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

[almstsobur]

"The Beast" is 1979 EUL-600 Fujitsu-Denso programable load rated for up to 120 amps and weighs in at about 60Lbs. It has 2 large AC 120v fans that draw air across the 6 full length heatsinks. My problem is the noise of those fans, they are full speed full time. The dogs start howling,  neighbors come knocking and the woman demotes me to the sofa when I need to use "The Beast" . 98% of the time this thing is overkill for my needs, it's rare I need to test more than  about a 6amp load. So I would love to be able to regulate the fan speeds to something reasonable under my typically use scenario. Ideally I would like to screw a thermal probe into one of the heatsinks and have the fan speeds fluctuate by temperature. I've scoured the web and literally found hundreds of write-ups, designs and circuits for DC fans, but these are 120v AC fans and I am not having any luck finding anything that will work for this scenario. Any input or a point in the right direction is appreciated.

Thank You,
-RP

[Messtechniker]

How about first replacing the fans with two 12 V DC types of the same size (and air flow rating).
And then get an off-the-shelf fan controller board. Should not be too expensive.

Yours - Messtechniker

[ZeTeX]

Get good quality PC fans and a small 12V supply, Your would not be able to even hear the fans from over 1 meter.
I would recommend "Arctic" fans, they are cheap and very quiet.

[David Hess]

Speed control of AC fans can be done by switching in different values of capacitance in series; film capacitors up to about 10 microfarads would be typical for a standard size tubeaxial fan.  We used to do this where commutation noise from a DC brushless fan or electronic speed control was unacceptable.  A variac can also be used with an AC fan but that is probably not what you want.

Maybe make a thermal circuit which switches a relay to select whether the capacitor is in series with the fan or not.  It would only be two speeds but still a major improvement if the load is idle or at low power most of the time.

[NiHaoMike]

Or since there are two fans, use a relay to select between series/parallel.

[JWoodrell]

here is a simple 3 component isolated ac switching circuit.  drive it with pwm and run the fans as fast or slow as you want. 

the VOM1271 MOSFET Driver is super easy to solder and provides isolation between mains and your microcontroller of choice, or even a 555 circuit.

pick the mosfets to have low gate capacitance, and a voltage rating to support mains.

[almstsobur]

JWoodrell, thank You. That certainly looks like it is worth a try.

I was considering the DC fan option, computer case style, ect. But you need to see the quality of these fans, it seemed like such a downgrade. They are pure aluminum, with aluminum housing and appear to be original. 37 years later they still run like a beast with no abnormal noise, vibration or other issues. You just don't see this kind of quality in a plastic DC fan. Also, you can see from the attached picture, they seem to be somewhat structural in nature. The fans are Prince Fan made by Tobishi.

[MagicSmoker]

For a quick-n-dirty hack requiring a minimum of effort you can use a fan speed control "dimmer" (intended for ceiling fans), usually located in the same area as incandescent lamp dimmers at home improvement stores. For example:

http://www.homedepot.com/p/Cooper-Wiring-Devices-5-Amp-Single-Pole-Variable-Fan-Speed-Control-White-TFS5-W-K/205618204

Don't use a regular incandescent dimmer because the significant inductance of a motor acts to prevent the triac inside the dimmer from ever turning off.

[almstsobur]

I ordered one of these off Ebay a couple of days ago. For $9 i figured I couldn't go too wrong http://www.ebay.com/itm/141409048579
It seems more or less like the dimmer principal, it hasn't arrived yet to test it. I know it does not act on a thermal response; which would be my ideal scenario if possible. But it might might keep me from sleeping on the sofa until I hopefully figure out a possible thermal controlled solution, maybe I will need to involve a Audrino, or something similar to do the thermal control aspect. I am completely new to this type of thing, my background is in repair as a hobby, not design.

[sarepairman2]

that thing looks like you can put some decent fans in it, not some cheap bullshit off ebay. Have a look at dB ratings.

