3458A replacement fan

[IanJ]

Hi all,

The fan inside both my 3458A & 3245A are working well.......but I'd like something a bit quieter as I film my videos nearby and am getting about 45dbA background noise which I'd like to eleminate as much as possible.

There are a few suggestions buried in the forum but just wondering whats the accepted best without losing any performance.

Ian.

[Dr. Frank]

Buried in my memories (of this forum and elsewhere) is, that there are no proper, more silent alternative fans available.

Whenever fighting for sub-ppm stability of DCV and OHM measurements, the thermal stability of the environment, and synchronized the internal TEMP? of the 3458A are already the limiting factor. Even those 0.2 .. 0.5°C changes, when manual-operating in front of the instruments pile create a shift of several tenth ppm.

In my use-case, I would not mess with the exchange of the fan at all.
Frank

[IanJ]

Thanks Frank......I figured as much.

Luckily, Davinci Resolve has a "voice isolation" filter which works very, very well........but sometimes I think my videos miss the odd 'clunk' of a tool on the workbench.....:-)

I have thought abour relocating the 3458A/3245A to the other side of the workshop.....but I'd miss them being so close to hand....:-)

Ian.

[Martian Tech]

You might want to check out iZotope RX 10.  It has a Voice De-noise feature that shouldn't eliminate odd clunks.  You can download a free trial version to see if it works for you...

[TheSteve]

The Noctua NF-A6x25 will reduce the noise quite a bit, but also raise the internal temps 3 degrees C or so.

[mendip_discovery]

What would you say is the normal temp offset?

[MiDi]

Mine has around 15K dT with cleaned filter, after 3 month it rises to ~18K.
Tried the Noctua and it is audible quieter, but ~3K higher than stock fan.
Maybe an outside fan duct with 80mm fan is quieter while maintaining dT.
Otherwise 3K is not too much to worry about...

[mendip_discovery]

Mine has around 15K dT with cleaned filter, after 3 month it rises to ~18K.
Tried the Noctua and it is audible quieter, but ~3K higher than stock fan.
Maybe an outside fan duct with 80mm fan is quieter while maintaining dT.
Otherwise 3K is not too much to worry about...

Well, I just had a look at the data I have and the average is 11.8°C between lab temp and TEMP?. Max was 12.6°C Min was 11.1°C. I am running essele's old 3458A so whatever fan he fitted seems to be doing a fine job.

[Dr. Frank]

The dT depends on the location of the instrument; inside a closed rack it's of course much higher, I think I have seen +15..20°C.
My 3458A sits free on top of another instrument of that size, both on a desk, so that air can flow freely from rear, bottom, top and from the sides.
If the fan filter (old tube style) is clean, it's as well around dT ~ 12°C, dirty maybe up to 14°C.
My usual room temperature is between 21°C and 23.5°C at most (during those few, very hot days approaching 40°C outside), plus maybe 0.5°C when I operate in front of the instrument. Therefore, max TEMP? is about 37°C.
That's important for me, as I reduced the oven temperature to about 65..70°C, instead of the default 90..95°C.
Secondly, I assume that this lower internal temperature of the whole instrument lets the fast comparators live longer.
Frank

[aronake]

I have been experimenting with temperature controlled fan in one of my 3458a with quite some success. I can remove around 70% ambient temperature change in the 3458a when the controller is set to operate at most optimal range.

One of these controllers:
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.78ef2e8d8IeuHf&id=688785421807&_u=62lgsmegbbc3

And one of these fans:
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.78ef2e8d8IeuHf&id=41203557490&_u=62lgsmeg650f

You can find these from other sources than taobao.

Then I have some paramters I use depending on use case.

One brute force no matter what sound level bring the 3458a to internal calibration temperature level. And one for more cassual usage where sound level is more of a concern.

At full speed this fan blows significantly more air then the original Papst, so can bring the 3458a to calibration temperatures with a much higher ambient temperature.

