EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: System Error Message on July 08, 2016, 02:23:01 am
-
After trying out the GW Instek i bought from tequipment there seems to be a few issues with noise which is very noticeable on the 1mV scale and annoying on the 1V scale. It also seems that the scope itself has a square wave generator.
The slow CPU is very apparent when using any of the apps or math. At one point i couldnt use my scope even if rebooting it since default wouldnt work when in advance math. Subtraction slows down the scope a lot more than addition and when using advance math theres a possibility of the scope hanging with the only way to solve it would be by spamming the default button when rebooting. After exiting a feature that is CPU heavy the scope still remains slow until reset to default.
With the noise on the inputs i cant tell how stable a signal/power supply is but i can still see spikes. If only i could set up another seperate trigger for spikes.
If i use the frequency filter to eliminate noise im not quite sure if it can catch spikes. If i use the math to subtract the noise the scope operates a lot slower. Capturing the noise and subtracting it does really help eliminating most of the noise but the performance hit is unbearable.
I am using firmware 1.09
Another issue is that voltage measurements are always a set .2V lower than what it should be despite doing SPC calibration.
I will post screenshots soon.
-
I don't have personal experience with that scope but based on your description it sounds you might have a defective unit. The GDS-2204E is based on a much faster platform than say the Rigols or Siglent scopes, and even these scopes don't grind to a halt when using math.
As to the noise you're seeing it's difficult to say without screenshots but be aware that a DSO will always appear somewhat noisy (especially when compared with an analog scope which shows a 'cleaner' looking waveform, but that isn't because the analog scope is less noisy but simply because the noise isn't visible on the analog scope).
-
I agree with Wuerstchenhund that your observations are in contrast to the rave reviews we have seen from other users on the responsiveness of the new GDS scope platform. However, I am not sure that this is due to a defect; you might just be focusing on a different aspect of performance.
The Xilinx Zynq FPGA/processor platform which GDS Instek use in these scopes does bring a very tight integration between the FPGA and the CPU. Hence I would expect very smooth performance in the fundamental scope operating modes -- high waveforms/s rate, fast display updates, smooth response to control changes like timebase, offset etc.. It seems that the good performance of the 2000E platform in this respect is indeed at the heart of the positive reviews.
On the other hand, I am not sure how powerful the actual CPU core in the Zynq is in comparison to conventional separate CPUs or SOCs? Operations which rely heavily on CPU computation, like measurements or Math traces, might be less than perfect on this platform. Even nctnico, in his otherwise very positive review, mentioned the word "slow" a couple of times, although he did not find anything that got in the way of using the scope productively. (https://www.eevblog.com/forum/testgear/gw-instek-gds2204e-(200mhz-4-channel-dso)-review/ (https://www.eevblog.com/forum/testgear/gw-instek-gds2204e-(200mhz-4-channel-dso)-review/))
I'm sure other users of the GDS scopes will chime in here!
(Edit: Typos)
-
It's a super old firmware. Update to 1.24 (need to update it twice, first time scope might not reboot by itself). Also install FFT app.
http://www.gwinstek.com/en-global/Download/DownloadFile/DownloadFile/download%23_%2301_Oscilloscope%23_%23GDS-2000E%23_%23Firmware%23_%23gds2000e_v1.24.upg.zip (http://www.gwinstek.com/en-global/Download/DownloadFile/DownloadFile/download%23_%2301_Oscilloscope%23_%23GDS-2000E%23_%23Firmware%23_%23gds2000e_v1.24.upg.zip)
http://www.gwinstek.com/en-global/Download/DownloadFile/DownloadFile/download%23_%2301_Oscilloscope%23_%2306_GDB%23_%23GDB-03%23_%23Software%23_%23Scope_Software%23_%23Function_Module%23_%23GDS2000E_FunctionModule.zip (http://www.gwinstek.com/en-global/Download/DownloadFile/DownloadFile/download%23_%2301_Oscilloscope%23_%2306_GDB%23_%23GDB-03%23_%23Software%23_%23Scope_Software%23_%23Function_Module%23_%23GDS2000E_FunctionModule.zip)
http://www.gwinstek.com/en-global/Download/DownloadFile/DownloadFile/download%23_%2301_Oscilloscope%23_%2301_GDS%23_%23New%20DSO_Firmware_Upgrade_Guide%2020150910.pdf (http://www.gwinstek.com/en-global/Download/DownloadFile/DownloadFile/download%23_%2301_Oscilloscope%23_%2301_GDS%23_%23New%20DSO_Firmware_Upgrade_Guide%2020150910.pdf)
-
About input noise, it's seen on any scope on the lowest ranges. In fact GDS-2204E has the lowest noise I've seen, my Rigol DS2000 has twice of that. Do you see that noise without probes attached? That may be something wrong with something around
https://www.youtube.com/watch?v=SJefUNAJZNA (https://www.youtube.com/watch?v=SJefUNAJZNA)
-
Thanks. For some reason the firmware didnt show here http://www.gwinstek.com/en-global/products/Oscilloscopes/Digital_Storage_Oscilloscopes/GDS-2000E (http://www.gwinstek.com/en-global/products/Oscilloscopes/Digital_Storage_Oscilloscopes/GDS-2000E) so i thought i had the recent firmware.
