I need to quiet "The Beast" AC fan control circuit options?

[almstsobur]

Nice rack off to the side too, been considering getting one since some of the old test gear is over 50cm deep and weighs over 15kg a piece (And i just got a deal on an old 50kg HP spectrum analyzer). I designed the equipment shelf on my bench for more modern 40cm deep gear nor did i specifically design it to hold over 100kg.

Yeah, the rack is very handy for the big stuff and less commonly used things.
Obviously I'm not sure what your lab/shop configuration looks like, but if you decide to do a rack and your floors will support it; I highly recommend doing a setup that will support casters/wheels. My rack has big casters and the floor is hardwood and even though I have a couple hundred pounds in the rack I can literally move it with 2 fingers. I have a rack mounted 12 plug power strip at the top and everything is wired within the rack, so it's just 1 power cable to the strip outside of the rack, makes it very convenient to move it where you need it to see screens or use a particular piece of gear without long cables all over, or just roll it off into the corner for more room when your not using that gear. Very slick and handy when you can easily move the rack.

[David Hess]

How about a suitably rated diode in series?
The fans will run quieter, and a tiny bit slower on the half-wave forward cycles.

We used to do this with desktop equipment with small high rpm fans that were too noisy - reducing the motors kickback on the negative cycles.

Not sure of the long term effect on the fan, but we never lost any.

Using a series diode or several will work for a low voltage DC brushless motor but not for an AC shaded pole motor.  Destruction of the AC motor winding is likely do to heating when the leakage inductance saturates.

[Berni]

Using a series diode or several will work for a low voltage DC brushless motor but not for an AC shaded pole motor.  Destruction of the AC motor winding is likely do to heating when the leakage inductance saturates.

Yeah this is like putting a diode in series with a transformer primary, The large DC component would likely saturate its core, at that point the only thing holding back the current flow is the winding resistance. So lots of current flows and it overheats.

This is a valid way of doing it in brushed AC motors. Most hair driers use a diode to get the slow speed setting on them. As you probably noticed hair driers are really loud as these types of motors tend to not run very quietly, also the brushes wear down but a hair driver is never run for very long. This is why fans tend to all be DC brushless motors or AC asynchronous (Offten the simpler inefficient shaded pole version).

[SL4P]

I take the learned advice with sincerity, and my experience was many years ago (when I was a nipper), but I recall the units/fans we did that to - were brushless AC in desk mounted consoles that ran 24/7 without any issues.
I was wondering about magnetic effects, but don't recall any negative symptoms.
I can only guess the fans were well over-specced in that application (and may have been 2x110v fans series - in a 220v device..
A long time ago, but it certainly worked!
I appreciate the technical criticism of the suggestion.

[nikodean1]

Fan technology has improved considerably over the years. Maybe you can find an equivalent pair of new 120 volt fans which are much quieter, and just swap them out without speed control.

I'm not sure what the CFM requirements for yours is, but I came across this 110 CFM, 120mm 115 volt model: https://www.amazon.com/AC-Infinity-AXIAL-Muffin-Cooling/dp/B009OWRMZ6/ref=sr_1_1?ie=UTF8&qid=1466559258&sr=8-1&keywords=AC+Infinity+120+mm+fan

The noise output is 47 dBA, which shouldn't be too loud. What size are your fans?

[Berni]

Its possible that some fans are so overbuilt that they don't saturate from DC that easily.

Once you get in to these high airflow fans the actual fan is only part of the story of making it quiet. These powerful fans create so much airflow that pockets of turbulence can form along the flow inside the case, causing extra noise. As i mentioned earlier i made the fan in a HP rack mount PSU way quieter by simply flipping it around so that it blows out of the case rather than suck in to it. That was mostly due to a heatsink being pretty close(cm or two) to the fan, this caused the blades to "wosh" past it and make a howling noise in sync with its rpm.

In another rack PSU i found a unused header on the CPU board with 12V on it so i simply replaced the 110V fan with a computer fan and plugged it in to there.

[almstsobur]

So far I'm thrilled with the simple series/parallel switch. I don't think I personally need to do anything else to the beast, though the additional input might help a future search'er. Last night a was able to load test in complete silence about ten 12v wall-worts to find one that truly met 12v at 2amps. 8/10 did not, 2 did and one is being used now to power my new self designed desoldering pump setup. I'm just happy to load test in peace 😉

[edpalmer42]

The only improvement I can think of for your series/parallel arrangement would be to replace the switch with a relay.  Power the relay through a normally open thermal switch bolted to one of the big heatsinks.  When the heatsink gets hot enough (maybe 60C?) to trip the thermostat, the relay operates and changes the fans from series (slow and quiet) to parallel (fast and noisy - but lots of  cooling). 

[almstsobur]

The only improvement I can think of for your series/parallel arrangement would be to replace the switch with a relay.  Power the relay through a normally open thermal switch bolted to one of the big heatsinks.  When the heatsink gets hot enough (maybe 60C?) to trip the thermostat, the relay operates and changes the fans from series (slow and quiet) to parallel (fast and noisy - but lots of  cooling).

Damnit Ed! I like that  I like it a lot. Now your going to get me to crack the beast open again.

[Berni]

Damnit Ed! I like that  I like it a lot. Now your going to get me to crack the beast open again.

Well using a switch does have an advantage in one area. This sort of automatic speed control will in most cases cycle on and off all the time. When the unit gets hot the fan kicks on and cools it down so its not hot anymore so the fans goes back to slow mode again, but then the heatsink slowly becomes hot again so it kicks on...etc This tends to be even more annoying to listen to than having it constantly be loud.

I guess you could fix that by adding a relay wired to self latch itself across the thermostat switch (or use a relay with a 3rd contact in it) so that when it kicks on it stays on until the instrument is power cycled when cold.

[edpalmer42]

Yes, cycling is a possibility.

But due to the mass of heatsinks involved and the built-in hysteresis of the thermostat, it might not be a problem.  The only way to find out would be to try it.  Attach a temperature sensor to one of the heatsinks and manually flip the switch when the heatsink reaches e.g. 60C.  How long does it take to drop by 10C?  Then look for a thermal switch with 10C hysteresis or more.  Extra points if you can log the data and show us a graph. 

If it looks like it's going to be a problem, you'd have to replace the thermal switch with a smarter circuit - or a switch! 

[edpalmer42]

On second thought, the first test you should do is to measure the heatsink temperature with the fans in series.  How much power do you have to dissipate to get them "too hot"?  If you'll never reach that point, you could consider the automatic switch from series to parallel as more of a safety feature like a flashing warning light instead of an operational feature.

[nikodean1]

The only improvement I can think of for your series/parallel arrangement would be to replace the switch with a relay.  Power the relay through a normally open thermal switch bolted to one of the big heatsinks.  When the heatsink gets hot enough (maybe 60C?) to trip the thermostat, the relay operates and changes the fans from series (slow and quiet) to parallel (fast and noisy - but lots of  cooling).

Damnit Ed! I like that  I like it a lot. Now your going to get me to crack the beast open again.

I like this idea!