U4 high pitch noise during operation?

[prl]

Have you enabled WOL?

I just did a search,if that's wake up with LAN ...

It is: WOL = Wake On LAN. Also one of Winnie-the-Pooh's friends

[HarryHoudini]

Have you enabled WOL?

Well done,great knowledge.
It was turned on,i don't know what the default is but i didn't turn it on but my Brother may have when he was mucking around trying to get it to connect to the Internet via the Computer.
Anyway end result after turning WOL off and disabling the auto Deep SB every night was that this morning the U4 is back to room Temp and the USB ports are off.
I'll re-enable Deep Sleep tonite and see what happens there so it was either that or the WOL or both that was causing it to stay active.
Thanks again,after years with Toppies heat is a killer.

[HarryHoudini]

Have you enabled WOL?

Well done,great knowledge.
It was turned on,i don't know what the default is but i didn't turn it on but my Brother may have when he was mucking around trying to get it to connect to the Internet via the Computer.
Anyway end result after turning WOL off and disabling the auto Deep SB every night was that this morning the U4 is back to room Temp and the USB ports are off.
I'll re-enable Deep Sleep tonite and see what happens there so it was either that or the WOL or both that was causing it to stay active.
Thanks again,after years with Toppies heat is a killer.

Edit....can confirm the power to the Tuners(causing the area to heat up during Deep SB) and USB ports was caused by WOL being switched on,everything back to normal with it switched off.Happy days.

[MichaelAtOz]

So, two months on, is there a H/W fix for this in the currently shipping models, so that fan speed is again related to temp and not at 100% all the time?

[prl]

So, two months on, is there a H/W fix for this in the currently shipping models, so that fan speed is again related to temp and not at 100% all the time?

There is a fix for the problem in the recently released official firmware (20180417). I don't know what the firmware is in the currently shipping U4s.

The fix for the noisy fan was achieved in firmware. The fan speed is not temperature-controlled. It's at a constant, user-settable speed, and has separate speed settings for standby and running. The T4 got the same fan speed control setup at the same time. The final fixes were made available in beta firmware in mid-March.

[MrQuade]

So, two months on, is there a H/W fix for this in the currently shipping models, so that fan speed is again related to temp and not at 100% all the time?

Just to add to prl's post, the fan never was temperature controlled in the first place either.

[sub3R]

And peteru gave a bit of background on the fan in the U4 in this post.

And I posted some information on fan settings vs. temperature on our U4 in that same topic two posts below peteru's post.

[BigBen]

G’day, new user here. Sorry my first post is about a problem.

Received my U4 on Christmas eve and have been loving it. It’s been brilliant except I’m getting a loud buzzing sound. It’s not high pitched and I’m on the latest firmware but I still think it’s the same kind of problem as others here because

• stopping the fan with my finger has no effect

• unplugging the fan stops the buzzing noise completely

I’ve tried

• Adjusting the fan speed - higher speeds make the buzz louder

• Reinstalled latest firmware

• echo >/proc/stb/fp/fan_pwm ff command - no effect other than increasing fan speed to maximum

Would like to try a capacitor, as peteru recommended, if someone can advise a precise model and give some tips on how to install.


On a probably unrelated question, is it normal for the HDMI out to stay active in standby and deep standby?
My receiver never stops reporting an active connection unless I completely switch off the U4’s power.
What’s really strange is if the U4 is in standby and I switch my receiver off & on, or switch the input & then back again (fresh HDMI handshake), my TV will start to receive an actual video signal. It’s just a black screen but if I push buttons on the U4 remote the “no action” symbol (square with a diagonal line through it) will display on screen.

Would really prefer not to have to post the U4 back for repairs so any advise would be much appreciated.

[Gully]

G’day, new user here. Sorry my first post is about a problem.

Received my U4 on Christmas eve and have been loving it. It’s been brilliant except I’m getting a loud buzzing sound. It’s not high pitched and I’m on the latest firmware

Welcome.

It does sound like the problem that was in early firmwares and then fixed.

Unfortunately, telling us you have the latest firmware is not very informative as there are official and beta releases, so you really do need to post the exact details of the firmware.

