Oscilloscope recommendation DS1054Z / SDS1104X-E / GDS-1104B / etc?

[corey]

Hi all, new member here. I've done a bit of hunting around on these forums and mm after some advice on currently available oscilloscopes, at the moment I'm in the market for my first benchtop one with a budget of AU$600-800. Just trying to find what's the best value in this range at the moment, including any hacking necessary. Not sure I'll need 4-ch but I've seen them recommended a few times as a general comment, "why not get a 4-ch?". Obviously looking at maximum BW/Gsps for the price/value.

Looks like the Siglent SDS1104X-E is a good option but can it be hacked to increase the BW to say 250MHz? It's a 1Gsps unit.

Also its datasheet says for the memory depth it has 7 Mpts/CH (not interleave mode); 14 Mpts/CH (interleave mode). So it's saying 14 Mpts per channel x 4 = 56 Mpts total?

The other I'm thinking about is the ol' Rigol DS1054Z. On TEquipment, it says they come with the mem depth upgrade: 24Mpts (1 CH)/12Mpts (2 CH)/6Mpts (4 CH). This implies a total of 24 Mpts.

Anything else I should be looking at?

(There's also the GWInstek GDS-1054B or 1104B, I don't know if the 1054B can be hacked to increase BW given it's a 1 Gsps)

Looks like forwarding would be around US$90 using Shipito.com (I've used them in the past).

Thanks!

[tautech]

Welcome to the forum.

FYI
SDS1104X-E uses two 1GSa/s ADC's, each with 14 Mpts memory support. Therefore we consider what channel pairs are active for the resources available to each channel.
They are a 200 MHz design and -3dB rolloff is in the region of 220 MHz for the 200 MHz model.

[ycui7]

The Siglent one has better user interface. The FFT function is very decent. It can do FFT at very low frequency.( You will be surprised a $24k 12-bit LeCroy scope cannot do FFT in low frequency region. Software screwed up.)

[KungFuJosh]

My first scope was the DS1054Z. I hated it. I didn't like the display, it was slow, and any remote connection was tedious to make and even slower. I returned it and got the SDS1104X-E.

The Siglent is a great scope, and the web server that's built in is excellent. You can also get a $9 dongle to get it to work on WIFI if you need to. Take a look at the threads on here for "improving" most of the Siglent stuff if you're inclined to.

[tatel]

I got GW-Instek GDS1054B. I'm happy with it, but if you want max bandwidth/sampling rate for the money, Siglent SDS1104X-E is what you want. As Tautech said it has two ADCs, so it effectively doubles what the Rigol and the Instek can do: 200 vs 100 MHz. Since each of these ADCs takes care of two channels, you can get 1 Gsa/s using channels 1 and 3 with 200 MHz bandwidth (sampling rate=5xBW=>really good). 500 MSa/s with all four channels active at 200 MHz (sampling rate = 2.5XBW => good enough). No doubt this is the most bang-for-the-buck you can get at this entry-level. At the end of the day, oscilloscopes are mainly about sampling rate and bandwidth.

In practice, 200 MHz will not get you much more further than 100 MHz however. If you want to do things with, say, really fast MCUs, 200 MHz probably isn't going to be good enough much longer than 100 MHz BW. But the price is not much higher, either, so why not? Unless you are on a short budget, of course.

The Instek has probably the better UI. Fast response, independent controls for all the 4 channels. But it comes with 70 MHz probes which aren't good enough after you hack it up to 100 MHz, so you'd be better getting new probes ipso facto. You would need at least 250 MHz probes to get total system bandwidth into 100 MHz range after hacking the Instek. That means the price could be near Siglent's (with original probes) but bandwidth can't go further than 100 MHz.

