5 Gsample/s, 1 GHz, 4 ch scope advise

[stejo780]

Hello,

First, I would like to thank Dave for his fantastic videos, which I have been following for a while and the rest of you for a great forum!

Secondly, I would appreciate some help when it comes to selecting a 5 GSample/s, 1 GHz, 4 ch scope. The bandwidth and sample rate is not written in stone, but around this range. So far, I have been looking at:

• R&S RTE1104
• Tektronix MDO4104B-6
• Keysight MSOX4104A

Which all of course have different pros and cons. But which one would you select and why?

Personally, I'm a bit sold on the R&S scopes, but I'm struggling a bit to find reviews on them. Dave sounded a bit disappointed in both the MDO4000 and the X4000 series review. How do you think the RTE series would stand up?

I have been using the MDO4000 series a bit and really like the pan & zoom functionality on these for example, and the spectrum analyzer is a nice add on indeed.

Thanks and Kind Regards,
Stefan

[AndyC_772]

I like my MSOX3104A. It's one of those tools that 'just works', and I rarely have to think about it.

The 4000X series is quite a bit more expensive, and really doesn't offer much more for the money. I suspect that's why it received a disappointing review; it's not that it's a bad scope at all, but there are few reasons to justify the cost over and above the cheaper 3000X series.

[Wuerstchenhund]

I would appreciate some help when it comes to selecting a 5 GSample/s, 1 GHz, 4 ch scope. The bandwidth and sample rate is not written in stone, but around this range. So far, I have been looking at:

• R&S RTE1104
• Tektronix MDO4104B-6
• Keysight MSOX4104A

Interesting list.

Any reason you omitted LeCroy?

Which all of course have different pros and cons. But which one would you select and why?

That is hard to say without knowing what it is for, but if it has to be from that list then I'd take the R&S RTE hands-down. The Keysight is a good scope but the DSOX Series is pretty expensive for what it offers and comes with some silly limitations (i.e. tiny sample memory). The Tek is pretty slow, the UI isn't great, plus you pay for a integrated spectrum analyzer with very poor specs. Plus based on my experience their support isn't as good as it once was.

Personally, I'm a bit sold on the R&S scopes, but I'm struggling a bit to find reviews on them. Dave sounded a bit disappointed in both the MDO4000 and the X4000 series review. How do you think the RTE series would stand up?

From what I've seen the RTE should stand up quite well against the other two contenders, although R&S' UI can take a while getting used to.

In any case, for that money you should get some loaners inhouse and try them out in your environment. Don't spend the money blindly.

[nctnico]

Don't spend the money blindly.

I agree. For this kind of money do thourough hands-on testing.

[stejo780]

I would appreciate some help when it comes to selecting a 5 GSample/s, 1 GHz, 4 ch scope. The bandwidth and sample rate is not written in stone, but around this range. So far, I have been looking at:

• R&S RTE1104
• Tektronix MDO4104B-6
• Keysight MSOX4104A

Interesting list.

Any reason you omitted LeCroy?

Which all of course have different pros and cons. But which one would you select and why?

That is hard to say without knowing what it is for, but if it has to be from that list then I'd take the R&S RTE hands-down. The Keysight is a good scope but the DSOX Series is pretty expensive for what it offers and comes with some silly limitations (i.e. tiny sample memory). The Tek is pretty slow, the UI isn't great, plus you pay for a integrated spectrum analyzer with very poor specs. Plus based on my experience their support isn't as good as it once was.

Personally, I'm a bit sold on the R&S scopes, but I'm struggling a bit to find reviews on them. Dave sounded a bit disappointed in both the MDO4000 and the X4000 series review. How do you think the RTE series would stand up?

From what I've seen the RTE should stand up quite well against the other two contenders, although R&S' UI can take a while getting used to.

In any case, for that money you should get some loaners inhouse and try them out in your environment. Don't spend the money blindly.

Thanks a lot for the feedback everyone.

