fan boosted radiator - how well does it work? Useful for heat pump?

[williamwright]

Ya boo sucks to you as well!

Bill

[Theo]

Nope, it's a high temperature output. 55C out of the ASHP at SCOP 3.15:

don't have the hybrid version with the oil boiler) Theo

[Fredxx]

Sigh - your usual sign of a lost argument.

[Ponyface]

I have toyed with the idea of fitting computer cooling fans to the bottom of my double radiator, they would run off a mains plug with a usb type socket, so not a lot of power use, but it would be possible to alter the speed of the fans easily.

[Rob Morley]

You won't get much out of a PC fan at 5V. The quiet fan hack used to be connecting the fan to +5V and +12V for an effective 7V which was about as low as you could go and have it run reliably.

[Paul]

Select larger fans, with known-good LV operation.

I use Vantec Stealth 12V fans, mostly run at 12V, but some have been put on voltage reducers. Unfortunately, I don't think Vantec makes those any more - they had excellent bearing life (none have died so far).

*******

Noctua as a brand, makes good fans. But this is a four pin PWM, and you cannot (or should not) voltage modulate these by running the 12V wire at 7V. These should be PWM controlled for best result. PWM is a digital signal (5V amplitude),

25KHz (inaudible side effects) where the width of the square wave tells the fan what speed to run at. Cheap PWM fans make a slight whine from the PCB circuitry in the hub. it was perhaps some Intel appnote that recommended not reducing the +12V on PWM fans (MOSFET problem).

+12V If fan has a PWM wire, do not reduce this voltage GND RPM Output signal, two pulses per revolution PWM 5V control signal, 25KHz, pulse width indicates speed

*******

This is an example of what one of the larger fans buys you.

"AeroCool Silent Master 200mm Blue LED Cooling Fan EN55642"

# Read the reviews !!!

"I was awoken to the fan blade wobbling and colliding with my cpu heatsink. Luckily it didn't break anything in the computer. The quality is pretty subpar but was the only option i had available at the time. If you plan on getting this, keep an eye on it."

Well, it's a sleeve bearing, innit.

It uses the same serrated edge as the Noctua. It even steals some of their datapoints.

"Airflow up to 76 CFM and low noise level at 18 dBA running at 800RPM"

The CFM rating when it is run at min possible speed, might be lower. You cannot have 18dBA noise rating at 76 CFM. Serrated edge or not.

Most fan tricks "make a 3dB difference". All that engineering work, does not magically make a fan quiet. It just makes it slightly less noisy.

For your silent rad project, you're going to object to the noise level, no matter what fan you use. That is just how the human ear works.

But big fan and low RPM, is the best you can do, at a measurable CFM value. Note that the slimmer dimensions of this design, is what endangered the bearing. The story behind the Noctua, they had to make a thicker bearing to reduce bearing load on the bearing surface, plus reduce blade mass.

Model: Silent Master 20cm Fan: 200 x 200 x 20 (mm) # 25mm is normal thick, 37mm is high pressure, 20mm or 15mm are thin Color: Blue LED / Red LED # we want NO LEDs, but if it must have LEDs, choose the RED model !!! Bearing: Sleeve # limited lifespan, the reviews read as expected Net weight: 155g Gross weight: 255g (Retail Box Included) Rated voltage: DC 12V Starting voltage: 9V (Power ON/OFF) # Now, this is a bit weird, and not a good sign Power Consumption : 5.1W Rated Current: 0.42A # It's a big fan, with a high CFM claim... Speed: 800±200 RPM Air Flow : 76.0CFM # The question is, how much less than 12V will it accept Air Pressure: 0.36mm-H2O Noise: 18.0dB(A)

Modern fans, dropping them on the table, dropping them on the floor, instantly ruins the bearing and makes them noisy. Low noise fans should be treated with respect when assembling them in your framework. When reviewers review products like this, they have to be very aware of whether they dropped the product while handling it for the review.

Sturdy fans, with BB motors, can take more abuse, but they are as noisy as hell. There are BB, FDB, Sleeve, Ceramic, Magnetic Tip... all sorts of motor/bearing types. FDB would be a Panaflo, for a famous example. The Vantec has lasted so long, it must be FDB or something. It can't be a sleeve, or it would have fallen on the floor by now.

********************************************************

Now that you know all the rules, here's a real fan supplier. I've highlighted two where you may be able to modulate the voltage. These don't have PWM pin, as interface is three pin. And if you really want it to run of +5V, there's a version for that.

+12V GND RPM Output, 2ppr

NF-A20 FLX 200x200x30 800 18,1 146,9 1,08 3-pin NF-A20 5V 200x200x30 800 18,1 146,9 1,08 3-pin

You could run the FLX off a 12V adapter. It has two speed settings.

Comes with kit.

User manual (if I got the correct file that is). Not a lot of help here.

That should not wobble or smoke, and uses real engineering.

For it to run at 550 RPM using LNA, the assembly would look like this. This is a best guess, as I cannot find any labeled pictures confirming it.

[Picture]

I keep male and female 3-pin fan parts here, so I can make my own adapter cables. If I need Molex adapters, I just buy a Y-cable at the computer store and chop it up to get males or females as appropriate.