I live in the tropics and not too keen on running AC 34/7 to bring my laboratory down to reference 21 degrees, so this has been very helpful. But ought to be helpfull for people living in less hot places too. For a normal north European home the temperature controlled fan ought to be able to bring the 3458a very close and stable to calibration temperature any time of the year, except for very hot summer days. 

[ap]

Keep in mind changing the fan and hence the air flow also has an impact potentially on calibration/MU. I would only change against standard fans as several have pointed out. Also, the official replacement fans (at least those examples I have tried are not quieter, it was rather the opposite). So when the original fan is still working in its normal loudness, dont mess with it. Also, have very rarely seen a bad fan on these.

[aronake]

Here is another still untested idea I got.

https://www.ebay.com/itm/284729465401

Replace the original fan with a 120mm fan and could then get same airflow, bring internal temperature to calibration temperature and the meter would be pretty much completely silent.

Even better then if also adding temperature control and the 3458a may keep very close to calibration temperature in a quite wide ambient temperature range.

[Kleinstein]

A variable speed fan is a 2 sided thing. It regulates the temperature at the sensor, but variable air flow means that temperature gradients can change. Probably overkill, but it could help for the initial warm up and later with regulation of the flow rate to compensate for filter clogging.

[aronake]

A variable speed fan is a 2 sided thing. It regulates the temperature at the sensor, but variable air flow means that temperature gradients can change. Probably overkill, but it could help for the initial warm up and later with regulation of the flow rate to compensate for filter clogging.

Good point on filter clogging.

My thinking on gradient:  Obviously best is to have a proper lab temperature, fan would then have same flow as the original fan and the gradient would be the same. At lower and higher temperatures there will be some gradient difference, but I find it verily likely that keeping the sensor temperature, located at the most temperature critical point in the 3458a (where could that be? the LTZ1000 is my assumption) closer to calibration temperature (at the sensor point) is more important than some slight gradient change.

[IanJ]

Am reading elsewhere that the Noctua NF-A6X25 FLX is a fan that others have used to replace the existing Pabst fan.
It's 12V wheras the Pabst is 15V I believe.
Might be worth acquiring one to test (outside of the 3458A):

- Adding a series 2.7V zener (not happy and long term driving a 12V fan with 15V).
- Measuring the airflow rate (I have a meter).
- Measuring the noise level (dBA) (I have a meter).

Ian.

[TheSteve]

The Noctua fans should include "speed/noise" reducers(they are diodes or resistors, never checked) with it - you can give them a try. Mine have lasted several years so far, no issues at all with the 15 volt drive.

I replaced the stock fans as they barely turned. Before the Noctua I used the proper Papst fans, they were so loud I could hear them throughout the entire house. So they had to go. However I have another 3458A with an original Papst fan and it isn't too loud at all.

[aronake]

Am reading elsewhere that the Noctua NF-A6X25 FLX is a fan that others have used to replace the existing Pabst fan.
It's 12V wheras the Pabst is 15V I believe.
Might be worth acquiring one to test (outside of the 3458A):

- Adding a series 2.7V zener (not happy and long term driving a 12V fan with 15V).
- Measuring the airflow rate (I have a meter).
- Measuring the noise level (dBA) (I have a meter).

Ian.

I have one of my 3458a stripped down in front of me, just checked and can confirm that the fan connector on the 3458a outputs 15V. Just backed up the calram with your great software for Dallas replacement. Thanks for that.

The Papst is rated for 12V nominal, but they also mentions a operational voltage range of 8 to 15V.

https://www.mouser.com/datasheet/2/120/620-276481.pdf

Fans tend to be quite OK with some ranges of voltage. I would not have any concerns run the noctura at 15V. And even at 15V it should be quite quiet.

Probably also good to measure dT vs papst with the noctua both at 15V and 12V and include in your decision process.

Or keep the Papst, add a switch hanging out on the back of the 3458a where you can introduce a zener to make it more quiet during your recordings and then go full speed when you do serious measurements.

You may know all these things already, in case not,,hope something helps.