If you are not already using it, install the latest beta firmware https://www.beyonwiz.com.au/forum/viewt ... 56&t=12441

Switch to the beta feed and update to 17.5.20181104 and see how it is then.

[BigBen]

Thanks. Yes should have said official firmware. Didn’t know how to try beta versions so thanks for the link. Will try tonight.

[prl]

On a probably unrelated question, is it normal for the HDMI out to stay active in standby and deep standby?
My receiver never stops reporting an active connection unless I completely switch off the U4’s power.
What’s really strange is if the U4 is in standby and I switch my receiver off & on, or switch the input & then back again (fresh HDMI handshake), my TV will start to receive an actual video signal. It’s just a black screen but if I push buttons on the U4 remote the “no action” symbol (square with a diagonal line through it) will display on screen.

When in standby, the U4's HDMI video and audio is shut down (the TV displays "NO SIGNAL") and timeshifting is stopped, so that the HDD will also spin down after 5 minutes unless a recording is being made. I can confirm, though, that if I force a HDMI handshake, I can see that the UI overlay displays the "UnhandledKey" popup on the TV screen when I press any button on the remote. That last ne is a bit odd, but not all that surprising since the U4 is in a state that's not a long way from fully running.

In deep standby the mainboard is almost completely powered down. The main system chip is still active to a small extent, but only enough to support startup. There will be no HDMI output.

As for the firmware version, there have been reports that some recently delivered Beyonwizes are running firmware a good bit older than the most recent official release (20180714) or a recent public beta (the most recent of which is 20181104).

The current firmware version is in MENU>Information>About>Revision.

[Grumpy_Geoff]

When in standby, the U4's HDMI video and audio is shut down (the TV displays "NO SIGNAL") and timeshifting is stopped, so that the HDD will also spin down after 5 minutes unless a recording is being made. I can confirm, though, that if I force a HDMI handshake, I can see that the UI overlay displays the "UnhandledKey" popup on the TV screen when I press any button on the remote. That last ne is a bit odd, but not all that surprising since the U4 is in a state that's not a long way from fully running.

I've had my U4 display output even though the box was in standby. I had turned on the TV and could see live TV playing, I used the remote and received the the "UnhandledKey" popup - I thought "what?". Then I realised the output to the TV was slightly darker than normal, and then further realised the front panel was showing the date/time as per standby mode. I put all this down to the standby background zapping of my provider visiting script and the quickly switching to running then standby it does at the end of the run.

[Grumpy_Geoff]

As for the firmware version, there have been reports that some recently delivered Beyonwizes are running firmware a good bit older than the most recent official release (20180714) or a recent public beta (the most recent of which is 20181104).

I know you meant to type 20180417

[BigBen]

Thanks for confirming HDMI is normal. It is strange that even in deep standby the link is kept open, but no actual video is sent so I guess it’s not a problem.

The initial order confirmation email from Beyonwiz recommended updating the firmware so I did that on day 1 from USB (and then again a couple of days ago) to 17.5-beyonwizu4-20180417.
Model: Beyonwiz U4
Bootloader: 2017-10-18 12:06:44
Kernel: 4.9.51
Drivers: 180411-16.4
Version 17.5
Build: 20180417
Revision: 20180417
Enigma2: 17.5-477-g41eefa7
GStreamer: 1.14.0
Python: 2.7.13

Have now changed to the beta feed
Build/Revision: 20181104
Enigma2: 17.5-800-g3d9a605
GStreamer: 1.15.0 (GIT)

It updated fine but no change to the buzzing noise sadly. Saw the CPU temp at 43C although I think settled at 41C.

[MrQuade]

Adjust the fan speed. Every unit is different and the fan will resonate at different speeds depending on the unit and environment. (And some users will be happier with different types of noise than others)

[BigBen]

Will give that another go. Might be able to put up with the noise as long as it’s not causing long term damage to anything?

Edit:
There’s really no sweet spot. Higher speeds are generally louder. Also varies a lot throughout the day, I think because of temperature — hotter = louder. At its quietest it’s not too bad but the other half of the time it really is frustrating.