[NiHaoMike]

Most likely a phase angle controller would make it noisier.

[Delta]

If they are "proper AC" motors (ie induction or synchronous motors) then you can't do any speed control without a Variable Frequency Drive...

[MagicSmoker]

I ordered one of these off Ebay a couple of days ago. For $9 i figured I couldn't go too wrong http://www.ebay.com/itm/141409048579
It seems more or less like the dimmer principal, it hasn't arrived yet to test it. ...

The main difference between one of these fan controls and a regular light dimmer is the addition of an RC snubber across the semiconductor switch (triac or anti-parallel SCRs) that temporarily diverts the lagging inductive current so said switch can turn off.

The ebay item you linked to is for 220V, and so the triggering voltage for the SCRs may be too high to give a useful adjustment range on 120V.

If they are "proper AC" motors (ie induction or synchronous motors) then you can't do any speed control without a Variable Frequency Drive...

Sure you can, under-volting an AC motor without changing the frequency essentially causes the flux to decrease and if that is insufficient flux to supply the torque demanded by the load then speed will decrease (ie - slip will increase). It's not terribly efficient or elegant, but it does work over a modest speed range with loads that have square or cubic law torque curves (ie - not constant torque or traction applications).

[almstsobur]

The main difference between one of these fan controls and a regular light dimmer is the addition of an RC snubber across the semiconductor switch (triac or anti-parallel SCRs) that temporarily diverts the lagging inductive current so said switch can turn off.

The ebay item you linked to is for 220V, and so the triggering voltage for the SCRs may be too high to give a useful adjustment range on 120V.

The Ebay controller says 220v in the title, but in the description it says it supports 120V?? roll of the dice I guess, i't only $9. So are you saying a light dimmer, or the Ebay gizmo is the better solution?(assuming it works at 120v)

Also, I happen to work at a place that has/uses a lot of VFD's. I might be able to repair or find a working one that's slated for scrap. But I know nothing about VFD's I think all of ours would be rather large, they control 480V motors, etc.

[Siwastaja]

I wouldn't use a standard phase-angle triac dimmer. When "dimmed", the overtones of the 50 or 60 Hz buzz are very audible in fan motors because of the sharp edge every mains period. The exhaust fan at my home has such a device and IMO it's less distracting at 100% than when slowed down.

A simple multitapped transformer (or autotransformer) is a rather good option if you have one.

I'd still consider replacing the fans because modern BLDC fans can be substantially quieter, even the fan blade design has improved in the recent decades.

[almstsobur]

I have a Variac that I use for testing. Could it be as simple as ....

Use the Variac directly on the fans, get 2 or 3 voltage points, say Low, Medium, High that are acceptable in sound and CFM, then purchase a multitap transformer that is roughly in those 3 selected voltages ranges and wire up a low, Med, High switch for the different taps? That's easy enough if that will work correctly. This thing is mounted in the bottom of my rack and it works flawlessly, it's not going anywhere, so I don't mind hardwiring a permanent solution. I can mount the 3 switches on the side of the rack in a neat little project box beside "The Beast" 

[Gyro]

I saw an project / application somewhere (might have been an EDN circuit idea) that achieved speed control by dropping one complete mains cycle in every so many (I think something like 1 in 3-5) to achieve speed control.

It was intended for desk / bedroom fans but might possibly be applicable to cabinet fans.

[Aodhan145]

What about some kind of triac or SSR phase control solution?

[MagicSmoker]

The Ebay controller says 220v in the title, but in the description it says it supports 120V?? roll of the dice I guess, i't only $9. So are you saying a light dimmer, or the Ebay gizmo is the better solution?(assuming it works at 120v)

I can't comment on the ebay gizmo as I have no experience with it, and I specifically said to NOT use a light dimmer, pointing out that a fan "dimmer" has a few extra components in it to help deal with inductive (motor) loads.

