Smart-UPS 2200 - Intolerable Fan Noise - Fan Mod Required

[Anonymous user]

I installed two SMT2200 tower units in a quiet office environment only to discover that the SMT2200's cooling fan is 3100 RPM, always on, and extremely loud and intolerable, like a hair dryer.  All the office PCs have multiple fans in each case, all engineered for quiet operation, using temperature to control fan RPM, so all PCs are extremely quiet, and cool.  Incredible that the Smart-UPS tower units, at this price point, were not engineered for quiet operation.  To keep the SMT2200 units, I need to modify the fans.

The Smart-UPS stock fans are Y.S. Tech (Yensun Technology Corp.) FD249225HB fans, and each UPS has one 140mm unfiltered intake vent (no fan) on the side and one 92mm fan exhausting to the rear.  The fan specs are - 24 volt, 0.16 amp, 92mm x 92mm x 25mm, 3100 RPM, 58.6 CFM, 5.2 mm-H2O (.205") max. static pressure, with a noise level spec of 40 dB(A) out of the case.  In the case, the fan noise is louder due to the airflow through the old-style exhaust vent that is stamp-cut in the rear case panel (a wire cage fan guard or honeycomb would have been a better solution, increasing air-flow and reducing noise).  The fan is two-wire (red/black) with a female 3-pin connector (black/null/red) connecting to a 3-pin fan header (J13) on the SMT2200 circuit board.

So what is the best modification option?

1.  Replace the stock fan with a much lower RPM fan (24 volt) that is engineered to be quiet?
One possibility:  Sunon ME92252V3-000U-A99  (24v, .056 amp, 2400 RPM, 39.5 CFM, 28 dB(A))

2.  Replace the stock fan with a quiet 12 volt fan, using an integrated circuit linear voltage regulator (e.g. LM7812) to step down the supply voltage from 24v to 12v?  Using 12v, the fan options then include the quietest PC case fans available from Noctua, SilenX, NoiseBlocker, Phanteks, Nexus, etc.

3.  Install a fan control circuit with thermistor so that the Smart-UPS stock fan RPM is variable with UPS internal air temperature?
I found a couple 24v controller circuit boards on eBay here.  And schematics are available online as well to DIY - example (example for 12v, not 24v supply)

4.  Using a 140mm intake fan and a 92mm exhaust fan, install two 12v fans in series (24v total) with zener diodes and capacitors?
Reference - http://forum.allaboutcircuits.com/threads/24v-12v-with-load-voltage-dividing.62024/
Reference - http://electronics.stackexchange.com/questions/178607/12v-fans-in-series-with-25-a-supply

5.  The UPS circuit board has a 3-pin header for the fan connection, so if the stock 2-wire fan (24v) is replaced with a 3-wire fan (24v), does the UPS circuitry already have a thermistor and ability to control fan RPM by voltage relative to internal temperature?  (And if not, why not??)

6.  Other?

Suggestions?  Has anyone satisfactorily fixed the noise problem with any of these fan modifications?  Something has to be done right away to quiet these UPS fans.  

Thanks in advance for any help.

[Terry_Kennedy_apc]

I'm just another user (I don't work for APC), but if these are newly purchased units I would suggest returning them for alternative models rather than modifying them and voiding the warranty.

Regarding your question about the 3rd pin on the fan header, I would guess that it won't do anything useful for you. If it is for a sensor from the fan (normally a yellow wire), then the UPS firmware is ignoring whatever happens on that pin (since the factory fan doesn't provide that signal). If it is a speed control signal to the fan (normally a blue wire) then the firmware probably doesn't generate it for the same reason (and related components on the UPS circuit board(s) may also not be installed).

If you are going to change the fan anyway, I'd suggest something like Nidec Option -10 or -60 where the fan handles all of its speed control itself via a thermistor.

[BillP]

I fitted #1, the fan should be running around half speed unless charging or on battery but yes it is way louder than is acceptable for a home office, or even a work office for that matter.

I doubt it is programmed to do the whole self adjust routine depending on the RPM signal received, it would seem it throws one voltage for idle and another when under load/charging etc.