There’s also a very quite buzz in deep standby that stops if the fan is unplugged.

Anyone know the precise specs of the fan? It just says FXDS 12V DC Brushless Fan

[sub3R]

There’s also a very quite buzz in deep standby that stops if the fan is unplugged.

Is something on your U4’s main board resonating? When I was trying to track the fan noise down I found if I pressed lightly various spots on the main board with the end of an insulated piece of thick plastic tubing I could make the noise louder or softer. But that was before peteru’s fix in f/w was implemented.

Anyone know the precise specs of the fan? It just says FXDS 12V DC Brushless Fan

I couldn’t find exactly which model fan we had (See here. I can’t get the links in that post to work anymore).

[Grumpy_Geoff]

There’s also a very quite buzz in deep standby that stops if the fan is unplugged.

Is it this noise. As peteru posted, my noise stops when I move the fan blade a fraction (but returns as soon as the blade drops back to its original position).

[Grumpy_Geoff]

I couldn’t find exactly which model fan we had (See here. I can’t get the links in that post to work anymore).

Be patient grasshopper, they opened for me

[BigBen]

Yes applying pressure at different places can reduce the noise a tiny bit.

I came to the same conclusions on the fan. It seems to be one of the 50x50x15mm 12V models from here (site can be slow/unreliable)
But there’s no model number so it could be any of the 4 or even a custom version.

While I’ve put together plenty of PCs and home entertainment systems all I know about electronics is red = positive and black = negative. Would the fan port support any similarly sized 12V fan?

[peteru]

I tuned the PWM output settings for one particular model of a fan - the one shipped in all U4s. It should work with any 12V fan with reasonably similar current draw.

The things you need to consider in order of priority:

* physical dimensions
* operating voltage
* operating current draw
* startup voltage / current requirements
* drop out voltage / current requirements

The last two are a bit tricky. If you have a motor that has high requirements and combine that with some extra friction in bearings, you could end up in a situation where low speed fans will result in a failure to spin the fan.

Having said that, as long as you satisfy the first two criteria on the list, just try the fan and see how well it works.

[BigBen]

That’s great. Thanks so much for such quick responses. Will try a 1000uF 25V capacitor tonight and maybe also a fan change.

[Paul_oz53]

There’s also a very quite buzz in deep standby that stops if the fan is unplugged.

Is something on your U4’s main board resonating? When I was trying to track the fan noise down I found if I pressed lightly various spots on the main board with the end of an insulated piece of thick plastic tubing I could make the noise louder or softer.
...

The old mechanics trick is to use a short bit of garden hose or similar as a stethoscope to narrow down the exact noise source. Careful not to put too narrow a tube in the ear though! Besides, a larger size works better to collect more of the noise.

Also, I have wondered about using an external 80-120 mm fan on top of the unit to provide deluxe cooling by plugging it in instead of the internal fan, but then the U4 seems not to have any heat issues in our air-conditioned lounge.

[BigBen]

There’s also a very quite buzz in deep standby that stops if the fan is unplugged.

Is it this noise. As peteru posted, my noise stops when I move the fan blade a fraction (but returns as soon as the blade drops back to its original position).

Moving the fan blades in deep standby can make the buzz much louder. It’s almost completely silent at room temperature but when warm, just after shutting down, it can be quite loud.

[sub3R]

I came to the same conclusions on the fan. It seems to be one of the 50x50x15mm 12V models from here (site can be slow/unreliable)
But there’s no model number so it could be any of the 4 or even a custom version.

Isn’t that website slow? Geoff, my links did eventually work.

I’m fairly certain the FXDS DC5015 [50*50*15mm] model 10 fan on that site isn’t the one in the U4. I previously measured the height of the one in the U4 at 12mm (the DC5015 model 10 is listed as 15mm).

An interesting test would be to unplug the U4 fan & connect it to a 12V DC supply & measure the current. Then looking at the Technical Parameters for the FXDS DC5015 [50*50*15mm] model 10 fan, one should then be able look at the 12V fan model with the closest current to get an idea of the air flow, rpm & noise. Even though I think the fan in the U4 is slightly different to the FXDS DC5015 [50*50*15mm] model 10.