I think i might be too picky about noise but i will update it and see if the update solves the issues i am having.
I know noise is unavoidable but if the firmware update makes the math fast than capturing the wave form and subtracting really helps.
-
You need to login to see the firmware for THIS model, but freely available for 1000B :palm:.
-
The slow CPU is very apparent when using any of the apps or math. At one point i couldnt use my scope even if rebooting it since default wouldnt work when in advance math. Subtraction slows down the scope a lot more than addition and when using advance math theres a possibility of the scope hanging
Another issue is that voltage measurements are always a set .2V lower than what it should be despite doing SPC calibration.
I will post screenshots soon.
Don't post them yet... I need to go out for extra popcorn. :popcorn:
-
Noise can also stem for a power supply. The cheap Chinese switching power supplies are pretty bad when it comes to injecting noise into a circuit.
But do update the oscilloscope first because the old firmware does contain some bugs. And yes it is annoying you have to login to download the latest firmware.
-
But do update the oscilloscope first because the old firmware does contain some bugs.
It's all coming out now. Pesky Truth. :popcorn:
-
But do update the oscilloscope first because the old firmware does contain some bugs.
It's all coming out now. Pesky Truth. :popcorn:
I'm not quite sure why you get so exited because it is all out in the open. Just read my review I posted at the beginning of this year: https://www.eevblog.com/forum/testgear/gw-instek-gds2204e-%28200mhz-4-channel-dso%29-review/ (https://www.eevblog.com/forum/testgear/gw-instek-gds2204e-%28200mhz-4-channel-dso%29-review/)
-
Updating firmware solved the performance issue but some areas are still laggy. At least the scope wont freeze again when i try to use advance math.
Regarding the noise im not sure if its the scope or some radio/wifi. Using the filters between either low pass with lowest hz or high pass at 50Mhz or higher completely eliminates the noise. If its not the oscilloscope than its probably the UK and that would mean the GW instek scope is really great. The noise seems to have its own waveform and by subtracting it most noise can be removed.
So i attached the probe to ch1 and the 2V waveform generator infront of the scope. While i didnt manage to capture any spikes where were a minor spikes that appear like bumps. Before attaching ch1 to waveform generator i saved the waveform from ch1 to ref1 and used that to subtract from ch1.
Screenshot 1 - input noise
Screenshot 2 - input noise and subtracting it (lowest line)
Screenshot 3 - a close look at the noise
Screenshot 4 - plugged in probe to ch1
Screenshot 5 - attached probe to 2V waveform generator infront of scope
Screenshot 6 - just another screenshot
Screenshot 7 - adjusted scales
the main thing i really like about this scope is that it is really sensitive. I doubt the rigol actually has the same sensitivity, performance (i can measure down to 1ns, independent channel memory), and noise, still this noise does bug me even when im just seeing the 2V square wave from the front panel on ch1. The math signal didnt seem to update as fast though so im not sure if its performance or just the noise consistent but slightly out of phase.
-
I don't see any abnormal input noise there. I see that you are using 1 Mpts memory, because of that there can be some lag in the controls while using math, still faster than most of the other scopes in the similar circumstances. If you set 100 kpts memory, any lag in the controls should be gone completely.
-
What noise? :-//
DSO's do not show fine traces like CRO's, learn and accept this.
Some study:
http://www.eevblog.org/video/EEVblog601-WhyDigitalScopesAppearNoisy-640x360.m4v (http://www.eevblog.org/video/EEVblog601-WhyDigitalScopesAppearNoisy-640x360.m4v)
http://www.eevblog.org/video/EEVblog610-WhyDigitalScopesAppearNoisyPart2-640x360.m4v (http://www.eevblog.org/video/EEVblog610-WhyDigitalScopesAppearNoisyPart2-640x360.m4v)
-
I don't see any abnormal input noise there. I see that you are using 1 Mpts memory, because of that there can be some lag in the controls while using math, still faster than most of the other scopes in the similar circumstances. If you set 100 kpts memory, any lag in the controls should be gone completely.
