How do i choose a replacement fan?

[JPortici]

In the saga of repairing a lecroy 7200A:
Mainframe is almost done, there is a little thing left to do: it is noisy as hell.
There are three fans, one on the power supply and one per plugin slot. I'd like to replace these 20-something years old fans, constantly running at full throttle, with something quieter.
Manual and service manual state that there should be a good airflow to remove heat from both the PS and the modules, as the ASICs in the modules will be damaged from eccessive heat.. and the plugins gets really warm during normal operation.

There is not a fan going directly in the mainboard but i don't think that's a problem, it's "just" a 386 machine. I remember the family computer from the windows 95 era to have only a small heatsink before and a very small fan after a power supply failure that killed the mobo (AMD K5)

Anyway, i haven't yet looked at the fan specifications, but what are the usual guidelines to get a fan that will be quieter while providing the same airflow, or better?
(Of course, the dimensions and voltages should be the same)



Previous episodes of the saga:
- Problem on intermittently blue CRT was identified as a broken leg or bad solder in the CRT driver board, will actually do the repair once i gain the courage do dismount the CRT to access the board (Fear of causing damages to the tube)
- Possible replacement pot from component manufacturer on its way to test if compatible

[OiD]

Hi,

First i'd check the power the fans draw and see if there is a spec that says how much air it moves and a sound level rating.
Then check for the same airflow fans at the lowest power and noise. I like delta fans but they can be pretty expensive. You can always check perfomance/quality pc cooling fans to see if some models offer more competitive prices.

[Armadillo]

Speed = Heat;
and those Giga Hertz ASIC are super expensive, so noise and airflow are part of the family, though we tried to reduce it and have got to think of the expensive chips cooling needs. Someone told me heat is the major killer.

So I broke down my works into 2 region;

Less than 100Mhz I use entry scope [some even don't come with fan].
>100Mhz no choice bear with the noise. Its an orchestra of noises in fact if you turn on few of these together.  symphony!

Just my opinion.

[JPortici]

I get your point, but if improvements can be made... plus I'd like to be ready for the day one of the fans dies

[OiD]

You could go for a better fan that moves more air and put a few deflectors inside and out to absorb the sound. If noise is an issue there are plenty of work arounds. If they are worth it is something else! But yeah, heat kills 

[Armadillo]

Failed bearings, misalignment, out of balance, poor design aside;

Airfoil or aerofoil blades, so much the current technology can do, regardless, You need to be at least 10 fold improved for the human ears to be able to even discern the minute of improvement.
The question is, is the fan noisy or the air channelling assembly through the panel, creating static regains inside, the enclosure amplifying the noise noisy?
Hence this fixed panel acoustic enclosure can be used as a meter to discern quantity [Qcfm] of airflow flowing through.
If suddenly you find it quite, not that its improved, it means...the airflow is actually low [fake fan parameter].
But hey, the commercial world still needs to revolve creating more jobs and we have things to sell and works to do.
Acoustic Engineer anyone?
So,
as long as you are happy with your personal touches. . Cheers. 

[drussell]

It really depends....

Are the original fans noisy because the motor structure and bearings are excessively noisy or is it the actual airflow that is causing most of the noise?

[alm]

What Armadillo said. Do not expect miracles unless the old fan was really worn or you lower the airflow. Especially at high airflow rates where turbulence noise tends to dominate any bearing noise. At low speeds a sleeve bearing fan may be significantly more quiet than a ball bearing fan (at the cost of a shorter lifespan, especially in hot positions or facing up or down). But low speed fans are not very common in test equipment.

Also keep in mind that the amount of air a fan moves is a function of the pressure it has to work against. Two fans that move the same amount of air in free air may behave very differently when screwed to a dense heat sink. Good fan manufacturers, like Papst, will have curves of airflow vs pressure in their datasheets. When replacing a fan, you want to match the airflow at whatever pressure the fan was working at. Comparing their datasheet curves is one way to do that. Fans designed for a high back pressure will generally be deeper (e.g. 120x120x25mm vs 120x120x38mm fans), have a fatter hub and stubbier blades.

[xwarp]

Since you are visually looking at a display, while listening to the noise that is made to create that display, you could always opt for these:



 

I'm kidding as I have a really nice HP sweep generator that is old, uses tubes and has a fan that is pretty loud.

[DaJMasta]

As said, you want to match or exceed as many parameters as you can, but it's rare that test equipment uses fans that are designed for sound reduction, so there is probably something available that will match size and airflow requirements that is at least a bit quieter.

Again, much of this has been said, but there are some other factors to consider:
If you have extra clearance for a thicker fan, going to a thicker one can mean lower RPM for the same airflow, so you can get a quieter rated replacement.
The turbulence generated by nearby parts or grills will often dominate fan noise - with a lot of manual labor you can try dulling edges on punched metal and such... but it's usually not worth it.  You probably want the shielding to be in place when it's on test gear, but a rubber or nylon washer or two worth of thickness between the fan and the vent holes can actually reduce turbulence considerably.
If the design doesn't already have them, using rubber mounts (washers, O rings through the mounting holes, or full sized gaskets) can help reduce any vibration noise from the fan.


With high airflow fans in test equipment not designed to be really quiet, you really can't expect silence, but you can generally  improve it considerably.  If you happen to be looking for 120mm fans without the space for thicker than 25mm and you need high airflow (100ish CFM), I can recommend the Scythe Slipstream series.  Fairly cheap compared to many quiet high airflow fans (<$15), but I swapped one into a recent repair and while it's far from silent, it is MUCH quieter than it was.

[orin]

Also keep in mind that the amount of air a fan moves is a function of the pressure it has to work against. Two fans that move the same amount of air in free air may behave very differently when screwed to a dense heat sink. Good fan manufacturers, like Papst, will have curves of airflow vs pressure in their datasheets. When replacing a fan, you want to match the airflow at whatever pressure the fan was working at. Comparing their datasheet curves is one way to do that. Fans designed for a high back pressure will generally be deeper (e.g. 120x120x25mm vs 120x120x38mm fans), have a fatter hub and stubbier blades.

This can't be stated often enough.

You have to match the airflow vs. pressure curves, not just the CFM rating.

If you are replacing a fan in a piece of equipment that was rated to run at 45 deg C ambient and your lab never goes above 25 deg C, then you could probably get away with a quieter fan that moves less air - but you are taking a risk...

[Armadillo]

If you have extra clearance for a thicker fan, going to a thicker one can mean lower RPM for the same airflow, so you can get a quieter rated replacement.

In most of the cases, because of space constraints, fans had been installed from other than ideal conditions so there is not a need to even make it adversely worse without due consideration for the "system effect factor". This is a deep science subject.

As I said above, if suddenly you find it's quieter.............. Yeap! don't even think that is better unless you have a system of measuring it and attesting to the performance "as-installed".

Will you risk your gears with words of mouth performance pieces.?

Well the truth is some actually would [they didn't read this forum discussion]. 

[nctnico]

I get your point, but if improvements can be made... plus I'd like to be ready for the day one of the fans dies

What you need to look for in the specs is the amount of airflow (volume per time like cubic feet/second, m3/minute, etc) and the maximum pressure the fan can produce which indicates how strong the fan is. From there select a new one with at least as much airflow and pressure with the least noise. What also can help is to use rubber fan mounting so the vibration of the fan isn't coupled into the case.