You should never, ever touch the BIOS with a tool like that unless you're absolutely certain what you're doing with it.

But how do I know that? My problem with flashing an unsupported BIOS is that I don't know how to assess the risks. Maybe it's essentially zero but I don't know.

They are essentially zero, yes. If the flash succeeds and the system boots normally, a BIOS modified in this way isn't going to somehow cause your computer to go up in flames somewhere down the line.

Thank You for the 120mm Fan Guide,

I recently bought a MicroServer and while the stock 120mm fan is quiet, i can still hear it, i bought a new 120mm PWM Fan and i could not for the life of my figure out why it wasnt working.

I almost sent it back for a refund until someone linked me to this guide, i will be trying the method later on!

Thanks!

Maybe a stupid question, but would a 3-pin fan work if i remove the guides on fan connector ? What i mean that 3-pin FAN has pins GND, +12V and Sense, while 4-pin on HP Microserver has Control, GND, +12V, Sense. If i don't connect the Control to the fan (has no use for it) and i connect the remaining 3 pins using the standard fan header with removed guides, the 3-pin fan should work and unless i slow down the fan using a Fanmate or another fan controller bellow the 500RPM limit it should work, no ?

UPDATE: So yeah, 3-pin fan connected to the front 3 pins of the 4-pin connector works fine, system doesn't shut down. Of course i guess the same minimum RPM limit is valid for 3-pin fans as well, but this brings up much more options, as the only thing you need to do with 3-pin fan is to get a 3-pin fan cable extension, cut the guides on the extension which allows you connect it to the microserver fan header - and then you can put a fan controller or any 3-pin fan to the other end of that cable.

Hi, I recently bought one of these and was wondering if anyone had tried connecting the 250GB drive that comes with the system into the SATA port on the m/b that is normally used for the optical drive? Will the BIOS cope with seeing a HDD on this port rather than an ODD?

I think the HDD could probably be mounted in the space where the ODD would normally go.

I am thinking about doing this so that I can free up the 1st drive bay for 4 x 1TB or 4 x 2TB drives.

Any thoughts?

Cheers,

Paul

I bought one of the N36 versions and have installed Freenas. I upgraded the RAM to 8GB primarily because the ZFS filesystem uses loads of RAM.

The reason I chose Freenas was because of the robustness and uncorruptability of the ZFS filesystem. Furthermore, Freenas can be installed on a USB key which is placed in the USB slot inside the chassis next to the drives. This allows all four drives to be used solely for data.

With Freenas, it's recommended that you don't use any form of hardware RAID, but instead rely upon the more robust RAID-Z configuration implemented in software.

I use this unit mainly for backup. Freenas can act not only as a NAS, but as a SAN because it supports ISCSI so that client machines can attach a drive as if it is directly attached. Sustained backup speeds from a Windows 7 client are 20MB/sec when configured as a NAS (using SMB/CIFS) and 40Mb/sec when configured as a SAN. This is using 4 x WD EARS 2TB drives configured as a single RAID-Z volume.

Hi,

I hope you can help me or make a suggestion or 2.

I bought one of these units and was very excited initially but that excitement is waning rapidly.

I have added the following:

Replaced 1GB DDR3 Ram with 2 X 4GB DDR3 RAM (purely because I got it stink cheap)
Added low profile Asus Radeon Graphics Card HD5450 - 512MB 64bit GDDR2 - PCI-E
Added Liteon Blu-ray Reader Combo Drive
Added Samsung EcoGreen F4 2TB 32MB Cache SATA hard drive
Added Seagate 2TB Hard Drive (don't have more detail with me)

Problem:
With these added the server won't even start up.

Steps taken:
Disconnected Bluray Drive - still won't start.
Remove Seagate hard drive - server starts.

The graphics card supposedly consumes very low power so this then means the server cannot handle a 3rd and 4th drive.
If this is correct it is a bit ridiculous seeing as it has 4 drive bays unless one is only supposed to use small hard drives.

Is it possible to replace the power supply with something more realistic?

Any suggestions would be appreciated.