Keep in mind the recommendation is to have probes with "at least" 3x the signal bandwidth you want to measure. You could do with less than that, but having 70 MHz probes on a 100 MHz oscilloscope somehow defeats your "upgrading". Both the Siglent and the Rigol came with better probes than the Instek (I'm told) so you could probably initially do with the original probes and delay the purchase of new ones. I'm not really sure which probes would you need to get Siglent's total system bandwidth into 200 MHz range however. Total system bandwidth = 1/sqrt((1/o'scopeBW²)+(1/probesBW²)). Things get crazy expensive quite fast when bandwidth increases. Surely others can give some advice here.

The Rigol is the cheapest I think. I'm told its user interface is the slowest but it can do all you need at the low-end MCU level. Which is probably the usual stuff when speaking about entry-level benchtop oscilloscopes. So it would be the obvious option if you are on a short budget. You are probably going to need more stuff than just an oscilloscope so that matter is worth some consideration.

If you are ever going to do anything with sensors/LCD displays, you'll be doing digital work. In that case you'll be happy with a 4-channel oscilloscope and a logic analyzer like, say, a DSLogic Plus or, perhaps even better, an oscilloscope with a good-working MSO option. If you are sure you'd never be touching any digital stuff, then a 2 channel oscilloscope would be enough, I think.

So, if you have the money and want to get the most bandwidth, get the Siglent. If you are on a short budget, get the Rigol. Leave the Instek for those that have the money and want a responsive UI, but don't care so much about the greatest bandwidth.

Unless you are on a really short budget, the 4-channel Siglent is probably the way to go. Just don't think its MSO option is any good and get a separate logic analizer like DSLogic Plus if necessary.
 
If you have deep pockets and really want to have lots of bandwidth, and good-working MSO option, it could be wise to save some more money (double/triple your budget) and go for an oscilloscope that could be hacked to 350 MHZ like Rigol MSO5074 or Siglent SDS2104X Plus.

Good luck

[nctnico]

About probes: recently somebody posted measurement results that show there really isn't a difference in bandwidth where it comes to probes.

Which scope is the best choice depends a bit on the purpose. For decoding the GW Instek is the best choice as it decodes the entire memory so you can inspect deep into decoded data without the decoded data dissapearing. Also the GDS-1054B has freely adjustable signal filtering which can be handy in some situations. It is possible to upgrade the screen to a larger sized one which makes the screen easier to read.

[Fungus]

I'm waiting on the new Rigol HDO800. It won't be here for another three or four months though. The next-gen UI and high definition is killer.

If you a 'scope right now and budget isn't a problem then the Siglent is probably the one to get.

All three are similar bang:buck so it's really just down to how much you want to spend. All three are very capable devices. The Instek is nice to use (separate channel controls, a proper menu select button instead of pushing the twisty knob, etc...) but Instek pricing varies wildly by country. If it's close to the Rigol then get the Instek for sure but I wouldn't pay Siglent prices for one.

I wouldn't obsess over bandwidth at this level. Real-life probing problems will swamp any difference between having 100Mhz and 200MHz. I'm not even sure I'd hack a Siglent to 200Mhz, it might be better to keep the higher sample-rate-to-bandwidth-ratio of a 100MHz/2GS 'scope. None of them have internal termination for coax setups.

[Fungus]

Which scope is the best choice depends a bit on the purpose. For decoding the GW Instek is the best choice as it decodes the entire memory so you can inspect deep into decoded data without the decoded data dissapearing. Also the GDS-1054B has freely adjustable signal filtering which can be handy in some situations. It is possible to upgrade the screen to a larger sized one which makes the screen easier to read.

Yep. The Instek has a lot of things going for it.

But as usual we don't know what the intended use for the 'scope is and the asker will be all surprised about not getting a definitive answer about what's "best".

