 1010Hz.

Have you actually tried it yourself?
Because here the frequency measurement jumps between 1000 and 998Hz.
And the hardware frequency counter says 1.00000KHz (doesn't jump).

[nctnico]

At it's absolute best could be compared with 3-digit multimeter accuracy in horizontal. So it could be 990 or 1010Hz.

Have you actually tried it yourself?
Because here the frequency measurement jumps between 1000 and 998Hz.
And the hardware frequency counter says 1.00000KHz (doesn't jump).

Perhaps it is better to continue this discussion in the appropriate thread and not repeat the same ad nauseum in every oscilloscope thread.

[MrWolf]

Have you actually tried it yourself? Because here the frequency measurement jumps between 1000 and 998Hz. And the hardware frequency counter says 1.00000KHz (doesn't jump).

Sry 1 beer to much  Original post fixed, indeed it will jump about -+2Hz at best timebase (single cycle occupies whole screen). Making it 3-digit at best. Hardware counter is more-less ok but it can be applied to one channel only and it's counter only. While on good scope you would have 4-digit / 10,000 count accuracy on every active channel and all horizontal measurements.

Perhaps it is better to continue this discussion in the appropriate thread and not repeat the same ad nauseum in every oscilloscope thread.

Well dunno, Z box does get recommended in every thread, no? Why not post relevant links to discussion in every thread also? It's all about getting a balanced view on the subject 

[Simon]

Well as the budget is looking like $1000 the 1054Z is out of the race so perhaps discuss the merits of a scope in the desired price range ? if you want to do a full technical reveiw on the rigol 1054Z your welcome to a whole thread on the topic, just not this one.

[Karel]

My advice if you want to spend more money: buy an R&S HMOxxxx. They start from 1800,=.
Most DSO's cheaper than the cheapest R&S HMOxxx have the same "quality" as the Rigol.

https://www.eevblog.com/forum/blog/eevblog-842-rohde-schwarz-hmo1202-oscilloscope-teardown/

So, either buy the DS1054Z or jump straight to an R&S HMOxxx.

[Fungus]

it will jump about -+2Hz at best timebase (single cycle occupies whole screen). Making it 3-digit at best.

Hardware counter is more-less ok

Yep, that's typical of Wolfy's definition of "unbelievably useless". 

Perhaps it is better to continue this discussion in the appropriate thread and not repeat the same ad nauseum in every oscilloscope thread.

 

[Simon]

it will jump about -+2Hz at best timebase (single cycle occupies whole screen). Making it 3-digit at best.

Hardware counter is more-less ok

Yep, that's typical of Wolfy's definition of "unbelievably useless". 

Perhaps it is better to continue this discussion in the appropriate thread and not repeat the same ad nauseum in every oscilloscope thread.

 

Which is what I just said so why continue yourself ?

[Fungus]

Well as the budget is looking like $1000 the 1054Z is out of the race so perhaps discuss the merits of a scope in the desired price range ?

OK. First you need to decide how many channels you need.

This will depend on what sort of work you do. For anything related to microcontrollers I'd say get 4 channels.

If you think 2 channels is enough then look at R&S HMO series.

If you need 4 then it's more difficult.

[MrWolf]

Why not 1104B from Gw Instek. Have not tested it personally but from this test seems to be proper 4-digit / 10,000 count scope:
https://www.eevblog.com/forum/testgear/testing-dso-auto-measurements-accuracy-across-timebases/?action=dlattach;attach=279233
Seems good oldschool engineering much like Tek.
BTW most  interesting thing - only scope seen so far in tests that calculates signal period with every timebase, not just when screen width > 1 period.

[nctnico]

My advice if you want to spend more money: buy an R&S HMOxxxx. They start from 1800,=.
Most DSO's cheaper than the cheapest R&S HMOxxx have the same "quality" as the Rigol.

https://www.eevblog.com/forum/blog/eevblog-842-rohde-schwarz-hmo1202-oscilloscope-teardown/

So, either buy the DS1054Z or jump straight to an R&S HMOxxx.

The R&S HMO (Hameg) series is way too expensive for what it offers. After the Rigol DS1054Z I'd look at GW Instek (GDS-1000B or GDS-2000E series) to go one step up and after that go into the >$5000 range probably from R&S (not Hameg!) if Keysight hasn't come up with scopes which offer deeper memory but even then I'd wonder if it really is worth spending so much money. Currently my Agilent DSO7104A (approx $25k list price) is sitting unused on a shelve because the GDS-2204E is so much easier to use and has some very neat features.

[rstofer]

Why not 1104B from Gw Instek. Have not tested it personally but from this test seems to be proper 4-digit / 10,000 count scope:
https://www.eevblog.com/forum/testgear/testing-dso-auto-measurements-accuracy-across-timebases/?action=dlattach;attach=279233
Seems good oldschool engineering much like Tek.
BTW most  interesting thing - only scope seen so far in tests that calculates signal period with every timebase, not just when screen width > 1 period.

Ok, the Instek 1104B is only $710 so it's a candidate.  Does it offer serial decoding?  If so, how much extra?  I couldn't immediately find the answer to these questions.

Most of the time, I work in the digital domain.  I ALWAYS know the frequency and the little frequency counters at the bottom of the screen are meaningless.  Further, I view the oscilloscope as a qualitative instrument, not quantitative.  I simply don't care if there's a digit error in a measurement based on the screen image.  The hardware counter works well enough should I somehow have a brain fade on frequency.

To me, decoding is an absolute requirement for a DSO.  Remember, I work only in the digital domain.  FFT is generally meaningless.  I have used it exactly once in a 30+ year career as a EE.  And that was on a study of the 11th and 13th harmonics of a 12 pulse rectifier driving a 20 MW heater.  Fun, sure, but that's the only time I have used it.