Maybe that is just not right for your project, but if I had to evaluate a fan, that's where I might start. You might already have enough regular computer fans to test with what you've got.

Paul

[Animal]

used 2 variants of it.

1970s Myson fanned rads. Worked, but background noise was enough to result in an unacceptable amount of 'what? didn't hear what you said'. They did save a lot of wall space, but didn't look good.

More recently added a single 4" pc fan under an unfinned rad, run on 6v. Much quieter, added 2-3C to the room.

Note that fan starting performance & noise on 6v varies a good deal between fans. This one was almost silent, some are irritatingly noisy and some don't start reliably. 5v is too low IME.

[charles]

Our village hall was fitted with these when a heating system was installed. The fans were fed from a Variac transformer so they could be turned down low to minimise the noise. We got rid of them and put conventional radiators in their place and turned them on half an hour before the booking started - 1 hour in really cold weather.

[williamwright]

Exactly my experience.

Bill

[Ponyface]

think you will get plenty of airflow with a comp fan, two fans attached to the bottom of the radiator would be more than enough, dont want a gale blowing through, just enough to increase the usual convection speed.

[PeterC]

I'd use the fans to blow downwards - heats the room more quickly, rather than the ceiling.

[alan_m]

Are assuming that the air will be entering from the top of the radiator, travelling down the radiator end exiting via the fan?

With a bottom mounted fan, unless the bottom of the radiator is completely sealed with the only opening being the fan the air is likely to be coming from the path of least resistance - coming in from the bottom of the radiator and straight into the fan - room air being circulated from the floor back to the floor without being heated.

Fitted to the top of the radiator and blowing down - hot air being input to the fan, being ducted down to heat a bit more and then rising with convection. Seal the top of the radiator and the fan will only blow air over a very small part of the length of the radiator.

I doubt is a PC type fan is even going to achieve much if you want it to be noise free.

[Paul]

You need a big fan and low RPM.

The example I picked out, is 76CFM at 18dBa. which is "unbelievable" as a spec, so should be treated with caution.

Part of noise, is the psychoacoustic "quality". I had a minicomputer in uni, and with locked doors in the way in the lab, the piercing sound of the fans could be heard at 200 feet from the computer. The weighted noise level of the fan, might have met the numbers quoted in the product spec, but in terms of the irritation level to humans, it was "off the scales". The fans in question, there were four of them and they were AC powered fans, not the brushless DC used on computers today.

Fans can make a tone. They can make white or pink noise. They can sound like "wind in the trees". Some of these qualities are more acceptable than others.

Placing the fan on "side-opposite" and using the radiator bulk to shield some of the noise. If the fan could be positioned at some distance, if the air could move through an absorptive labyrinth, that can help too. Most of the time, there isn't room for fancy airflows.

Paul

[Animal]

With air fanned up, convection helps. Down it hinders. Obviously radiators don't just heat the ceiling.

[Animal]

I guess imagination & confusion beats asking those of us that have done it.

[Animal]

Some 12v fans are fine at 6v, many aren't. IMLE of a few dozen anyway.

[Animal]

If you want to maximise airflow while remaining silent, that is the way to go. If you only want to up output a bit there are simpler cheaper options.

[Rob Morley]

But you said USB PSU - that's only 5V which won't do much with a 12V PC fan.

[Paul]

You do occasionally find 5V fans. The claim is, with the 8" (200mm) side-PC fans, there are no standards for hole placement. The standard fans are 120mm or 140mm (as seen in ATX PSU designs for example). The even-larger sizes can be non-standard, because it's harder to make a good/large fan (bearings stink), so not as many suppliers make them.

NF-A20 PWM 200x200x30 800 18,1 146,9 1,08 4-pin PWM 12V @ 0.08A NF-A20 FLX 200x200x30 800 18,1 146,9 1,08 3-pin

NF-A20 5V PWM 200x200x30 800 18,1 146,9 1,08 4-pin PWM 5V @ 0.17A NF-A20 5V 200x200x30 800 18,1 146,9 1,08 3-pin

I have a PC fan here, which draws 12V @ 1A or 12W, and if they were to make a 5V version, the current flow required would be obnoxious. A cell phone charger adapter might not be enough, and you might need a powered USB hub wall supply to get the amperes. That fan is loud enough, you can't sit in the room with it, and even with a speed reducer fitted, it was still too loud. So it's retired. That one does 110 CFM at 12V.

Standard PC fans, come in four speed ranges. Low, medium, high, ultra. Low and Medium are typical home PC choices. High or Ultra are good for server rooms (maybe).

The highest air velocity, is around 800 LFPM. Above that speed, heat transfer is asymptotic (cannot be improved on). We did one product at work, running air that high, and you had to wear ear protectors on your head in the lab setting :-/ But you could change out the fans while the machine continued to run. The fans had their own "box" for containment, and there were holes in the machine to slide the cooling fan into. As soon as the power contacts touched, it would spin up. A louver scheme prevented backflow when the box was pulled out. But that probably rates as my worst cooling experience. You would not want those on your rad.

Paul

[Animal]

Do yourself a favour and don't use a 5v fan on 5v.