The Siglent SDG2042X Thread

[DL2XY]

...
Perhaps you could show it with a more faster modulation? 1 Hz is quite slow you know and superimposed on a square wave is hard for me to visualise on a scope.
...

The spectrum clearly shows 1kHz modulation so i think 1Hz is a typo.
Have run this on a SDG 6052 for about an hour - no probs.

 

[hpw]

...
Perhaps you could show it with a more faster modulation? 1 Hz is quite slow you know and superimposed on a square wave is hard for me to visualise on a scope.
...

The spectrum clearly shows 1kHz modulation so i think 1Hz is a typo.
Have run this on a SDG 6052 for about an hour - no probs.

just read as posted:

- FM 1kHz = modulation
- Freq. Dev 1Hz

and than wait about 1 minute, sometimes the mod button has to be pressed twice, so the modulation starts again

[Blue]

As I have changed the fan to a quiter version, I wanted to analyse the temperature of the components using my flir camera.

Quite shocking is the extreme heat produced by some resistors.

For example in the power supply:

[Blue]

The DAC area

[Blue]

another very small resistor:
 20210329_204840_0.jpg (83.06 kB. 384x288 - viewed 160 times.)

[Blue]

[Blue]

some more

[Blue]

The main problems are in the powersupply:

D18 (Shottkey?) diode @ 102 °C
Three 2002 Resistors R29,  R25, R44 @ 170 °C

I have an input of 230.4 V AC right now!

[Blue]

[Blue]

What do you think of these findings?

Is there anyone who had a failing powersupply?

[TurboTom]

...
Perhaps you could show it with a more faster modulation? 1 Hz is quite slow you know and superimposed on a square wave is hard for me to visualise on a scope.
...

The spectrum clearly shows 1kHz modulation so i think 1Hz is a typo.
Have run this on a SDG 6052 for about an hour - no probs.

The SDG6000X has got its own share of problems under these condintions as well. See here.


@Blue: The temps are okay except for the 160C resistor... That's smoking hot, it will tan the PCB in no time, make the copper clad adhesive brittle and even foul the solder eventually. I guess these are the startup resistors of the PSU that get that hot. If your in the Netherlands as shown in your avatar (?) -- high five to our neighbours  --, your SDG will run on 230VAC, causing these dropper resistors to dissipate a considerable amount (four times as much) of power in comparison to countries that run a 115VAC mains (but that's less than half as much fun "to the touch"  ). I may check temps in my SDG6000X that's probably equipped with the same PSU, and maybe replace the resistors with such of twice the resistance. My SDG will probably never run off 115V...

[tautech]

What do you think of these findings?

Is there anyone who had a failing powersupply?

Not here in NZ or anywhere else that I can recall.

[Blue]

@Blue: The temps are okay except for the 160C resistor... That's smoking hot, it will tan the PCB in no time, make the copper clad adhesive brittle and even foul the solder eventually. I guess these are the startup resistors of the PSU that get that hot. If your in the Netherlands as shown in your avatar (?) -- high five to our neighbours  --, your SDG will run on 230VAC, causing these dropper resistors to dissipate a considerable amount (four times as much) of power in comparison to countries that run a 115VAC mains (but that's less than half as much fun "to the touch"  ). I may check temps in my SDG6000X that's probably equipped with the same PSU, and maybe replace the resistors with such of twice the resistance. My SDG will probably never run off 115V...

Do you think that these resistors were designed for a powersupply running at 110V?
Would replacing these 20K Ohm to e.g. 40 Kohm still work?

I bought them at Batronix, will they complain about me opening the SDG?
What shall I do?

[TurboTom]

Use it as it is. The resistors will run at this temerature, no problem, but the PCB will suffer. Nevertheless, I haven't heard of any of these suppies fail. So probably, just "don't worry, be happy" and deal with the problems if they should arise. Still, IMO, it's not a good design practice by the manufacturer to allow components to run that hot in some parts of the woods...

[Blue]

Apparantly my powersupply in the SDG is different than show in Dave's teardown

They have changed things and botched up

[tautech]

Apparantly my powersupply in the SDG is different than show in Dave's teardown

They have changed things and botched up

The complete lack of PSU warranty claims suggests otherwise.

[Roger Need]

As I have changed the fan to a quieter version,

Blue, 

What fan make/model # did you use for the replacement?  Did it reduce the noise level much?

Roger

[exe]

@Blue: The temps are okay except for the 160C resistor...

From the picture it's not obvious to me if it's a peak temperature on the image, or it's just a range indicator. I'd expect central marker to show the peak temperature.

So... what model of thermal camera is that?

[Blue]

In the left upper corner you weill see two numbers with an arrow (solid triangle). They correspond to the temperatures of the triangles in the picture.
They are also min and max of the whole picture when the temperature range is not locked (see left lower corner while using the instrument - is not recorded in these pictures).

In addition, there is a centre cross with it's own temperature.

I'm using a Therm-App made by Ogal.

[blurpy]

So... what model of thermal camera is that?

Wondering that as well, looking into one and the pictures looked pretty good.

[exe]

In the left upper corner you weill see two numbers with an arrow (solid triangle). They correspond to the temperatures of the triangles in the picture.

Ah, missed triangles on the picture. Oh shi~, that's bad. The temperature here is during the spring, and with open enclosure. I can imagine during the summer it'll be significantly hotter.

[Blue]

I checked the readings from the flir with a ultra small thermocouple made from GG-KI-36-SLE and an AD595 / 9V battery
It is a bit difficult to get a 0.3 mm thermocouple attached but I managed.
The readings from the thermocouple attached to the resistor with Kapton tape was around 120C. Was not perfectly connected. Still way too high.



Few uncertainties remain. Emissivity used in the flir was 0.95. Are the components being measured correctly?
Thermocouple was not calibrated above 100C.

Anyway, I remain confident that the resistors and diode are being overstressed.
And for those who think it is ok, it would be nice if you power your Siglent on for at least 3 months, 24/7. And then report back here what happened. Would be nice if you could take pictures.

Does anyone dare?

[JohnG]

I would consider the following possibility:

If you are doing thermal images, that implies the cover is off. This is likely to affect the airflow substantially, and often makes the cooling worse because the air can take the easy path out the open box, rather than the harder path along the circuit board and components.

When you use the thermocouple, you can put the cover on and see if it makes any difference.

Cheers,
John

[Blue]

Hi JohnG,

No difference measured with or without plastic cover (Its not possible to slide the metal cover back without slicing the thermocouple)
Hardly any airflow over the components due to their location and shielding.

[tautech]

Hi JohnG,

No difference measured with or without plastic cover (Its not possible to slide the metal cover back without slicing the thermocouple)
Hardly any airflow over the components due to their location and shielding.

Yes well if you have changed the fan and run it on low speed as you reported here:
https://www.eevblog.com/forum/testgear/the-siglent-sdg2042x-thread/msg3535016/#msg3535016

I'm sorry Blue but we can't take your reports seriously when you're not running the OEM fan.