So, what do I get for the extra $310 dollars?  Another digit in a ginned up frequency count?  No decoding?  Pretty simple choice...

[nctnico]

I wouldn't call FFT meaningless but then again I'm regulary working on projects involving digital signal processing and (fast) FFT on a scope is very helpfull to look at signals to see if filtering in the digital domain is an option or the analog frontend needs to be better. For one of the projects I'm currently working on FFT has been on constantly.

[tautech]

Like the old saying, buy a old scope but make sure you have another to keep it in working order.

My first answer when someone asks me what oscilloscopes are good for is that they are good for fixing other oscilloscopes.

For the record: I'm suggesting to use only ONE scope to measure itself using the working channel. I've done that before...

I have done it also.  It works great *if* you tie the probe's ground lead down preventing it from shorting out on something.

Risky to say the least, much better if it's disconnected and the probe tip sleeve is installed to prevent the Gnd ring near the probe tip from making contact with any part of the DUT.

[MrWolf]

To me, decoding is an absolute requirement for a DSO. 
...
So, what do I get for the extra $310 dollars?  Another digit in a ginned up frequency count?  No decoding?  Pretty simple choice...

Well such attitude can easily end in tears, perhaps well deserved AFAIK decoding does not work properly in Rigol for exact same reason that it's only "3-digit" scope. But if entry level Instek does not have it at all then yea... Guess it's the case when you pay more or suffer. I have no interest in decoding so perhaps better informed will take over.

[Fungus]

So, what do I get for the extra $310 dollars?  Another digit in a ginned up frequency count?  No decoding?  Pretty simple choice...

If decoding is a must the you won't get anything below about $1500.

Apart from a hacked DS1054Z that is.

What Wolfy's great discovery is that the DS1054Z (and many other 'scopes) does is make calculations based on screen pixels. The DS1054Z has 800 pixels horizontally, minus the menus at the sides. Calculating a very exact frequency of 1kHz using that information obviously isn't very precise.

His life's ambition (it seems) is to find out all the ways working with on-screen information will make readings inaccurate.

Fortunately the DS0154Z (and many other 'scopes) have separate dedicated hardware for counting frequencies and most of the other readings don't really matter because people naturally zoom in on the area of interest and therefore get more pixels for measuring things like rise/fall times.

Getting back on topic: You're finding out that a hacked DS1054Z beats most other scopes in your target price range. You probably have to spend $1200+ to get something significantly better in terms of features. Serial decoding in particular is expensive.

Serial decoding is also done better by dedicated logic analyzers, eg saleae.

[MrWolf]

Serial decoding is also done better by dedicated logic analyzers, eg saleae.

Yea but my old 400€ scope does serial (lots of protocols) off 48k buffer, not screen
900€ scope from same company does 1-Wire, ARINC 429, CAN, DCC, DMX512, FlexRay, Ethernet 10Base-T, USB 1.1, I²C, I²S, LIN, PS/2, SPI, SENT, UART/RS-232 off 128M buffer. But psst... that technology is not supposed to exist, so it is "despised around here" as per herr Fungus. 

[AndyP]

There are a raft of 2Gsps scopes, as a digital scope user one of these would be my best value for money guess, possibly the Rigol 1074B if all the usual hacks work. 
Siglent also do a 2Gsps scopes with a deep 140Mpts memory the SDS2000, as do Micsig with the 10 inch tablet scope, TO202a, 90Mpts, 500K wfm/s.
With that kind of budget, it's the 2G sample convertors I'd aim for. Of course other features are important depending on use case. Some hardware debug requires high waveforms per second, and software dbug often benefits from deep memory.

[Karel]

If you can connect the DS1054Z to a pc/laptop, you can do the serial decoding there
at the highest samplerate possible (up to 1 GHz) and using the whole waveform memory (up to 24Mpts).
So, the limitation that it decodes only what's on screen, is not valid anymore (if you connect it to a pc).

[nctnico]

If you can connect the DS1054Z to a pc/laptop, you can do the serial decoding there
at the highest samplerate possible (up to 1 GHz) and using the whole waveform memory (up to 24Mpts).
So, the limitation that it decodes only what's on screen, is not valid anymore (if you connect it to a pc).

With what software would that decoding happen? And how about browsing through the results? Think about segmented acquisitions, triggering on certain packets, etc. If you go that route you are way better off using a USB logic analyser. In other words: doing the decoding offline defeats the purpose of decoding in the oscilloscope because you'll be creating a PC based oscilloscope in the end. Just buy an oscilloscope which has decoding implemented properly.

[Howardlong]

One thing that the Rigol does to do well is serial triggering, despite not doing a great job at decoding.

[AndyP]

If your doing digital work on a budget, use the scope to check signal integrity and the USB logic analyzer with serial decoders for the serial deciding. (I've been very pleased with the DS logic 16 channel 100MHz, with 16M sample buffer).

Cheaper scopes often have poor ppm xtal for frequency counting.
Micsig TO1104 ~20ppm
Keysight MSO 3000T ~20ppm
Keysight MSO 3000X~ 2.5ppm
Rigol MSO 1104z ~25ppm
R&S RTM2000 ~3.5ppm

Just something to keep an eye on when reading those 6+ digit frequency counters. A similar problem occurs with some dedicated frequency counting hardware.

R&S HMO suffered from a poor memory depth last time I looked. This may or may not be an issue depending on the problem. Even the Keysight MSO3000 depth is poor for a modern scope.

Any thoughts on the Rigol 1074B (2G sample rate 4 chan, 28mpt) or the Micsig TO202 (2Gsps, 90M pt buffer.)

[nctnico]

2Gs/s is nice for a 500MHz scope but total overkill for anything 200MHz and lower.