GDS-2000A series New economic oscilloscope by GW Instek comes to market

[marmad]

(It is possible that I did something wrong, but I don't think so.)

Well, I'm not sure if I see the frequency correctly displayed on the Hameg - is that 271Hz it's showing? But in any case, if you were trying to see if the GW-Instek would do 80k waveforms per second, you have it set to the wrong horizontal timebase. No DSO does it's maximum wfrm/s at every timebase setting - it varies dramatically. According to the GW-Instek documentation, you have be at the 500ns setting to get the 80k rate - and it seems you have it set to 10ns. So there's no way to know - without running tests - what the average wfrm/s update rate is at that setting - unless it's ~270.

[grego]

Does anybody know if all models have the 80,000 waveform update rate? I read through the manual and didn't see where it says. I did notice however that the logic analyzer comes with I2C, SPI, and UART decoding only. Nothing else.

Yeah that was disappointing to see.  I got a little excited about this scope until I saw that - I kind of need/want CAN too so if I'm going to go with a MSO I want to it be able to handle that as well.  Too bad there are no reviews anywhere on this thing so far.

[marmad]

Does anybody know if all models have the 80,000 waveform update rate? I read through the manual and didn't see where it says. I did notice however that the logic analyzer comes with I2C, SPI, and UART decoding only. Nothing else.

GW-Instek have published documents stating that the scope can reach 80k wfrm/s - and I would suppose that it's true (the company seems to be careful about publishing specs they can hit). And why wouldn't it be? It's not unbelievable given the current new generation of DSOs (i.e. Agilent InfiniiVision X series [50k - 1M], Rigol UltaVision series [50k - 110k], etc). Besides, as mentioned many times before - that would be the FASTEST the DSO could capture - at just ONE particular timebase setting. As shown in this chart I created for the Rigol DS2000, waveform update rates change constantly based on timebase settings, channels turned on, etc.

[grego]

GW-Instek have published documents stating that the scope can reach 80k wfrm/s - and I would suppose that it's true (the company seems to be careful about publishing specs they can hit). And why wouldn't it be? It's not unbelievable given the current new generation of DSOs (i.e. Agilent InfiniiVision X series [50k - 1M], Rigol UltaVision series [50k - 100k], etc). Besides, as mentioned many times before - that would be the FASTEST the DSO could capture - at just ONE particular timebase setting. As shown in this chart I created for the Rigol DS2000, waveform update rates change constantly based on timebase settings, channels turned on, etc.

Marmad - I wasn't commetning about the wfm/s - I'm assuming you just hit "quote" on my post rather than reply.

I'm tempted to suck it up and order one of these to be the guinea pig for everyone - but I'd hate to get screwed in some aspect of it.  I've been looking for a good bench scope to complement/replace my Picoscope 3206A as I don't like having to lug my laptop to my bench all the time and, on paper at least, the price/performance/features on this are looking good.  I'm slightly cautious based on the interface, I'm not sure I'm loving it (compared to the Rigol/Agilent interfaces) from what little I've seen.

[marmad]

Marmad - I wasn't commetning about the wfm/s - I'm assuming you just hit "quote" on my post rather than reply.

Ha ha - yeah, sorry, I just noticed the mistake I made with the quote - and went back and edited it to the correct one. 

I'm tempted to suck it up and order one of these to be the guinea pig for everyone - but I'd hate to get screwed in some aspect of it.  I've been looking for a good bench scope to complement/replace my Picoscope 3206A as I don't like having to lug my laptop to my bench all the time and, on paper at least, the price/performance/features on this are looking good.  I'm slightly cautious based on the interface, I'm not sure I'm loving it (compared to the Rigol/Agilent interfaces) from what little I've seen.

Well, of course, I would be happy if you would be a guinea pig and review one of these  - but I understand your reticence. I agree with you in all your points; I actually think between the three fairly similar options (this scope, a Rigol UltraVision, and an Agilent X), perhaps the GW-Instek is the best bang-for-the-buck (although the current version of firmware on Rigol DS2000s allows you to keep the 'options' indefinitely - so that adds a lot of value), Rigol has the best interface, and Agilent wins hands down in the support department (not only service but external software, etc).

[grego]

A couple of other concerns I have without more information.

