Everyday bench scope, a 3 way race, EDUX1002A, GDS-1054B, SDS1202X-E

[TK]

I applied Gated measurement and now the reported frequency is correct

[TK]

All these extensive tests paid off... I found a bug in the serial trigger function.  The Data bits you set for triggering get reset to all 0 when the scope is turned off, so you turn the scope on to continue your work and you need to set again the data bits.

[Rerouter]

Out of curiousity TK if you save it as a setup, does it recall correctly after startup?, Currently looking into how and where the siglent preserves settings over power cycles.

[TK]

Saving it as a setup, does NOT recall it correctly.  Recalled value is all 0s.

[TK]

I modified the code to generate an SPI packet with single byte 0x75 when a button is pressed, and asked the Siglent to trigger on this value.  My test setup is generating 280,000 SPI packets per second and the scope was able to trigger on every single byte value 0x75 I put out with the button.  I am trying to see if it can still trigger for 0x75 (and do a SINGLE capture) while at the same time is triggering (NORMAL mode) on byte value 0x74, but I am not finding an easy solution to do it.

The test case is getting complex and I don't think there a real life use case for it, so I am stopping my tests for now.

[TK]

OK, thanks to a suggestion from user 2N3055 to use SPI trigger on 0x74 and a mask on 0x75, I was able to test the new challenge on the Keysight and the Siglent

I had to change the bytes to 0x37 and 0x3F (because the Keysight expects a trigger on all the bits of the SPI frame), but you can indicate only the first byte if you are triggering on the first byte, but need to indicate all the bits of the frame if it is the last byte.

So I changed the code to generate 0x3F when a button was pressed.  Defined the Pass/Fail mask on 0x37 and set to trigger on 0011 X111 to trigger on 0x37 or 0x3F. 

When I pressed the button several times on the Keysight, it catches 50% to 75%.  So far, I cannot get the SDS1104X-E to catch a single one.  It catches 100% if I set the trigger on 0011 0111, but does not catch 0011 1111 when the trigger is set to 0011 X111.

[alpher]

Jeez guys you're still going at this?
Quite frankly I felt even dumber (if that's possible) after reading first couple of pages of replies.
So I just went with my first instinct, bought an EDUX, quite busy right now getting my RV ready for the season, so any hackin' has to wait.
Thank you anyway, very informative reading.

[TK]

I increased the SPI clock speed to 45MHz.  It means 50,000 x 8 = 400,000 SPI messages per second.  50,000 messages containing the byte 0x37.  Siglent SDS1104X-E can trigger 3,000 per second.  It does not detect the random glitch (byte value 0x3F) when the button is pressed.  Keysight EDUX1002G can trigger all 50,000 per second!.  It still can detect the random glitch, but around 10-30% of the time.

[BillB]

Jeez guys you're still going at this?
Quite frankly I felt even dumber (if that's possible) after reading first couple of pages of replies.
So I just went with my first instinct, bought an EDUX, quite busy right now getting my RV ready for the season, so any hackin' has to wait.
Thank you anyway, very informative reading.

You've chummed the waters and are surprised there is a frenzy?  I can't even think of a topic that would garner more activity than a scope n-way p*ssing contest...

Anyway, congrats on making a decision and moving on with life!  Remember, those afflicted with TEA don't have it so easy.   

[Old Printer]

I spent 2 years making my decision for my first digital scope, though I had picked up an AD2 to go with my Tek 2225 so I wasn't naked while I deliberated. A few weeks ago I pulled the trigger on a SDS1104X-E and so far very happy. Then I see TK's tests and my heart sank, my choice was not doing well, but then I realize I don't really know what he is doing and it is unlikely to ever have any impact on me - I hope Have not even bothered to hack it yet.

[TK]

I spent 2 years making my decision for my first digital scope, though I had picked up an AD2 to go with my Tek 2225 so I wasn't naked while I deliberated. A few weeks ago I pulled the trigger on a SDS1104X-E and so far very happy. Then I see TK's tests and my heart sank, my choice was not doing well, but then I realize I don't really know what he is doing and it is unlikely to ever have any impact on me - I hope Have not even bothered to hack it yet.

Please don't feel bad about your decision on the SDS1104X-E.  I like it very much.  As I said before, I worked long time on performance tuning on large computing systems, so that is why I run these tests, just out of curiosity.  I am not sure if it has any real use case... and the Keysight 1000X does not have serial decoding list function and the sample memory is only 1Mpts, but has a very fast scrolling horizontal knob 

[tautech]

I spent 2 years making my decision for my first digital scope, though I had picked up an AD2 to go with my Tek 2225 so I wasn't naked while I deliberated. A few weeks ago I pulled the trigger on a SDS1104X-E and so far very happy. Then I see TK's tests and my heart sank, my choice was not doing well, but then I realize I don't really know what he is doing and it is unlikely to ever have any impact on me - I hope Have not even bothered to hack it yet.

