Agilent DSOX3014A vs Tek DPO3014 vs Rigol DS4014

[kek]

Funny thing is, just today Rigol were going to send me a DS4000 by mistake, they were supposed to be sending the DS2000.
I told them to send the DS2000 instead.

They should send you both!
I'm really interested in a new 4 channel scope myself. The Rigol DS1052E has served me well, but would like to upgrade in the next year or two to something with 4 channels and a bigger screen. Any scope maker with a good 4 channel scope would be crazy not to send you an evaluation scope, how many DS1052E scopes have you helped move in the last couple of years?

Ken

[EEVblog]

They should send you both!
I'm really interested in a new 4 channel scope myself. The Rigol DS1052E has served me well, but would like to upgrade in the next year or two to something with 4 channels and a bigger screen. Any scope maker with a good 4 channel scope would be crazy not to send you an evaluation scope, how many DS1052E scopes have you helped move in the last couple of years?

They might send me both if I ask.
The issue is time to review and tear down all this stuff.
I already have a big backlog.
They were going to send me the 4000, but I asked for the 2000 instead as I think there would be more people interested in that.

Dave.

[snoopy]

The Rigol 2000 series looks quite impressive according to its datasheet if you just want 2 channels and 200MHz bandwidth max. http://emona.com.au/Datasheets/2%20Oscilloscopes/DS2000.pdf

With a 50,000 waveforms per sec performance on the Rigol 2000 the low end dso market is slowly moving the performance/price goal posts and are encroaching on the big boys domain. Healthy competition can only mean good news for the consumer What will we see in 5 years time ??

regards
davo

[T4P]

Checklist :
Rigol DS4034 and many more ... lol

[Smokey]

It would be cool if you reviewed whichever model is more directly a competitor with the agilent 3000x series scopes since the 3000x series seems to be the general purpose leader in your reviews so far.  It sounds like either rigol will give a good idea of their new stuff in that class though.  Ill take what Ican get.

Could you maybe fast track the scope review?  It seems like more people would be interested in that than a spectrum analyzer.

[EEVblog]

What will we see in 5 years time ??

Hopefully a new design Tek scope! 

Dave.

[firewalker]

I am guessing Tek would love to poke you with a razor sharp probe! 

Alexander.

[mikeselectricstuff]

What will we see in 5 years time ??

Hopefully a new design Tek scope! 

Dave.

By which time Agilent will have something new. Unless Tek are already a long way into development of something good, they will be  dead in the low to midrange scope market soon. There's only so much they can cut prices on their current creaky old  product line. 

[KTP]

Does the Tek or Rigol have similar software development support for external control of their scopes as Agilent?  We have already started work on the software that will turn a 2000x/3000x into a bode plotter for filters and a RLC meter.  I am very impressed with the amount of work Agilent has done documenting the VISA libraries, including example code in C#.

Just another factor to consider if someone is choosing between these three scopes.

[Smokey]

Ktp, that is too funny.  Making automated measurements to do bode plots is exactly the first application I need to do.  If you search the web for dedicated bode hardware, its like 12k usd for a box that just dumps out bode plots.   

Edit to add:
How long does it take to dump the agilent memory to a pc?

Are you getting good results from the agilent taking automated measurements?   Does anyone know if the rigol uses the full memory for calculations like the tek?

When I asked the rigol guys about remote pc control they said it wasn't a problem and I didn't need anything extra to do it, but I still haven't played with one yet to confirm. 

[KTP]

Ktp, that is too funny.  Making automated measurements to do bode plots is exactly the first application I need to do.  If you search the web for dedicated bode hardware, its like 12k usd for a box that just dumps out bode plots.   

Edit to add:
How long does it take to dump the agilent memory to a pc?

Are you getting good results from the agilent taking automated measurements?   Does anyone know if the rigol uses the full memory for calculations like the tek?

When I asked the rigol guys about remote pc control they said it wasn't a problem and I didn't need anything extra to do it, but I still haven't played with one yet to confirm.

We really just started, so I can't answer all of this right now.  The first thing my wife wants is a full written spec (uhg, software engineers...can't they just start hacking it like we do?).

Just from playing around though, I found it quite fast to do things like change the waveform frequency.  You can almost make it look like a regular sweep function generator...it is that fast.  Memory dump to pc is fast also, but don't have a specific time measurement yet.

I will let you know more as we progress.  I want to make it super easy to use for beginners, but have all the features for advanced users.

[Smokey]

Well I just had a really interesting on site demo of the Tek scope.  (Thanks Tek guy, if you happen to read this.) The guy had a pretty good sales pitch.   