A) 8x10 divisions kind of bothers me - scopes in this price range should be at least 12 horizontal
B) Sample rate is listed as "Max 2Gs/s" -- but it doesn't say if that's per channel, interleaved, what?  If you have a 4 channel scope does that mean if you're using all 4 channels you're now at 500Ms/s, or are the separated? 

[Hydrawerk]

According to the specs, the GDS-2000A series has 1GSa/s per channel, but if you use half channels, it will be 2GSa/s. So I think that If you use all four channels, each one will have 1GS/s.
The Trig out signal was very unstable and yes, it was about 270 Hz... What was wrong?? As you wrote, the Rigol's update rate varies a lot, but it is still a few thousands waveforms/sec. Worse than Agilent, but still not bad at all.

Here are the results of a quick test I did measuring the Rigol DS2072 waveform update rates at all timebase settings and memory depths (also attached in Excel format). If you compare these to the Agilent 2000X series published rates, it's obvious the Agilent is the clear winner - although it doesn't have anything close to the possible memory depths.

Single channel - 1MHz square wave to Channel 1 - Trigger Out to frequency counter. I took best-case rate when it was fluctuating:

	14kPts	140kPts	1.4MPts	14MPts	56MPts
5ns	15,000	13,150	1,412	142	36
10ns	9,400	9,400	1,412	142	36
20ns	50,012	13,515	1,416	142	36
50ns	25,003	13,515	1,416	142	36
100ns	17,859	13,159	1,412	142	36
200ns	11,365	11,360	1,408	142	36
500ns	5,434	5,435	1,336	142	36
1us	5,263	2,890	1,126	139	35
2us	5,054	1,506	846	133	35
5us	4,425	1,176	733	130	35
10us	3,789	1,157	720	130	35
20us	2,945	992	442	117	34
50us	1,326	639	414	114	34
100us	683	421	306	94	32
200us	347	245	200	69	28
500us	140	109	97	39	21
1ms	70	56	52	29	15
2ms	35	29	27	19	10
5ms	~14	~13	~11	~9	~6
10ms	~7	~6	~6	~5	~3
20ms	~4	~4	~3	~3	~2
50ms	~2	~1	~1	~1	~1
100ms	~1	~1	~1	~1	~1

The DSOX-2000 update rate does not go lower when you set a faster timebase...

[Hydrawerk]

Well, the waveform update rate of your former Owon SDS7102 was slow, but the Trig out signal was stable. Why not at GDS-2000A??? What else to say? The plug-in modules with 3MHz waveform generator and probably even the logic analyzer cannot be bought today, because they are still under development. They were not shown at the trade fair. Unfortunately there was no complex signal source (e. g. video signal) to check the digital phosphor (or visual persistance) capabilities. The autoset process is rather slow. There is no protective glass in front of the display, that you marmad, do not like. 

[grego]

Well, the waveform update rate of your former Owon SDS7102 was slow, but the Trig out signal was stable. Why not at GDS-2000A??? What else to say? The plug-in modules with 3MHz waveform generator and probably even the logic analyzer cannot be bought today, because they are still under development. They were not shown at the trade fair. Unfortunately there was no complex signal source (e. g. video signal) to check the digital phosphor (or visual persistance) capabilities. The autoset process is rather slow. There is no protective glass in front of the display, that you marmad, do not like. 

The 8 channel LA module is available for sale from tequipment.net -- the 16 is a "contact us" thing.  The DDS is also listed for sale from tequipment ($245 USD).

[marmad]

The DSOX-2000 update rate does not go lower when you set a faster timebase...

Although, to be objective, these are Agilent's own published stats. Probably (more or less) accurate, but I wouldn't mind seeing a third-party with no financial incentive run those tests on a DSOX-2000.

Well, the waveform update rate of your former Owon SDS7102 was slow, but the Trig out signal was stable. Why not at GDS-2000A???

Well, to be fair, you were at a trade show. Are you sure the cable wasn't faulty - or some other minor thing? Hard to know for sure under those conditions.

There is no protective glass in front of the display, that you marmad, do not like. 

That's good - I'm glad they ditched that. The display does look nice.

The 8 channel LA module is available for sale from tequipment.net -- the 16 is a "contact us" thing.  The DDS is also listed for sale from tequipment ($245 USD).

BTW, Rigol is currently developing LA and DDS modules for the UltraVision line as well. I doubt they will be backward compatible with older models (like mine) but I don't really care since I prefer a separate AWG and LA. But I thought it might be interesting for people thinking about purchases.

[Hydrawerk]

The 8 channel LA module is available for sale from tequipment.net -- the 16 is a "contact us" thing.  The DDS is also listed for sale from tequipment ($245 USD).

I don't know... Maybe it is an preorder only.

[Hydrawerk]

marmad I think that the cables were OK. The other Demo signals were good. It looks like there was an error in the GDS-2000A.
The Agilent DSOX-2000 should have a good update rate, because they use a special custom designed ASIC, that does all the stuff. Rigol and probably GW Instek have nothing like that.

[marmad]

A) 8x10 divisions kind of bothers me - scopes in this price range should be at least 12 horizontal

This brings up one of the things that bothers me about DSO manufacturers - a way they can 'juke' the stats. They mention the screen size in pixels - but they don't bother to mention what portion of screen real-estate the waveform is actually mapped to. The Agilent X2000 series is 10x8 (non-square) divisions mapped to only 640x400 pixels of screen space. I'm happy that the Rigol UltraVision series at least increased that to 700x400 (with 14x8 square divisions).

BTW, the number of horizontal divisions is also a factor in blind time calculations. So if you calculate their fastest waveform update rate by divisions, you get the following:
Agilent DSOX2000: 54k * 10 = 540k
Rigol DS2000: 50k x 14 = 700k

So, at least at the 20ns timebase setting, the Rigol DS2000 series has a 23% less average time to capture a glitch than the Agilent DSOX2000 series.

Edit: Although to be fair, the Rigol is not displaying a chunk of the vertical scale (and gaining some corresponding speed because of that) - I couldn't say what the Agilent does in this regard, but it's another way in which manufacturers can bend statistics in a desirable way.

[Hydrawerk]

New videos made by ITTSBEurope.

[marmad]

One other point I want to mention connected to my post above (since this is a thread about the new GW-Instek): since the number of divisions affects the blind time calculations, you might notice that the Rigol DS2000 and the GW-Instek 2000 series are closer in this respect than you might first think:

Rigol DS2000: 700k divisions per second maximum.
GW-Instek GDS-2000: 800k divisions per second maximum.

[Hydrawerk]

The method of measuring the waveform update rate according to GW Instek presentation. Well, nothing special at all.

[grego]

New videos made by ITTSBEurope.

Thanks Hydra - that was more than we've really seen anywhere else.

I will admit I'm really turned off by that interface.  I don't know what it is but for some reason it just looks (to me anyway) extremely ugly. 

Marmad the Rigol 4104 is the one I'm comparing this against (and one I'm looking at as well, either that or a used Agilent MSOX2024A).  Rigol just seems, for the most part, have their act together.  It's anecdotal for me, and it's not like I need anything crazy so usability is a huge factor (mainly model railroad electronics work with square-wave AC stuff and CAN bus).

I'm with you though - knowing how the manufacturers actually use the screen real-estate is something sorely lacking in all the specs.

[marmad]

I will admit I'm really turned off by that interface.  I don't know what it is but for some reason it just looks (to me anyway) extremely ugly.

Yes, I'm not thrilled by their design aesthetic either - although I do give them kudos for using 750x480 of the 800x600 screen for waveform mapping. But unfortunately, like the Agilent, they go with a non-square graticule, with 75x60 pixel divisions. What is up with equipment manufacturers doing that?

Instead, they should have increased the horizontal divisions to 12 - and used 720x480 with square divisions. This would have decreased their publishable waveform update rate but it would have, imo, increased the usability of the scope.

[grego]

Bottom line until someone buys one and shares their experiences with it we're going to keep guessing.

Part of me wants to go "sure, I'll do it" but for my first bench scope replacing my USB scope I'm a bit nervous about jumping in.  Honestly if the interface didn't look like it was made 10+ years ago I'd probably be more willing to be the guinea pig but I'm waffling back and forth right now.

Come on Instek - send Dave a demo unit so we can find out about this thing!

[marmad]

Bottom line until someone buys one and shares their experiences with it we're going to keep guessing.

