Siglent SDS1104X-E and SDS1204X-E Mixed Signal Oscilloscopes

[rf-loop]

rf-loop, would you be willing to go into exactly what bytes need to be changed for the longer modes? I can only see 1000s for the upper and 200ps and 500ps unused for the lower, as strings in the app? while I'm curious about those lower time bases, I would love to know what the modification procedure is.

Individuals mods are very much other things. And as told previously these things are better to discuss in dedicated thread for these...  example here : https://www.eevblog.com/forum/testgear/sds1104x-e-hack-to-200mhz-and-full-options/

or here: https://www.eevblog.com/forum/testgear/unlocking-siglent-sds1104x-e-step-by-step/msg1788704/#msg1788704



Firmware development is Siglent work.  It need add to product features.
It is Siglent work to develop products better and better and add features or develop some features better. Of course in priority order. First "immediately" fatal bugs if exist
Then 2nd level bugs
Then some finishing/grinding UI and also adding some features or change some limits (as example add 200, 500 and 1000s/div and also wish they add one more short max memory limit selection (now they are 14/7M, 1.4M/700k, 140k/70k and 14k/7k) and I like if there is also 1.4k/700 selectable. (it give more freedom for history/segmented)

I do only that I write wish list to "Santa Claus" (in this case Siglent) and then waiting christmas.

[tinhead]

i found small bug, only CH1 and CH3 affected, only when CH2 and CH4 are off. In fine adjust mode, 302mV/DIV is on wrong position. Seems to be only the case as well memory ans timebase dependent (14M 1ms->1ns, 1.4M 100us->1ns, 140k 10us->1ns, 1.4k 1us -> 1ns).

Tested on 8.1.6.1.26, hw1.3


EDIT: hmm, there is actually more like that, 1.54mV/DIV to 1.50mv/DIV, then very strange jumps/no move from 1.36mV to 780uV/DIV, then the wave jumps down by 2 DIVs down between 770uV/DIV and 750uV/DIV. Same as above, dependent on timebase and memory settings, this time CH1 more affected than CH3, no problem with CH2 and CH4.

Can someone check that? My DSO is for sure properly calibrated and still under quick warranty, so maybe i have to replace it.

[Rerouter]

Around the weird steps would you be able to run the scpi command c1:vgac? This will return the setting for the variable gain amplifier. And will give insight into if its the adc scaling at that point or the vga

[rf-loop]

i found small bug, only CH1 and CH3 affected, only when CH2 and CH4 are off. In fine adjust mode, 302mV/DIV is on wrong position. Seems to be only the case as well memory ans timebase dependent (14M 1ms->1ns, 1.4M 100us->1ns, 140k 10us->1ns, 1.4k 1us -> 1ns).

Tested on 8.1.6.1.26, hw1.3


EDIT: hmm, there is actually more like that, 1.54mV/DIV to 1.50mv/DIV, then very strange jumps/no move from 1.36mV to 780uV/DIV, then the wave jumps down by 2 DIVs down between 770uV/DIV and 750uV/DIV. Same as above, dependent on timebase and memory settings, this time CH1 more affected than CH3, no problem with CH2 and CH4.

Can someone check that? My DSO is for sure properly calibrated and still under quick warranty, so maybe i have to replace it.

Tst 7.1.6.1.26 (and siglent have told to me that FW  first number 7 or 8 is just "do not care" when update FW if it is 7 or 8 but not told what is real difference (if any). My HW 00-03: Nothing like this.

My first opinion is: return it with note "fault on arrival" and short info about problem so that if seller is very bad like amazon or equal they do not sell it to next victim.

[Rerouter]

Yep, looks like there is something weird about the calibration for certain values, and seems to be related to VGAC it rolls over at 750uV, 1.50mV, 3.50mV, 7.50mV etc,

However the first 2 seem to have an error at the roll over. Equally it hints that it should be able to allow for as low as about 350uV / div on a hardware perspective. and up to about 12V/div on the upper,

To make clear, the C1:VGAC command is a com/query that you can use to manually set the variable gain amplifier, values between 15 - 199 seems to be the valid range used during device calibration, though it may be able to use the full 0-255, which would possibly allow all the way down to 300uV/div. room for imagination for anyone else who actually has gear to test accurate AC amplitude at crazy low amplitudes.