FXDS DC5015 [50x50x15mm] model 10 fan

One good part about the U4 is the fan speed adjustment & the CPU temperature indication in the menu. The CPU temperature indication can be used to see if a replacement fan is doing its job. Both very nice features.

The old mechanics trick is to use a short bit of garden hose or similar as a stethoscope to narrow down the exact noise source.

Yep. That’s what I was doing with the plastic tubing to track down where the fan noise was coming from. A trick I’ve used many times in the past for tracking down car engine noises & tuning twin & triple carburettors on cars.

[BigBen]

Yes you’re right, it’s not 15mm although it looks like a 15mm would fit in the case.

Just tried a 0.7 m3/min, 16.6 CFM, 2.28 watt, 190 mA, 5500 rpm, fan.
It ran ridiculously fast by comparison unless I set the speed to the minimum. Every shutdown and boot up the speed runs at maximum for a few seconds as well so I’m not going to leave that one in. Will order a much slower rated model to try.

Also trying 100uF - 1000uF 25V capacitors but not with much luck except maybe in combination with the new fan.

[peteru]

A 50mm fan running at 5,500 RPM will sound like a model aeroplane taking off. You don't want that. A sensible speed would be somewhere under 1,500 RPM. You want to cool the SoC, not suck it off the motherboard.

[sub3R]

Yes you’re right, it’s not 15mm although it looks like a 15mm would fit in the case.

Just tried a 0.7 m3/min, 16.6 CFM, 2.28 watt, 190 mA, 5500 rpm, fan.
It ran ridiculously fast by comparison unless I set the speed to the minimum. ...

The fan in the U4 draws 30.5mA at 12.0V pure DC.

I think the FXDS-DC5010[50*50*10mm] model 9 (part number D5010L12TB) fan shown here may be closer to what we have in the U4. Even though it is called a 10mm, the drawing shows the height dimension as 12mm (same height as the one in the U4). The only differences that I can see are; no enclosed sides & only 3 motor mounting supports (the U4 fan has enclosed sides & 4 motor mounting supports), but the DC5015 [50*50*15mm] model 10 fan also only has 3 motor mounting supports with enclosed sides.

I don’t know why two fans with the same part number in those Technical Parameters lists have different specs (see the D5010M05TB, the D5010M12TB & the D5015M12TB in their respective tables).

Perhaps the one in the U4 is a custom version or an older model.

FXDS-DC5010[50x50x10mm]-DC-fan

BTW - Did you notice peteru’s last post here about using one or two series diodes?

[BigBen]

It has to be a custom model. Its max speed is nowhere near 3500rpm and nobody seems to sell anything slower which is a real pain. The unusual 50mm size doesn't help either.
The only option I can think of without over complicating things is to use a high quality fan designed for low noise and then set both the Normal and Standby speeds to the minimum.

The other thing I thought of was to use the spare rear USB2 port to power the fan instead. As long as deep standby cuts the power to the USB ports.

[sub3R]

It certainly looks like a custom model. From what I've noticed, 3500rpm & higher seem to be fairly common.

I was surprised by such a low current. With just the fan running flat out at 12V pure DC I couldn't even hear it, but I have far from perfect hearing.

A high quality fan designed for low noise could be worth a try, providing you can get one that is suitable. Noctua have a pretty good name for quality & low noise.

[BigBen]

Noctua don’t make a 50mm model. It’s why I say it’s an unusual size, 40 & 60mm models are a lot more common.

Had another search and settled on this:
Noiseblocker NB-BlackSilentFan XS-1 50mm (Data sheet)
3000rpm, 7.5–13.8V, 0.48 Watts, 11.5 m3/h (0.19 m3/m), 16.3 dB/A

Will need to convert from 3 to 2-pin but that’s easy. Couldn’t find it in Oz either so will be waiting a while but might turn out to be a nice upgrade since the U4 does get very warm to the touch.

The other option I thought about was the Fractal Design Silent Series R3 50mm (3500rpm, 19 dB/A).

