Replace fan on 53131A counter with quieter version?

[motocoder]

engiadina -

I forgot to say thanks earlier. I appreciate you taking the time to look up the fan model number.

Thanks,
matt

[G0HZU]

I assume the effect of any unwelcome thermal gradients will be more noticeable with the cheaper timebase options

I only can tell from my experience. When I had one counter running without case for a while you could even see the effect of just blowing against the OCXO. So I decided to run my own experiments and packed the OCXO into some thick foam. That stabilized the counter noticeable.

So when the Power Supply gets warm, why to distribute that into the case? The OCXO is heated anyway. I think gradients are less problematic than changes in temperature.

And at least, the new fan has identical parameters compared to the old one, except the noise.

The sentence in the manual would not worry me too much. Sometimes manuals contain stuff because of "... oh, we screwed that part. Let's state in the manual, it's serving a special purpose so users don't complain ..." 

I don't own one of these counters but looking at the images of the chassis on google the vent hole patterns in the case suggest that a fan system would pull in cool air up through the unit so it should vent the warmed air out of the back. But that's just a guess based on how I would expect Agilent to do it.

So maybe your fan is fitted backwards?

[engiadina]

No, my fan is in the right direction. But you made a good point. Of course the fan is sucking hot air out of the device.

Anyway, when you see how the OCXO is mounted, sideways at the frame, directly behind the front panel, I do not know how much the little fan really influences that.

but if the replacememt fan has the same params, why bother. Should make no difference but less noise.

[G0HZU]

Can we now agree that the fan isn't just there for the PSU? If not then read below:

Agilent say the fan helps with timebase stability and this makes sense when you look at the 30W power consumption. That's a lot of heat to get out of the box if you want to maintain a steady internal temperature when you leave standby mode.

Also the (minimal) vent hole pattern indicates a system that requires a fan to suck cool air through the whole counter. Look at the vents in the enclosure. There aren't many exit vents apart from the entry vents underneath (where the fan draws in cool air). Without the fan the enclosure would get hotter once there was 30W of active high speed silicon.

It isn't a case of the OCXO being directly in line with the airflow. It's the fact that the fan unit has to suck the waste heat energy out of the enclosure to minimise the overall internal temperature from slowly rising over time after the counter goes from standby to normal operation. An overall change in internal temperature will not be a good thing for timebase stability.

By contrast, I have a slightly older counter from Anritsu that is also a high performance reciprocal/vernier counter and it consumes about 22-24W in normal mode and maybe 4W in standby. So it isn't much different but it doesn't have a fan but instead it has a LOT of holes in the flat top of the top cover and in the underside and the flanks to try and get this unwanted heat vertically out of the box because hot air rises vertically. It has a specific pattern of holes in various places to try and govern the airflow up through the unit.

The 53131A has no holes in the flat top of the top cover and so it needs the fan to suck this heat out of the unit sideways via a defined route from the cool air intake on the underside and out through the fan. Otherwise the OCXO/timebase would be in a box that slowly got hotter and hotter once you took it out of standby.

Without the fan, I think there aren't enough vent holes (in the right places) to get this heat out of the box without having an undesirable rise in internal temperature.

[motocoder]

The fan on my 53131A pulls air through the case and blows it out the back.

I used the Sunon whose part number I posted earlier. It is nearly silent - definitely quieter than the PCs in the room, and no longer making those "I am going to fail soon" noises.

Based on the permanent marker marks and flux residue on the PCB, it looks like this fan was already replaced once. What was installed is a Delta DFB0412M, 12V 0.08A. This is a little higher CFM than what I replaced it with, but based on the comments from engiadina, I think it's comparable to the OEM model.

[billy_gr]

For reference (and providing some hits for google) replaced a FONSAN DFB0412M with a MF40201V2-1000U-A99 21dBA

Thank you !!!