Siglent SDG 1032X Hot Resistors in Power Supply

[sump]

Hi everyone, lets see if I get my first post right...

My SDG 1032X is plagued by quite a bit of fan noise due to its case design. I wanted to do the most simple fix and add a resistor to the fan.
Being the cautious guy that I am, I first checked temperatures with an IR camera. Well, everything seems pretty cool (up to mid 40°C), except for three 20k resistors in the power supply section, which heat up to 106°C.
They can be seen in the top middle of the picture, just below the red line:



It seems to me the resistors supply the switching circuitry with its operating power. Their relatively huge (surface) size and the fact that there are three of them tells me some heat is to be expected here.
But I am not sure if 106°C is to be expected and if even more would still be ok.

My research only revealed that a typical temperature limit for resistors could be between 105 and 150°C. So it is unclear to me if this is already maxed out or not.
The resistors are not directly in the air stream (see air inlet on top left side), but there might be turbulences providing cooling.
My measurement was done with the case open (therefore no air stream from fan.)

What do the experts here say? Is the temperature already in the red and I should not touch the fan or is there is still room to go higher - and quiter?

Thanks.

[tautech]

Welcome to the forum.

Those resistors look like they're for the VCC supply of the SMPS IC and form the series dropper from the rectified mains HV DC.
Their current should be very low when the AWG is OFF and little when the AWG is ON due to the SMPS IC getting the bulk of its VCC current requirements from the flyback winding on the SMPS transformer.

The bulk capacitor on the IC's VCC supply should supply sufficient current at startup to exceed the UVLO for the first few pulses of the switching silicon then the flyback winding will provide the IC's current requirements.
Download the IC's datasheet and check if it has a typical application design where I think you may find they're actually 200k resistors. Be careful if you are to measure them unless you discharge the HV cap.

SDG airflow comes via the fan over the digital PCB and then over the power supply and out the side vent so yes there will be some flow past these dropper resistors to assist with their cooling.
So measuring their real operating temp will be challenging with the case removed.

[bdunham7]

What do the experts here say? Is the temperature already in the red and I should not touch the fan or is there is still room to go higher - and quiter?

Let's assume, as it appears to be at first glance, that these are 5025-sized 1W@70C ambient 20K resistors with a typical max operating temp of 150C and therefore an anticipated temperature rise of 80C at full power.  And that the three of them are in series and subjected to approximately 380 volts across all of them (you should carefully measure that if you can).

So 380V and 60K ≈ 2.4W, so 0.8W each or 80% of their rating, but they show an approximate temperature rise of 80C (just guessing your ambient) so perhaps my guess as to the specs is off or the installation (circuit board and proximity to each other) increases the temperature rise a bit.   In any case, the resistors should be fine with a 10C increase in temperature, but the circuit board might be getting a bit uncomfortable.  However, the fan probably reduces that temperature a fair amount and you would also think there would be some margin in your case because the device is rated to work in temperatures up to 40C.  So I think you would be safe to slow the fan down a little bit.

[sump]

Download the IC's datasheet and check if it has a typical application design where I think you may find they're actually 200k resistors. Be careful if you are to measure them unless you discharge the HV cap.

I measured them and they appear to be 20k as indicated by their marking. The chip used is a L6599AD. The resistors appear to be directly connected (0 Ohm, no voltage drop) to Vcc of this chip. Therefore I would expect the voltage drop across the resistors (and heat) always to be the same when the device is turned on.

The reference design(1) I found seems to take a more elaborated approach to generating Vcc, probably similar to what you had in mind. It seems Siglent went for a simpler solution. (No blame, its a cheap device after all.)

Let's assume, as it appears to be at first glance, that these are 5025-sized 1W@70C ambient 20K resistors with a typical max operating temp of 150C and therefore an anticipated temperature rise of 80C at full power.  And that the three of them are in series and subjected to approximately 380 volts across all of them (you should carefully measure that if you can).

The voltage across all three resistors turned out to be 300V. Using some 3d printed dummies I narrowed the resitor form factor to be 2512/6332. Google tells me they make those from 0.5 to 1.5 Watts.  The ones used appear to be more on the thinner side. So maybe even just 0.5W parts? If Pmax specified is typically associated with a 80°C temperature rise - as in the example you gave - everything would fall into place.

So I think you would be safe to slow the fan down a little bit.

Already done. And thank you both again for taking the time to help me out.

1: https://www.st.com/resource/en/datasheet/l6599a.pdf, Page 31

[tautech]

Download the IC's datasheet and check if it has a typical application design where I think you may find they're actually 200k resistors. Be careful if you are to measure them unless you discharge the HV cap.

I measured them and they appear to be 20k as indicated by their marking. The chip used is a L6599AD. The resistors appear to be directly connected (0 Ohm, no voltage drop) to Vcc of this chip. Therefore I would expect the voltage drop across the resistors (and heat) always to be the same when the device is turned on.

The reference design(1) I found seems to take a more elaborated approach to generating Vcc, probably similar to what you had in mind. It seems Siglent went for a simpler solution. (No blame, its a cheap device after all.)

https://www.st.com/resource/en/datasheet/l6599a.pdf, Page 31

Thanks, quite interesting and I presume if indeed a flyback supply wasn't used it was for reduced EMI so to keep the AWG's output cleaner.
Never had one apart however if I do I'll take a good close look at the SMPS implementation.