There is something else also being used during those step changes, but I have not yet found a command for it.

Below is the VGAC values for each vdiv step on my scope, you can see it goes a bit wonky at 750uV and 1.50mV

Code: [Select]

500uv 183
510uV   182
520uV   181
530uV 180
540uV 179
550uV 178
560uV 177
570uV 176
580uV 175
590uV 174
600uV 174
610uV 173
620uV 172
630uV 171
640uV 171
650uV 170
660uV 170
670uV 169
680uV 168
690uV 168
700uV 167
710uV 167
720uV 166
730uV 165
740uV 165

750uV 154
760uV 154

770uV 199
780uV 198
790uV 197
800uV 196
810uV 196
820uV 195
830uV 194
840uV 193
850uV 192
860uV 192
870uV 191
880uV 190
890uV 189
900uV 189
910uV 188
920uV 188
930uV 187
940uV 186
950uV 186
960uV 185
970uV 184
980uV 184
990uV 183
1.00mV 183
1.02mV 182
1.04mV 181
1.06mV 180
1.08mV 179
1.10mV 178
1.12mV 177
1.14mV 176
1.16mV 175
1.18mV 174
1.20mV 174
1.22mV 173
1.24mV 172
1.26mV 171
1.28mV 171
1.30mV 170
1.32mV 170
1.34mV 169
1.36mV 168
1.38mV 168
1.40mV 167
1.42mV 167
1.44mV 166
1.46mV 165
1.48mV 165

1.50mV 154
1.52mV 154

1.54mV 199
1.56mV 198
1.58mV 197
1.60mV 196
1.62mV 196
1.64mV 195
1.66mV 194
1.68mV 193
1.70mV 192
1.72mV 192
1.74mV 191
1.76mV 190
1.78mV 189
1.80mV 189
1.82mV 188
1.84mV 188
1.86mV 187
1.88mV 186
1.90mV 186
1.92mV 185
1.94mV 184
1.96mV 184
1.98mV 183
2.00mV 183
2.06mV 181
2.12mV 180

2.18mV 178
2.24mV 177

2.30mV 175
2.36mV 174
2.42mV 173
2.48mV 172
2.54mV 171
2.60mV 170
2.66mV 169
2.72mV 168
2.78mV 167
2.84mV 167
2.90mV 166
2.96mV 165
3.02mV 164
3.08mV 164
3.14mV 163
3.20mV 162
3.26mV 162
3.32mV 161
3.38mV 160
3.44mv 160

3.50mV 190
3.56mV 189
3.62mV 188
3.68mV 188
3.74mV 187
3.80mV 186
3.86mV 185
3.92mV 184
3.98mV 183
4.04mV 182
4.10mV 181
4.16mV 181
4.22mV 180
4.28mV 179
4.34mV 178
4.40mV 178
4.46mV 177
4.52mV 176
4.58mV 176
4.64mV 175
4.70mV 175
4.76mV 174
4.82mV 173
4.88mV 173
4.94mV 172
5.00mV 172
5.10mV 171
5.20mV 170
5.30mV 169
5.40mV 169
5.50mV 168
5.60mV 167
5.70mV 166
5.80mV 166
5.90mV 165
6.00mV 165
6.10mV 164
6.20mV 163
6.30mV 163
6.40mV 162
6.50mV 162
6.60mV 161
6.70mV 161
6.80mV 160
6.90mV 160
7.00mV 159
7.10mV 159
7.20mV 158
7.30mV 158
7.40mV 158

7.50mV 187

[bluejedi]

Some suggestions to Siglent about messages and captions on the screen.

1. The captions on the screen should be in front of the waveforms and not transparent or almost not transparent. Currently they often become unreadable when they are crossed by waveforms.

Clear suggestions for improving readability of text overlays were made before (July 2018) here:

https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1713224/#msg1713224

Unfortunately @Siglent appears still not to care about this. The same holds for Wifi related bugs e.g. like described (July 2018) here:

https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1708814/?topicseen#msg1708814

[bluejedi]

I do only that I write wish list to "Santa Claus" (in this case Siglent) and then waiting christmas.