Part of me wants to go "sure, I'll do it" but for my first bench scope replacing my USB scope I'm a bit nervous about jumping in.  Honestly if the interface didn't look like it was made 10+ years ago I'd probably be more willing to be the guinea pig but I'm waffling back and forth right now.

Come on Instek - send Dave a demo unit so we can find out about this thing!

Honestly, I would be more worried about GW-Instek's seemingly complete lack of ability to promote, market, followup, and (according to some) service their scopes then I would be about the interface. I wrote them two detailed emails (cc'ed to various GW-Instek facilities around the globe) expressing interest in the 3000 series without the tiniest response from them - and I've heard many similar reports from others. Just look at the 3000 series - it was out for a couple of years and the only 'reviews' of it on the web were videos created by GW-Instek employees. I would be seriously worried about ever getting firmware updates from them to fix bugs.

[grego]

Honestly, I would be more worried about GW-Instek's seemingly complete lack of ability to promote, market, followup, and (according to some) service their scopes then I would be about the interface. I wrote them two detailed emails (cc'ed to various GW-Instek facilities around the globe) expressing interest in the 3000 series without the tiniest response from them - and I've heard many similar reports from others. Just look at the 3000 series - it was out for a couple of years and the only 'reviews' of it on the web were videos created by GW-Instek employees. I would be seriously worried about ever getting firmware updates from them to fix bugs.

A completely valid point.  And a good one too.  It's one thing to get no support on a $400 piece of kit, but if I'm spending $2000+ on something I'd like to know it'll be supported.  Makes you wonder how they stay in business.  I may stick with my plan of a Rigol 4014 or the used Agilent MSOX2024A.

[Hydrawerk]

Honestly if the interface didn't look like it was made 10+ years ago I'd probably be more willing (...)

What exactly don't you like?? I think that this scope is easy to use, I tried it... But yes, it has no pushable knobs and it even has no fine vertical scale setting. But on the other hand, Tektronix also has almost no pushable knobs. Maybe Tek guys think that pushing knobs are unreliable...

[marmad]

A completely valid point.  And a good one too.  It's one thing to get no support on a $400 piece of kit, but if I'm spending $2000+ on something I'd like to know it'll be supported.  Makes you wonder how they stay in business.  I may stick with my plan of a Rigol 4014 or the used Agilent MSOX2024A.

I'd go with the Rigol DS4014 (if I could have afforded the 4 channels I would have bought one of those instead of the DS2072) - I couldn't live with the miserly 100k of memory on the X2024. Have you seen the review and teardown of the DS4014 that was posted in the last week?

[grego]

What exactly don't you like?? I think that this scope is easy to use, I tried it... But yes, it has no pushable knobs and it even has no fine vertical scale setting. But on the other hand, Tektronix also has almost no pushable knobs. Maybe Tek guys think that pushing knobs are unreliable...

This one is simply personal preference but I don't like the Tek interface either.  Too much blue, the fonts look dated, etc.  Again, highly subjective.  Sometimes you like what you like.  I personally prefer the "cleaner" (again, opinion here) look of the Agilent and Rigol (black backgrounds, etc).  That's simply esthetics, and I get it, but it's something individuals have to keep in mind.  The pushable knobs I don't really care about overall - that's pretty minor (and can be annoying if selections change as you're pressing).

I look at a 'scope like a car - everyone has things they want, everyone has things they like.  No one will ever agree about everything.   I am, however, glad to hear that you found it pretty easy to use at the show.  At least someone here on the forums has touched the beastie! Out of curiousty how would you compare the output (screen display) in person compare to, say, a Rigol or Agilent?  The pictures posted are all a mess - considering that the screen is 800x400, the pictures are larger than that so you can't use them to identify pixel doubling and whatnot.  I ask because it's sometimes hard to get an idea from the straight video.

Marmad: yeah I hear you - someone on this forum said something that I try to live by: buy once, cry once.  I'd rather wait an extra month or so and buy something that has more headroom for me to "grow into" since it's hopefully a piece of kit I'll have around for years and years.

[Hydrawerk]

Well, the Agilent interface is also nice... The Tektronix is dated in general... 
The display of GDS-2000A looks good. It has better viewing angle than Rigol DS6000. The Rigol DS2000 was not presented at the trade fair. I don't know why. The GDS-2000A has even the color gradation, is it useful? It surely is a not common feature among middle class scopes...