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

[grego]

Well the 2204A is in transit now.  I should have it on Friday.  More to follow once it's up and running.

[marmad]

Well the 2204A is in transit now.  I should have it on Friday.  More to follow once it's up and running.

Nice to hear - you'll be having some fun soon    Any chance you 'll do a small written or video review here?

[grego]

Well the 2204A is in transit now.  I should have it on Friday.  More to follow once it's up and running.

Nice to hear - you'll be having some fun soon    Any chance you 'll do a small written or video review here?

Absolutely - I'm not really set up for video but I will see what I can do.

[grego]

It has arrived... and yes, that green tape is our friends at Homeland Security and Border Protection having opened my box when it hit the shores in Long Beach.  Go-go government bureaucracy!



(more coming later once I get home and play with it over the weekend)

[Hydrawerk]

Well, it's a decent scope with great screen. It looked good at the trade fair in Czech Republic. But I bought the DSOX2002A instead, it is a sweet scope especially if you need a 20MHz function generator. Moreover, it is very easy to download and install the firmware.

Please check if the set/clear button works on your GDS-2000A . It should place marks. But there seems to be a software bug.
Then there might be a problem with the Trig Out signal, it is too low. It might be impossible to use it for synchronising with other test equipment...

[grego]

Set clear works.  I just grabbed the square wave demo signal in single shot, turned on search and started setting/clearing marks.  You get "not supported in this mode" if you don't have search turned on.

I'll check the trig out on mine when I get it home.  Interestingly the FW version is already up to 1.08.

[grego]

Ok - played with it for about 30 mins or so now - I'm at my office so can't generate any signals other than the demo signals on the scope so that'll have to wait until I get home.

As mentioned I had no problem with set/clear with search turned on.

It seems pretty responsive when changing settings, etc.  Demo signal is a simple 1kHz square wave (see pic).  I will admit I prefer the "look" of the screen on the Rigol but I will also say this screen is huge - at least it feels like it.  There is no pixel doubling as far as I can tell.  What's interesting is the saved picture doesn't look as good as the image on the screen.  Go figure.  The screen is bright, clear and crisp.

A couple of things I wasn't thrilled with, and I knew these going in but it is a little annoying - since the knobs are not pushable I you have to default a fair amount if you get all sorts of stuff on your screen (marks, cursors, settings, measurements, etc).  I'd love to be able to "push to center" like on the Agilent. 

The fan is noticeable but pretty quiet.  Not too bad.  There don't appear to be any "warranty void" stickers so I'll open mine up this weekend and take some pictures.

Unfortunately I only have a 1Mhz func gen at home right now (damn you Rigol for not getting me my 4062!) but I'll run some tests with various signals with that.  Including marmads infamous "wfm/s" test -- after I check the trig out voltage to make sure my fluke can read it.

I should add that the default is sin x/x, but you can turn it off if you want under display options.

[Hydrawerk]

So probably there is no bug with the set/clear button. It can be used only in search mode. Couldn't those business guys at the trade fair tell me that??? https://plus.google.com/photos/106264218831814439783/albums/5857196858625060337
Anyway, this scope might be not so easy to use as my DSOX-2002A, but this GW Instek is much more feature rich. Ir has longer memory, segmented memory function and even pass/fail. On the other hand, the function generator is not free and it is only up to 3MHz.

[grego]

Yeah but to be honest I always planned on getting an external func gen anyway (Rigol 4062) to supplant my crappy 1Mhz generator.

I'm going to start digging into this thing this evening.  Gotta start making my bug/feature request list for the Instek folks.

[Hydrawerk]

The 20MHz generator in my scope is not arbitrary... But new firmware brings new functions! They recently added FM, AM and PSK modulation.  The arbitrary function may appear in two years or so. Who knows??

[grego]

Well someone followed Dave's advice *ahem* and "took it apart".  Aforementioned someone has some pictures, one below.  Wasn't a full tear-down but the skirt was lifted and some photos taken.

More coming as it gets put together.  There didn't appear to be any dodgy solder jobs, no flux I could see on the board, everything looked pretty tight actually.  At least that's what "someone" told me.

[JimmyMz]

Well someone followed Dave's advice *ahem* and "took it apart". 

I'm trying to research the ICs, and even with a photo magnification program, I'm having major issues viewing the model numbers. Would it be possible for "someone" to take some quality close-up photos, aiming for clarity? I'd really appreciate it. By the way, I grew up in LaGrange, also right outside Chicago. 

[EEVblog]

Interesting, AD Blackfin DSP just like in the Rigol.
I'd bet U1400 (the big one) is a Xilinx FPGA. Perhaps the other two heatsunk devices PLD's or smaller FPGA's.

[Hypernova]

The wavgen signals are coming from the removed cards on the PIC-E slots? Got any pics for the generator cards?

[Hydrawerk]

I am not sure if grego actually bought the signal generator plug-in module...

[grego]

The wavgen signals are coming from the removed cards on the PIC-E slots? Got any pics for the generator cards?

No - those slots are for expansion (LAN or the logic analyzer).  I think the generator is a software key upgrade.  I didn't really look into it since it's only a 3Mhz generator.

I'm trying to research the ICs, and even with a photo magnification program, I'm having major issues viewing the model numbers. Do you think you can't take some closer photos, aiming for clarity? I'd really appreciate it. FYI, I grew up in LaGrange, also right outside Chicago.

"Someone" has some more photos so I'll see about getting them up.  Shoddy camera work I know.

[grego]

There are a couple of bugs I have found.  Well one I'm not sure is a "bug" but might be a poor design decision.

BUG (and this is pretty damn minor):

- The graticule dots are on top of the image - e.g. they overlap the waveform.  Really the graticule should be the lowest layer and the waveform should overlay it.

BUG/POOR DESIGN?:

The trig out, as had been mentioned elsewhere, is pretty damn low voltage - ~180-220mVAC as far as my Fluke 289 can tell me.

[marmad]

There are a couple of bugs I have found.  Well one I'm not sure is a "bug" but might be a poor design decision.

BUG (and this is pretty damn minor):

- The graticule dots are on top of the image - e.g. they overlap the waveform.  Really the graticule should be the lowest layer and the waveform should overlay it.

BUG/POOR DESIGN?:

The trig out, as had been mentioned elsewhere, is pretty damn low voltage - ~180-220mVAC as far as my Fluke 289 can tell me.

Thanks for the continual updates, Greg. Hopefully those 'bugs' you found will be the extent of it 

[Hydrawerk]

The DDS 3MHz generator is a plug-in module according to the datasheet. But it is difficult to find any closer information. The Trig Out signal is damn low. The Taiwanese businessman at the trade fair told me not to use a multimeter! Use another scope instead. It is crazy.

[grego]

Some more pictures.

Just the back case off - power input module visible and heavy shielding.


Channel input stages:


Broad shot:


Closeup of the main DSP:

[Hydrawerk]

The scope looks nice. But the power supply seems to be somehow cheap. It has quite small heatsinks. On the other hand there is a real 230V power switch and easy replaceable glass fuse. And it seems that the whole scope is like 50% full of air...  The fan is quite small.

[grego]

In complete silence you notice the fan - it's not loud but it's a small fan so you do notice it.  With the normal noise I have around me (computer, music, etc) you don't notice it at all.  That's my general rule with fans - if I don't notice it when I have the stuff I normally have on it's all good.

The power supply seemed pretty solid to me - good components, no weird solder joints, silastic where it's needed, etc.  No clue what the MTBF is on it though.  I've checked the scope a couple of times while running and it runs quite cool - the fan is constantly on mind you but I've never noticed any hot spots but you're right - there's a lot of "space" inside the case -- which is a good thing!

[grego]

Ok - finally got my scope stuff straightened out.  I hooked the trig out to my picoscope and pulled a trace - 80k wfm/s as specified @ 500ns - and the output is actually fine voltage wise as you can see with the attached.  I think it's just too fast for my Fluke to be happy with it.

I'll run it through various time-bases and collect the data.

[grego]

Metrics of wfm/s.  Please note that this was done via my Picoscope and I'm not 100% happy with how it was reading some of the lower timebases so take it with a SLIGHT grain of salt.

All measurements taken with a 500kHz sine wave.  All results are in 1000s of wfm/s.  Scope set to short memory depth.                  
                  
Timebase     Min     Max     Avg     Notes      
10us         6.85    8.4     7.88         
5us          7.81    15.79   14.19         
2us          19.23   34.25   22.81         
1us          38.46   53.81   40.7         
500ns        35.71   83.33   77.23         
200ns        28.17   83.34   57.16         
100ns        23.15   83.34   47.08         
50ns         28.02   83.34   58.79         
20ns         18.72   48.9    27.48   Picoscope started having problems calculating here      
10ns         13.97   53.66   25.39   Picoscope started having problems calculating here      

[marmad]

Metrics of wfm/s.  Please note that this was done via my Picoscope and I'm not 100% happy with how it was reading some of the lower timebases so take it with a SLIGHT grain of salt.

Thanks, Greg, for finally (after the whole Kiriakos thing) doing this correctly    Respectable numbers across all ranges. I'm curious why 500ns is the sweet spot for the Instek - I wish I knew more about modern DSO design to comprehend what that reveals about the nitty-gritty of the circuit/firmware choices.