Rigol 1052e Fan Replacement

[Mechatrommer]

I added 120mm intake fans to the front of the computer case resulting in the case being slightly pressurized.

thats the idea. i have mind the same, fan blowing in, but gotta put air filter before it (outside casing). but i dont think there will be cd-rom in the rigol , unless you want something (filter) poking out side of your unit.

[Russel]

[...]thats the idea. i have mind the same, fan blowing in, but gotta put air filter before it (outside casing). but i dont think there will be cd-rom in the rigol , unless you want something (filter) poking out side of your unit.

Apologies for straying off topic.

The dust in the shop environment that I was talking about was so bad that adding filters to the fans did more harm than good. Unless you cleaned the filters daily, the filters would clog up and actually hinder the airflow to the point that the CPU would begin to overheat. Having more fans blowing into the computer case (without filters)than out was a strange and counter intuitive solution to CD-ROM destruction, but it worked. I really went against the grain with me, using a computer in such an environment.

[Mechatrommer]

i dont understand why would people call this off topic. btw my pc mod fan that blowing in got very dirty filter now, i dont have time to clean up. so now it works as normal like before it was modded, ie the psu fan is blowing out, negative pressure inside and my cdrom start to collect dust now. and i'm not talking cpu fan mate, i dont touch that bundled original fan, if i touch that, i think i'm going to screw the cpu. i'm talking ventilation fan. so back to rigol, if somebody want to make the fan blowing in, then they have to make a filter and necessary step in case it got clogged up. that why in my topmost post said... "you have to know what you are doing".

[BobD]

Now done in Germany, works beautifully.  The Rigol can hardly be heard with no other noise in the room.

[RCMR]

Hmmm.. my Rigol is also very quiet.  No obtrusive fan noise -- although if I listen carefully I can tell it's turning.

[ciccio]

Fan blowing in or out has been alway a big debate.

When designing BIG power amplifiers, my colleagues and I discussed a lot about this, and resorted to test temperatures inside  the amplifiers, with both airflow directions.
The result was this: the fan must BLOW IN , if we are searching for the better heat transfer.
The reason are two:
first: the heat removed is proportional to air MASS, not to air VOLUME, cold air is denser than hot hair, so the fan must move cold (external) air inside the hot internal  space of the equipment.
We measured temperatures  5 to 10 °C lower in a 2 KW amplifier.
second: the fan's motor and bearings will live longer if their  temperature is lower, so it must move cold air.

I see two (lower value) advantages for fan blowing out:
first: a lower noise level, due to the fan's blades being farther from case's grille when blowing out.
second: dust, cat's fur and other pollution are evenly dispersed inside the case, and not clogging the intake fan, but this depends on intake fan's speed: a low-speed, large fan remains usually  more clean than a fast, small one.

 
 

[alm]

Funny, with computers with low-speed fans negative pressure usually works better. Of course these are complex systems with often multiple fans and many obstructions in the airflow path. It probably depends on the design of the equipment and the temperatures involved.

I'm sure both arguments you state are valid, but I would expect the airflow design to be more important. Blowing in is easy: the air stream is focused, so you just point it at the hottest component. It also introduces more turbulence. Blowing out requires attention to the airflow so the hot components receive enough airflow. For this reason, I would be careful changing the direction without careful analysis.

[Mechatrommer]

blowing in directly to the subject surface is called "spot" cooling, you cool that particular spot the most at times, as opposed to "ambient" cooling (blow out, not sure the exact name) where you will stabilized temp throughout space (averagely). it will depend on where you fan is blowing and what application.

[vtl]

I've seen most people are replacing the fan which I think is quite puzzling. Theres actually nothing wrong with the stock fan, I took mine out and its actually pretty quiet. Its just the poor design of the vents that create a lot of turbulence and noise.

You can buy fan adapters that just uses a series resistor to drop the voltage and hence speed of the fan, typically on the order of <100Ohms

I measured the fan voltage to be around 10v (not sure if the fan is variable speed or not) and hooked the fan to a bench PSU. After playing with some resistor values I notice that something like 20 ohms would give a decent drop down to 7 or so volts. Its surprising how quickly 1/4 watt resistors can heat up so I put 3 68R resistors in parallel to give a value of around 23 Ohms and share the current between them. Heatshrinked to make it look somewhat neater and now the fan is quiet. Measured the temps and all is well.