Some of the cool new stuff and info:
1) The Tek 3000s were introduced in 2008, so it's not really that old.  The previous generation that the 3000 replaced is about 15 years old now.
2) The newer 3000 series tek scopes are software bandwidth upgradable all the way to 500mhz.  The older ones were not calibrated for anything other than the set bandwidth, but the newer ones, starting with SN C02 are factory calibrated for all bandwidths and just require a software upgrade.  No sending it back required.  They just started doing that, probably to compete with the Agilent's upgradability.  There is no upgrade path from DSO to MSO for these models though.  He confirmed a very small percentage of people actually ever upgrade anything.  That is at least an option now though.  I guess if you have an older serial number you can still send it back and have them upgrade it with the new calibration.
3) The decode expansion modules, which are those stupid physical hardware key token things to plug into the side by the top, can be stored digitally in the scope so you don't have to have them physically in there to use them and potentially get them stolen if you are in that sort of environment.  The hardware token thing is deactivated when you load them on the scope though, so you can't install the same token in a bunch of scopes.  Not that there is much he could do about it, but I did complain that the modules are ridiculously priced for the cheaper units, being over 25% of the price of the whole scope, per expansion module.
4) He said that they recently did a detailed side by side comparison with the Agilent 3000x series and a bunch of that info will be coming out soon. 
5) He made a point of Tek having some big company syndrome when it comes to advertizing and they are just now starting to learn how to use things like youtube to help advertize their stuff to the middle and lower markets.  So I guess we should be expecting to see more stuff like that.
6) The scope comes with a lite version of labview to help automate stuff.  That could be super useful if you had labview experience and wanted to set something up quickly.

Some impressions, understanding that I still haven't played with an Agilent or Rigol yet, so this is just my impression of the Tek having played with it for about 30 minutes.
1) I didn't have any of the digital mixed signal stuff going on during the demo and I didn't see much lag in the interface.  It seemed pretty responsive, even on a long timebase at full memory.  Could the digital+MDO stuff have been causing the lag?
2) As far as usability of the interface, I thought it was fine.  There are a lot of deep menus, but there are also a lot of features.  Nothing was too outrageously placed, menu or button wise.  Things that were non-intuitive like having to turn a knob instead of push a button to set something was pretty well noted what you had to do on the screen menus.  The panel buttons were fine as far as amount of them, placement, and feel.  Nothing that would drive me crazy on a daily basis.  Using anything this complicated is going to take some getting used to but I didn't see anything that would be long term a deal breaker.
3) I'm not sure how much I like the wide screen non-square boxes, but it didn't seem like something that would drive me crazy.
4) The triggering looked pretty nice.  Lots of options for setting up ranges and conditions.  I could see that being useful.
5) Having the measurements use the full memory instead of just the screen data (I think he said like 1000 points on the Agilent) would be nice.  It seems that once you have all that data in a digital format, which these scopes do anyway, it would just be a software thing to do math over the full memory.  I wonder what Agilent was thinking restricting themselves like that, and why they don't write some new code and patch in those features since it's something their competitor is rightfully advertizing as an advantage.

So to sum it up, first of all I'm not a Tek fanboy.  This is also the first scope in this class that I have physically played with.  I'll hopefully get the others here soon.  On a purely technical basis, I didn't see too much wrong with the Tek in the brief time I played with it, and I do like the fact that the memory/sample rate isn't split between the channels and the math is over the whole memory.  I may request a demo unit to keep for a while if I'm still undecided after seeing the Agilent and Rigol, but I don't think I'd regret going with the Tek and using it on a daily basis.  If I consider the equivalent upgrades to the Agilent to bring it up to about the Tek specs (500USD for the memory upgrade to 4meg, and 400USD for the lan/VGA module) and ignore the wavegen, the Tek list price of 4050USD isn't too outrageously priced.  I haven't got their actual quote yet, but it should be better than that. 

When I mentioned EEVBlog, he only rolled his eyes a little.  He did say he knew of it and that he was impressed at Dave's ability to review things like the zipper on the case the scope came in and make it sound interesting.  I had to chuckle.

Just some more info for everyone.

[KTP]

Can you give a couple of examples of when you might find it useful to do math functions over the entire memory and not the displayed data?  Just curious because I have not thought about this being much of an issue.

