Siglent SDS1104X-E noisy fan replacement

[Bob Sava]

For anyone contemplating fan replacement, here are specs of the fan used in SDS1104X-E:

Sleeve bearing, 60mm DFM6025S, 0.11A, 3300RPM, 17.41CFM, 2.56mH2O, 24.64dBm (yeah right  )


http://www.szdosense.com/eng/products-detail.asp?cpid=177


EDIT: as I was testing few replacement fans I realized that a lot of noise is from enclosure and vent holes, not the fan itself.  Well rated and reviewed Noisebreaker XR2 (15dBm) was still too noisy, and even 11dBm XR1, while totally silent outside, it can be heard once mounted.

[TWMIV]

Slap this in there and enjoy the silence then!
https://noctua.at/en/nf-a6x25-flx

This is something that I will probably hold off on until the factory warranty expires. The fan is quite loud though...

[Bob Sava]

I think two low RPM fans, one blowing (at PSU and Zynq) and other exhausting would be more optimal and pretty much silent.

As to warranty, from what I read, at least in USA, Siglent cannot invalidate warranty w/o proving it was caused by modification.  And, if scope has been running fine for months already the only thing I can think of that could break it are firmware upgrades.

[Fungus]

As to warranty, from what I read, at least in USA, Siglent cannot invalidate warranty w/o proving it was caused by modification.  And, if scope has been running fine for months already the only thing I can think of that could break it are firmware upgrades.

I definitely regret not upgrading my Rigol fan sooner. If it's six months old then it's unlikely to fail due to hardware.

(unless you put in a completely unsuitable fan and it overheats, obviously)

[Bob Sava]

I ended up installing two Noisebreaker XR1 fans, one blowing in, the other exhausting (pic).  I  also tried more powerful XR2, which is close to original fan in specs, but two were just too noisy and likely too much.  XR1 attach with no tools, with rubber mounts that came in the box (blue part in the pic) and piece of doublesided foam under blowing fan to keep it in from sliding up and down.

Two fan push-pull configuration seems to work well and is quiet.  I keep wondering why Siglent engineers did not put their single fan on the left side, instead of right, blowing air in - it makes a lot sense to have one on the left since PSU's heatsink, Zynq processor and ADCs are on that side.

[sibir]

Just got my SDS 1104X-E (HW version 01-04, whatever that means) and I am seriously bothered by the vent noise.

My Kill-A-Watt tells me that it is using 23 W. Not a whole lot. So I wonder how much air flow is actually needed. Has anybody looked at the running, open unit with an IR camera (I am aware if the emissivity caveat)? How warm do the various parts get? And the Zynq SoC's heatsink? Which component is most critical? Maybe a thermistor based fan speed controller is the solution? Maybe a tiny fan mounted on the heatsink?

If the airflow through the vents is causing the noise and not the fan, I don't understand how having to fans would make it quieter unless the flow is lower despite having 2 fans.

I also came across this interesting comparison of different vent pattern designs with regard to noise:
https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/

Judging by the stacked vent patterns and the partially occluded fan I have the impression Siglent has not paid any attention to the noise issue

My warranty period is apparently 3 years, so I am reluctant to open up my unit just to try things...

[tautech]

Just got my SDS 1104X-E (HW version 01-04, whatever that means) and I am seriously bothered by the vent noise.
..........................

Judging by the stacked vent patterns and the partially occluded fan I have the impression Siglent has not paid any attention to the noise issue

Welcome to the forum.

Well that might be your impression but reality is different.
Designers must consider all environments where an instrument might be used and provide adequate cooling to ensure the stated specifications are met. Period.
If a quieter fan can be used in SDS1004X-E models they already would be using one as has already been done in SDS1202X-E units that are somewhat quieter than in early units.

Fan mounting OTOH can be improved so to transfer less harmonics into the chassis so it doesn't reverberate noise such as a loudspeaker might and if you are to attempt to reduce noise I recommend modifying the fan mounting and nothing else. 

My warranty period is apparently 3 years, so I am reluctant to open up my unit just to try things...

Wise.
If you can leave it for a year to lower the infant mortality risk the need to rely on warranty support should be greatly diminished.
FYI, of the many dozens I've sold just one had a very rare dodgy fan that occasionally would squeal and it was replaced under warranty without issue. Siglent wanted to know everything about that fan as I don't believe they have had other issues with them.

[bluejedi]

Another option to further reduce fan noise and improve airflow would be to remove the punched (circulair) grill fom the inner metal casing where the fan is mounted to (using a dremel tool or similar).
The plastic outer casing already contains a (striped) grill that mechanically protects the fan.

Similarly the punched (striped) metal grill on the right side (without fan) will also increase noise and reduce airflow, but probably (much) less substantial.
Removing it may further reduce noise and increase airflow.