Thanks

Marius

An ATI HD5450 may be low power compared to other graphics cards but it still uses 150 watts at idle. I have my doubts that it really makes sense to use anything other than the built in graphics with the HP MicroServer.

Jack

If you just want to install a graphics card so you can add an extra monitor - and not for gaming type use there are some cards that you could use.

Matrox P690 Plus LP is 11 watts, fanless and is supposed to be able to drive two monitors

150W? Surely a typo.

I don't see why the power supply couldn't support this configuration.
I don't know what was meant by "start" exactly but in any case the "steps taken" above are not sufficient to establish that the power supply is defective.
If it's defective, HP should replace it.

That seems quite way too high. Typo? I recall <10W idle. Still, I agree that....

Also, mariusjk should provide clearer details of what's happening with his setup. It may have nothing to do w/ the PSU, even the 150W version should be enough. Maybe the MB doesn't like the RAM.

Should be enough ? I am not so sure. Spinup current of the Ecogreen F2's and Barracuda Greens is 2A@12V (unfortunately there is no datasheet for F4), so the 5 hard drives use 10A@12V at spin up. That itself is 120W at system startup. Then you need to ad ~10W for the graphics card, and we are already at 130W. Now considering the 30-50W consumption of the board itself (and you won't have an idle load at startup), i wouldn't be sure that the 150W PSU is enough for that. What he needs is probably the staggered spin-up, but i am not sure how to do it in the microserver, if possible at all.

AFAIK, mariusjk added 2 drives. The Bluray drive should not draw any power at startup.

So with the new RAM, video card, bluray & only the original HDDs, does the machine boot? If so, pull the video card & try adding the other drives, one by one?

My mistake then, i somehow did read 4 Samsung F2's .

i had that, the psu that it came with (1U) was quite noisy, I replaced it with a 160w picopsu.. I would have got a cheaper 90w version had I known the true usage (35w).

I would recommend a pico rather than an internal 1u psu, because I'd imagine a passive psu will get hot and raise the internal terms.

Which will require faster fan rpm to cool the internal.

My Microserver boots fine with four 5400RPM drives, one 7200RPM 2.5" drive, and two extra SATA controllers.

Hi All,

Thank you for the comments.

Turns out the Seagate drive I put in is faulty. Server starts up with Bluray, 1 Samsung 2TB and internal 25GB drive and graphics card.

Sadly the drive is less than a year old but I don't have a warranty - can't find my invoice.

Marius

You do not need an invoice. Just feed Seagate's site the serial number and they'll deal with it.

I just got one of these and can't reproduce the review's results at all. Mine is far, far louder than the review sample and the best idle power consumption I can get is 45W. That is with a 2.5" Western Digital HDD and the 2GB of RAM it came with. The stock Seagate 250GB HDD adds another couple of watts.

I'm debating what to do, send it back or keep it. Average power consumption at the Windows 7 desktop with nothing else attached (no network, keyboard, monitor etc) is about 50W. Twice what SPCR managed with two drives attached. Even with no drives it idles at 45W in the BIOS.

I get the feeling that once again SPCR has been given a special review sample with extra quiet fans and extra efficient PSU. All I can think of doing is shelling out more cash or a PicoPSU.

This is so unlikely that I doubt it has ever happened... well maybe at some point in the distant past with a much smaller company.

What I would suggest is to examine your machine carefully to check whether it is the same model as our sample. There has been a change to a 150W PSU (from the 200W which the Microserver began with) -- it's possible that is noisier than the original. Are there any other differences in your machine?

There are only 2 fans. The little one in the PSU should be much quieter than the main 120mm fan. Can you monitor the latter's speed? I measured 1080rpm.

You say you get 45W at idle -- this is almost the same as the 46W I measured with Prim95 (and 2 drives). What's your AC meter? Are you sure of its accuracy? I measured 34W at idle with the two supplied 250gb Seagate 3.5" -- no way you should get 11W higher w/ a single 2.5" drive. Check the CPU load. The CPU & the board just can't draw an extra 11W out of the blue, there's got to be some other process causing the extra draw.