No other reason to omit LeCroy than that I'm not familiar with their products, I have never seen them or worked with them. And I don't know if there's a local supplier.

R&S are coming here for a demo and we have some older (but similar) Tek and Agilent scopes in house to play with. But the order is a but "urgent" for reasons that some of might guess this time of year.

The x4000 seems very expensive indeed. Almost double the price if I compare the offers I've have received on the two other scopes listed. Seems lite Tek and R&S are able so give me some pretty good deals (demo equipment, year-end sale and so on) but probes and options will pretty much rocket the price anyway.

[robert_]

I recently had a MSOX4000T as a demo unit, and i was quite disappointed mostly about it lacking 1 and 2mv/div. The range exists, but its just digital zoom of 5mv/div, resulting in about 5 steps per div. at 1mv/div, which is close to useless. I do need these ranges however... Apart from that, it doesnt look too bad, much better user interface than anything Tek currently offers. Other weak points: limited FFT size, non-adjustable record length, noise.

[stejo780]

I recently had a MSOX4000T as a demo unit, and i was quite disappointed mostly about it lacking 1 and 2mv/div. The range exists, but its just digital zoom of 5mv/div, resulting in about 5 steps per div. at 1mv/div, which is close to useless. I do need these ranges however... Apart from that, it doesnt look too bad, much better user interface than anything Tek currently offers. Other weak points: limited FFT size, non-adjustable record length, noise.

Yes, I absolutly agree. That Keysight don't have a 1mV or 500uV range is almost a showstopper for me. I think this will be very useful when measureing low noise power supplies for example and if I go for a R&S scope I will seriously consider the HD 16-bit option.

[stejo780]

That is hard to say without knowing what it is for, but if it has to be from that list then I'd take the R&S RTE hands-down. The Keysight is a good scope but the DSOX Series is pretty expensive for what it offers and comes with some silly limitations (i.e. tiny sample memory). The Tek is pretty slow, the UI isn't great, plus you pay for a integrated spectrum analyzer with very poor specs. Plus based on my experience their support isn't as good as it once was.

If you forget about my list and select freely, but in the same range, what would you chose?

Thanks,
Stefan

[AndyC_772]

What application do you have in mind that requires measuring noise down in the mV, but a bandwidth in the GHz?

I'm struggling to think of the use case. If you're measuring power supplies on high speed digital logic, then I completely agree that you need the bandwidth, and quite likely an active probe too. But you'll have noise margins in the 10's of mV at least, so I don't see how a 1mV range is really necessary.

Or, if you're doing analog design, and need to measure much lower noise levels, then it's unlikely you'll need the bandwidth. A lower bandwidth scope and a pre-amplifier would be a better combination, and probably cheaper too.

Do also bear in mind that the apparent noise of a scope is a function of bandwidth; the phrase "noise bandwidth" doesn't seem to get mentioned anything like as often as it should. A GHz scope will always look noisier than, say, a 100 Mhz scope, simply because both are showing all the noise from DC-100MHz, but the faster scope is adding the noise from 100 MHz - 1 GHz as well.

[tggzzz]

If you're measuring power supplies on high speed digital logic, then I completely agree that you need the bandwidth, and quite likely an active probe too.

Sometimes spectrum analysers are better for that, especially for EMI/EMC compliance testing.

[joeqsmith]

I recently had a MSOX4000T as a demo unit, and i was quite disappointed mostly about it lacking 1 and 2mv/div. The range exists, but its just digital zoom of 5mv/div, resulting in about 5 steps per div. at 1mv/div, which is close to useless. I do need these ranges however... Apart from that, it doesnt look too bad, much better user interface than anything Tek currently offers. Other weak points: limited FFT size, non-adjustable record length, noise.

Yes, I absolutly agree. That Keysight don't have a 1mV or 500uV range is almost a showstopper for me. I think this will be very useful when measureing low noise power supplies for example and if I go for a R&S scope I will seriously consider the HD 16-bit option.

When I need to measure down this far, I shove a preamp in front of it.    Maybe you could get away with this for your measurements.