I wouldn't use a standard phase-angle triac dimmer. When "dimmed", the overtones of the 50 or 60 Hz buzz are very audible in fan motors because of the sharp edge every mains period. ...

It depends on the fan and how it is mounted, but, yes, phase angle control does tend to create some harsh sounds from most motors. I suggested this mainly because it is cheap and easy to try. Your suggestion to use an variac to find a voltage that gives a good compromise of airflow vs. noise will likely result in a superior solution, but is likely to be more difficult to implement (though, see below).

I have a Variac that I use for testing. Could it be as simple as ....

Use the Variac directly on the fans, get 2 or 3 voltage points, say Low, Medium, High that are acceptable in sound and CFM, then purchase a multitap transformer that is roughly in those 3 selected voltages ranges and wire up a low, Med, High switch for the different taps?

Yes, it is indeed as simple as that, but the chance of you finding a multitap transformer that fits the bill are somewhere between slim and none.

However, if you find that you only need to lower the line voltage somewhere between 6 and 36V you can wire up a standard voltage transformer to buck the line voltage down by the amount of its secondary (e.g. - a 24V transformer would subtract 24V from ~120V, giving you 96V). You do this by winding the secondary in opposition to the primary. Scroll to the bottom of this page for a good explanation: http://www.allaboutcircuits.com/textbook/alternating-current/chpt-9/winding-configurations/

[almstsobur]

Sounds good! Thank you everyone for the input. I think I will start with the Variac and see where I need to be voltage wise, figure out if a transformer is feasible. The load is weak, the fans are only 15 watts a piece. That may open up some more options for getting the correct one. That's where I'll start anyway.

[David Hess]

If they are "proper AC" motors (ie induction or synchronous motors) then you can't do any speed control without a Variable Frequency Drive...

They are not; they use shaded pole AC motors and changing the input voltage without changing the frequency is completely acceptable to control their speed.

https://en.wikipedia.org/wiki/Shaded-pole_motor

They may have multiple taps near one electrical end of the winding, which provides variable speed and power via selection of one tap at a time, as in ceiling fans. Moreover, they are compatible with TRIAC-based variable-speed controls, which often are used with fans.

We did not use a TRIAC based speed control because of the increased electrical noise.  The fans were part of an environmental chamber adjacent to a pair of 8.5 digit voltmeters.

[almstsobur]

Tested the fans out on the Variac alone after I got home from work. Seemed perfect, no odd noises, nothing unusual. I had full stop to max control with about 65v being a perfect "low" setting. I wonder if this means that the ebay gizmo might work out just fine, or if I should be hunting for a 60, 90, 120 transformer.

[NiHaoMike]

A relay to select between series/parallel should work quite nicely, then.

[Whales]

A relay to select between series/parallel should work quite nicely, then.

How would the load created by the fans react to being in series?  It's definitely worth trying, but make sure you check how they behave if you gently apply friction to one of the fans.

I know that for brushed motors you generally can't put them in series (unless mechanically linked).  If you physically stall one then it stays stalled after you let go, with most of the voltage drop appearing across the other instead.  Atleast IME with small ones.

[Siwastaja]

AC induction motor variant typically used in most fans have slip characteristics without the horrible kind of torque breakdown point like "proper" AC induction motors.

These always run at quite a lot of slip and poor efficiency, but are designed to do so. Hence, dropping the voltage works just fine; the efficiency will get even worse, but the motor efficiency is not usually important in small fans, because the delivered power (power to move the air) drops exponentially, too, when slowed down.

Any method of dropping the voltage is equally good in principle, as long as you are aware of the audible noise caused by triac phase control. Skipping full mains periods would work quite well most likely.

But if you mostly need it for low-dissipation tasks, and full power every now and then, do you really need a multi-tapped transformer? How about a simple 1:2 transformer that can be bypassed? That would give you two fan settings - quiet and loud - and it's easy to source.