As manufactured there are double grills both on the left (fan) side and right side of the scope: both in the metal inner casing and in the plastic outer casing.
Placing one grill behind another increases noise & turbulence and reduces air flow. This will be worse when the grills have different patterns (more obstruction of air flow), which is the case here on the fan side.

No-grill will provide the least noise and best air flow but is not safe. Therefore placing a single grill is standard. The double grills on the SDS1x04X-E however are far from optimal.

Based on my logic and limited knowledge, removing the punched grills from the metal inner casing should not cause any safety issues because the outer plastic casing already provides grills and also the power supply is already shielded with its own metal cover.
If removing the punched metal grills would be an issue (for whatever reason) then they can be replaced with wire grills which are more optimal.

Questions:

• Does removing the grills from the metal inner casing have any negative impact on safety (regulations), even while the plastic outer casing already provides grills?
• Could removing the metal grills have any substantial negative impact on RF/noise shielding?

[bluejedi]

I ended up installing two Noisebreaker XR1 fans

Do you mean Noiseblocker BlackSilent XR-1 ?

[Bob Sava]

You're right: I used two NoiseblockerXR-1 1600rpm 60mm fans.  Later, I also added some copper tape around the fans, just because I could.

[imk]

Have spent some time testing PC fans and much quitter fans can be bought for not a lot.
But maybe adding a fan speed controller would help, as from my experience you only have to reduce the power a little for a quitter fan.
I guess many of us have temperature multimeter probes so would be interesting to see how hot it gets inside the scope in a warm sunny environment with fans running flat out.

[tautech]

Another option to further reduce fan noise and improve airflow would be to remove the punched (circulair) grill fom the inner metal casing where the fan is mounted to (using a dremel tool or similar).
The plastic outer casing already contains a (striped) grill that mechanically protects the fan.

Similarly the punched (striped) metal grill on the right side (without fan) will also increase noise and reduce airflow, but probably (much) less substantial.
Removing it may further reduce noise and increase airflow.

As manufactured there are double grills both on the left (fan) side and right side of the scope: both in the metal inner casing and in the plastic outer casing.
Placing one grill behind another increases noise & turbulence and reduces air flow. This will be worse when the grills have different patterns (more obstruction of air flow), which is the case here on the fan side.

No-grill will provide the least noise and best air flow but is not safe. Therefore placing a single grill is standard. The double grills on the SDS1x04X-E however are far from optimal.

Based on my logic and limited knowledge, removing the punched grills from the metal inner casing should not cause any safety issues because the outer plastic casing already provides grills and also the power supply is already shielded with its own metal cover.
If removing the punched metal grills would be an issue (for whatever reason) then they can be replaced with wire grills which are more optimal.

Questions:

• Does removing the grills from the metal inner casing have any negative impact on safety (regulations), even while the plastic outer casing already provides grills?
• Could removing the metal grills have any substantial negative impact on RF/noise shielding?

I haven't noticed any resulting from an experiment with my SDS1104X-E completely removing the fan side metal grille which BTW didn't offer any noticeable improvement in fan noise.

Have spent some time testing PC fans and much quitter fans can be bought for not a lot.
But maybe adding a fan speed controller would help, as from my experience you only have to reduce the power a little for a quitter fan.
I guess many of us have temperature multimeter probes so would be interesting to see how hot it gets inside the scope in a warm sunny environment with fans running flat out.

This I believe is the only solution by slowing the fan slightly which must be done very carefully in a scientific manner while monitoring IC and/or exhaust gas temps along with ambient temp changes.

[1audio]

I think this https://www.quietpc.com/ac-arctic-f12-tc would be a better choice. Slow unless the sensitive part gets hot. I have used similar in other instruments and it works very well. Key question- where to sense?

These are all quiet compared to a Tek 547. It could be much worse.

[Neutrion]

Did anyone ever actually measure, how off the measurements can be with lower airflow? Because I think if the temp is going slowly up and down again, in case of a not properly done temp. control, the measurement values could start oscillating. Was removing the grill causing more measurement noise?
Because from air cisrculation and noise perspective (in case of a properly designed fan of corse.) removing anything downwards from the traiiling edge of the fan blade would absolutely make sense, even in changing efficiency. I would try a Papst with the same dimensions, and pressure curve first to be on the safe side. But don't want to loose the warranty...

[pope]

I ended up installing two Noisebreaker XR1 fans, one blowing in, the other exhausting (pic).  I  also tried more powerful XR2, which is close to original fan in specs, but two were just too noisy and likely too much.  XR1 attach with no tools, with rubber mounts that came in the box (blue part in the pic) and piece of doublesided foam under blowing fan to keep it in from sliding up and down.