I still found it noisy with the new fan so I fitted an old rheostat from a very old  heatsink I had in my box of bits so it idles around 900  ish RPM, I spent a lot of money making my games pc near silent so the UPS remains the loudest thing in the room now along with my microserver.

I didn't care for the 1st four  years of ownership as I had it fitted in my loft but child #2 saw me moving into the dining room so the quest for quiet was on.

Before you go for it check for firmware downloads as it may enable 2 speed fan control if it doesn't already have it. http://www.schneider-electric.us/sites/us/en/support/faq/faq_main.page?page=content&country=US〈=en&i...

that may be enough but chances are it's already set to 2 stage fan control?

I did this about a year ago so I don't have all the details, obvious things are APC staff will warn you off for H&S reasons, warranty void etc. I managed not to blow myself up but I am sure you know about discharging everything after shutting off and removing the battery packs. I don't think the pins are the same as a normal fan but you can reuse the plug from the original fan as they are push down wire crimps like network/phone sockets. Take the case off to see what's what and you can remove the back panel where the fan mounts to get round the lack of clearance. 

Oh and I  used rubber fan mounts   http://www.quietpc.com/na-sav2?gclid=CjwKEAiAvauyBRDwuYf3qNyXmW4SJACX9-fXn93s62rrorzqiH6B8PB5roV7GeE...

[BillP]

In response to Terry, These units will generate a lot of heat very fast during discharge so you may damage the unit using a thermistor setup as it may not get up to speed fast enough.

As I was well out of warranty and ready to retire it I decided that this UPS like rack mounts and all the ones shoved into a tiny electric cupboard with an ambient of 40 degrees (changed battery packs in enough of those in my time) was the reason why APC decided to run the fans so fast instead of a sensible option of temperature .

Hope that helps.

[Anonymous user]

Before buying the SMT2200 tower units, I checked all similar APC and competitor models, and the SMT2200 tower is the best solution for my purposes, so I really don't want to return them.  I emailed APC support about any possible existing fan control via the 3rd pin on the 3-pin fan header, but there's no reply yet.  Maybe APC already has a solution on the shelf using a 3-pin fan, and the 3rd pin, and a firmware upgrade, and it all can be done at an authorized service center - who knows.  Unfortunately, unless there is an immediate authorized fix offered, I have to fix the noise problem, warranty or no warranty.

Thanks very much for that Nidec fan link.  I am now looking at that (24v fan with thermistor and thermal speed control circuit mounted on the fan housing) as the first option to try.  I don't know about Nidec's sleeve bearings - usually sleeve bearings get noisy faster than any other type.  Will call Nidec tomorrow.

[Anonymous user]

According to the product documentation, the SMT2200 fan is "variable speed", but per actual operation, that just means it has two fixed speeds - 100% and something lower, and the lower speed is not quiet.  At one or the other speed, the fan is always on.  These are brand new units out of recent stock, and by their two-speed operation I'd guess these have the updated firmware, although I have not checked yet.  

How does that manual rheostat control work when the fan header voltage switches high/low?  Similarly, if I install a thermistor controlled fan such as that Nidec suggested by Terry (maybe with the thermistor located closer to the batteries), I'm curious how it will play with those 2-speed voltage changes.

Good call on those anti-vibration fan mounts - recommended.

[BillP · Accepted solution]

When I fitted my new fan on the mounts it was still louder  than I wanted, all I did was fit it inline  by putting one end on the plug and soldering the other to the fan wire, when it ran on low speed I adjusted it down to the point where it was only just running as that still provided decent airflow (it could hold a sheet of paper on the inlet)

When it ramps the fan up under load it probably runs about 2000rpm, it provides a lot of airflow and the discharge test I did saw the internal temp sensor hit about 30 degrees.

Then it was time for a beer as it was a job well done 🙂

In a normal office the ambient temps will never realistically exceed 22 degrees, so I doubt my mod decreased the expected battery and UPS life.

As for the firmware I have learnt over the years that you should always check as they could be old stock sat on a shelf in the main APC  warehouse that has a recent set of batteries put in prior to shipping.

Or worst case you find the battery printout shows the batteries are old stock, then you need to kick off with the supplier 🙂