Please don't feel bad about your decision on the SDS1104X-E.  I like it very much.  As I said before, I worked long time on performance tuning on large computing systems, so that is why I run these tests, just out of curiosity.  I am not sure if it has any real use case... and the Keysight 1000X does not have serial decoding list function and the sample memory is only 1Mpts, but has a very fast scrolling horizontal knob 

TK, you know a fairer comparison would be between similar Sa/s rate DSO's eg. SDS2000X-E models that have the same sampling rate as your KS.
Double the sample passing through the scope will impact on signal triggering.
We saw how much higher frequency it could accurately resolve vs a 1 GSa/s DSO in your BW testing thread.
Unfortunately they're only available in 2ch models. 

[2N3055]

I spent 2 years making my decision for my first digital scope, though I had picked up an AD2 to go with my Tek 2225 so I wasn't naked while I deliberated. A few weeks ago I pulled the trigger on a SDS1104X-E and so far very happy. Then I see TK's tests and my heart sank, my choice was not doing well, but then I realize I don't really know what he is doing and it is unlikely to ever have any impact on me - I hope Have not even bothered to hack it yet.

What TK said..

Despite fact that it has slower triggering frequency, I still firmly believe SDS1104X-E is much better all rounder and much better bang for the buck than Keysight 1000.
If you just want digital replacement for your analog scope Keysight is great at that.
But if you use new digital scope as a new digital scope with all advantages that carries, SDS1104X-E is more capable in several areas.

There is no practical use for 50000 triggers per second on SPI packets.
If you wanted to catch elusive random glitch (byte value 0x3F) in a sea of 50,000 messages containing the byte 0x37, you would just set trigger to byte value 0x3F and catch them 100% of the time until there is several thousands of them per second. In which case I wouldn't call them elusive.

If anything, this kind of investigations show how advanced these inexpensive instruments are.

[TK]

I spent 2 years making my decision for my first digital scope, though I had picked up an AD2 to go with my Tek 2225 so I wasn't naked while I deliberated. A few weeks ago I pulled the trigger on a SDS1104X-E and so far very happy. Then I see TK's tests and my heart sank, my choice was not doing well, but then I realize I don't really know what he is doing and it is unlikely to ever have any impact on me - I hope Have not even bothered to hack it yet.

Please don't feel bad about your decision on the SDS1104X-E.  I like it very much.  As I said before, I worked long time on performance tuning on large computing systems, so that is why I run these tests, just out of curiosity.  I am not sure if it has any real use case... and the Keysight 1000X does not have serial decoding list function and the sample memory is only 1Mpts, but has a very fast scrolling horizontal knob 

TK, you know a fairer comparison would be between similar Sa/s rate DSO's eg. SDS2000X-E models that have the same sampling rate as your KS.
Double the sample passing through the scope will impact on signal triggering.
We saw how much higher frequency it could accurately resolve vs a 1 GSa/s DSO in your BW testing thread.
Unfortunately they're only available in 2ch models. 

Actually my test is making the KS to sample at 1 GSa/s because I am using the external input.  It keeps 2 GSa/s when using up to 2 analog channels, but goes down to 1 GSa/s when the external input is enabled.

[Old Printer]

I spent 2 years making my decision for my first digital scope, though I had picked up an AD2 to go with my Tek 2225 so I wasn't naked while I deliberated. A few weeks ago I pulled the trigger on a SDS1104X-E and so far very happy. Then I see TK's tests and my heart sank, my choice was not doing well, but then I realize I don't really know what he is doing and it is unlikely to ever have any impact on me - I hope Have not even bothered to hack it yet.

Please don't feel bad about your decision on the SDS1104X-E.  I like it very much.  As I said before, I worked long time on performance tuning on large computing systems, so that is why I run these tests, just out of curiosity.  I am not sure if it has any real use case... and the Keysight 1000X does not have serial decoding list function and the sample memory is only 1Mpts, but has a very fast scrolling horizontal knob 

TK, you know a fairer comparison would be between similar Sa/s rate DSO's eg. SDS2000X-E models that have the same sampling rate as your KS.
Double the sample passing through the scope will impact on signal triggering.
We saw how much higher frequency it could accurately resolve vs a 1 GSa/s DSO in your BW testing thread.
Unfortunately they're only available in 2ch models. 

Not to worry, really. It was just a second of second guessing myself before it registered that these are just hobby machines. We have come so far in what is available and at what prices, that at least I tend to get greedy. I would have cherry picked options from four different machines to get all the stuff i wanted, individual channel controls, HDMI output etc. At some point you just have to draw a line and say enough is enough, unless you have some particular needs but I really did not.