Oh, please not Christmas because then we will have to wait for (almost) a whole year. 

[Rerouter]

Ok, can confirm VGAC works from 1-255, with a code of 255 does make it display a hair shy of 300uV/div,

And a code of 1 on the 10V/div range makes the signal display at about 100V/div, dont have a signal source that high, so its approximate, no sure to what end someone would want to compress the signal at the higher end, but it exists, 

[rf-loop]

Yep, looks like there is something weird about the calibration for certain values, and seems to be related to VGAC it rolls over at 750uV, 1.50mV, 3.50mV, 7.50mV etc,

However the first 2 seem to have an error at the roll over. Equally it hints that it should be able to allow for as low as about 350uV / div on a hardware perspective. and up to about 12V/div on the upper,

To make clear, the C1:VGAC command is a com/query that you can use to manually set the variable gain amplifier, values between 15 - 199 seems to be the valid range used during device calibration, though it may be able to use the full 0-255, which would possibly allow all the way down to 100uV/div. room for imagination for anyone else who actually has gear to test accurate AC amplitude at crazy low amplitudes.

There is something else also being used during those step changes, but I have not yet found a command for it.

Below is the VGAC values for each vdiv step on my scope, you can see it goes a bit wonky at 750uV and 1.50mV

Code: [Select]

500uv 183
510uV   182
520uV   181
530uV 180
540uV 179
550uV 178
560uV 177
570uV 176
580uV 175
590uV 174
600uV 174
610uV 173
620uV 172
630uV 171
640uV 171
650uV 170
660uV 170
670uV 169
680uV 168
690uV 168
700uV 167
710uV 167
720uV 166
730uV 165
740uV 165

750uV 154
760uV 154

770uV 199
780uV 198
790uV 197
800uV 196
810uV 196
820uV 195
830uV 194
840uV 193
850uV 192
860uV 192
870uV 191
880uV 190
890uV 189
900uV 189
910uV 188
920uV 188
930uV 187
940uV 186
950uV 186
960uV 185
970uV 184
980uV 184
990uV 183
1.00mV 183
1.02mV 182
1.04mV 181
1.06mV 180
1.08mV 179
1.10mV 178
1.12mV 177
1.14mV 176
1.16mV 175
1.18mV 174
1.20mV 174
1.22mV 173
1.24mV 172
1.26mV 171
1.28mV 171
1.30mV 170
1.32mV 170
1.34mV 169
1.36mV 168
1.38mV 168
1.40mV 167
1.42mV 167
1.44mV 166
1.46mV 165
1.48mV 165

1.50mV 154
1.52mV 154

1.54mV 199
1.56mV 198
1.58mV 197
1.60mV 196
1.62mV 196
1.64mV 195
1.66mV 194
1.68mV 193
1.70mV 192
1.72mV 192
1.74mV 191
1.76mV 190
1.78mV 189
1.80mV 189
1.82mV 188
1.84mV 188
1.86mV 187
1.88mV 186
1.90mV 186
1.92mV 185
1.94mV 184
1.96mV 184
1.98mV 183
2.00mV 183
2.06mV 181
2.12mV 180

2.18mV 178
2.24mV 177

2.30mV 175
2.36mV 174
2.42mV 173
2.48mV 172
2.54mV 171
2.60mV 170
2.66mV 169
2.72mV 168
2.78mV 167
2.84mV 167
2.90mV 166
2.96mV 165
3.02mV 164
3.08mV 164
3.14mV 163
3.20mV 162
3.26mV 162
3.32mV 161
3.38mV 160
3.44mv 160

3.50mV 190
3.56mV 189
3.62mV 188
3.68mV 188
3.74mV 187
3.80mV 186
3.86mV 185
3.92mV 184
3.98mV 183
4.04mV 182
4.10mV 181
4.16mV 181
4.22mV 180
4.28mV 179
4.34mV 178
4.40mV 178
4.46mV 177
4.52mV 176
4.58mV 176
4.64mV 175
4.70mV 175
4.76mV 174
4.82mV 173
4.88mV 173
4.94mV 172
5.00mV 172
5.10mV 171
5.20mV 170
5.30mV 169
5.40mV 169
5.50mV 168
5.60mV 167
5.70mV 166
5.80mV 166
5.90mV 165
6.00mV 165
6.10mV 164
6.20mV 163
6.30mV 163
6.40mV 162
6.50mV 162
6.60mV 161
6.70mV 161
6.80mV 160
6.90mV 160
7.00mV 159
7.10mV 159
7.20mV 158
7.30mV 158
7.40mV 158

7.50mV 187

Nice findings.


