3458A replacement fan

[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.

[dlebed]

I wanted to add my two cents here, maybe that will be helpful for someone.

I've overhauled a few 3458a recently and one of the problems was the stock fan noise was too loud.
I tried some new Papst fans as well, but they're still a bit noisy.

I finally settled with Sunon HA60251V4-1000U-A99 (aka CUI CFM-6025V-125-107-20): https://www.digikey.com/en/products/detail/sunon-fans/HA60251V4-1000U-A99/6198731
Essentially the same fan from CUI https://www.digikey.com/en/products/detail/same-sky-formerly-cui-devices/CFM-6025V-125-107-20/7605547

It's pretty inexpensive and very quiet, only 10.7dBA at 12V.
With 68 Ohm resistor soldered in series, it gets about 13.5V (it is rated for 13.8V max) when powered from 3458a's 15V fan connector supply.

Temperature difference with the stock one is about 1C, which is negligible to me.
It also could be tweaked by changing the series resistor value.

I don't see a real reason to go with originalistic approach here and only use the same type of Papst fan.
Clearly, in 80's when this DMM was designed, there were no sophisticated thermal design or simulation in place, add a "dirty filter" effect to that, that itself could raise the temperature by 3-5 degrees and I don't see how 1-2C difference from a different fan could be a real problem.

As long as you keep you filter clean and the fan you use provides temperature difference within 1-2C from the original one that seems to be good enough to me, unless you're running it in metrology-grade temperature controlled environment with no dust in the air.