Low Noise (Sound) Oscilloscope

[Fungus]

Of course it will reduce the lifetime on paper but will it make a noticeable difference in real life if there's only 10% less airflow?

First explain from which dark hole the number '10% less' is coming from.

The static pressure of the stock-vs-replacement fans is 25% different, the max airflow is exactly the same. There's not much blocking the fan, so... 10% ballpark, maybe 15%. Use 20% if you want, I don't really care, I don't believe it will be reduced by the full 25% though, not with the fan pressed right up against the output grille.

Let's see what the numbers say.

[Andreas]

I really can't understand why all scope is so noisy (at least the one I got).

I have a PicoScope 5444A USB Scope -> fanless design (except the laptop where it is attached).

with best regards

Andreas

[DaJMasta]

Yeah, generally my approach has been to look for the fan with the higher rated airflow and less noise, a lot of times the parts they're cooling already run warm and prolonged heat will eventually make them fail, so in the interest of having stuff stick around, I try to do at least as good of a job moving air as the OEM fans. 

I don't know about Siglents, but ... as I pointed out in the "Rigol fan replacement" thread, a Rigol is designed to work in 50 degree Celsius ambient temperatures. I imagine Siglents aren't much different.

That means there's quite a lot of wiggle room if you're using it in a 25 degree office.

Sure, but the wiggle room is a benefit for test equipment.  Their overall lifetime is higher because of that sort of tolerance, and since I do occasionally upgrade my kit and sell them, I don't want to put in some dramatically quieter but just barely performing fan that will just be a timebomb in any space that isn't an air conditioned office.  I also end up getting used equipment pretty commonly, so to try to get a nice long lifetime out of a piece of gear that may have already been heavily used, getting a closely rated fan is a precaution I have no issue taking.  I've used my scope and counter outdoors, I'll occasionally carry stuff to somewhere else, and if it's possible (and it usually is), I'd like for them to perform well in any situation they normally would even if I modify them.

Sure, it's gilding the lily, as Dave would say, but so is input protection and extra EMI filtering on the power cable..... it may not be required for normal operation, but it sure comes in handy sometimes!

[Fungus]

I don't want to put in some dramatically quieter but just barely performing fan ...

Sure, but I was only talking maybe 10-15% worse. The fan I put in my Rigol (for example) was chosen for noise/performance ratio, not because it was the quietest possible.

[GregDunn]

It surely is complex and not linear.  Analog electronics and computer hard drives also have been shown to exhibit extended lifetimes with relatively small-looking decrease in operating temperatures.  A significant amount of writing has been done on the subject:

https://www.electronics-cooling.com/2017/08/10c-increase-temperature-really-reduce-life-electronics-half/

https://ieeexplore.ieee.org/abstract/document/6740182/

https://www.extron.com/company/article.aspx?id=thermalmgt1_ts

[exe]

That's why I think it's better to measure temperature of critical components before doing any conclusions about fans (if this is feasible).

I personally often trade cooling efficiency for silence. May be I'm accelerating ageing of my equipment, but that's deliberate choice (but my equipment is not too expensive). Having said that, I did kill a CPU and a video card, but they were severely abused and had many hours of work well above recommended temperature.

There is another risk of getting the instrument out of spec. But I think it's more about precision bench DMMs.

[Bob Sava]

RTB2000 seems like a well designed scope (@20:58) that is supposedly very quiet:

https://youtu.be/mcgJSKxj0i0?t=20m58s

[Fungus]

There is another risk of getting the instrument out of spec. But I think it's more about precision bench DMMs.

Precision bench DMMs usually have internal heaters to run the voltage reference at a temperature well above 'normal' ambient just so they can stay in spec even on hot days.

[exe]

Rumors say expensive DMMs can compensate temperature drift in software. So, they rely on fans to have certain known performance, otherwise software model will not work as expected. Don't know if this is true or not, that's what I read in a volt-nut thread.

[Fungus]

Rumors say expensive DMMs can compensate temperature drift in software. So, they rely on fans to have certain known performance, otherwise software model will not work as expected. Don't know if this is true or not, that's what I read in a volt-nut thread.

Maybe so but it's much easier to heat the reverence voltage to a known temperature.

[glarsson]

Rumors say expensive DMMs can compensate temperature drift in software. So, they rely on fans to have certain known performance, otherwise software model will not work as expected. Don't know if this is true or not, that's what I read in a volt-nut thread.

But the cooling performance also depends on the room temperature and any object blocking air from entering or leaving. They have no external sensors for this and internal sensors will se the combined result of fan performance, room temperature and any blockage. That means that fan performance (within reason) can't be that important for the software.

[Kleinstein]

With a change in fan speed, there will be a change in temperature gradients, not just the overall temperature. A change in room temperature is more like changing the overall temperature if the fan's speed stays the same.  It depends on the design on how important the air speed is - ideally it would not matter, but ideally temperature would not matter either.

With modern DSOs things are usually way less sensitive than low DC voltage parts like a DMM. So in an DSO a spped controlled fan is usually not a problem, just like with an PC.