[Professzore]

Hi, and welcome!
As a long-time Rigol user (MSO1074-Z here from around 2016), so... Votes against Rigol:

• FFT is something, which shouldn't be even noticed – the worst in class,
• UI may be a bit awkward at the first, but you may become familiar with it after a few hours of use,
• literally no practical decoding option, even with the 16-channel probe,
• without practical decoding, it's a waste to have 16 digital channels on MSO models,
• a bit old, compared to the currently available competitors,
• fan is noisy, whatever you do,
• no RTC, so every screenshot starts from the lowest available ID,
• I've never ever used pass/fail and record during "live" measurement processes, they're both awkward to set-up and barely produced the expected results

What I actually like (not really love):

• fair quality for the price,
• regular models are all easily hackable (MSO doesn't, but it's another story),
• fairly wide range of trigger options (for example RUNT),
• meet the specs, even near the extremes,
• lightweight and compact.

It's good to know, that back in 2016, this was the „benchmark” beginner scope with 4 channels, a relatively big display, and all the really useful trigger options – there was no real competitor then. Within these 7 years, the whole market changed, and all competitors came up and made even better scopes.

Ideas to consider:

• 4 vs 2 channels: clearly 4, in some cases, may be enough to decode an SPI line, and there are many measurements, where 2 channels are not enough,
• digital channels: avoid it, it's way better to have a sealea, if necessary -- though, the price is nearly comparable
• strong front-end is a must,
• Bandwith, sample rate, and sample memory are all important, but you must exactly know what are you looking for, for example, it may be seriously more important to have a decent capture rate on relatively low bandwidth to catch a glitch, or to find a missing /ACK in i2c communication, so you must know what are you looking for
• built-in waveform generators are usually very compromised and limited,
• in the entry-market, there is no perfect scope, there are scopes which may have less limitations,
• Your decision must be made based on your measurement needs, and not tech specs – if you know what are your goals/measurement plans, then, we can introduce you to the key factors you must deal with. In general: for some measurements, an analog scope is much better, in the digital domain, it's possible that you must have a scope capable of handling signals up to 500 Mhz, or even more. Some description of your ideas/plans may be very useful.

If you have any question, just feel free to ask!

[tatel]

About probes: recently somebody posted measurement results that show there really isn't a difference in bandwidth where it comes to probes.

Well I can assure you that my oscilloscope has more bandwidth with 150 MHz Testec probes than with the original 70 MHz ones. Like some 20 MHz more. After that I got 250 MHz a couple of weeks ago, yet I don't have any spare time right now to test this new batch so I will remain silent about them. They will have to wait at least a couple weeks more

[Fungus]

in 2016, this was the „benchmark” beginner scope with 4 channels

Yep, and the price difference with the competition was much larger.

FWIW: It looks like Rigol is about to create a new "benchmark" with their DHO800 series but we're only at "preorder" status right now.

In the entry-market, there is no perfect scope, there are scopes which may have less limitations,

I'll go further and say there's no perfect 'scope at any level. 

But yeah, don't expect any of these 'scopes to be 100% amazing and have no problems.

(Think: If they were then there'd be no $1000 or $2000 'scopes on the market.  )

[tautech]

That means the price could be near Siglent's (with original probes) but bandwidth can't go further than 100 MHz.

FYI
SDS1104X-E is supplied with PP510 100 MHz rated probes whereas 200 MHz SDS1204X-E is supplied with PP215 probes.
Their frequency responses were checked by member Performa01 and PP510 was down 1dB at 200 MHz compared to PP215.

These results indicate PP510 will give satisfactory accuracy for most work at frequencies well above their rating.

[Fungus]

What I actually like (not really love):

• fair quality for the price,
• regular models are all easily hackable (MSO doesn't, but it's another story),
• fairly wide range of trigger options (for example RUNT),
• meet the specs, even near the extremes,
• lightweight and compact.

How about:

It has a row of buttons down the left side for turning on measurements very quickly/easy (something which you do a lot...)

Siglent's system for turning on measurements is a real pain in the ass by comparison.

there is no perfect scope

QFT.

[corey]

Thanks all! I'm quite surprised at the level of responses - some long and detailed. I acknowledge I didn't mention the intended use, and hence it's somewhat of a difficult question to answer! But thanks to all, this is quite the helpful community.