And then one small pick-up: "....has gear to test accurate AC amplitude at crazy low amplitudes."

100uV class is not crazy low even if it is peak to peak (around -76dBm what is quite strong signal (in radio). Normal good rf generators goes usually down to well below -127dBm (100nVrms) This my one old "33kg junk" goes down to -140dBm (23nVrms)

One tiny experimental playing like kid.
Using higher level as external trig Ch4 and measured in Ch1 just using 500uV/div and 129MHz signal -72dBm, nicely  without big error (256 acquisition  average) Gen out -72dBm and SDS1204X-E display signal level Stdev (ac rms) 53uV. This trigger method need because of course it can not trig directly this kind of low signal what was measured in Ch1. 
Also this need high amount of trace averaging for get any useful signal out from noise to display, but with continuous signal it is possible.



But then bac out from "just fun"  playing.
Also it need think reality. Reality is that this kind of oscilloscope input and front end have quite LOT of noise if think these low signals. It is never designed as SA or radio front end. First thing is that even when there is AC/DC selection front end amp is dual pathway DC coupled and LF (specially 1/f) noise is huge due to many reasons. Then it is also for 1Mohm input.  Even when these cheap scopes have 50ohm inputs front end is still 1M. How much is 1Mohm Johnson noise alone.
Even this 500uV/div is borderline but still useable. If we amplify this noise more... what we get.
Front end is totally different animal if it is designed for -80 or -100dBm range...
500uVp-p is roughly -62dBm in 50ohm system.

Perhaps it is more wise to use preamplifier if oscilloscope really need look this kind of low level signals. Most oscilloscopes normal  most sensitive range is around 2 - 5mV/div (full resolution) and with reason(s).

[Rerouter]

I chose my wording poorly, but for me that is low amplitude in the realm of a scope. And while I have plenty of signal sources, I lack any direct way to measure amplitudes that low with any hope of accuracy.

I agree its pretty noisy down at the 500uV range, however It works on all attenuation ranges. so you can abuse it to view smaller signals on the larger offset ranges, e.g. 8mV / div at a +-21V offset range. or 70mV/div at a +-210V offset range, this is where I see it being much more useful, 

[rf-loop]

I chose my wording poorly, but for me that is low amplitude in the realm of a scope. And while I have plenty of signal sources, I lack any direct way to measure amplitudes that low with any hope of accuracy.

I agree its pretty noisy down at the 500uV range,

We can say it is very npisy. But also we can say it have very low noise. It depends context. When we talk roughly this class of oscilloscopes it can say it have low noise. Compare to Rigol, compare to Keysight, compare to... ~30uVrms 100Mhz BW is not at all bad.


@srce: https://www.eevblog.com/forum/testgear/oscilloscope-input-noise-comparison/msg1964747/#msg1964747


200MHz model


Compared 100MHz vs 100MHz, settings are comparable.

however It works on all attenuation ranges. so you can abuse it to view smaller signals on the larger offset ranges, e.g. 8mV / div at a +-21V offset range. or 70mV/div at a +-210V offset range, this is where I see it being much more useful, 

Can you explain it bit more. Perhaps I now have a brain short circuit and I don't understand.

[Rerouter]

Ok, the simplest way to demonstrate it for yourself would be, set your scope to 200mV/div, this will click over the attenuator so its in the +-21V offset range, now at this point, the VGA for the front end is barely amplifying at all, with a VGA (variable gain Amplifier) code around 30,

The higher the code, the higher the amplification,

Now with it like that, sending "C1:VGAC 255" over SCPI (or webserver) will set the VGA to maximum amplification, which is about 8mV/div for my device, As I dig into the device calibration, Its possible we can figure out exactly what it is for each code, I should point out its more noisy than the actual 5mV/div range, but if you need it on a bigger offset, then the scope can indeed do it.