[Wuerstchenhund]

No other reason to omit LeCroy than that I'm not familiar with their products, I have never seen them or worked with them. And I don't know if there's a local supplier.

I see. I don't know about local suppliers (but I'd avoid resellers and deal directly to the manufacturers anyways, no matter which of the big brands) but I'd recommend to at have a look at them, considering that they make some of the most advanced scopes on the market and that their pricing is often better than for Keysight and Tek.

All scopes are roughly in the $17k-$19k range ('naked') so lets assume that as a ballpark figure. In that price range LeCroy offers the WaveRunner 610zi:

http://teledynelecroy.com/oscilloscope/oscilloscopeseries.aspx?mseries=352&capid=102&mid=504

Datasheet:
http://cdn.teledynelecroy.com/files/pdf/waverunner-6zi-datasheet.pdf

That's a 1GHz 20GSa/s scope with up to 128M sample memory, a large touch screen that can be used in landscape and portrait mode, and a really huge range of options.

However, if we just take your basic specs (1GHz, 4Ch, 5GSa/s) then there's also the LeCroy WaveSurfer 10:

http://teledynelecroy.com/oscilloscope/oscilloscopeseries.aspx?mseries=470

Datasheet:
http://cdn.teledynelecroy.com/files/pdf/wavesurfer10-datasheet.pdf

This one only comes with 10GSa/s and 16M memory, however at $10k it is even cheaper than a 500MHz Keysight DSOX3054T.

If you forget about my list and select freely, but in the same range, what would you chose?

I'd take one of the LeCroys, simply because in my opinion they offer the most bang for the buck, plus even the WaveSurfer 10 offers more capabilities than the Keysight or the Tek (aside from the spectrum analyzer). There's also support, LeCroy gurantees support for all their scopes for 7 years after end of production, and on a best effort basis after that (they still repair 9300 Series scopes, and these are from the '90s).

My second choice would be the R&S RTE. Good scope with a versatile but often somewhat confusing UI and decent specs.

The Keysight is pretty much a blown up DSOX3000A, and while not a bad scope for what it can do (and what not) it is in my opinion vastly overpriced. Tek, well, you pay a lot for the spectrum analyzer and the name (which they build up during the analog scope days), and in my opinion the rest of the scope isn't exactly spectacular. What would also worry me with them is the support.

[branadic]

I recommend HDO6000 series, I'm very very happy with my HDO6054-MSO and it's strong Matlab based math. Write your own Matlab skript and let it running on life data. It's an absolut winner 
I had R&S, Agilent and LeCroy here on the same exercise and was disappointed by R&S and Agilent. The 12bit vertical resolution is brilliant.

[stejo780]

Thanks a lot for the feedback!

The increased resolution is nice two have when qualifying power supplies to high-speed digital, or to find when small discrepancies on a high voltage signal. A preamp can solve some of the cases I guess, at least the power supply measurement, but it is just more convenient to have it directly available.

Perhaps the spectrum analyzer of the MDO4000 is not the best, but the nice thing about it is the time correlation between analog/digital/RF...

The Lecroy scopes do look interesting I have to admit! I have requested some prices to see if they are competitive.

[robert_]

I just got a HDO6054 as a demo unit on my desk today...
Definitely the best scope i have ever used.
Huge screen, FAST (rather responsive even at 250Mpts, really fast FFT) , low (electrical) noise and best user interface ive seen yet, surprisingly easy to use even though it does have a lot of features. Native 12bit IS an improvement, no "HiRes" mode comes close. LabNotebook is a nice feature too, as is the graphical math programming.

Its quite loud and power hungry, and needs a while to boot/shutdown though. Also, 2.5Gs/s is all you can get with an 12bit ADC at the moment. Below 10ns/div its good to have a stable trigger and enable RIS.

Price: 12k-ish for the basic 350Mhz version is what i heard. The HDO4k isnt much cheaper.