Two fan push-pull configuration seems to work well and is quiet.  I keep wondering why Siglent engineers did not put their single fan on the left side, instead of right, blowing air in - it makes a lot sense to have one on the left since PSU's heatsink, Zynq processor and ADCs are on that side.

Hi, just wondering whether your scope is still working OK after 3 years? I'm thinking of doing a similar mod so I thought I'd ask for your experience 

[AdiGital]

It’s been a while since the last post but I hope this might be of use to some.

After a few months  I finally gave up and opened the scope. My idea was to measure critical temperatures and characterise the fan performance based on its voltage – when the scope is fully closed. I attached 3 probes (DS18S20) to the power supply heatsink, FPGA heatsink, and one “loose” inside the case. The probes are attached using double sided heatsink thermal tape, which is quite strong. Additionally, I placed one more probe a few cm from the inlet fan.

Next, I hooked up the SLA1016 logic analyser to give the setup more power draw. I also switched all channels, analog and digital on.
I programmed the SPD3303 to 5 1h slots. Started with 8V increasing it by 1V every hour. Below you can see the graph. I also attached the scope placement. There is not much air movement in the corner.

Looking at the data I take it’s safe to under-volt the fan. I believe I’ve read the operating temperature is up to 40 deg. Here in North-East England, where I live there is rarely above 25 deg outside. Plenty of headroom in my view.

I’m settling with powering the fan with 9V. The noise is much-much lower and bearable. I can barely hear it and it’s not annoying. Besides, it's easy to stick a standard, linear regulator without risking noise from a buck regulator.

I’m planning to do similar experiment with the SDM3055 and STD1062X soon. Can’t stand them fans!

P.S.
I just spotted a mistake. The last slot should be 9V.

[pope]

Is there any chance that although the scope would be functional, the measurements won't be accurate?

[Fungus]

Is there any chance that although the scope would be functional, the measurements won't be accurate?

Nah.

People worry too much. I don't know about the Siglent but my Rigol DS1054Z was specified to work in ambient temperatures of 50 degrees Celsius so I didn't worry too much about the fan when I swapped it here in my 24 degrees room.

I don't imagine the Siglent's spec is much different.

[Domitronic]

The SDS1104X-E has an operating temperature range from 10°C to 40°C. So not as much as the Rigol. But also more than a typical ambient temperature in most labs.

[AdiGital]

The SDS1104X-E has an operating temperature range from 10°C to 40°C. So not as much as the Rigol. But also more than a typical ambient temperature in most labs.

Looking at my chart, when the fan is on full blast (12V) FPGA heatsink is about 31°C above room temp (26°C). Applying that to the max operating temp of 40°C it's safe to assume FPGA at 70°C should be OK.

Yesterday I had it all day on 9V with the logic probe attached and it never went above 63°C. So by limiting the fan voltage I lowered the max operating temperature by 5-6°C (12V=57°C -> 9V=62°C). OK, giving a huge margin, let's say 10°C.

If I ever have 30°C in my room I won't be playing with the scope and be outside with a frosty glass of beer in my hand 

[Neutrion]

Hi AdiGital!
Do I see it right that the Rubycon caps are gone in the power supply? When was your scope built? What kind of caps do you have?

When I measured the FPGA heatsink with the original cooling in a closed house, the FPGA temp was much higher than that (Don't remember right but around 50-60 degree Celsius) and the ADC also around 60.
Are you sure that you attached the probes right? I measured it with a BM789 with original probes, and didn't even stick it to the metal correctly.
Would be an interesting experiment to see whether the ADC would might be less noisier with a heatsink and much lower temp. Did you try to not to destroy the warranty sticker?

[ceut]

Hello !
Thanks to @tautech, I have bought a brand new Siglent SDS1104X-E but super noisy fan 
I use it at home for hobby, I live in a small apartement without any garage or lab, so this one is really annoying me 

He gaves me the link to this topic, so I have just tearing down my DSO to lower the fan voltage which seems to be the only solution available.

Here are some photos of the beginning of the teardown.
>I use my (great and beautiful  ) Xiaomi Soccas hairdryer to remove the sticker 
>Then doing some maths to get the 9V power with my Siglent silent PSU

Now thinking about which solution I will use inside it, without doing any irreversible mods for the 3-years warranty.
So I will post my thought later....

[tautech]

At just 57mA fan draw use a LM7809 or LM78L09.
Best to KISS. 

[Muttley Snickers]

Thanks to @tautech, I have bought a brand new Siglent SDS1104X-E but super noisy fan 

[ceut]

Thanks to @tautech, I have bought a brand new Siglent SDS1104X-E but super noisy fan 

I was saying Thanks for Tautech to helping me to choose between Rigol anns Siglent models, and give me this topic link 
Not for the noisy fan: that was a big (and bad) surprise when I switched it on yesterday for the first time ...