I'm leaning toward the Siglend SDS1104X-E, which is available at AppVision for $693 incl GST currently. https://appvision.com.au/index.php?id_product=119&id_product_attribute=0&rewrite=siglent-sds1202x-e-digital-oscilloscope-200mhz-2-channel-1gss-with-serial-decode&controller=product

Second in line is a GWInstek. Sounds like it is stronger in the decoding from memory, which I'm also likely to need.

Intended use is SDR (low RF up to 1GHz, IF), ADC/DACs, as well as digital interfaces (to an FPGA). RF/IF is why I was going for highest bandwidth possible. I'm thinking I'll get a separate logic analyser (e.g. Saleae? Haven't looked at pricing) but would be great to go for a MSO, I just don't think I have the budget. So I'm also interested in the FFT capability, which sounds like is good on the 1104X-E.

Must say, though, we're spoilt for choice. When I was at university 15-20 years ago, CRO/DSO's were $$$.

[Professzore]

Well,

Yes, sure. On the other hand, you can be easily messed up when you fill up all the free places, and you're going to change one less used to a new one. I use cursors in about 95% of all measurements.

[Professzore]

Hi,

At this moment, I have some home-brew SDRs on my bench to be assembled and tested.
From a broader view, with designated frequencies near 1 GHz, you'll have serious limitations even with a scope with a higher bandwidth of 250 MHz (both HW frontend, A/D, FPGA, and software interpretation). These scopes are barely capable to support your measurement needs.

Planning your measurements is a must at this point. With an SDR, you'll have to check the signal of the Tayloe driving circuit (precise phase shift) for example. I'm about to use it only on HF bands (up to 60 MHz), so a simple scope is capable to handle the signal. But as soon as you plan go up to the GHz range (or near to it), your chances to make key measurements will be significantly reduced (if not zeroed). Again, it's a question of your measurement goal.
Catching up a digital feed from an FPGA is also a question of bandwidth and capture rate. It depends on the clock frequency of the line (if exists).

Saleae is way over your budget at the moment... :-/  https://eur.saleae.com/products/saleae-logic-pro-8

[tatel]

That means the price could be near Siglent's (with original probes) but bandwidth can't go further than 100 MHz.

FYI
SDS1104X-E is supplied with PP510 100 MHz rated probes whereas 200 MHz SDS1204X-E is supplied with PP215 probes.
Their frequency responses were checked by member Performa01 and PP510 was down 1dB at 200 MHz compared to PP215.

These results indicate PP510 will give satisfactory accuracy for most work at frequencies well above their rating.

I mean GW-Instek GDS1054B can't go further than 100 MHz, even using better probes than originally supplied, which will make GW Instek GDS1054B about the same price than Siglent SDS1104X-E, which does go up to 200 MHz: I didn't knew with which probes.

Just saying it as clearly as I could, because perhaps your assertion seems to imply I said Siglent SDS1104X-E couldn't go further than 100 MHz with the original probes? Not really sure about it. You know, not a native english speaker here.

It turns out, from Performa01's information, Siglent SDS1104X-E can go up to 200 MHz even with the original probes. Good to know. I think this information will be useful to the OP.

But I forgot that in Europe right now the Siglent is cheaper than the Instek: €429 vs €465, + VAT, in Eleshop.eu

In the US, pricing seems to be like it was in Europe two years ago: cheapest one is the Rigol, then some 25 bucks more buy you the Instek, then about another 75 bucks more buy you the Siglent. Or so it was some weeks ago. IIRC the Instek was the cheapest one in the US two years ago.

So the point I made is still valid in the US, but not in Europe anymore.

Is this the post from Performa01 you are refering to?
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/msg1434665/#msg1434665

Quite interesting, mainly because I got some Testec probes recently: Testec TT-LF-312, Testec TT-MF-312, Testec TT-HV-250 (because it's high voltage, not sure 300 MHz bandwidth matters) and finally Testec TT-HVP-40 (multimeter HV probe).