[Wuerstchenhund]

The increased resolution is nice two have when qualifying power supplies to high-speed digital, or to find when small discrepancies on a high voltage signal. A preamp can solve some of the cases I guess, at least the power supply measurement, but it is just more convenient to have it directly available.

The HDO Series is great, plus it's a true 12bit ADC not some oversampled ones like in the current R&S scopes or the olde Keysight DSO9000H.

If you want to qualify switch-mode power supplies then have a look at LeCroy's PMA2 option (which is available for the WaveRunner 6zi):
http://cdn.teledynelecroy.com/files/pdf/lecroy_pma2_datasheet.pdf

There's the PMR package (slightly less functionality than PMA2) which is available for WaveRunner 6zi, WaveSurfer 10 and the HDO Series:
http://cdn.teledynelecroy.com/files/pdf/power_analyzer_datasheet.pdf

Perhaps the spectrum analyzer of the MDO4000 is not the best, but the nice thing about it is the time correlation between analog/digital/RF...

True, and if you need the full bandwidth that the built-in spectrum analyzer offers all in one unit then the MDO4000 is the scope to buy. However, considering that you're looking at a 1GHz scope, the FFT functionality that comes in these scopes might even be enough for your purposes (and there's also the Spectrum Analyzer option for the LeCroy scopes which turns them into a VSA).

[Wuerstchenhund]

BTW: please let us know when you've decided what you finally bought and what made you buy it.

[stejo780]

Had a guy from R&S in last week and he lended me a RTE1104 to play around a with. Looks pretty good to me 

[robert_]

So, i sadly had to return the HDO6k this week. We had R+S show up, and now i have a RTE1054 to try out for a while.
Impression is so-so, i really dislike the fact that i have tons of small boxes etc covering up waveform display area, and some of them cant be moved out at all, with the channel/timebase boxes not even being transparent. From experience (DPO3k/4k) this will end up blocking exactly the part of the waveform i want to see later... Also opening some menu results in several semi-transparent windows covering each other, making it really hard to read and tell apart which text belongs to which layer. While i could adjust to both Agilent and LeCroy UI within minutes, i still struggle with the R+S way after a whole day.
I ran the same tests i did on the HDO and well, its a plain 8bit system, a good one but still 8 bits. 1Mhz 1Vpp Square plus 10mVpp 20Mhz sine is a nice test... On the HDO in its standard mode its instantly visible there is something, and after zooming in theres a nice clear 20Mhz sine to see. Any 8bit system shows the usual noisy square, if you dont know theres something to look for, you miss it. If you know what to look for, with added lowpass filtering ("HD mode" "16Bit" marketing BS) i can see the sine, but only with intens/colour grading, it still is mostly buried under noise on a single shot acq. If its not synchronous (in real life some ripple rarely will be), a single shot acq is your only chance to get it.
Oh by the way, 500uV/div isnt. I counted approx. 125 steps for full 8div, so its just software zoom on 1mv/div.
RTE doesnt have any nice spectrum app, which the lower-end RTM does have (paid option). I really cant explain that decision from RS to leave out this app on the higher end models... With stgandard FFT as math function, its about as hard to use as on most older scopes, with turning knobs for timebase, memory depth and fff options at the same time, iteratively until the display looks right. NOT good.
Spectrum mode on the HDO is so much easier to use, just like a spectrum analyzer. After all im not interested in the FFT as math function, but rather in doing spectrum analysis.
Good points for the RTE is price, lower (acoustical) noise, colourful leds everywhere (TOO bright though!), 5Gs/s, support by R&S, price, better triggering and umm, price again.
Im trying to justify the HDOs extra price to my boss, but it will not be easy...

[EEVblog]

I like my MSOX3104A. It's one of those tools that 'just works', and I rarely have to think about it.
The 4000X series is quite a bit more expensive, and really doesn't offer much more for the money. I suspect that's why it received a disappointing review; it's not that it's a bad scope at all, but there are few reasons to justify the cost over and above the cheaper 3000X series.