Is it possible that some of these problems are partly caused by the fan being so unusually low powered? If adding diodes in series with the fan can stop the buzzing in deep standby, could using a fan capable of drawing more current also work?

Also wondering if the fan_pwm ff command completely turns off PWM or if changes to the drivers (by “the Koreans”?) could mean that the maximum is ~95%?

[MrQuade]

Had another search and settled on this:
Noiseblocker NB-BlackSilentFan XS-1 50mm (Data sheet)
3000rpm, 7.5–13.8V, 0.48 Watts, 11.5 m3/h (0.19 m3/m), 16.3 dB/A

I spotted those in the T4 fan thread here:
https://www.beyonwiz.com.au/forum/viewt ... 38#p172954

The Amazon reviews were not favourable in terms of their claims to being quiet though, despite the claims of 16dB.
It might have just been some bad ones, but it could also be that they start quiet, but get noisy shortly afterwards as the parts wear.

If you do go for it, then let us know how it goes.

[BigBen]

That’s a pretty old review. They’ve changed bearings and drive since then. The other bad review was after 2 days so either a faulty unit or something else was causing their problem.

[sub3R]

That’s a shame about Noctua.

Both those fans you linked to look fairly good to me, but the NB-BlackSilent XS-1 50mm has pretty good noise figures.

If you are going to run the fan at the minimum speed setting the Fractal Design Silent Series R3 50mm may have a slight edge over the NB-BlackSilent XS-1 50mm as far as the minimum start-up voltage goes. The Fractal Design one has a minimum start-up voltage of 7V whereas the NB-BlackSilent one has a minimum start-up voltage of 8V. When hooked up to the U4 I don’t know how relevant that would be though, but worth considering.

Out of the two I would favour the Fractal Design, but I’m only saying that because I like their computer cases. I have a new Fractal Design Define R5 in the cupboard for my next build & I’m very impressed with the design & quality. I don’t know how good their fans are though.

I don’t know the answer to the fan_pwm ff command, but peteru probably would.

It sounds like you could be doing a bit of experimenting to see which way to go, so the following could be worth a try first to see how your existing U4 fan sounds without being PWM controlled. Do you have access to a 12V battery? If so, wire the fan to it via a fuse with adequate insulation at all the connections & see if you can hear the fan when using the U4. I’m not sure if you would need to use the command to turn the fan off to try this or not. You may need to load the battery up a bit to get it down to 12V from its open circuit volts.

I would be interested in how you go with any of this.

[BigBen]

Yes I’ve never owned any Fractal Design parts but always heard good things.

Don’t think startup voltage should be an issue because the U4 seems to leave PWM off for the first few seconds of boot up.

I’ve actually ordered a very cheap USB to 12V molex cable/adapter just to make it easy to test things, but I might try connecting the fan to my motherboard when I get a chance. Although no matter how it runs I’m going to want to try a different fan in the U4 anyway so I’ve ordered the NB-BlackSilentFan XS-1 — I really want it to be as silent as possible.

It’s going to be a while shipping but will keep updating.

[peteru]

Is it possible that some of these problems are partly caused by the fan being so unusually low powered? If adding diodes in series with the fan can stop the buzzing in deep standby, could using a fan capable of drawing more current also work?

I don't think power draw has anything to do with it. If anything, higher current demands are likely to make things worse. Experimentation would be required to be sure.

As far as the diodes in series are concerned, that is just an untested theory. That theory is based on what I know about the circuit that drives the fan. There is PNP transistor with emitter connected to 12V and collector connected to the fan pin. The base has a 1k resistor to the emitter and 18k resistor to the collector. As a result, you'll always have some current flowing through this 19k path. Clearly it's not enough to turn the motor of the fan, but it is sufficient to induce enough of a magnetic field to cause the rotor to oscillate. The idea with the diodes in series is that they could possibly act as open circuit while the voltage / current is under a certain threshold in deep standby. Maybe.