I agree. The traces look completely normal when it comes to noise (and there are DSOs out there with 3 times more noise on the traces!).
-
Thanks Tautech for the video, it explains quite a lot. So i guess the scope is fine. Its a good thing i didnt get the rigol, it would've been less sensitive.
I still wish you could use both the low pass and high pass filter at the same time for a range in between (or not in between) but i guess i would have to use 2 channels and some math.
So how do i use the scope to measure current?
-
Thanks Tautech for the video, it explains quite a lot. So i guess the scope is fine. Its a good thing i didnt get the rigol, it would've been less sensitive.
I still wish you could use both the low pass and high pass filter at the same time for a range in between (or not in between) but i guess i would have to use 2 channels and some math.
There are several things you can do to "clean" up a trace.
Have a fiddle in the Acquisition menu and for repetitive waveforms that are not suspected of containing glitches then Averaging can be useful.
But mostly it comes down to experience, knowing how to setup your scope, trusting it and interpreting the display, only some of which can be learnt from a book. ;)
-
So how do i use the scope to measure current?
For that you'll need a current probe, a current transformer (AC only) or a current shunt resistor.
A current shunt is cheap but has the limited that the ground of the shunt must also be the ground of the oscilloscope. A way to circumvent that is to use a differential pre-amplifier which allows the current shunt to float a few volt. A current transformer is also cheap but no DC current. A current probe can be used fully floating but these start from a couple of hundred dollar and have noise and offset issues. Bandwidth can also be a problem with current probes.
-
Thanks Tautech for the video, it explains quite a lot. So i guess the scope is fine. Its a good thing i didnt get the rigol, it would've been less sensitive.
I still wish you could use both the low pass and high pass filter at the same time for a range in between (or not in between) but i guess i would have to use 2 channels and some math.
There are several things you can do to "clean" up a trace.
Have a fiddle in the Acquisition menu and for repetitive waveforms that are not suspected of containing glitches then Averaging can be useful.
But mostly it comes down to experience, knowing how to setup your scope, trusting it and interpreting the display, only some of which can be learnt from a book. ;)
If i want to see the waveform without anything else than i could use averaging but i want to see spikes and glitches so i find using the math to be quite effective. I just realised the math function on this scope is still slow. After unplugging the probes the math was still displaying the wave for a few seconds, not sure if this is intentional or just slow. I know the FPGA is a massively parallel CPU but GW Instek with the updated firmware did get the priorities right with input not lagging anymore when using other functions and other functions actually not lurking about when you turn them off. Using the low pass filter is effective in measuring DC voltage without the background noise.
If i dont connect the earth of the oscilloscope to mains could i measure currrent with it?
-
I just realised the math function on this scope is still slow.
Don't use large sample memory if you want blazing fast math. You just haven't seen what is slow.
-
Also note, if you select 1Mpts acquisition memory, you will have real 1Mpts math on this scope. Rigol for comparison cheats and there is no such thing as 1 Mpts math on it, it's speed (and accuracy) is not great and remains roughly the same be it 7 kpts or 28 Mpts which tells a lot.
-
If i dont connect the earth of the oscilloscope to mains could i measure currrent with it?
:--
This is the dangerous territory of floating a scope, a practice highly publicly frowned upon and a definite no-no for the inexperienced.
Think of a scope like a tractor, not much good for anything without a range of accessories in order for various tasks to be accomplished.
So current must be converted to a voltage and to do that the most safe universal method is with a current probe. Yeah I know...more cost.
-
Its a good thing i didnt get the rigol, it would've been less sensitive.
the main thing i really like about this scope is that it is really sensitive. I doubt the rigol actually has the same sensitivity
How can you be so certain about that without trying one? :-//
(I'm not saying it is or it isn't - I don't own an Instek to put side by side with a Rigol, I just wonder how you can be so certain about that)
-
It's all coming out now. Pesky Truth. :popcorn:
I'm not quite sure why you get so exited because it is all out in the open.
Frankly I'm shocked that Instek released a buggy 'scope to market. Bugs which took a firmware update to fix. :palm:
Just kidding: I'm more interested to see how calm and objective the Rigol-bashers suddenly become when you change the label on the front of the box to "Good Will Instruments".
-
Just kidding: I'm more interested to see how calm and objective the Rigol-bashers suddenly become when you change the label on the front of the box to "Good Will Instruments".