I would concur with that. You need to have a use for one of the decoders in the 4000X to make it worthwhile.
The 3000T has the nice touch zone triggering.

The MDO scopes are nice bang-per-buck with the spectrum analyser, but they are as slow as wet week. I would not want to use one for everyday use.

[robert_]

I like my MSOX3104A. It's one of those tools that 'just works', and I rarely have to think about it.
The 4000X series is quite a bit more expensive, and really doesn't offer much more for the money. I suspect that's why it received a disappointing review; it's not that it's a bad scope at all, but there are few reasons to justify the cost over and above the cheaper 3000X series.

I would concur with that. You need to have a use for one of the decoders in the 4000X to make it worthwhile.
The 3000T has the nice touch zone triggering.

The MDO scopes are nice bang-per-buck with the spectrum analyser, but they are as slow as wet week. I would not want to use one for everyday use.

Wasnt too thrilled about the MSOX3k/4k. No real 1/2mV/div, quite short memory on the 3k (2M/ch for 4Ch), FFT length limited to 64k, etc. They seem to be developed purely for waveforrms/sec and nothing else.

MDO3k have the same annoying issiue with the encoders reacting in strange ways as on most DPO3k. You turn the knob one step clockwise, you sometimes get 3 steps backwards. As if the decoding misses one edge. On some instruments its really annoying with some knobs doing this every other time, on some it only fails a few times a day. Is it too hard to get the debouncing done properly?
Plus, Tek comes with Danaher grade non-support. Just forget about reporting firmware bugs, they DONT CARE at all, and they wont fix anything.
Only good point about DPO/MSO/MDO 2k/3k/4k is you get the fully optioned highest-bandwidth model for the base price, as all of these are fully hackable with minimum effort.

[KE5FX]

Wasnt too thrilled about the MSOX3k/4k. No real 1/2mV/div, quite short memory on the 3k (2M/ch for 4Ch), FFT length limited to 64k, etc. They seem to be developed purely for waveforrms/sec and nothing else.

In the real world, 2 MS is a lot of sample data, 64K is a long FFT record, and sub-millivolt signals are often best observed with said FFT.

And at least Keysight's firmware devs know how to read an optical encoder.     I can't imagine anything that would get a scope put back in its shipping box and returned to the seller faster than that.   I'd even pay extra for FedEx, just to get it out of my sight.  Sic transit gloria Tektronix if their scopes are behaving that way.

[Someone]

Wasnt too thrilled about the MSOX3k/4k. No real 1/2mV/div, quite short memory on the 3k (2M/ch for 4Ch), FFT length limited to 64k, etc.

It may not be ideal if you're limited by the memory depth but the averaging and hires acquisition modes both top out at 12 bits and give a very useful 10uV resolution in the memory which would be around 200uV per division equivalent.

[Wuerstchenhund]

In the real world, 2 MS is a lot of sample data, 64K is a long FFT record, and sub-millivolt signals are often best observed with said FFT.

Not in 2015, no, and even less so for a 1GHz+ scope with sample rates in excess of 4GSa/s. 2Mpts memory was pretty good (not stellar, though) 15 years ago, however even back then 64k FFT was pretty poor.

To offer a scope like the Keysight DSOX4k in the price range it sits in with 4M memory and 64k FFT is a pretty bad joke in this day and age.

And at least Keysight's firmware devs know how to read an optical encoder.    

I guess you haven't had to use their high end scopes (i.e. the DSO90k) where some firmware versions suffered from jumping encoders, wrong indications (i.e. showing channels as set to GND when they weren't) and other silly things. Granted, these problems were have been fixed, but Agilent didn't exactly rush out to squash these bugs.

[Wuerstchenhund]

So, i sadly had to return the HDO6k this week. We had R+S show up, and now i have a RTE1054 to try out for a while.

Interesting choice to say at least, as the RTE isn't a very attractive offering. It's *very* expensive (500MHz 4ch for $15k, 1GHz 4ch for roughly $18k) which is pretty close to the much more capable RTO (600MHz 4ch for $18k, 1GHz 4ch for $21k), but the RTE is 5GSa/s only and pretty much lacks all the analysis features of the RTO (including VSE support which is R&S' equivalent to Keysights 89600B VSA software).

