Fan noise in test equipment

[LaurentR]

Unlike PCs that can now be easily made silent/quiet, test equipment is all over the place. One can read the datasheet, but it's always a surprise, when getting a new piece of equipment, whether it's going to be quiet or ridiculously noisy.

I found myself getting tired of the noise and swapping fans in several equipment pieces I own (DS1074Z, 33522B) and considering it on some other ones (DP832, 3446XA). Fan noise swapping is well documented in this forum for several devices like the DS1000Z and the DP8XX.

There are several aspects to fan swapping:

• Reducing mechanical/switching noise.
• Reducing airflow noise
• Reliability/specs
• Warranty

Mechanical/switching noise is the easiest one. Most of the industrial fans use have terrible noise signatures with lots of whining and clicking. OTOH, there are plenty of brands of PC fans (my favorite is Noctua) that have models with essentially no mechanical/switching noise (at least inaudible compared to the airflow noise). When a model with comparable airflow and static pressure is available, it's a no-brainer. E.g. a Gelid Silent 5 for the DS1074Z.

Reducing airflow noise is tricky as normally mean reducing airflow altogether and potentially getting the device out of spec or exceeding its max internal temperatures. Swapping for a high-end PC fan (like a Noctua) can help as they often has a lower free airflow signature. However, most of the time, the airflow noise is due to the airflow design of the case and not much can be done at equal air flow. For instance, the 33522B has an internal airflow duct along the side and this creates a significant amount of noise.

When voluntarily reducing the airflow, reliability and specs come into play. Some pieces of equipment have a fixed airflow regardless of ambient and are supposed to run at 55C operating ambient. For those, reducing airflow when placed in a human setting (usually closer to 25C, maybe peaking at 35C) can seem reasonable. For those rare pieces of equipment that have thermally-controlled fans, this is generally ok as the closed loop will maintain the appropriate airflow. The fact that the max airflow is limited is probably not an issue if the operating ambient is much lower than the max spec.

Of course, reducing airflow can also cause the instrument to not be in spec or be further from its best precision. For instance, the fan on the 3446XA is not too noisy but its whiny signature could be improved at equivalent airflow. Given that it's an instrument that is most definitely sensitive to temperature/airflow, is it worth taking the risk?

The warranty is the last issue. Some Rigol gear has a little sticker as warranty-breaking evidence with awesome methods to work around. The Keysight gear has nothing and can usually be opened without any issue. Whether changing the fan (with same or lower airflow) will make any warranty claim harder is an interesting question (provided the original fan is put back in place of course  ).

[dr.diesel]

Yup, me too.  I can't stand loud stuff, I've probably swapped fans in 50%+ of my equipment, warranty void stick or not. 

[alterbaron]

I used to care about fan noise, that is, until I got my HP 8566B. Now everything else seems whisper quiet in comparison. 

[tautech]

Reliability/specs should be classed as individual factors in TE IMHO.

Study of a Cal cert shows temperature and humidity when the unit was calibrated.
If these parameters are dramatically changed how can you be sure your TE is still within spec?