EEVblog #985 - Siglent SDS1202X-E Oscilloscope Teardown

[NivagSwerdna]

Nice review.  I find the multi-layer board completely mind boggling.  Is that routed by human, computer or magic?   It's also completely awesome that those many layer boards are becoming such commodities.

As far as the scope is concerned it is looking like interesting times... just need some manufacturer to have 4 channels for the price of 2 with serial decode and I'll be up for an upgrade (almost tempted for the Rigol but it is starting to look like an old model now).

[GreggD]

R185, Why is this oversized novelty resistor doing in the oscillator over-thruster ?

[bittumbler]

The capacitors in the power supply have that white goo over their vents.
Is this o.k? Will the vents still work?

[Fungus]

Cheaper then decode options on some scopes

Maybe this will the the straw that breaks that particular camel's back.

(Fingers crossed...)

[bitwelder]

What a boring scope!
Already comes with full bandwidth, serial decoding is included, UART connector nice and easy, just waiting to be plugged, console that boot into root shell
Nothing left to hack 

  to Siglent

[jimdeane]

I'm happy with my purchase back in late '15, but I've kept my eye on scopes with the thought of "what scope would I buy right now if I was in the market for my first scope". Answer has stayed 1054 for a long time, but I'm thinking it just changed.

[Fungus]

I can see it taking quite a few sales away from Rigol at that price, even with only two channels.

[PA0PBZ]

Could they have put the caps in the power supply any nearer to the heat sink?

[nctnico]

Notice Siglent roll their own PSU, not some off the shelf item that they can't maintain the spec and QC of.

I don't want to derail but that is a slightly surprising and suspiciously silly slippery slope argument.  It's not like they depend on the performance of any other third party components in this design amirite?

Only offered as an observation.

Over the years there have been a # of threads here with failures of scope PSU's, in many cases they have not been made by the scope manufacturer. Could you call this modus operandi a weak link ? I do.

That is mainly due to way more units sold and older equipment. Rolling your own PSU is just a waste of resources and there are many ways to screw up reliability (like having electrolytics close to heatsinks as seen on the video). Also it will be easier to get a replacement for a generic PSU than a proprietary one.

[tautech]

When will the scope actually be available?

To dealers in the next week or two, then with the time to have them shipped some of us should have them before the end of this month.

[tigrou]

Hi. I have two questions related to the PSU :
- Why do we have two six pins (usually they are 4 pins) optocouplers ? Isn't one enough ?

- What is the purpose of the two mosfets (with big heatsinks) on the left side (low voltage board side) ? The one on the right (high voltage board) is acting as a switch for the SMPS I presume.

[rf-loop]

- Why do we have two optocouplers ? Isn't one enough ?

A little tip. See the directions of information transfer.
 

[nfmax]

- Why do we have two optocouplers ? Isn't one enough ?

A little tip. See the directions of information transfer.
 

I doubt it - normally only one signal, sensing the secondary regulated voltage, crosses the barrier back to the primary-side control IC. The second optoisolator is probably carrying the 'line trigger' signal from primary side to secondary side.

[rf-loop]

- Why do we have two optocouplers ? Isn't one enough ?

A little tip. See the directions of information transfer.
 

I doubt it - ....
The second optoisolator is probably carrying the 'line trigger' signal from primary side to secondary side.

This is not information transfer??

[nfmax]

- Why do we have two optocouplers ? Isn't one enough ?

A little tip. See the directions of information transfer.
 

I doubt it - ....
The second optoisolator is probably carrying the 'line trigger' signal from primary side to secondary side.

This is not information transfer??

Certainly it is; but it is not, as the original question asks, 'part of the PSU', even though it is on the PSU PWA.

[gnif]

You had a full blown shell on the prompt there, ash is the shell. Try run some of the following to get some interesting information out:

Code: [Select]

cat /proc/cpuinfo
cat /proc/meminfo

And you can list all the files, provided it has the tool installed

Code: [Select]

find /


[Fungus]

You had a full blown shell on the prompt there, ash is the shell. Try run some of the following to get some interesting information out:

Code: [Select]

cat /proc/cpuinfo
cat /proc/meminfo


From the video:

And you can list all the files, provided it has the tool installed

Code: [Select]

find /


It also has the OS, kernel version, etc.

[gnif]

You had a full blown shell on the prompt there, ash is the shell. Try run some of the following to get some interesting information out:

Code: [Select]

cat /proc/cpuinfo
cat /proc/meminfo


From the video:

And you can list all the files, provided it has the tool installed

Code: [Select]

find /


It also has the OS, kernel version, etc.

1) they were examples
2) cpuinfo will show more then the boot up, such as supported feature set
3) meminfo will show how much memory the OS is consuming
4) find will list all files, among which some will stand out as custom things silent have written and might make for an interesting peek at.

[Stephan_T]

Hi Dave,

you may also try the commands "mount" and "df" while a USB memory stick is connected. This way you can find out where it is mounted and than you can copy config files from the Scope to the stick for further external investigation.

For example with

Code: [Select]

cp -R /etc /whatever-path-the USB-drive-is-mounted-toyou could copy the entire /etc directory to the stick.

[bitwelder]

BTW hopefully the scope comes with the necessary GNU license and an offer to receive the source code of the GPL-licensed parts, which it won't spill all the secret Siglent sauce but it could reveal something interesting anyway.

[ali_asadzadeh]

Rigol Should build a 4 channel one using similar hardware to this, why they did not do it 4 channel, they have the ADC and the ZYNQ, they just needed to add the front end

[Monkeh]

Nice review.  I find the multi-layer board completely mind boggling.  Is that routed by human, computer or magic?   It's also completely awesome that those many layer boards are becoming such commodities.

They're not that hard to make...

There are much more complex boards in your computer and phone. They've been relatively low cost, commodity items for many years.

[Fungus]

Nice review.  I find the multi-layer board completely mind boggling.  Is that routed by human, computer or magic?   It's also completely awesome that those many layer boards are becoming such commodities.

They're not that hard to make...

The whole point of having more layers is to make it easier to route.

Can't find a route for a wire? Add another layer!

[Brumby]

... beats a bodge wire any day of the week - and twice on Sundays!

[mash107]

Rigol Should build a 4 channel one using similar hardware to this, why they did not do it 4 channel, they have the ADC and the ZYNQ, they just needed to add the front end

at a $370 that's a tall order. You'd need another dual channel 1GSPS ADC ($100-$200 itself) and 2 front ends to make a four channel scope.

I think if Siglent offered a similar four channel 200MHz scope for less than $500 (or even slightly above), they'd obliterated the ciompetition (e.g. DS1104Z)