Products > Test Equipment

A review of the GWInstek 1054B

<< < (12/16) > >>

akimmet:
I can also recreate this issue on my GDS1102B.

I can't complain too much since I have gone 2 years occasionally noticing those gaps in dot mode, and not thinking much of it. :palm:

saturation:
@maxwell3e10:  I'd be interested in your thoughts on what you think this anomaly does to affect the accuracy of the DSO or performance.  I do not know what practical significance it has after the re-tests I ran.  Screen caps to follow.

Here are my thoughts:

It does not affect the published spec sheet vertical accuracy of ~3% across all its ranges and frequencies, both visually measured as well as using automated values

The 'anomaly' does not contribute more distortion or artifacts than the 8-bit CPU can resolve, [ beyond the gaps shown ] and performing FFT on the test waveforms do not show extraneously spurs above the noise floor of ~50dB.

If the ADC were effectively sampling using less bits, the noise floor would rise.

Its unclear if this is from the HMCAD1511 ADC, Zynq SoC or Instek's firmware.

Finally, there is no practical lag in the response to the controls, but if you have some objective measure I can retest it for you.  Suffice to say, you move a knob the screen responds instantly.  The lag however is worse as you turn on more channels, increase the memory depth and turn on the math.  However you can control the response by simply turning off what you don't need at the moment, the fastest speed up is reducing the memory depth, then turning them back on or increasing the memory depth in sequence to get best resolution, or use the "STOP" mode and examine the issue after data has been acquired.  The good news with this design is the speed vs resolution tradeoff is under user control.

maxwell3e10:
I noticed it right away because I haven't seen such gaps in other scopes. I don't entirely understand the effect. It looks as if they do some vertical scaling of the ADC values. But the scaling depends on how many channels are being sampled, it makes no sense. It does have effectively fewer bits, but only about 25% fewer, so the increase in noise is not significant.

I got the scope specifically for an application where I want to display 10M points per channel and all 4 channels. So perhaps its not surprising that it is a little slow.



maxwell3e10:
I got a confirmation from Instek that this is in fact part of the design:
"It’s because the dynamic range of A/D converter can’t 100% match the front end attenuation circuit .
 Thus we use digital amplifier"

saturation:
Thanks, good and new info which I think no one has tried to figure out.  So kudos to you.  And that makes sense as an 'economy' move.  That GWInstek was able to squeeze this type of performance from the low cost platform is, IMHO, another plus.

I recall the Hittite ADC has a built in "digital zoom" function that the spec sheet claims would not marked affect the SINAD and I presume the 'accuracy', but naturally ENOB is reduced because it simply scales to the amplitude without resampling.  They say ENOB of > 7.5 from 8 bits, which sounds like what might happen given what you found, but your data seems the ENOB is far less than 7.5 bits.

Relevant items from the data sheet:

http://www.analog.com/media/en/technical-documentation/data-sheets/hmcad1511.pdf

8-bit High Speed Single/ Dual/ Quad ADC
 Single Channel Mode: FSmax = 1000 MSPS
Dual Channel Mode: FSmax = 500 MSPS
Quad Channel Mode: FSmax = 250 MSPS
• 1X to 50X Digital Gain
 No Missing Codes up to 32X
• 1X Gain: 49.8 dB SNR. 10X Gain: 48 dB SNR
• Internal Low Jitter Programmable Clock Divider
..
• Coarse and Fine Gain Control
• Digital Fine Gain Adjustment for each AD

...

. Internal 1 to 50X digital coarse gain with ENOB > 7.5 up to 16X gain, allows digital implementation of
oscilloscope gain settings. Internal digital fine gain can be set separately for each ADC to calibrate for gain errors.




--- Quote from: maxwell3e10 on February 21, 2018, 05:43:17 pm ---I got a confirmation from Instek that this is in fact part of the design:
"It’s because the dynamic range of A/D converter can’t 100% match the front end attenuation circuit .
 Thus we use digital amplifier"

--- End quote ---

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod