dfars vs 24hr fars

[C5250]

This tech says the fan is just an off the shelf fan, nothing special, ...

I'm not so sure about that, while the motor may be an off the shelf part and the impeller is made by Advanced Air, I believe the housing is a part molded for Davis to their spec.

The one I took apart was the original for my circa 2006 VP2, so the current part could be something different. Does your motor look like this one:

[johnd]

It doesn't surprise me that you got 2 different responses from Davis T/S, the same thing has happened to me. They have so many new, perhaps not as capable techs as they used to. I do know, though, that they lost a couple of the older, better techs to the 'Great Beyond'.

Have to agree with this. I think that you have to remember that Davis frontline support deal with pretty much a continuous stream of calls/emails from users who (i) have a problem/question which is answered in the manuals but which they haven't taken the time to read in sufficient detail; (ii) have a software installation issue; or (iii) have a relatively familiar problem with an older system such as a failing temp/hum sensor or wind speed reading. So I suspect that the vast majority of their time and attention is taken up with very repetitive bread and butter issues. But typically frontline support won't always _know_ the answer to a really detailed specification question though with the confidence of youth they may well hazard an assured guess. Obviously there are people at Davis who know the answer to most questions (insofar as the answer is known at all, which I'm not 100% sure it is say around the fan motor) but getting in touch with them via frontline support can be tricky.

So I'm not sure we're really any further forward on the issue of whether operating the fan continuously at lower voltage (eg 1.5v) eg from the mains offers a significantly longer fan life while still giving adequate ventilation to the sensor. Common sense says to me that the answer is yes either because of lower bearing wear or because the lower airflow causes slower gumming up of the bearing. But maybe I'm wrong.

[dalecoy]

Data points:
1.  Computer fans of similar construction (sleeve or other bearings) run for years at much higher speeds.
2.  My old dfars fan has been running on 3V for roughly 6 years.

Suspicion: I suspect that the Davis dfars fan mechanism is actually a "commercial" design for 5-volt service, rather than a special 2.5-volt model.

But obviously, the way to prolong life of the fan is to use software (or an alarm module) to turn the fan off when the wind speed exceeds some threshold (like 10 kph), or even by also considering the solar radiation, time of day, etc.

 

[johnd]

Data points:
1.  Computer fans of similar construction (sleeve or other bearings) run for years at much higher speeds.

Agreed, but I think that the obvious problem with FARS/DFARS is that they run outdoors and often experience both condensation and particulate matter in the airflow. So I suspect that its the gumming up issue that dominates the FARS fan life.

Perhaps those in warmer, drier climates might experience much better fan life than those in cooler damp climates where condensation must surely encourage gumming up. Perhaps what's really needed is a fan type that can run with really sealed bearings or at least with a construction that's more resistant to gumming up. I'm sure these must exist but maybe Davis's cost-benefit decision was that it's cheaper to replace the fan at intervals. (Remember that the original Davis recommendation was that the FARS fan be replaced annually, though I see that has been subtly reworded now.)

[dalecoy]

One other factor, just to make things complicated.

Arguably (I'm not totally convinced), running the fan on 2.3 volts might slow down the progress of "gumming up" of the fan bearing.

However, the amount of "gumming up" required to prevent the fan running on 2.3 volts is less than that required to prevent it running on 3 volts.

In other words, 3 volts will run the fan with more "gum" than 2.3 volts.

I certainly don't know what the correct answer(s) are, and we can discuss the various factors for a long time, of course.  But I doubt that anybody is going to do an exhaustive controlled years-long test to find out.

But I'm really glad that jerryg is doing some performance tests and reporting the results.

[graculus]

Computer fans are brushless, which is hard to do with a very low power motor.

[johnd]

A small addendum to this that explains why I'm being a nuisance on this thread: Some users here in the UK running the FARS on mains/DC power (which is arguably the only real option here if you want good 24hr FARS data) are finding that they need to replace the motor after 15-18 months. So if it were reliably possible to get 2-3 years of fan life by running on eg 1.5-2.0v then that would be information of definite value.

[dalecoy]

A small addendum to this that explains why I'm being a nuisance on this thread: Some users here in the UK running the FARS on mains/DC power (which is arguably the only real option here if you want good 24hr FARS data) are finding that they need to replace the motor after 15-18 months. So if it were reliably possible to get 2-3 years of fan life by running on eg 1.5-2.0v then that would be information of definite value.

Yes, it certainly would.  Perhaps you would be in a good position to encourage Davis to undertake such a study or test?

I've done a bit of research, and the engineering subject of predicting "bearing life" is not an exact science - more of a black art, deep in personal opinion and proprietary claims.

[SLOweather]

Comments on a couple of posts...

It wouldn't surprise me at all if Davis has used fans with both types of bearings over the years, for engineering, cost, or availability reasons.

In my experience, "computer" fans used outdoors DO have a lesser life, probably for the reasons JohnD mentioned. I have a stash of new,removed from equipment, 1.5" 12vdc fans. I was using them in a FARS I built to put in the greenhouse for a set of Adicon Bobcat Temperature and Humidity modules.

In that service, I couldn't get a year out of them, due to the temperature swings, dust, humidity and condensation.

[jerryg]

I noticed on my fan that on the hub it is stamped with exhust. Must be an exhust fan in real life lol. Well i have been waiting for the sun to shine so i can do some comparisons, but, alas it will not shine lol. Yesterday it was cloudy with 6 to 8 mph winds and the standard shield tracked right along withthe dfars and 24 hr. When the winds laid in the evening the main difference was  the ss would lag behind the dfars and 24hr by a few tenths of a degree and when the temp change slowed down it caught up ok. Now if i can just get a day of sun so i can compare how they work with the heating and wind speed i will let you know how it checks out.

[jerryg]

Just some more info, rain shower moved overhead and temps dropped to 81 degrees. When the sun came out the wind was calm and the two fars temp climbed to 85 degrees while the ss was at 82.5. The wind picked up to 3 to 4 mph for awhile and the ss started to catch up to the fars when the wind laid again and the fars went up to 87 degrees while the ss was at 85, then the wind started to blow again around 7 to 10 mph and the ss caught up real quick and is tracking with the fars again. This sort of shows how much the wind is needed with the ss. Oh yes, ss is standard shield, just do not like typing the long form of it lol. Now just some sun shine and i am set.

[jerryg]

Finally had some sunshine and light winds to do alittle comparing. In general with winds less than 2 mph the ss was showing 1.2 degrees warmer than the fars. With the wind between 2 to 5 mph it was around .8 degrees higher and with the wind 5 or more it was running around .5 degrees higher. One big thing is a cloud moved overhead and the fars started to drop right away while it took a couple of minutes for the ss to respond and it was over 5 minutes before the ss got down to where the fars where showing and when the cloud moved out the fars increased in temp almost immediately and was back up to the high temp in a couple of minutes whereas the ss stayed low for several minutes before it started to climb back up. The lag time below 5 mph is pretty long. Once the windspeed gets up above 5 mph or so  the response time improves quite a bit. When i did this test awhile back the temp in the ss was 2.5 degrees higher than the fars but that was with no wind, dead calm. The main thing i have seen so far is how close the dfars with the fan running all the time compares to the 24hr fars with fan running all the time, they are almost the same all the time. I guess if you are in an area where the wind blows all the time a ss is ok with only a slightly higher daytime temp. The fan helps a bunch in the daytime with light winds and helps with respose time. When its cloudy or night time i get a better response time with the temp and humidiy readings.

[jerryg]

Well this is the final report, had no wind and clear skies this am and some fog. It looks to me like the bad thing about the standard shield is the lag time when there is no wind.. After the sun was up and low to no wind the temp was lagging behind the fars by a degree or so and the humidity was showing 97 percent while the fars were showing 85. When the wind picked up to around 5 mph plus the ss stared to catchup with the fars and finally did. So far the first couple of hours of temp/hum change the ss was reporting it wrong compared to the three stations with fans running and the local metar. So i guess if you live in an area with light winds a fars is important to get accurate air temp. I changed the shield back to a dfars and using the 3 volt wall wart i did put a 10 ohm half watt resistor in series with the fan and now the voltage is running 1.9 volts, so might extend the life of the fan. At that voltage it was tracking along just fine with the other fars, so no need for any extra speed on the fan.

[jerryg]

Well i have been doing some more playing around with the shields and found something interesting to share. I had several standard shields lying around and decided to build me a 24 hr type shield, this was before i got the 24hr shield, i used the same number of plates as the 24 hr and used 2 inch sked 40 pvc for the chamber because the davis sensor fits perfect inside the tube on edge. I had about a half inch sticking out the bottom two plates which are solid ones with holes that you have to just force the tube through to hold it in place and the fan chamber of a dfars was sealed at the bottom with a disc cut to fit in the bottom and over the tube so air is drawn up the tube. When i ran it the temp was running just under a degree hotter than the dfars. After pondering it for awhile i decided that the dfars and standard shield had solid plates on bottom and i might be getting reraidated heat from the ground so i took a solid plate and mounted it about a half inch below the opening of the tube and wow the temp read just like the dfars. Ok got the 24 hour and was disappointed the temp was running higher than my homebrew and the dfars so i took a solid plate and mounted it below the opening on the 24 hr and again the temp came right in line with the other two. well yesterday i decided to put the 24hr back on solar/battery to compare it to the dfars especially at night when the fan cuts back to half of the fps. The temp in the 24hr fell alot slower after the fan slowed down but i had the big idea of removing the bottom plate i had hung on the shield. Today i watched the temp running high again in the 24 hr like it did when i first got it so i put the bottom plate back on and again the temp was in line with the others. Will see tonight if the temp drops like the others with the slower fan speed.