[Smokey]

If I was using the built in frequency measurement, or peak-peak measurement, or some measurement of a stable repetitive signal like a sine wave, the more periods of the signal you do the calculations on, the more accurate the measurement will be.  Say I only wanted one period of the sine wave on the screen so I could see the quality of the signal, but I also wanted accurate frequency and peak-peak measurements, it sounds like it would only do the measurements on the one period on screen vs the average of a bunch of periods that are technically in the scope's memory.  It sounds like that as you scroll through the memory of a capture, the agilent's measurement results change as the screen image changes. 

The std deviation and min and max and all the other statistics that these scopes show these days should take care of displaying the variation from period to period that the scope used to get the average measurement value.

I could be wrong though, it has happened before.

[mikeselectricstuff]

3) The decode expansion modules, which are those stupid physical hardware key token things to plug into the side by the top, can be stored digitally in the scope so you don't have to have them physically in there to use them and potentially get them stolen if you are in that sort of environment.

I wonder if like the older scopes, these modules are still just eeproms, with the license name in ASCII. If so give me that any day over real security!

[EEVblog]

Well I just had a really interesting on site demo of the Tek scope.  (Thanks Tek guy, if you happen to read this.) The guy had a pretty good sales pitch.   

Some of the cool new stuff and info:
1) The Tek 3000s were introduced in 2008, so it's not really that old.  The previous generation that the 3000 replaced is about 15 years old now.

Whilst that is true, I think it was still based on the same processor and architecture platform as the previous model. Can anyone confirm?

2) The newer 3000 series tek scopes are software bandwidth upgradable all the way to 500mhz.  The older ones were not calibrated for anything other than the set bandwidth, but the newer ones, starting with SN C02 are factory calibrated for all bandwidths and just require a software upgrade.  No sending it back required.  They just started doing that, probably to compete with the Agilent's upgradability.  There is no upgrade path from DSO to MSO for these models though.  He confirmed a very small percentage of people actually ever upgrade anything.  That is at least an option now though.  I guess if you have an older serial number you can still send it back and have them upgrade it with the new calibration.

Agilent have told me a similar thing in terms of bandwidth, few people actually take up the option.
I can imagine the LA upgrade to be a bit more popular, as it gives you different capability, not just more performance.

Some impressions, understanding that I still haven't played with an Agilent or Rigol yet, so this is just my impression of the Tek having played with it for about 30 minutes.
1) I didn't have any of the digital mixed signal stuff going on during the demo and I didn't see much lag in the interface.  It seemed pretty responsive, even on a long timebase at full memory.  Could the digital+MDO stuff have been causing the lag?

Yes, the more stuff u turned on, the slower it got.
Although the MDO is the 4000 architecture, not the 3000.

5) Having the measurements use the full memory instead of just the screen data (I think he said like 1000 points on the Agilent) would be nice.  It seems that once you have all that data in a digital format, which these scopes do anyway, it would just be a software thing to do math over the full memory.  I wonder what Agilent was thinking restricting themselves like that, and why they don't write some new code and patch in those features since it's something their competitor is rightfully advertizing as an advantage.

Perhaps it's an architecture thing.
Agilent use the ASIC to push the waveform direct to the screen (and possibly do the measurements I think), so it might require changing the ASIC instead of the processor software?

When I mentioned EEVBlog, he only rolled his eyes a little.  He did say he knew of it and that he was impressed at Dave's ability to review things like the zipper on the case the scope came in and make it sound interesting.  I had to chuckle.

Once again, I didn't get a single Tek person post a response in any way to my review of the MDO4000 or the 2000 teardown. Only a personal email to mention that I got some nit-picky things wrong like it takes 4 modules instead of two etc.
Why is that I wonder?
It's because they know I'm right and they didn't have a response. The MDO4000 was as slow as a wet week.

If they thought their regular scopes were hot stuff, they would be sending them to me for review. But they haven't, they only sent their new MDO scope. And they forget that I praised them for it, the MDO part was excellent, I told people to rush and out buy one if they needed MDO capability. But as an everyday use scope it kinda sucks, and I said so. They didn't like that, but deep down they know I'm right.

Dave.

[Rufus]

5) Having the measurements use the full memory instead of just the screen data (I think he said like 1000 points on the Agilent) would be nice.  It seems that once you have all that data in a digital format, which these scopes do anyway, it would just be a software thing to do math over the full memory.  I wonder what Agilent was thinking restricting themselves like that, and why they don't write some new code and patch in those features since it's something their competitor is rightfully advertizing as an advantage.

I think there is some confusion here. The 2/3000X capture and display is always the 'full memory'. You have the option of making measurements on the display or the (mega)zoomed window.

The Agilent megazoom could be criticized for splitting the display vertically between the capture and zoomed window which makes both cramped and low res. If it had the option of only displaying the zoom then it would be capturing more than it displays (which is probably what the Tek can do).