[IanJ]

Hi all,

Here's my testing on the Noctua NF-A6x25 FLX versus the Papst 812L (from my No.2 3245A).
I held each of the fans up against my 3D printed jig so that all airflow is directed through the anenometer.

NOCTUA:
12V = 2.9 m/s blowing through jig
15V = 3.8 m/s blowing through jig
15V = 2.9 m/s suction through jig
61dBA (@ 15V) which includes background noise.

PAPST:
12V = 1.5 m/s blowing through jig
15V = 1.8 m/s blowing through jig
15V = 2.9 m/s suction through jig
64dBA (@ 15V) which includes background noise.

There must be an airflow/obstruction/resistance thing going on, because the Papst seems almost the same if not slightly more powerful on an in-the- face test, but the blowing-through tests don't seem to back that up......but the suction tests do!

3dBA difference so exactly half the volume for the Noctua (for the human ear). Value measured at 1" distance from side of fan (thus not affected by airflow)

The Papst fan seems to have a lot more torque/weight/inertia, I guess thats why we hear that 3458A whine on power up as it gets up to speed.

Ian.

[IanJ]

I wonder if I should increase the distance to the anenometer......perhaps a 6" length pipe between the two?

Ian.

[Kleinstein]

The rotation in the air flow can effect the anenometer. Some pipe with fins to slow down the rotation may be a good idea. More air resistance also would bring the working point closer to the use an the instrument.

[IanJ]

Tries same with an 8" length pipe between the anenemoter and DUT.

NOCTUA:
12V = 2.6 m/s blowing through jig
15V = 3.1 m/s blowing through jig
15V = 2.7 m/s suction through jig

PAPST:
12V = 1.5 m/s blowing through jig
15V = 1.7 m/s blowing through jig
15V = 2.9 m/s suction through jig

Not much difference.

[IanJ]

The rotation in the air flow can effect the anenometer. Some pipe with fins to slow down the rotation may be a good idea. More air resistance also would bring the working point closer to the use an the instrument.

And I guess adding the filter on the back of the fan also......to properly simulate working conditions.

Will try both these soon.

Ian.

[IanJ]

With 8" pipe, air rotation eliminator and 3458A filter fitted. Blowing through.

NOCTUA:
15V = 1.6 m/s

PAPST:
15V = 2.2 m/s

So, it has somewhat reversed the results.....but a bit better aligned now.

I think fitting to my 3245A will be the next test. I have two 3245A's so could run them both side by side and measuring the internal temps at multiple spots.

Ian.

[IanJ]

Full Results:

Noctua NF-A6X25 FLX versus Papst 812L 60x25mm Fans




So the Noctua drops quite a bit with the pipe, air-straightener and filter fitted which I guess is because the air rotation has been negated which was contributing to the air speed previously.
The Papst on the other hand, which had the air rotation going in the opposite direction and was previously slowing the measured air speed, is now raised.

I think 1.6 m/s versus 2.2 m/s is the true comparison of the two fans, both at 15Vdc.......within the confines of my experiments anyway!
And therein lies the question......by using the Noctua, is 1.6 m/s enough for the 3458A.......?

Ian.

[MiDi]

As the original Papst fan in my unit had quite some vibration due to imbalance it was replaced with Sanyo Denki San Ace 60 9R 109R0624D4011 60x60x25mm 24V 2.88W.
It gives same dT and is not significantly quieter, but w/o the annoying metal scratching sound (old HDD type).
See 3458A Worklog

[alligatorblues]

The largest fan makers are Nidec  and NMB-MAT. The way I match up fans is by testing. Chassis cooling fans many times are VDC input, but they have inverters, so actually run on AC. This makes it tricky to measure current consumption. I use a Fluke 87-V, which does so accurately.

Once I have my current figure, I know how high I can go. So, then I just look at Nidec and NMB-MAT, filter on dB, and look for the lowest I think will do the job. Noctua are quiet, but limited selection, and not the highest quality. Nidec and NMB-MAT are made to last indefinitely.

I have some NMB-MAT 12V 80mm that have been going continuously since the 90s. But the sheer number of fan models is daunting.