Year 2013 Oscilloscope Choices...

[Teneyes]

[..] Otherwise, get a Siglent. Solid FW, decent build, god features, etc...

Which button does that? I wouldn't mind one or two of those 

 It is the Button above the100 Souls /Devil 

[echen1024]

Exactly. Enter single shot, trigger level 680mv, and turn the Ch. 1 time base all the way up. Now, feed a 100uHz, 679mv p-p square wave in. While rapidly pressing the "Math, Trig. Menu, and Run/Stop buttons , put one hand on your ear and jump on one foot at exactly 16/5 jumps/sec. Keep doing this for 1 minute and 24 seconds, and you should see a new menu on the scope, with various god features, such as instant circuit fix, and 10,000 volt out on the BNC probes. 

[Hydrawerk]

 

[uboot]

Hi folks,

I looking for an entry-level DSO too and after reading through many threads on this excellent forum, I'm still considering the Uni-T UTD2102CM. There has not been updated infos / posts on that particular device for some time, so I'd like to add my findings and maybe your responses will help me with my decision:

The claim of 150k wfrms/s is real. It's not a typo as stated by some people. The shop that once wrote 15k corrected it to 150k meanwhile: http://www.pinsonne-elektronik.de/pi1/pd125.html Also, the manual available on that website not only states 150k and it also points out that this a 75x increase compared to other DSOs. (Well - it might still be a typo, maybe they're referencing a 20 wfrms/s scope, which would equate to 1500. On the other hand, 2k*75 equates to 150k, so I'm quite sure they are serious. There's also real typos in the manual like "Cutshort", "Vedio", etc.  )

Another detail that can be found in the manual: there's a hardware frequency counter. It also says, that this is equal to the trigger frequency for edge or pulse trigger mode. Maybe this is how wfrm update rate could be measured!


Unfortunately, there's no explanation on "Fast/Normal Acquisition Mode". Seems that the firmware at the time of writing the manual didn't have this mode switch.


Another finding: on page 81 is says "Waveform capturing rate >= 150k." In words: it's claimed to be greater than or equal to 150k. In other words: 150k is claimed to be the minimum rate 

Question: 150k wfrms/s, that is processing 150k*6k*8bits=900 Megabytes per second  Could it be that they just implemented a fast trigger which is able to interrupt / restart acquisition? On the other hand: a fast FPGA with fast but expensive FIFO-buffers might do the trick.


Other scopes on my list are Siglent SDS1072CML and hackable Hanteks and I am pretty undecided atm. Rigol is too pricy here in Germany.

Twintex TSO1062 - ever heard of that?


Edit: just forgot to mention as this has been asked in some places - the UTD2102CM has a shielded PSU: https://www.mortoncontrols.com/blog/files/category-oscilloscopes.html

[marmad]

The claim of 150k wfrms/s is real. It's not a typo as stated by some people.

Although this might be true, I don't think it's actually been confirmed by any independent source - so I'm not sure where you're getting your info that it's "real" from. From Uni-T's manual and published specs?

But a bigger question might be: if the Uni-T actually does 150k wfrm/s - how is it getting the extra captured information to the display? 120Hz is the upper end of refreshing an LCD, so DSOs with higher waveform capture rates generally use a Z-buffer and intensity-graded display in order to convey the extra waveforms captured between refreshes.

The Unit-T has no intensity-graded display - so, for example, if it's capturing 150k waveforms per second and refreshing the LCD 100 times per second, what is happening to the 1500 waveforms it captures between refreshes? Are all 1500 waveforms being combined into a single, uniformly-colored waveform for display?

[Hydrawerk]

UNI-T is a weird brand... Well, they have never sent a scope to Dave, while others did. (Owon, Tekway, Tektronix, Rigol, Agilent)

[uboot]

But a bigger question might be: if the Uni-T actually does 150k wfrm/s - how is it getting the extra captured information to the display? 120Hz is the upper end of refreshing an LCD, so DSOs with higher waveform capture rates generally use a Z-buffer and intensity-graded display in order to convey the extra waveforms captured between refreshes.

The Unit-T has no intensity-graded display - so, for example, if it's capturing 150k waveforms per second and refreshing the LCD 100 times per second, what is happening to the 1500 waveforms it captures between refreshes? Are all 1500 waveforms being combined into a single, uniformly-colored waveform for display?

Jup - that's what I'm asking myself, too.


Even if it manages to _trigger_ with 150k wps and fill the 6k buffer at that rate, in order to actually _notice_ what has been captured, it has to be visualized.

The only way to achieve this is persistent display mode.

But I really doubt that this works with 150k wps.

For this to work, you have to decide if you want to apply persistence only to the visible-on-screen-portion of the 6k sample buffer or to the whole 6k buffer.

Then you have to transfer the samples to another buffer at high speed (sort of a frame buffer, let's call it persistence buffer or DPO buffer). This buffer is bigger than the source because it has to store more than a single pixel position per time step: You can either pack this info into 256bits=32Bytes per time step or you can store a certain number of samples (8bits=1 Byte each) per time step (this allows to delete old samples / make them fade out, but how many samples to store?).

Applying persistence to the whole 6K sample buffer is very demanding, so assuming it is applied only to the visible area, which shows at most - say 512 timesteps , it may be possible to achieve 150k wps.

Now the persistence  buffer has to be transferred to the screen. The rate at which this is achieve does not matter.


EDIT: simple math: 1GSa/s divided by 6k buffer size equals ... guess what???  O0  With that in mind, their claim of >=150k waveforms makes sense....

But still I question that they manage to visualize it properly...


So, Hydrawerk is right: They are weird for not sending Dave a scope because he would definitely uncover the true performance of it....


EDIT2: I'm getting distracted from Siglent/Atten/Twintex due to the trigger issues reported for the LeCroy WaveAce which is said have same internals (https://www.eevblog.com/forum/testgear/more-waveace224-issues!/ https://www.eevblog.com/forum/reviews/dso-triggering-issues/)    It looks like sooner or later I well end up with a Hantex....