[Hypernova]

Having used Tek's MSO3000's at my last job I have nothing good to say about them. The moment you try to do anything with 5Mpt the thing freezes like it's in Antarctica. God help you should you turn on the digital channels and have a sleeping bag ready should you have the Math active too.

The most retarded thing is the way the digital channel work. Turn them on and you can't use Rolling display, you are stuck with normal view, good luck trying to sample digital channels over long periods.

For the same BW 3000X's are faster with more features, and did I mention that due to the wavgen integration you can copy acquired wfm from your input and recreate it with the arb gen?

[T4P]

Plus the DPO3 is very slow compared to the X3000, so the agilent wins in this aspect by far bounds
I don't see how nobody mentioned this that the X3000 has 1Million Plus waveforms/sec waveform acquisition rate

That alone makes me buy a Rigol or Agilent instead of a Tek
The rigol has 110kwfrms/sec and that alone makes me buy the DS4014 instead, and it's being much cheaper on top of that

[EEVblog]

Plus the DPO3 is very slow compared to the X3000, so the agilent wins in this aspect by far bounds
I don't see how nobody mentioned this that the X3000 has 1Million Plus waveforms/sec waveform acquisition rate
That alone makes me buy a Rigol or Agilent instead of a Tek
The rigol has 110kwfrms/sec and that alone makes me buy the DS4014 instead, and it's being much cheaper on top of that

It's not just the waveform update rate (which changes a lot depending upon timebase), but the Tek slows down a lot when you try and do other things, just like Hypernova said. I found that on the Tek 4000, and also on the older Tek 3000 I've used.

Dave.

[tnt]

I wonder if like the older scopes, these modules are still just eeproms, with the license name in ASCII. If so give me that any day over real security!

Ahaha, "Real security", right ... good one

[KTP]

The Agilent AWG does update fairly fast over USB, as I tested this morning.

I wrote a simple program to sweep a RC filter (10K resistor, 100pF capacitor) with a 2vpp sine wave from 1Khz to 1Mhz in 100 steps per decade.  I also added a 10uS 3V pulse at the start of the sweep to allow me a trigger point (I am sure there is some better way to do this).  At any rate it looks neat 

Corner frequency should calculate to about 160khz, but the 100pF cap is surely getting some loading from the 10x probe!

[Smokey]

So a Tek rep contacted me about how the demo went.  I pointed them to this thread.  I wonder if they will actually read it.

[H.O]

Hi everyone,
New forum member here though I've been following Daves video blogs since the beginning.

I've had a Rigol DS4014 for a couple of months now. I'm not in a position to do a full review of it and I don't have any hands on experience with the "rival" Teks or Agilent units discussed here so I can't comment on hands on differences either. However I can give you my overall impressions for what it's worth.

My use of the scope is for general purpose stuff in my home "lab". I tend to fool around a bit with motor controllers, quadrature encoders and such and wanted a 4 channel model. I originally gravitated towards the Rigol DS2000 series but was a bit bummed since there's no 4 channel model available. That forced me into raising my budget (this is payed for with my own money so price IS unfortunately a factor) and turning towards the DS4000 series. I did briefly concider the Agilent3000 series but to me the Rigol looked better on paper - at a lower price.

Basically there are two things so far that I don't particullarily like:
I have not been able to get the PC-software for it to work. This isn't something I'll likely use and I've only played with it for any hour or so. I'm sure it's something I'm doing wrong and I need to RTFM properly. What I did find a bit strange though is that you have to download third party software from National Instruments in order to install the Rigol software, this was somewhat cumbersome since the file that Rigol points you to in the documentation doesn't exist on the NI site, I had to ask support for a direct link.

The second and more annoying thing for is the fan. In rbola's video review he says that he doesn't find the fan very noisy. On that point I have to disagree, I actually find it quite annoying. I can't give you a number but compared to any of my 4 desktop PCs around the house and my Tek2465B scope the Rigol is MUCH louder.

Apart from those two things it's been a pleasure to use. The screen is bright and clear (much better than on my old GWInstek 806C) and the userinterface is intuitive and responsive.

I ordered mine without any options, I definitely could use both the RS232 and SPI decoding module but I concider them too expensive for me at this point.

Like I said, not much of a review, just a couple of observations after having it for a while. If you have any questions ot if there's anything in specific you want me to test I'd be happy to do that if I have the equipment to do it.

/Henrik.

[jimjam]

Hi H.O. I'd be interested in hearing your opinion on Rigol DS4014 now that you've been using it for a while. Thanks