[ceut]

At just 57mA fan draw use a LM7809 or LM78L09.
Best to KISS. 

Yes! It was my first idea 

But as I do "strange electronics" as some friends of mine tells me, I have made a schematic with 2x diode-bridge to have sufficient voltage drop, and a 45°C Thermal Switch to put the full power to Fan when Temperature is above 45°C inside.

I have just closed my Siglent to check what noise it does...cross finger...

Edit: I have made a photo to show that the green led is ON a very very very low value when the Siglent is powered down, and only visible at night. Also about 4.4W at the outlet shows on my UT230B wattmeter, and 26W full powered ON without any input signal.

[Muttley Snickers]

@ceut
No worries, the face palm was simply a tongue in cheek poke at tautech who happens to be a dear friend of mine. Your post almost made it sound like he forced you into purchasing this particular scope which we all know he would never do.   

[ceut]

@ceut
No worries, the face palm was simply a tongue in cheek poke at tautech who happens to be a dear friend of mine. Your post almost made it sound like he forced you into purchasing this particular scope which we all know he would never do.   

Great to read that   

[tautech]

@ceut
No worries, the face palm was simply a tongue in cheek poke at tautech who happens to be a dear friend of mine. Your post almost made it sound like he forced you into purchasing this particular scope which we all know he would never do.   

You'll never know how many of those curses in your profile were mine.   

[AdiGital]

Hi AdiGital!
Do I see it right that the Rubycon caps are gone in the power supply? When was your scope built? What kind of caps do you have?

When I measured the FPGA heatsink with the original cooling in a closed house, the FPGA temp was much higher than that (Don't remember right but around 50-60 degree Celsius) and the ADC also around 60.
Are you sure that you attached the probes right? I measured it with a BM789 with original probes, and didn't even stick it to the metal correctly.
Would be an interesting experiment to see whether the ADC would might be less noisier with a heatsink and much lower temp. Did you try to not to destroy the warranty sticker?

Sorry I didn't read your message before and have already removed the probes, put the fan on LM7809 and closed the scope - but if you want I'll re-open and verify - just let me know.

The FPGA Version from the system screen shows 2021-07-12, hardware version 01-05. I can also verify the probes with a thermal camera but I trust the measurements. I had all 4 probes showing exact (air) value before I put them inside. They were all firmly attached (by flat face of TO-92) with pretty strong thermal tape that's used for mounting heatsinks.

P.S.
Already silenced the SDM-3055 and SDG-1062X. The meter's fan is on 6V (LM7806). I didn't log the measurements but checked the temps with thermal camera. Not much heat inside but just in case I added small heatsinks to the warmest components (controller chip and +/-15V regulators on the analog board). The meter is now practically silent. The generator runs on ~8V as its switching power supply does get quite warm.

[Neutrion]

So I should learn to read...."above room temperature". So that means we are at around the same ballpark.
(If your text is not edited but it doesn't seems so.)

But still if you would happen to have time and motivation would be nice to know whether the caps were changed.

Isn't the fan on 10v normally? That stands in the SM for the scope.

Did you check if your temp. regulation might causing a cycling internal temp. change of the scope, throwing off the measurements as the fan rpm goes up and down?

Oh I am so fed up with that sticker...

[pope]

Maybe there's some replacement fan that would do the job? A noctua or some other brand?

I have both a 1104x-e and a sdg1032x and they are really noisy and annoying.

[Neutrion]

That is what I am thinking of as well,(before after 3 years getting rid of some plastic on the exhaust side as a minimum)  so now I checked the overhyped Noctua brand... they don't even provide the pressure-airflow chart of the fan, the bare minimum technical parameter. (Even the current cheap chinese has it in its doc.)
It seems to be the Dyson of the fans.

So I would rather try it with a Papst instead, usually they are very silent not just on the db scale but on the "annoyingness of the noise" scale as well. At least many years ago I did a lot of comparison of 80x80 fans, and nothing come close to them within the same airflow level. Also no commutation noise with them. (The roaring sound some slower turning fans have.)

[AdiGital]

[...]
But still if you would happen to have time and motivation would be nice to know whether the caps were changed.
[...]

As requested, the caps are identified along with the PSU version - see pics. After you mentioned ADCs I measured them with IR camera when disassembled. They showed about 50°C. So just in case I now added heatsinks to them as well.

Isn't the fan on 10v normally? That stands in the SM for the scope.

Double checked - the fan is definitely powered with 12V. I think it's the SDM3055 that's about 10.8V.

Did you check if your temp. regulation might causing a cycling internal temp. change of the scope, throwing off the measurements as the fan rpm goes up and down?