If you don't want to butcher your stock fan then you can make up some kind of fan adapter if you like. The plastic fan header you can just lift out and put any 3 pin fan on it.



The 7905 gets pretty hot because of no heatsinking and is placed right next to a bunch of caps. Too bad it was not designed to just be bolted to the chassis for heatsinking.

[A Hellene]

Nicely done, vtl!


Though it is not directly relevant to this thread, I think I should cite one of my replies regarding the high amounts of heat generated by the PSU losses within the device:

The cooling fan is running on 12V unregulated, directly from the -12V rail of the internal switching PSU. Should you decide to silence the fan, I think it would be better to insert an 8V regulator (like the old LM7808/LM7908) to decrease its rotational speed and airflow.

But, if you asked me, I wouldn't touch the cooling fan since it is needed for the good operation of the unit. That is because almost half of the power consumed from the mains (12.87W measured out of the 23VA totally consumed) goes to the actual oscilloscope circuitry, and what remains becomes heat within its casing; heat that has to be let out or it will shorten the life of the device or even destroy it.

The right answer would be a redesigned or a modified PSU that does not produce those amounts of heat within the unit.


-George

(Emphasis is mine)

In short, the major heat sources of the PSU are:
4.5W by the TOP247 and the flyback coil (the "transformer") of the PSU
3.6W by the 317T regulator on heatsink
1.7W by the Schottky rectifier on heatsink
1.1W by the LM7905 regulator ((-11V-(-5V))*(-335mA-(-150mA)) = 1.11W)
0.4W by the MUR460 rectifier
... totaling in an amount of 11.3W dissipated heat that continuously builds up within the case and does not produce any useful work other than heating the device up.

These 11.3 Watts of heat are generated by the PSU components ONLY (meaning that there is another 11W dissipated by the actual oscilloscope circuitry, like the other on-board linear voltage regulators and the components running at 100MHz++ frequencies) and they should somehow be forced out of the DS1000 casing before doing any damage.


-George

[Salas]

Yet another silent running one. 80X80X25mm 1600 RPM 12V 80mA 18dBA GELID fan. Got it from EBAY a long time now but I did not bother the PITA modding till my last long session with the 1052E that its hairdryer howl while I was troubleshooting a high voltage circuit trying to focus not to fry myself, brought me near to really smashing up the bugger. The noise reduction result is dramatic. From annoying level to not noticing level. This one is 100 MHz also of course. If one's got to void it, then void it in stereo modding I say.



Thanking all for the tips & guides.

[SeanB]

Not owning one of these ( and i would like to) I woyuld also recommend cutting out that stamped grille, and using the existing holes for the old fan to mount a wire grille.That will increase airflow a lot, and if there is a external grille then the internal one is not needed. This works for me on PC cases, as they generally are quieter with a wire grille and have more airflow.

With the 7905 on the board a small tag will make it run cooler, with no other mod required. Just cut a strip as wide as the casing and about twice as long, and bolt it to the regulator.

[Salas]

30C tops with the laser aim handgun thermometer straight and close to the plastic shell's exits after an hour of both channels running is the result of the above mod. Since the 7905 is sideways located in front of the larger fan now, I hope it benefits a bit better than before also.

[Salas]

For the those who had asked me, the original fan has this spec: GLOFN GFA06015H12H, 5500rpm @12V DC, 34.2m3/h. 36.5dBA noise.
1 m3/h = 0.5885745 CFM. So 34.2m3/h translates to 20.13 CFM.

[AndersAnd]

Great mod, the original 60 mm fan is way too noisy. 

The pictures embedded in the OP has gone missing at the image hosting site and I can see "phunk" hasn't logged in to this forum since January 2011.
But fortunately I saved an offline copy of this topic a long time ago. So here's the OP with the pictures inserted again and attached to this post, so they should be staying this time regardless of any image hosting site.

Hey guys,

i´ve been visiting this site from time to time and bought a Rigol 1052e lately
due to the blog episode, the many good reviews and other recommendations.
Like most people i am pretty happy with it. One big thing that bothers me though
is the annoying fan noise, that is IMO quite distracting/disturbing in a quite
DIYers paradise.

Long story short, i decided to take the thing apart and replace the fan without
worsing the overall cooling of the scope and i think that worked out very well.