A simplified circuit diagram for the fan drive looks something like this (there's a lot more to it, but this should be enough to get the idea)

Simplified fan circuit

Screenshot_2019-01-18_Fan_circuit_simplified.png (8.15 KiB) Viewed 5089 times

Also wondering if the fan_pwm ff command completely turns off PWM or if changes to the drivers (by “the Koreans”?) could mean that the maximum is ~95%?

The old "Korean" code had some bizarre value transformations with all sorts of weird discontinuities. I ditched a large part of the original driver code and re-implemented using something more sensible. A value of 0xFF will program the PWM hardware with 100% duty cycle, as you would expect. The only limit is at the low end. Any value in the range 0x01-0x10 will be translated to 0x11. This is to prevent motor stalls with a correctly functioning factory installed fan.

[peteru]

I just checked some old comments in the code and it appears that at some stage, some design used a fan that had a maximum speed of 4500 RPM at 12V. I don't have enough code history to conclude if the U4 fan would bear any similarities to what I found in the code - it could have been for a completely different piece of hardware. However, knowing that a lot of the time designs and components are reused with minimal changes, it would be reasonable to assume that the U4 fan is a model that has a max speed of 4500 RPM.

[BigBen]

Wow, thanks for all the info.
I should have realised ff was hex and I could try other speeds with the same command.
Wouldn’t a 4500rpm fan be extremely loud at 100% duty? The 5500rpm fan I tried (the only 50mm model I could quickly find) felt and sounded like it was 10x faster than the OEM.

[sub3R]

The other thing I thought of was to use the spare rear USB2 port to power the fan instead. As long as deep standby cuts the power to the USB ports.

For future reference; across pins 1 (VCC +5v) & 4 (GND), the top rear USB 2.0 port measures 5.13V DC with my U4 in normal standby or ON, & 0V DC in deep standby. This is with a ‘Bluey’ USB tuner plugged into the bottom rear USB 2.0 port.

[BigBen]

Good to know. Yes you would either need a 5V fan or something like this:
USB 12V Step-Up Power Cable

[sub3R]

An interesting device. I wouldn’t call it ‘Absolute genius!’ though like they did. It probably has something like a LM2577ADJ at its heart like this project. A LM2577 pdf data sheet is here.

The unknown is how much output filtering is used in the one you linked to. If not much you may still get fan noise. And as you would probably know, the output power will naturally be less than the input power (the U4’s USB 2.0 port should be rated at 500mA).

Still, it is cheap enough to try seeing you have ordered a 12v DC fan, & you can probably just duck up to your closest Jaycar store & pick one up for the same price (it looks identical to the one you linked to).

But if I didn't have a 12V DC fan on order I would prefer to run a 5V DC fan from the USB 2.0 port. But having said that I don’t know what is available in 5V DC fans compared to 12V DC models.

[BigBen]

I think I’m out of ideas.

Tried one of these:
Scythe Mini Kaze 50mm Quiet Fan
4500rpm, 9.42 CFM, 16 m³/h, 0.06 A, 26.09 dB/A, 2-pin

...but still had the same buzzing and even the clicking from the fan. Adding a 100uF or 1000uF 25V capacitor seemed to just increase the fan speed.

I should also mention that the U4’s fan port is not standard ATX. I think it’s a 2.0mm mini 2-pin port usually used by GPUs. This was a bit of a pain although you could try one of these.

Don’t think I’ll bother trying the USB route, it was just an idea. Trusting a component’s only method of cooling to a USB port doesn’t seem right somehow.

Might still try the Noiseblocker fan when it arrives but I don’t hold much hope.

[sub3R]

I think I’m out of ideas.

Using the fan supplied with the U4, here are a few you could temporarily try for testing purposes only to see if the noise is reduced. If you haven’t done so already:

1. Acoustically decouple the fan by removing the fan mounting screws & sit the fan on four small pieces of thin foam rubber (don't put the screws back in until after testing & removing the foam rubber). I remember having to make a permanent job of something like this on a previous PC to eliminate fan noise resonating the PC case.
2. Tap into the 12V DC U4 power supply plug pack at the U4 & run the fan from that to see if you hear any noise.
3. Try the series diode/s peteru suggested. Any of the 1N series silicon diodes should do (1N4001 to 1N4007).
4. Run the fan at full speed connected to its normal supply.