The fan isn't thermally regulated so there is no deviation. It's fixed speed.

I also included a shot of my mod. I know under-volting the fan isn't the best method but since the scope can work in 40°C environment, I assumed there is enough room to lower the air flow. It's been working for a couple of days and no issues noticed. As I mentioned earlier it's much quieter. Now having all 4 instruments on isn't a problem anymore.

[Neutrion]

Thank you for the pictures!
Nice little heatsink on the ADC. Did you check whether there is any difference in RMS (or rather standard deviation) noise level before and after the "heatsinking"?
How did you get them BTW?

It is quiet disappointing that the Rubycon caps are gone, and also begs the question wheter Dave might gets "prepared" examples sent to him for testing.

[AdiGital]

Thank you for the pictures!
Nice little heatsink on the ADC. Did you check whether there is any difference in RMS (or rather standard deviation) noise level before and after the "heatsinking"?
How did you get them BTW?

It is quiet disappointing that the Rubycon caps are gone, and also begs the question wheter Dave might gets "prepared" examples sent to him for testing.

I didn't do any other analysis than the temperature plots. In my opinion, the mod should not affect any dynamic parameters in the slightest. The only result is that I raised the "environment" temperature the scope operates in. The highest temperature increase is the FPGA heatsink which raised by 5°C and the other components even less than that. My room temperature is in the range of 22-26°C. So after the mod it's like it was working in 27-31°C without it. Way below 40°C max stated by Siglent. Also, since I used a linear regulator there is no additional noise introduced (like switching from buck converter).

The blue heatsinks are for stepper motor drivers. I bought them a while ago when building 3D printer. I believe these are the ones I used:
https://www.amazon.co.uk/Easycargo-Heatsink-conductive-adhesive-9mmx9mmx12mm/dp/B07G12KXH6/ref=sr_1_7?crid=1KGHHTW5RY9XA&keywords=stepper+motor+driver+heatsink&qid=1648118712&sprefix=stepper+motor+driver+heatsink%2Caps%2C69&sr=8-7

[Neutrion]

I was thinking about a possible much cooler ADC actually performing a bit better, but not a lot is told in its tech doc about this topic.. But in your case the difference is not much anyway, as for the cooling fins you also have  a slightly higher inside temp.

[ceut]

I also included a shot of my mod. I know under-volting the fan isn't the best method but since the scope can work in 40°C environment, I assumed there is enough room to lower the air flow. It's been working for a couple of days and no issues noticed. As I mentioned earlier it's much quieter. Now having all 4 instruments on isn't a problem anymore.

Hello,
Thank you for sharing your mods !
I have ordered this one: https://fr.aliexpress.com/item/32954848967.html 
and select "10pcs Type A".

As I have 2 bigs 12cm Noctua Fan in my computer since some years (on the CPU heatsink, in push-pull configuration) , works very well and I don't hear them at all, I also ordered the Noctua NF-A6Fx25 FLX to try it.

[MathWizard]

I'm going to tinker with my sds1104 and sdm3065, when 1 or both are on, next to my PC, it's too loud. And I don't like turning off the DMM since it has no memory of my settings.

I live in a fairly cold place, so I'll be fine turning it down a little, or trying some old PC case fans I have.

[tautech]

I'm going to tinker with my sds1104 and sdm3065......And I don't like turning off the DMM since it has no memory of my settings.

Oh yes it does !
Goto:
Utility (Shift + Dual) > Store/Recall > Power On = Last. < Last used settings saved for next boot.

[AdiGital]

[...]
Oh yes it does !
Goto:
Utility (Shift + Dual) > Store/Recall > Power On = Last. < Last used settings saved for next boot.

Any chance I've missed "Power On Line" setting, that powers on the meter after power loss/line on? I use it with SDS1104. All gear is connected to remote power plug. It's the only device that doesn't "Wake up" is the SDM.

[coppice]

As I have 2 bigs 12cm Noctua Fan in my computer since some years (on the CPU heatsink, in push-pull configuration) , works very well and I don't hear them at all, I also ordered the Noctua NF-A6Fx25 FLX to try it.

Did you try a Noctua fan, and was it effective?

[BillyO]

Just a rhetorical reply here.

In my lab I usually have at least 3 computers, a power supply, a function generator, and a benchtop DMM and sometimes multiples of each of those along with a DC load and some other equipment on with the scope(s) (I have 3 workstations) so I have never really noticed how noisy my SDS1104X-E is.  So I just compared it with the other two scopes in there and it is the quietest of the lot.

It makes about the same noise as the little Lenovo X230 I have at my soldering station when its fan is on.  From my perspective, it's pretty quiet.

[tautech]

As I have 2 bigs 12cm Noctua Fan in my computer since some years (on the CPU heatsink, in push-pull configuration) , works very well and I don't hear them at all, I also ordered the Noctua NF-A6Fx25 FLX to try it.