1. You void your Warranty and 120V/240V is lethal, so do this at your own risk

The factory mounted fan that came with my Rigol was a "GLOFN GFA06015H12H".
That is a 60x60mm running with 5500rpm @12V DC. According to the manufacturers
website it has an air flow of 34.2m3/h  and makes 36.5dBA noise, thus no wonder
it´s getting a little louder in the room when powering on the chinese friend...

Replacing the factory fan with just another 60x60mm fan, that has
the same air flow specs wouldnt make much sense IMO, since you probably
wont find one that is significantly quiter because the downside of high air flow is high revolution
relating to the area.

But the people over at Rigol constructed the inner metal case in a very interesting
way. It is exactly 80mm wide and thus could be a perfect housing for a 80x80mm
fan. Infact it cant be mounted flat on the side like the 60mm fan due to the
bend of the case but it doesnt have to since the airflow must go out through
the fan hole...and the shape of the case supports that like a tunnel.







The nice thing about 80mm fans is that they can have equivalent
or more air flow with less rpm. I chose a "Noiseblocker M8-S2" (Germany based company www.noiseblocker.de)
with 47m3/h @1700rpm and 14(!) dBA. The construction features
4 rubber edges to prevent vibrating.



So how do you mount it? Well...simply squeeze it in. Thanks to the
80mm case it fits perfectly tight and is fixed without any screws, hotglue, double sided tape or whatever.
You could do that of course if you feel more comfortable ...








Anything else i did was shortening the cable and soldering a jack fitting the pcb.





If your fan is a little smaller or not totally even, just put some foam
in between (not needed with the NB Fan though)




The result is a considerably more quiet fan with better air flow specs.
Not unhearable but more like a whispering than a hairdryer.
The scope is running fine for hours now. I´ve not measured temperature
so far but its definately not getting hotter than before.

The next step would be a temperature controlled fan regulator.
One could realise it with just a few parts and a NTC mounted on
the heatsink of the LM317 (definately the hottest part in the scope).
That would probably further reduce rpm. IMO Not needed though.

I recorded a comparison with a voice mic i had laying around here.
Nothing normed or so... just 15cm clearance from the mic to the scope.
1st: with noiseblocker
2nd: factory fan
3rd: both crossfaded
http://rapidshare.com/files/436125565/Rigol_Noise_Comparison.mp3

------------------
2 hours running, air temperature measured directly at the outlet 30.5 grad celsius:









Have fun,
Alex

[Drewbie]

Considering the outer case plastic case has a grille of its own, that inner stamped steel grille seems unneeded and looks like it would limit air flow quite significantly.

Before going out and buying a new fan, for a quick fix that might see a reduction in noise level (enough to make it bearable) would be to just cut that metal grille out of the steel case?

Anyone got a sound level meter and a pair of tinsnips handy? 

[alm]

Wouldn't that affect radiated emissions (EMI)?

[hammy]

I replaced the fan with a new one I had bought at Conrad Electronic in germany.

NB-BlackSilentFan XR1
ultra quiet
60x60x25 mm
9-12V
18 m3/h
11dB/A

I also used the "NB-slics" for vibration-free assembly.
No problems since several month.

[Drewbie]

Wouldn't that affect radiated emissions (EMI)?

Good point.

From http://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/ it appears that cutting out the current grill and using a separate wire grill would help noise levels, and maintain EMI protection.

[Woodside]

I made up short photo-story on case opening and fan replacement; it may be valuable for any newbie (like me) who opens this box first time: http://amichalec.net/2011/03/rigol-ds1053e-silenced/

cheers,
-andy.

Thanks for that, very useful.

I replaced my fan with the NB BlackSilentfan XR2 and am really please with the noise reduction.  I used the NB slics for mounting, with thin foam around the fan edges to seal against the case.  I also  taped the holes as discussed by LvdK in your post.

I now see temperatures of 54 deg C directly on the LM317 heatsink (thermocouple poked through the case onto the heatsink) and 28 deg C directly in the fan airflow outside the case (19 deg C ambient).  Unfortunately I forgot to measure the LM317 temperature before doing the mod, so don't know if it has gone up.

Has anyone measured the LM317 heatsink temperature prior to modification?

[hammy]

I replaced my fan with the NB BlackSilentfan XR2 and am really please with the noise reduction.

I see, the X2 is the faster one:
XR-1   / 1600rpm / 11 dBA / 18.0 m3
XR-2   / 2200rpm / 15 dBA / 24.0 m3

I get ~30 deg C airflow temp with the X1 fan. (22C room temp)