Hereby some company names that didn't make it through the Trademark Office :)
- Most Bang For Buck Instruments
- Big Memory Instruments
- Highest Waveform Update Rate Instruments
- We Do Our Best Instruments
- We Should Try Harder Instruments
- Don't Try This At Home Instruments
- Don't Measure High Voltage Instruments
- Slow Fourier Transform Instruments
Other suggestions about company names that didn't make it through the Trademark Office are welcome! :)
-
Frankly I'm shocked that Instek released a buggy 'scope to market. Bugs which took a firmware update to fix. :palm:
Just kidding: I'm more interested to see how calm and objective the Rigol-bashers suddenly become when you change the label on the front of the box to "Good Will Instruments".
- Most Bang For Buck Instruments
- Big Memory Instruments
...
At least it not:
Cannot fix the bugs for years instruments
Introduce new bugs instruments
Cannot properly design Vreg and PLL instruments
-
It's all coming out now. Pesky Truth. :popcorn:
I'm not quite sure why you get so exited because it is all out in the open.
Frankly I'm shocked that Instek released a buggy 'scope to market. Bugs which took a firmware update to fix. :palm:
Just kidding: I'm more interested to see how calm and objective the Rigol-bashers suddenly become when you change the label on the front of the box to "Good Will Instruments".
As you can read in my review GW Instek fixed the bugs within less days than I have fingers. I have not seen Rigol or Siglent do that.
-
Just kidding: I'm more interested to see how calm and objective the Rigol-bashers suddenly become
FYI there are also some "anything that is not the holy R...." haters on this forum too who often loose their cool to quickly ::).
-
To further complete my list with company names, hereby a name suggestion for a company that produces the best T&M equipment in the world! :)
Wuerstchenhund Instruments Ltd.
Just kidding: I'm more interested to see how calm and objective the Rigol-bashers suddenly become when you change the label on the front of the box to "Good Will Instruments".
Hereby some company names that didn't make it through the Trademark Office :)
- Most Bang For Buck Instruments
- Big Memory Instruments
- Highest Waveform Update Rate Instruments
- We Do Our Best Instruments
- We Should Try Harder Instruments
- Don't Try This At Home Instruments
- Don't Measure High Voltage Instruments
- Slow Fourier Transform Instruments
Other suggestions about company names that didn't make it through the Trademark Office are welcome! :)
-
At least it not:
Cannot fix the bugs for years instruments
Introduce new bugs instruments
Cannot properly design Vreg and PLL instruments
Calm and objective as usual...
PS: You've yet to demonstrate that anything about the Vreg or PLL design causes a problem.
-
As you can read in my review GW Instek fixed the bugs within less days than I have fingers. I have not seen Rigol or Siglent do that.
I was always told GW Instek didn't have any bugs, that their quality control made it impossible.
-
You can't expect the same quality control from GW-Instek as from Wuerstchenhund Instruments Ltd. :)
-
well now that the issue is solved i would like to see more apps on the scope itself or perhaps a way to make your own apps to run on the scope. Theres a programming manual but im not sure if its to run as a program on the scope itself or just to issue instructions from a PC.
So if i cant measure current easily i guess its not really an issue as i can use my multi meter for calibrating my PSU for the current limits and could just use those simple segment displays with a circuit to measure current. As long as i can measure the voltage output well and at least hope that the SPC calibration works as before i was always getting -0.2V all the time.
i install the demo app but it keeps asking for demo boards. Is there a way to change the wave from the front wave generator?
-
i install the demo app but it keeps asking for demo boards. Is there a way to change the wave from the front wave generator?
Because it is intended to be used with demo board. You can change frequency of the compensation output in 1-200kHz range (Utility->Probe Comp.)
-
i install the demo app but it keeps asking for demo boards. Is there a way to change the wave from the front wave generator?
The demo app is to show the capabilities of the scope at tradeshows so it is not very useful unless you are an equipment dealer. AFAIK you can't change the type of waveform from the calibrator. It is always a square wave. I think one of GW Instek's customers liked to be able to use the calibrator output as a simple square wave generator so they implemented it.
-
Is there a way to change the wave from the front wave generator?
Yep. buy a DSO with an inbuilt AWG.
SEM, you might have been better to have bought a cheap DSO for your first one, you would have had some $ left over to buy accessories to expand the measurement capabilities of your scope then when many of the capabilities of a DSO are learnt then focus on a unit that better serves your needs. Maybe the GW would be the same choice, maybe not.
-
I do have money to buy the accessories but right now i dont need serious waveform generation or current probes, im just asking what the scope is capable of doing and how to do it as i have gone through all the menus on the scope.