Did you try a Noctua fan, and was it effective?

Our customer did with a Noctua with very similar spec to OEM without gaining any significant difference.
Investigating further for him I cut the metal fan grille from mine and this too didn't gain a significant difference.

IMO the only way to address this is to undervolt or change the fan to one of a different spec, however then throw the datasheet accuracy spec out the window.

[coppice]

Just a rhetorical reply here.

In my lab I usually have at least 3 computers, a power supply, a function generator, and a benchtop DMM and sometimes multiples of each of those along with a DC load and some other equipment on with the scope(s) (I have 3 workstations) so I have never really noticed how noisy my SDS1104X-E is.  So I just compared it with the other two scopes in there and it is the quietest of the lot.

It makes about the same noise as the little Lenovo X230 I have at my soldering station when its fan is on.  From my perspective, it's pretty quiet.

Is there something seriously wrong with your notebook? I wouldn't call the SDS1104X-E noisy by test equipment standards, but those standards are pretty low. I have an X220 beside the scope right now, and the scope completely swamps the notebook.

Most of the electronics work I do these days is fairly low power, and pretty much silent. When everything I worked on was high power and really loud I wouldn't have noticed a scope like this at all.

[coppice]

As I have 2 bigs 12cm Noctua Fan in my computer since some years (on the CPU heatsink, in push-pull configuration) , works very well and I don't hear them at all, I also ordered the Noctua NF-A6Fx25 FLX to try it.

Did you try a Noctua fan, and was it effective?

Our customer did with a Noctua with very similar spec to OEM without gaining any significant difference.
Investigating further for him I cut the metal fan grille from mine and this too didn't gain a significant difference.

IMO the only way to address this is to undervolt or change the fan to one of a different spec, however then throw the datasheet accuracy spec out the window.

That's very useful input. Thanks. It sounds like you just have to live with this. Its not horrible, but it is annoying.

[tautech]

As I have 2 bigs 12cm Noctua Fan in my computer since some years (on the CPU heatsink, in push-pull configuration) , works very well and I don't hear them at all, I also ordered the Noctua NF-A6Fx25 FLX to try it.

Did you try a Noctua fan, and was it effective?

Our customer did with a Noctua with very similar spec to OEM without gaining any significant difference.
Investigating further for him I cut the metal fan grille from mine and this too didn't gain a significant difference.

IMO the only way to address this is to undervolt or change the fan to one of a different spec, however then throw the datasheet accuracy spec out the window.

That's very useful input. Thanks. It sounds like you just have to live with this. Its not horrible, but it is annoying.

The same form factor 2ch SDS1202X-E has a much quieter fan but also less than half the HW inside also and that is where they would be great to be but I think the 4ch X-E would run too hot.
The FPGA generates most of the heat as I checked on this recently when doing a warranty repair but the fan is some distance from it so maybe directed ducting and a quieter fan can do the job adequately but we mustn't overlook cooling the PSU also. The mainboard is fixed to the internal metal chassis in numerous places so that too will be adding to the overall heat dissipation picture.

To do a proper job IMO exhaust heat and FPGA temp need be monitored while progressively slowing the fan in 30 minute steps to find where temps start lifting and the corresponding fan RPM. However results must be related back to ambient and of course only be valid at those temps but for those working in climate controlled workspaces it could be done successfully.

[BillyO]

Is there something seriously wrong with your notebook?

No.  It's quiet when the fan is off, but when the fan kicks in it's noticeable.  Maybe a tad quieter than the SDS1104X-E, but not a ton.  If they are both on together they just seem to blend into each other.  My UNI-T scope makes the Siglent veritably noiseless.  Let's not even talk about my Tek 475 let alone the KORAD KA3305P power supply which re usually on too!  I honestly don't know what it would be like to actually do work in lab where the sound of the Siglent scope would be the slightest concern.  I guess the point I'm trying to get across is, if your doing any work, you will have stuff turned on.  There will be noise.

Even sitting in my office right now the sound of the PC and the NAS would drown out the Siglent scope.

[balnazzar]

Is there something seriously wrong with your notebook? I wouldn't call the SDS1104X-E noisy by test equipment standards, but those standards are pretty low. I have an X220 beside the scope right now, and the scope completely swamps the notebook.