[2N3055]

Not to worry, really. It was just a second of second guessing myself before it registered that these are just hobby machines. We have come so far in what is available and at what prices, that at least I tend to get greedy. I would have cherry picked options from four different machines to get all the stuff i wanted, individual channel controls, HDMI output etc. At some point you just have to draw a line and say enough is enough, unless you have some particular needs but I really did not.

I wouldn't even say that they are hobby machines. Pretty much all other scopes except Megazoom based Keysight scopes will have same or worse retrigger rate than lowly SDS1104X-E.
In real world work it doesn't matter. Professionals don't spend time benchmarking scopes, they use them to design and troubleshoot stuff they work on.

In real life superfast rettriger rate of my Keysight 3000T is rarely useful, and even more rarely necessary.
On a Picoscope with 512 MPoints mem, I can grab ALL of 50000 packets in one go, and then make detailed analysis of data and timings.
I can create analysis of packet time distribution, data distribution, and try to correlate problematic data with outside world...

With  SDS1104X-E you can do more of that than on a scope with 28 times less memory and only 50 segments....


 

[branchmispredictor]

On a Picoscope with 512 MPoints mem, I can grab ALL of 50000 packets in one go, and then make detailed analysis of data and timings.
I can create analysis of packet time distribution, data distribution, and try to correlate problematic data with outside world...

Would a Picoscope be a good supplement to a SDS1104X-E then, for capturing high speed data? Which one would you recommend?

[TK]

On a Picoscope with 512 MPoints mem, I can grab ALL of 50000 packets in one go, and then make detailed analysis of data and timings.
I can create analysis of packet time distribution, data distribution, and try to correlate problematic data with outside world...

Would a Picoscope be a good supplement to a SDS1104X-E then, for capturing high speed data? Which one would you recommend?

If you want to supplement an SDS1104X-E, the best option is to get a logic analyzer.  I use the Zeroplus which offers over 100 protocols, but the software only runs on Windows.  It has it's own sample memory and triggers in hardware.  You can get the basic one with 16 channels and 64K memory for around $130, you have other options with more channels and more memory, up to 32 channels and 2Mb per channel, but price jumps over $1000.  You can use the SDS1104X-E for signal integrity and trigger the zeroplus or viceversa (it has trigger in and trigger out).

Saleae runs on windows, linux and mac os x... but I consider the hardware to be pretty limited and overpriced, usually captures to PC memory and does post processing, with no or limited on device sample memory.

Another option is an open source logic analyzer with pulseview.

[Old Printer]

I wouldn't even say that they are hobby machines. Pretty much all other scopes except Megazoom based Keysight scopes will have same or worse retrigger rate than lowly SDS1104X-E.
In real world work it doesn't matter. Professionals don't spend time benchmarking scopes, they use them to design and troubleshoot stuff they work on.

In real life superfast rettriger rate of my Keysight 3000T is rarely useful, and even more rarely necessary.
On a Picoscope with 512 MPoints mem, I can grab ALL of 50000 packets in one go, and then make detailed analysis of data and timings.
I can create analysis of packet time distribution, data distribution, and try to correlate problematic data with outside world...

With  SDS1104X-E you can do more of that than on a scope with 28 times less memory and only 50 segments....

I understand what you are saying, I just meant it is a hobby machine for me and at my level of experience I could have bought a much less capable tool, but I like my toys to be as good as I can afford.

[jemangedeslolos]

Hello TK,

I wanted to test quickly high speed SPI decoding on my SDS2204X ( I don't have a SDS1000X-E ).
So I made a program on my NucleoF446 with 45Mhz SPI clock.....and I have very strange result.

With my SP2030A, the clock signal is...ugly. I have much better result with my cheap P6100 100Mhz probe !
I never used SPI clock that high, maybe I made something wrong with my measure but with the P6100, the clock signal integrity looks OK.

I saw post from Performa01 who test different probe and it seems that there is no issue with the SP2030A.
Maybe you use a specific technic to measure 45Mhz clock ?

[jemangedeslolos]

Hello,

My daugther is sick, so I took with me the oscilloscope and now I have no issue with the same SP2030A on the same channel...
Very strange as I double check connections on my lab at work.

So I configure my NucleoF446 and now I wondering how do you compute the triggering rate ?
With the time on the decode list or there is hidden option are a specific technic ?

Thank you

I have a SDS2204X and a Rigol MSO5074 so maybe my result will be usefull for someone.

[TK]

To measure the trigger rate, you connect the trigger out from the Siglent to another scope input and just measure the frequency.

[jemangedeslolos]

Ah ok, I don't know why, I was convinced that the SDS1000X didn't have a trigger out BNC   
 
Thank you TK