Low noise fan for HP 3458A ?

[carl_lab]

I want to renew standard fan of that old 3458A because its noise level is painful to my ears...

Does anybody know what low noise high quality fan I could look for?
Ball or sleeve bearings?

How to determine correct airflow?

[pelule]

In some other thread for the 3458A someone mentioned:

For reference: Keysight currently sells ás FAN for the 3458A the Papst 622L
- easy to get at Digikey/Mouser etc and it will assure proper air flow and is rated to handle 15 volts.

[Theboel]

in my day works I use many papst 622L but I replace the with Noctua NF-A6x25 its very quite the datasheet say it's 10dB quieter but don't shoot me caused I don't measure it but from what I hear it very quite.
the only problem is Noctua fan use 5V instead of 12V in some application its a problem.
http://datasheet.octopart.com/622L-EBM-Papst-datasheet-62110739.pdf
http://noctua.at/en/products/fan/nf-a6x25-5v/specification

[TheSteve]

The 622L is reasonably loud in my 3458A's. I can't say I am a fan of the noise but if it what Keysight uses it is what I will keep using, I want the same airflow they expect the 3458A to have.

[carl_lab]

in my day works I use many papst 622L but I replace the with Noctua NF-A6x25 its very quite the datasheet say it's 10dB quieter but don't shoot me caused I don't measure it but from what I hear it very quite.
the only problem is Noctua fan use 5V instead of 12V in some application its a problem.
http://datasheet.octopart.com/622L-EBM-Papst-datasheet-62110739.pdf
http://noctua.at/en/products/fan/nf-a6x25-5v/specification

Thank you.
I found 12V version of this fan here:
https://www.reichelt.de/Case-Fans-40-70-mm/NO-NF-A6X25-FLX/3/index.html?ACTION=3&LA=446&ARTICLE=179658&GROUPID=7315&artnr=NO+NF-A6X25+FLX&SEARCH=fan%2Bnoctua
Sadly I couldn't find an airflow diagram to compare with original.

[dacman]

Airflow
622L: 12.4 cubic ft/min
NF-A6x25 w/L.N.A.: 13.8 cubic ft/min (converted from 23.5 cubic m/h)
NF-A6x25 w/U.L.N.A.: 9.8 cubic ft/min (converted from 16.7 cubic m/h)

[rastro]

What about induced electrical noise? 
I know this isn't specified but the 3458A was built and probably characterized with a specific fan. 
It also seems like HP equipment seems to favor Papst fans on their test equipment - perhaps for  a reason other than air flow. 

[carl_lab]

Airflow
622L: 12.4 cubic ft/min
NF-A6x25 w/L.N.A.: 13.8 cubic ft/min (converted from 23.5 cubic m/h)
NF-A6x25 w/U.L.N.A.: 9.8 cubic ft/min (converted from 16.7 cubic m/h)

Airflow depends on velocity pressure. These values apply at what pressure?

What is (U)LNA (ultra low noise adaptor)? Is it just an adaptor from 12V to 7V (=12V-5V) or a voltage regulator?

It also seems like HP equipment seems to favor Papst fans on their test equipment - perhaps for  a reason other than air flow.

I think they've chosen Papst for it's reliablity, not for extraordinary airflow, nor it's silentness.

[Kleinstein]

Free air flow is not the only parameter to look at. The important parameter is the air flow with the specific pressure needed. So one should also make sure the maximum pressure is similar. To a certain degree the fans can be designed for low pressure and high flow or the other way around. So just looking for a fan with highest (free) flow might very well end up with one that can no deliver much pressure.

Unless really bad, electrical noise should not be an issue, with the fan powered from the ground referenced part of the circuit.

[rastro]

Free air flow is not the only parameter to look at. The important parameter is the air flow with the specific pressure needed. So one should also make sure the maximum pressure is similar. To a certain degree the fans can be designed for low pressure and high flow or the other way around. So just looking for a fan with highest (free) flow might very well end up with one that can no deliver much pressure.
Unless really bad, electrical noise should not be an issue, with the fan powered from the ground referenced part of the circuit.

How would the systems pressure/airflow be determined? 
Would you put a barometric sensor inside the system at key locations?  Maybe a temp sensor as well.  Then compare to the new fan.

Also is fan speed any indicator of the pressure/airflow performance.  The Pabst data sheet shows it over 1000 rpm slower than the proposed replacement.

[Kjelt]

Noctua is pricy but very good, I exchanged all the fans in my managed pro 10Gb switches with Noctua's.

[carl_lab]

How would the systems pressure/airflow be determined? 
Would you put a barometric sensor inside the system at key locations?

That's not a bad idea!