It's something in between a hair dryer and a vacuum cleaner. I regularly use SDS5000s and they are *much* more silent than the 1104x-e.
Today, I tinkered with an old 1 GHz Agilent scope at the university. It's as deep as a CRT scope despite being digital, weighs a ton, and needs a lot of cooling. In terms of sheer sound pressure, maybe it's even a bit over the 1104x-e, but since the pitch is much lower (big fan), one finds it to be more bearable (but it worn me off after a while nonetheless).
The 1104x-e exhibits both high sound pressure and high pitch.
Pity, because it's a good scope. However, consider that the fan always run at maximum rpms, and is made to cool the scope under stress with 40C.
From my experience with small fans, if you run them at 9V rather than 12V, the difference in airflow and static pressure won't be dramatic, but the difference in acoustical noise could be substantial.

[balnazzar]

My UNI-T scope makes the Siglent veritably noiseless.  Let's not even talk about my Tek 475 let alone the KORAD KA3305P power supply which re usually on too!  I honestly don't know what it would be like to actually do work in lab where the sound of the Siglent scope would be the slightest concern.

Now I'm really thinking that maybe I got a defective 1104x-e specimen.

[balnazzar]

And consider that I work regularly in a lab with many scopes, multimeters and AWGs (~20 stations, mainly Siglent stuff, but some old Agilent/HP things too).

My 1104x-e would be by a far stretch the noisiest device in there.

[tautech]

Is there something seriously wrong with your notebook? I wouldn't call the SDS1104X-E noisy by test equipment standards, but those standards are pretty low. I have an X220 beside the scope right now, and the scope completely swamps the notebook.

It's something in between a hair dryer and a vacuum cleaner. I regularly use SDS5000s and they are *much* more silent than the 1104x-e.

SDS6000A are even quieter again ! 

From my experience with small fans, if you run them at 9V rather than 12V, the difference in airflow and static pressure won't be dramatic, but the difference in acoustical noise could be substantial.

Actually with all the concern over 4ch X-E fan noise I'm surprised no one has done a fan controller PCB with a temp sensor, picking supply from the fan header and supplying the fan from the controller PCB.

What would ppl pay for such a controller ?

[BillyO]

Now I'm really thinking that maybe I got a defective 1104x-e specimen.

Does it vibrate at all?  I once had a PC power supply that had a fan with a small piece missing from one blade.  It made enough vibration that there were strange sounds coming from nearly everything in the PC.

[BillyO]

I just did a little test.  Turned everything off in the lab then did a sound test on the UNI-T and the Siglent.  The UNI-T was 6dB louder than the Siglent.  That is quite a bit.  The Siglent was just under 3dB louder than the Lenovo X230 notebook with the X230 blowing as hard as it could.  So it is noticeably louder than the notebook, but very much quieter than the UNI-T.  The UNI-T sound is not only louder but has more high frequency noise so it even more annoying.

Anyway, I guess I'm just used to working in noisy environments as the noise it makes is generally not noticeable to me when things are normal (lots of stuff blowing).  Maybe I'm going deaf.

[balnazzar]

Now I'm really thinking that maybe I got a defective 1104x-e specimen.

Does it vibrate at all?  I once had a PC power supply that had a fan with a small piece missing from one blade.  It made enough vibration that there were strange sounds coming from nearly everything in the PC.

It doesn't vibrate at all.

[coppice]

I just did a little test.  Turned everything off in the lab then did a sound test on the UNI-T and the Siglent.  The UNI-T was 6dB louder than the Siglent.  That is quite a bit.  The Siglent was just under 3dB louder than the Lenovo X230 notebook with the X230 blowing as hard as it could.  So it is noticeably louder than the notebook, but very much quieter than the UNI-T.  The UNI-T sound is not only louder but has more high frequency noise so it even more annoying.

Anyway, I guess I'm just used to working in noisy environments as the noise it makes is generally not noticeable to me when things are normal (lots of stuff blowing).  Maybe I'm going deaf.

The fans in the Uni-T scopes I've used seem to get the chassis into resonance, which can be quite unpleasant for long periods.

[sja]

Tautech:

What about something like these cheap fan controllers: https://www.aliexpress.com/item/4000120607047.html?spm=a2g0o.productlist.main.5.3e033c8bpmZxY2&algo_pvid=6ce9d057-c345-46c6-bd57-ee2536774044&algo_exp_id=6ce9d057-c345-46c6-bd57-ee2536774044-2&pdp_ext_f=%7B%22sku_id%22%3A%2210000000332499657%22%7D&pdp_npi=2%40dis%21EUR%213.35%212.88%21%21%211.72%21%21%402100b1a616664275073707418d0764%2110000000332499657%21sea&curPageLogUid=gWuAS4LC4PNt

I have only used them in manual mode to provide cooling for devices without fan headers.

In the model that I have, the temp sensor cable is 9.5 cm from the tip to the top of the connector. Would that be long enough?