This will revert if you change the channel offset or vdiv manually, The increased offset range comes from the "C1:DACC" command I documented earlier, you have a bit of extra offset range at both ends,

This just leaves some other attenuation / amplification happening in the chain when the VGAC code rolls around, I suspect its to do with the "C1:DAC_ATTEN_SET" command, Haven't documented it, but its in the same decision tree as the other 2, so once I have him nutted out, there may be a little more we can abuse.

[Rerouter]

More digging, the VGAC command directly sets the fine gain register for the VGAC which is an AD8370, And some further digging confirms there is a roll over in the middle which is the switch from low gain mode to high gain mode.

So my example of 8mV/div on the 20V offset range is amplified to 52 V/V, and does show similar noise to the 500uV range, which makes sense when its amplifying that much.
Now that we know this, Its not hard to extrapolate out any intermediate V/V of gain if you need an odd measurement

https://www.analog.com/media/en/technical-documentation/data-sheets/ad8370.pdf

[Rerouter]

And finally, an explanation for the quirk with the fine gain calibration, at 750-760uV and 1.50-1.52mV its an overflow based on how they limited there calibration routine, At these points the VGAC code is meant to be greater than 199, but in how they have implemented things, that results in it defaulting to 154 when it cannot find a matching record.

[tmbinc]

Not sure if this had been posted here, but the analog frontend was reverse-engineered:

https://github.com/360nosc0pe/siglent_hardware/tree/master/sds1202xe/siglent_AFE

The gain setting is a combination of the VGA setting, the ADC settings and the shift register:


### AFE (analog front-end) setting
Each channel of the scope contain one 74HC595 shift out register powered by 5V.
Five of the eight outputs are used to change hardware settings in the AFE:

    595_QB:  First divider, 10:1, active high
    595_QC:  Second divider, 10:1, active high
    595_QD:  AC coupling, low = AC, high = DC
    595_QE:  PGA enable, active high
    595_QF:  BW limit, low = 20MHz, high = full



From my notes: (Sorry about the formatting)

ADC is the write to reg 0x2B of the HMCAD1511 ADC (Datasheet: https://www.analog.com/media/en/technical-documentation/data-sheets/hmcad1511.pdf), the "coarse gain" register. Value is in dB (actually). ShiftReg is the value written to the frontend shift register. VGA is the value set to the VGA.



         ADC ShiftReg ADC VGA
  5mV   2B 09   78   +9  B9
 10mV   2b 06   78   +6  B9
 20mV   2B 04   78   +4  B9
 50mV   2b 02   78   +2  AD
100mV   2B 00   78   +0  AD
200 mV  2B 04   78   +4  27
500 mV  2B 00   78   +0  3f
  1v    2B 00   78   +0  1F
  2V    2B 04   7A   +4  29
  5v    2B 00   7A   +0  41
 10V    2B 00   7A       21
 20V    2B 04   7E       28
 50v    2B 04   7E       41
100V            7E       20

[tautech]

Not sure if this had been posted here, but the analog frontend was reverse-engineered:

This is a better thread for SDS1202X-E stuff:
https://www.eevblog.com/forum/testgear/the-_official_-siglent-sds1202x-e-thread/

[tinhead]

Not sure if this had been posted here, but the analog frontend was reverse-engineered:

https://github.com/360nosc0pe/siglent_hardware/tree/master/sds1202xe/siglent_AFE

i don't think this has been properly done, look at this part:

[Rerouter]

Thanks for that tmbinc. That explains the 2B value I keep seeing around ADC calibration. Now to just determine how to change it via scpi.

As it notes its calibration as X1 X10 and X100. That explains some of the weirdness I have been seeing. If i can read or write them. Well there is a part of the calibrarion routine that gives up on correcting the Dac offset. Amd that would let me verify if the values end up used.

[Andreax1985]

I need to measure AC ripple on a 320V DC signal using AC coupling mode. SDS1004X-E input is rated for 400Vpk. Is this enough for my needs? (My worry is due to the fact that 400Vpk correspond to 280Vrms...)

[Martin72]

At this voltage rate you should use a differential or a 100:1 probe.