[TheSteve]

The Papst fan is 4 times the price of other fans as it is. Are Noctua priced even higher?

[Kleinstein]

A complete data-sheet would ideally have a curve of air flow vs pressure difference. Likely the shape of the curve would be similar (like a more or less straight line between the extremes), but with different scales. So it is about the two points of free air flow and the peak pressure. The interesting part would be the airflow at an somewhat intermediate point.

I don't think there will be simple relation with rotational speed to the pressure. It is also the number, size and shape of the blades and the angle.

Measuring pressure difference or temperature (difference) sound like a good idea, especially if you can still run a test with the original fan. I am not sure about the pressure range, a simple version with water in a tube could be good enough.

For the DMM the important part is to get a similar airflow, as there is not only one temperature, but also temperature differences that depend on the air flow. So to get the ultimate precision the airflow should also be constant - thus the usual procedure of cleaning the filter. A temperature regulated fan is not such a good idea, as it might initially look. This would only work well if the critical parts where enclosed without direct air flow.

[Kleinstein]

I just looked at the DS of the fans (Pabst PCL0412F and  Coctua NF-A6X25 FLX):
They both seem to have a peak pressure of a little over 20 Pa. So it looks like a suitable replacement.

[carl_lab]

Noctua spec is better than Papst over the complete range:

Max. voltage of Noctua is 13V vs. Papst 15V. That has to be checked before replacing!

Got the noctua fan yesterday.
Opened 3458A an measured voltage: 15V. Nevertheless installed new fan @15V. It runs very silent. 
Had to cut Noctua connector (3 pins) and soldered on original Papst 2-pin-connector.

After running the 3458A overnight, temperatur inside (TEMP?) is 37.0°C. So I think it' s OK.

[Ero-Shan]

Sorry for bringing this up from the dead.

But thanks for the recommendation - the Noctua is already on my list.

I also just couldn't resist ...

It also seems like HP equipment seems to favor Papst fans on their test equipment - perhaps for  a reason other than air flow.

Or maybe the person doing the mechanical/thermal design at HP was just a Papst fan.

[texaspyro]

Screw trying to compare data sheet specs...  the safest way to do a fan change is to put a temperature probe (or probes) in the box and see what the temperature is (with the case on).  Then swap the fan and repeat the test.  If the new temps are close to the old ones, you are good to go.   Of course, this does not work if the old fan is dead.

I put a Noctua fan in my HP 53132A freq counter.   It is damn silent.  Almost too silent.  I worry that if it fails, I'll never know it until the counter frys.  The power supplies in the HP counters are always on and are known to die if the fan craps out. 

The datasheets show the Noctua fan probably has around 2/3 the airflow of the original fan, but the final temperature in the unit was pretty much the same.  You could hear the old fan three counties away.  You can only hear the new fan if you are a couple feet from to the box and nothing else if running.  Also remember that most test equipment is spec'd to run in an industrial environment and most people will be using it these days in a comfy home/office environment.

[Ero-Shan]

Also remember that most test equipment is spec'd to run in an industrial environment and most people will be using it these days in a comfy home/office environment.

That was what I wanted to answer. If the device is spec'd for 40°C, and you run it below 30°C, that's a neat safety margin. I am definitely not afraid of overheating. Much more of becoming deaf.

My major reason for posting, however, was my love of (awful) puns. 

[B1nary]

+1 for adding a Noctua Fan. I got so annoyed at the fan noise on my HP 53131 counter I added a 40mm Noctua - they come in 5V and 12V so be sure to order the right one.  I measure the internal temp with the old and new fan. It made less than 1degree difference and the Noctua was the cooler option. They advertise a ultra quiet option (a 100ohm resistor in series with the power supply). I left the resistor in and my counter has been sitting on the bench (as noted by a previous poster - so d*m quiet you have to put your ear next to the counter to hear it) for more than 1 year now. No regrets.

[Armadillo]

The instrument should best be doing some intensive maths, measurement or calculation or try during the summer time to be somewhat indicative judgement.
If in air-conditioned environment, then mostly no worries.     

[TheSteve]

A week ago I swapped out the Papst 622L for the Noctua NF-A6x25 in both of my 3458A's. The noise of the Papst fans was just too annoying in my house. They had a bit of a "whir" sound to them that I could hear a long distance away. The Papst fans are only a year old or so and are the proper spec'd fan however I am pleased with the Noctua fans so far. They have only a nice "shhh" sound to them. The 3458A's internal temp sensor does report temps 3 to 3.5 degrees C warmer with the Noctua fans so they clearly move less air. During a bench test they also drew more power then the Papst fans. My ambient temp is usually around 20 degree C so temp isn't an issue at all, if you are in a hotter area you might want to stick with the Papst fans.