[tautech]

Tautech:

What about something like these cheap fan controllers: https://www.aliexpress.com/item/4000120607047.html?spm=a2g0o.productlist.main.5.3e033c8bpmZxY2&algo_pvid=6ce9d057-c345-46c6-bd57-ee2536774044&algo_exp_id=6ce9d057-c345-46c6-bd57-ee2536774044-2&pdp_ext_f=%7B%22sku_id%22%3A%2210000000332499657%22%7D&pdp_npi=2%40dis%21EUR%213.35%212.88%21%21%211.72%21%21%402100b1a616664275073707418d0764%2110000000332499657%21sea&curPageLogUid=gWuAS4LC4PNt

I have only used them in manual mode to provide cooling for devices without fan headers.

In the model that I have, the temp sensor cable is 9.5 cm from the tip to the top of the connector. Would that be long enough?

Depends on where the sensor is best to be located as there are a # of things to be considered:
PSU heat and the need to keep its caps cool.
Mainboard temps, to keep operating temp stable (after warm up) and system accuracy temperature stable.
Cooling the FPGA in particular where all the hard work goes on and it's the only device sporting a heatsink.

To do this properly it's within the realms of a thermal engineer and temp stable environs so to test different ambients to see how well the cooling system can keep the system temps stable.

By far the simplest solution is to throttle back the fan a little to reduce noise by how much, that is the issue.
I don't have the equipment to test temperatures accurately and all the locations needed and the fan adjustment could be quite simple with a 1W wirewound pot as we are only dealing with ~60mA fan current.
Maybe a 200 Ohm pot has the range to make adjustments.

A noncontact RPM meter can give us a OEM measurement and post mod measurement that we should be able to correlate to fan charts for flow and pressure results that might also point toward a better fan replacement.
It's just not a 5 minute job to do properly.

[chemary]

I bought this scope 4 years ago, it's a great scope more than enough for my hobby projects but it's loud fan is really annoying (sometimes I want to leave the data-logging overnight but my flat is small and I have it close to my bed).
While trying different solutions I noticed that closing the plastic enclosure increases the noise. The air exhaust is really bad designed as it's partially obstructed by the plastic enclosure and probably it creates some turbulence that increases noise. I have put some anti-vibration foam between the metal and the plastic and the noise has drop a little, I have not measured it so it's just my impression.

I tried different fans I have lying around from old computers but the original is by far the less noisy, then as suggested on this forum I was trying different fan voltages I added two 1N4148 diodes to drop the voltage 1.2V as it's hot where I live and dropping the voltage to 9V seems too much.

[LogicalDave]

@chemary Did the two diode-drops reduce the noise significantly?

[chemary]

Yes @LogicalDave, the noise now is still audible but tolerable.

[tjookum]

Sorry for digging up this old thread. Since the beginning of owning my SDS1104X-E I have always been annoyed by the fan noise. It's not overly loud but just generally annoying for a small home lab and I found myself switching it off as soon as I could in between measurements.

Well, fast forward almost 4 years and since I well and truly own this piece of equipment now I decided to do something about it.

It’s been a while since the last post but I hope this might be of use to some.

After a few months  I finally gave up and opened the scope. My idea was to measure critical temperatures and characterise the fan performance based on its voltage – when the scope is fully closed. I attached 3 probes (DS18S20) to the power supply heatsink, FPGA heatsink, and one “loose” inside the case. The probes are attached using double sided heatsink thermal tape, which is quite strong. Additionally, I placed one more probe a few cm from the inlet fan.

Next, I hooked up the SLA1016 logic analyser to give the setup more power draw. I also switched all channels, analog and digital on.
I programmed the SPD3303 to 5 1h slots. Started with 8V increasing it by 1V every hour. Below you can see the graph. I also attached the scope placement. There is not much air movement in the corner.

Looking at the data I take it’s safe to under-volt the fan. I believe I’ve read the operating temperature is up to 40 deg. Here in North-East England, where I live there is rarely above 25 deg outside. Plenty of headroom in my view.

I’m settling with powering the fan with 9V. The noise is much-much lower and bearable. I can barely hear it and it’s not annoying. Besides, it's easy to stick a standard, linear regulator without risking noise from a buck regulator.

I’m planning to do similar experiment with the SDM3055 and STD1062X soon. Can’t stand them fans!

P.S.
I just spotted a mistake. The last slot should be 9V.

After seeing the excellent post by AdiGital, I added a LM7809 on a small board with some JST XH connector so it is easily revearsable. There is plenty of space next to the fan so I mounted it with some high strength double sided tape. It rarely gets above 30C here and when it does both me and my trusty SDS1104 will be out of commission anyway.

The fan is still audible but sounds a lot more like a normal PC idling instead of the RGB gamermode beast it was before. I can definitely recommend it if your SDS1104 is out of warranty anyway and you don't live in a tropical climate. Thanks to AdiGital for doing the measurements and giving me the confidence to do something I should have done years ago.