Impression is so-so, i really dislike the fact that i have tons of small boxes etc covering up waveform display area, and some of them cant be moved out at all, with the channel/timebase boxes not even being transparent. From experience (DPO3k/4k) this will end up blocking exactly the part of the waveform i want to see later... Also opening some menu results in several semi-transparent windows covering each other, making it really hard to read and tell apart which text belongs to which layer. While i could adjust to both Agilent and LeCroy UI within minutes, i still struggle with the R+S way after a whole day.

Yes, the UI of RTE/RTO isn't very intuitive, and a bit cumbersome (that's pretty much the case on most R&S gear). It works great if you do signal analysis where you can drag the various windows around to give you all the information you need. It makes sense on the RTO because of its capabilities but I can't see the point on the RTE.

I ran the same tests i did on the HDO and well, its a plain 8bit system, a good one but still 8 bits. 1Mhz 1Vpp Square plus 10mVpp 20Mhz sine is a nice test... On the HDO in its standard mode its instantly visible there is something, and after zooming in theres a nice clear 20Mhz sine to see. Any 8bit system shows the usual noisy square, if you dont know theres something to look for, you miss it. If you know what to look for, with added lowpass filtering ("HD mode" "16Bit" marketing BS) i can see the sine, but only with intens/colour grading, it still is mostly buried under noise on a single shot acq. If its not synchronous (in real life some ripple rarely will be), a single shot acq is your only chance to get it.

Well, that's not surprising, since as you say their "HD mode" is simple software (oversampling, the same Agilent tried did its DSO9kH). What's worse is that R&S offers that as a paid-for option, which isn't worth the money.

Oh by the way, 500uV/div isnt. I counted approx. 125 steps for full 8div, so its just software zoom on 1mv/div.

That is indeed very disappointing. 

Considering how much R&S loves to tout its low noise ADC.

RTE doesnt have any nice spectrum app, which the lower-end RTM does have (paid option). I really cant explain that decision from RS to leave out this app on the higher end models... With stgandard FFT as math function, its about as hard to use as on most older scopes, with turning knobs for timebase, memory depth and fff options at the same time, iteratively until the display looks right. NOT good.
Spectrum mode on the HDO is so much easier to use, just like a spectrum analyzer. After all im not interested in the FFT as math function, but rather in doing spectrum analysis.

I'm actually surprised that a 'spectrum analyzer UI' isn't available for more scopes, as I guess it's much more straight-forward to use than the traditional FFT math function.

Good points for the RTE is price,

Is it? The RTE seems to be very expensive for what it offers:
http://www.testequity.com/products/5188/

Now compare that with the RTO:
http://www.testequity.com/products/4980/

The RTO is priced in the same range as it's competition but the RTE is pretty expensive.

I'm sure you can get some discounts from R&S but then you should get the same from any other vendor.

lower (acoustical) noise, colourful leds everywhere (TOO bright though!),

If I remember right the brightness of the LED indicators can be dimmed so that it doesn't burn your retina 

5Gs/s, support by R&S, price, better triggering and umm, price again. Im trying to justify the HDOs extra price to my boss, but it will not be easy...

Well, if you can live with 8bit, how about a WaveSurfer 10 then? 1GHz 10GSa/s for S10k:
http://teledynelecroy.com/oscilloscope/oscilloscopeseries.aspx?mseries=470

Even the WaveSurfer 10M is cheaper:
http://teledynelecroy.com/oscilloscope/oscilloscopemodel.aspx?modelid=9615&capid=102&mid=504
(and it comes with spectrum analyzer mode)

In the same price range as the RTE1054 there's also the WaveRunner 606zi:
http://teledynelecroy.com/oscilloscope/oscilloscopeseries.aspx?mseries=352

Considering the available alternatives, not sure why one would want to buy a RTE.