[Performa01]

I need to measure AC ripple on a 320V DC signal using AC coupling mode. SDS1004X-E input is rated for 400Vpk. Is this enough for my needs? (My worry is due to the fact that 400Vpk correspond to 280Vrms...)

It heavily depends on the exact waveform what 400Vp actually corresponds to. For DC however, peak and rms are identical.
So with 320VDC you got 80Vp headroom left. That means that the ripple must not exceed 80Vp, but I would think if it is anywhere near that much, then there would also be no need to maximize the sensitivity (and use x1 probe factor) either.

[Rerouter]

I should also point out probes have there own voltage ratings, looking on the side of a random one I have on hand (X1: 150V, X10: 300V) its not a siglent probe, so check yours first.

If the probe cannot handle the voltage, but the scope can, it means you will just have to use a coax cable to measure it.

[rf-loop]

I should also point out probes have there own voltage ratings, looking on the side of a random one I have on hand (X1: 150V, X10: 300V) its not a siglent probe, so check yours first.

If the probe cannot handle the voltage, but the scope can, it means you will just have to use a coax cable to measure it.

Siglent example PP510 and PP215 and PB470
10x  600VDC  (or AC peak)
1x    300VDC  (or AC peak)

[plurn]

I should also point out probes have there own voltage ratings, looking on the side of a random one I have on hand (X1: 150V, X10: 300V) its not a siglent probe, so check yours first.

If the probe cannot handle the voltage, but the scope can, it means you will just have to use a coax cable to measure it.

Siglent example PP510 and PP215 and PB470
10x  600VDC  (or AC peak)
1x    300VDC  (or AC peak)

Just checked one of my siglent PP510 probes and it says this on the probe itself:

1x   150V
10x  300V
CAT II

user guide for that probe says:
max input voltage: 1X:150V RMS CAT II&10X:300V RMS CAT II

[rf-loop]

I should also point out probes have there own voltage ratings, looking on the side of a random one I have on hand (X1: 150V, X10: 300V) its not a siglent probe, so check yours first.

If the probe cannot handle the voltage, but the scope can, it means you will just have to use a coax cable to measure it.

Siglent example PP510 and PP215 and PB470
10x  600VDC  (or AC peak)
1x    300VDC  (or AC peak)

Just checked one of my siglent PP510 probes and it says this on the probe itself:

1x   150V
10x  300V
CAT II

user guide for that probe says:
max input voltage: 1X:150V RMS CAT II&10X:300V RMS CAT II

You are right.
I chek PP215 and PB470 from real probe in my hand.
PP510 I check from Siglent list what I have get from Siglent and it say:

PP510 1X/10X attenuation, 85/18pf, 1M/10Mohm, 6M/100MHz, 300V/600V,
operating temperature -15 to 75 degrees

Same as PB470 and PP215 exept bandwidth.

Now I open one scope carton and take PP510 my hand and.....    ....and it is just as you told 150/300V.

Except...
PP215 is 300/600V DC and  PK AC

PP510 probe read 1x 150V  10x 300V CAT II and nothing more
but in probe pouch small sheet reads
1x  150V  RMS, 10x 300V RMS

Ok if I think 300V AC RMS. Then Vpk is  424V rounded 400Vpk.
But then, what is AC without frequency. Oh yes, one mHz and uHz is also AC. All except theoretical DC is AC. And it do not exist in nature. So, only conclusion is that it can also turn to 300V AC RMS, 400Vpk, 400V "DC".

Then for more mess...

Attached part of Siglent's probes data sheet. There all these probes have same info exept BW and it is also strange. I do not anymore believe it..

What to believe now. How it can be this kind of mess.

PP215 probe pouch small info sheet: 1x: 300V&10x:600V Working voltage (Vp-p)

I have not previously looked these so carefully due to fact that no one have asked anything about over some tens of volts because it is not so common that serious users use this kind of probes for high voltage things and who ever perhaps use they are normally enough educated so that they do not need ask anything. One I have told to some people, it last normal 230Vac when 10x.

But I must say,  the information is rather confusing and contradictory. I guess those who are writing these papers don't even really know what these are and how the parameters should be written so that the interpretation is unambiguous and appropriate.

So, what is final conclusion...
 
I can say "Who knows" until smoke come out.