The PSU is a 200W model, but of course it could be different inside. I'm on 240V as well so it should be more efficient if anything. Mine came with 2GB of RAM and 1x 250GB Seagate HDD but appears to be the same model mobo and CPU wise.

My meter is accurate. I checked it with a couple of lightbulbs just now, but previously I have tested it with a true RMS Fluke multimeter and found it to be within 1-2W of the expected value. I know you use some quite expensive equipment, which you have calibrated presumably... Even so you got 25W at idle, nearly half what I measure, which is too much to explain that way.

CPU usage is 0-2%, basically idling at the desktop. Standard Windows 7 install with disk defragmentation disabled and all power saving settings on max. 120mm fan runs at 1020 RPM.

MoJo --

It's inexplicable to me how the power could be so different. It is the same CPU, motherboard & RAM -- how could it vary so much? Sample variance cannot explain it, and even if an entirely different board was used, with the same chipset & onboard components, power simply cannot differ that much. I'd still examine the setup in detail.

If your fan is running at 1020rpm, that's slightly slower than our sample -- and its noise signature should not be that different, because this fan is the loudest thing in the box -- aside from the Seagate HDDs. Perhaps you are too optimistic about how quiet 23 dBA @ 1m is -- this is what I measured of the stock Microserver w/o any drives. I wrote then --"The overall noise level is modest, and the noise quality is not bad. Though it has distinct tonal aspects and is several decibels louder than the noisiest PC in the lab (~16 dBA@1m) we could live with the noise. "

But we'd prefer not to, which is why I modded it: http://www.silentpcreview.com/article1193-page7.html It is working now with 4 5400rpm 2TB drives (a mix of WD and Samsung) as our central server in a wall-shelf in a closet. It sits atop a 2" thick pad of foam to minimize HDD noise transfer into the wall, which keeps it from being audible once the closet door is closed.

The fan noise is much more than that of the HDD. That isn't all that surprising to me, reading many threads on many forums a lot of people say the fan isn't particularly quiet. I'm not really worried about it though, power consumption is my concern.

Tomorrow I'll see if I can try a DC-DC PSU. As you say, if the mobo and CPU are the same then the PSU must be the variable.

My N40L with 4xWD10EADS and 8GB RAM is idling at 48W from socket at BIOS boot screen (no OS installed right now).

Looking at your location you are on 230V, right?

Yes, 230V. N40L, P/N: 66447-425 :
http://h20195.www2.hp.com/v2/GetPDF.aspx/c02511742.pdf



As i said, i added another 4GB DDR3 ECC module from Kingston and 4xWD10EADS.

I took the DC PSU out of my Atom system which idles at 22W. In the Microserver it idles at 28W in the BIOS with no HDDs (so obviously can't boot up). The HDD makes about 5-6W difference so we can say about 34W with it, and then drop down 8-10W once booted due to power saving in Windows. As it happens that is around 25W.

HP are terrible for these kinds of antics. Ship something really good at first so all the reviews look wondering, then silently switch to something terrible and hope no-one will notice. I'm debating if I should send it back or not, just on principal. I suppose the difference only equates to about £2.50/month in electricity, but even so... Combined with the poor fan I'm really not impressed.

Edit: This is a more accurate representation of the noise: http://www.youtube.com/watch?v=8RtzRtUvtLQ

Edit2: Installed a Scythe 120mm fan. It is a 3 wire connector, just had to remote one of the posts to make it fit. Didn't need to remote the old fan wiring, just put it in the 5.25" bay for now. The new fan runs at 1100 RPM and is much, much quieter. I tried reducing the speed with a Zalman Fanmate but even on the highest setting the BIOS wasn't happy. It doesn't really matter as the 120mm fan is now no longer the loudest thing in the case, the PSU fan is.

With a PicoPSU this thing might actually be a reasonable low power low noise server.

If you got the noise like in the video, then your fan is running at 100% PWM, which means your temperature sensor is failed. You should send the microserver it back and request another.

The temperature sensor is working, the fan kicks in at 100% and then drops down to around 1100 RPM and stays there.

There is no way that even the PSU fan on its own is anywhere near the 22dB HP claimed.