TT-LF-312 is a real improvement over the original, 70 MHz GW-Instek probes supplied with my GDS1054B.

Quite disappointing to see TT-MF-312 at the bottom of the graphs. I still have to test it. I hope they are an improvement over TT_LF-312.

Well at least it turns out Testec probes are relatively cheap. But after the comments about Testec on multiple posts in this forum, I wasn't expecting having to say this. Perhaps they are becoming a cheap chinese brand?

Maybe I'll also get Testec TT-HF-212 to see if it's any better. But this will be for another thread.
 
 

[KungFuJosh]

Probe Master makes awesome probes, and they have a dealer in Europe too.

[tautech]

Is this the post from Performa01 you are refering to?
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/msg1434665/#msg1434665

Yes.
And this one:
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/msg1435196/#msg1435196

[tatel]

Is this the post from Performa01 you are refering to?
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/msg1434665/#msg1434665

Yes.
And this one:
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/msg1435196/#msg1435196

Yeah, I saw it. Where would be useful for that probe to have 300 MHz bandwidth? I was under the impresion that at those voltages, one would be dealing with some mains-related thing. So probably 50-60 Hz? I don't think a CR tube goes at 300 MHz. One could use that probe to look at the CRT of some Hameg oscilloscopes because it's 2.5 kV, not because it has 300 MHz bandwidth.  Multimeter probe Testec TT-HVP-40 could look at almost any CRT and it has just 300 Hz BW, am I right? Even a VFD would usually be under 500 Hz, it isn't?

Well I guess there will be something where a 300 MHz HV probe could be useful, but right now I can't figure what it could be.

[tautech]

Well I guess there will be something where a 300 MHz HV probe could be useful, but right now I can't figure what it could be.

Resolving fast edges.

FWIW all 100x HV probes we sell are 300 MHz rated not because we need such but that's what the manufacturer offers. Here like most, maybe all of EU mains is 230VAC and 3ph 415VAC between phases and being low impedance IMO it's wise to protect the instrument and yourself with a HV rated probe with better breakdown specs rather than any 600V rated probe.

[corey]

I’m starting to think, for my SDR work and working with IF and FOGAs, maybe I should stretch my budget up to an MSO5000 series. 😅 Probably save me on a nicer signal generator and logic analyser etc. TEquipment sell a 5074 with logic probe. But it’s looking expensive, even from TEquipment, about AU$2000 landed.

[mwb1100]

I’m starting to think, for my SDR work and working with IF and FOGAs, maybe I should stretch my budget up to an MSO5000 series. 😅 Probably save me on a nicer signal generator and logic analyser etc. TEquipment sell a 5074 with logic probe. But it’s looking expensive, even from TEquipment, about AU$2000 landed.

A couple things to think about that might save you some money on the logic probe:

  - there are people selling a non-Rigol version of the logic probe (based on designs that were developed in discussion threads on EEVBlog).  On US eBay they run about $80 USD shipped to the USA.  I'm not sure how the quality compares or what feature differences there might be (like supported signal voltages) different from the Rigol logic probe.

  - unless you need to have your logic waveforms displayed with and synchronized with analog signals, one of the many inexpensive USB logic analyzers might save you a fair bit     

[bdunham7]

Well I guess there will be something where a 300 MHz HV probe could be useful, but right now I can't figure what it could be.

It isn't a 300MHz HV probe--it can be used as a 300MHz probe or a HV probe, but not both at the same time.  It's main feature will be a lower input capacitance, the 100M input resistance is mostly a sideshow.

[tautech]

I’m starting to think, for my SDR work and working with IF and FOGAs, maybe I should stretch my budget up to an MSO5000 series. 😅 Probably save me on a nicer signal generator and logic analyser etc. TEquipment sell a 5074 with logic probe. But it’s looking expensive, even from TEquipment, about AU$2000 landed.

Don't ever imagine an inbuilt AWG will meet your every need.
Typically their drive levels (p-p) are lacking for a multitude of uses.....just so you know.