Obviously you would need to leave the U4’s cover off during some of these tests.

... Adding a 100uF or 1000uF 25V capacitor seemed to just increase the fan speed.

That sounds like the electrolytic caps you used could be too large a capacitance value. I would suggest trying smaller capacitance value caps & see what happens. Say, a range from 1µF up to 25µF. (Just a reminder that capacitors work opposite to resistors in series & parallel if trying to get the right value from two or more).

[BigBen]

...here are a few you could temporarily try...
...1µF up to 25µF...

I’ve now tried all these things, including every combination I can think of, except for the diode(s) which I thought was only for buzzing/squelching in deep standby. Is there any chance it could help when the fan is running?

[sub3R]

I’ve now tried all these things, including every combination I can think of, except for the diode(s) which I thought was only for buzzing/squelching in deep standby. Is there any chance it could help when the fan is running?

Well that’s disappointing. I thought you would hear a minor improvement with one of those, especially item 1 & 2.

That’s what the diodes were intended to do in deep standby according to peteru’s untested theory, but I thought you had this problem as well. By adding just one diode the current through the 18kΩ + 1kΩ resistors would be dropped to 5.94µA from 6.3µA without the diode [i.e. (12V – 0.7V) / 19kΩ], so you may need two or more in series. The only thing they would do in normal standby or ON would be to reduce the voltage to the fan by 0.7V for each diode in series. Enough diodes in series may reduce the voltage to the fan enough to make a difference when the fan is running, but I don’t know.

Are you able to test the fan on a 12V battery as suggested here? I can’t hear my fan at all when I tested it that way, that’s why I’m assuming it is the PWM that is causing the problem with your fan.

Without any decent test gear, a DSO (or CRO) & better hearing, I can’t run any useful tests for you.

Hopefully the NB-BlackSilentFan XS-1 fan you have on order will be the answer.

[BigBen]

Yes I do sometimes get a buz in deep standby but it’s not such a problem. If It continues after solving the main issue then I’ll definitely give the diodes a try (already bought a 4-pack while in Jaycar).

Actually didn’t try tapping the 12V, I’m not really sure how, but since the problem didn’t change with the new fan installed (which works fine in my PC) I think it must be the PWM circuit. It’s just strange that running it at 100% duty doesn’t stop the buzzing.

[sub3R]

Actually didn’t try tapping the 12V, I’m not really sure how, but since the problem didn’t change with the new fan installed (which works fine in my PC) I think it must be the PWM circuit. It’s just strange that running it at 100% duty doesn’t stop the buzzing.

A test of the fan connected to a 12V battery with the U4 switched OFF will conclusively prove that it is the PWM (maybe even resonating something on the main board) or if the fan is a bad fan.

If you aren’t sure about tapping into the 12V DC U4 power supply plug pack at the U4 it would be best not to try. It could be done near the rear of the 12V DC IN socket on the U4 but without decent test leads & knowledge of what you are doing, it could get messy.

[ETA_2016]

Interestingly, you can now buy a replacement fan for the T4 & U4 for $13.95 + delivery.

[sub3R]

Interestingly, you can now buy a replacement fan for the T4 & U4 for $13.95 + delivery.

The one we received was slightly different to the original fan in our U4. It required the mounting holes in the bottom of the fan case to be deepened with a suitable size twist drill bit to allow the fan to sit flat against the heat sink (the holes are stepped). I did this by hand with the drill bit in a tap wrench, it would be easy to damage the plastic fan case if an electric drill was used.

The plastic header for the +ve & -ve wires is different to the header on the U4’s original fan & won’t fit to the mother board. The instructions provided with fan suggests removing the plastic from the mother board pins & pushing the fan header onto the bare mother board pins, but I decided to swap the ends by cutting & soldering the wires then insulating with heat shrink.

It sounds a touch noisier than the original fan in our 2nd U4 but works fine. It appears to cool slightly better than the original by one or two degrees using the same settings.

Image shows the original fan on the left & replacement on the right.