HP MicroServer

DDC · **Joined:** Thu May 05, 2011 1:23 am **Posts:** 1

Hi,

in the article you were mentioned some power draws with different HDD-s. I am interested in the ones with WD Caviar Green:

idle: 25W
Prime95: 37W

Are these power draws are calculated by a program or it was measured by an external power draw tool?

MikeC · **Posted:** Sun May 08, 2011 9:30 am

DDC wrote:

Hi,

in the article you were mentioned some power draws with different HDD-s. I am interested in the ones with WD Caviar Green:

idle: 25W
Prime95: 37W

Are these power draws are calculated by a program or it was measured by an external power draw tool?

All SPCR power measurements are done with external tools. In this case the AC power was measured using an Extech power analyzer of excellent accuracy -- it sells for $5~600.

TigerUK · **Joined:** Fri Jul 01, 2011 5:33 pm **Posts:** 29

The article on quietening the micro server fan was decent enough, it went to great lengths to measure the difference in noise between the stock and scythe fans.

My only gripe with is is that the owner/writer seemed to have missed out on the major noise producer on this device, I believe the PSU fans make far more noise than the case fan.

I think there could be potential to revisit this guide and looking into passive PSUs, obviously for a server which is on 24/7 the biggest concern with passive PSUs is reliability from indefinite use. Can the PSU stay cool throughout and whether or not it has a fail rate at some point.

I've done crude tests, by putting a wooden toothpick through the fan grills into the psu fans [from a turned off position] and powered up the microserver, with an akasa apache fan installed, it was near silent. 5 seconds later pulled out the toothpick from the psu and let the fans spin and they sounded like a jet engine in comparison.

I have a picopsu coming on the way, which I'm looking forward to, and perhaps I'll be able to build up a silent rig.

The downsides of this microserver though is it uses unconventional pin configurations which makes fan speed control impossible using software and you really need a manual one with a knob to vary the RPM.

The akasa Apache fan runs 500rpm which is a constant 49 degres idle, 55 fully loaded. The system isn't going to last long at these temps.

I was wondering if there were any heatsink mods floating around which improve the passive cooling ability of the heatsink.. I can't see how a larger heatsinks could be fitted in there if I'm honest, but never say never.

MikeC · **Posted:** Sat Jul 02, 2011 2:32 pm

TigerUK wrote:

My only gripe with is is that the owner/writer seemed to have missed out on the major noise producer on this device, I believe the PSU fans make far more noise than the case fan.

Perhaps you missed the point I made about the PSU fan: Our sample makes very little acoustic contribution, even after the 120mm fan is replaced or quieted, and especially if all four drive bays are used. The 4 HDDs and the 120mm fan drown out the contribution of the PSU fan. Sure, you could replace the whole PSU but I really can't see the benefit when 4 HDDs are used. If your PSU really is louder than the 120mm fan, then it sounds like your PSU might be a lemon... or maybe HP snuck in a change w/o telling anyone.

HFat · **Posted:** Sat Jul 02, 2011 3:43 pm

I haven't noticed much noise from the PSU fan but perhaps it ramps up and makes a lot noise if it's hot or if you let the PSU heat up like this:

TigerUK wrote:

by putting a wooden toothpick through the fan grills into the psu fans [from a turned off position] and powered up the microserver,

TigerUK wrote:

The akasa Apache fan runs 500rpm which is a constant 49 degres idle, 55 fully loaded. The system isn't going to last long at these temps.

Why? How do you know?
Lots of laptops run hotter than that.

MikeC wrote:

maybe HP snuck in a change w/o telling anyone

HP indeed changed the PSU and let us know by changing the part number of the Microserver and the specification of its PSU. But the old version is still available in some shops. I don't know which version you got. I got mine early after the release so I have the 200W PSU.

TigerUK · **Joined:** Fri Jul 01, 2011 5:33 pm **Posts:** 29

HFAT I just bought my MS recently, perhaps i got a louder one, 200W, hows the noise on yours/?

TigerUK · **Joined:** Fri Jul 01, 2011 5:33 pm **Posts:** 29

HFat wrote:

I haven't noticed much noise from the PSU fan but perhaps it ramps up and makes a lot noise if it's hot or if you let the PSU heat up like this:

TigerUK wrote:

by putting a wooden toothpick through the fan grills into the psu fans [from a turned off position] and powered up the microserver,

TigerUK wrote:

The akasa Apache fan runs 500rpm which is a constant 49 degres idle, 55 fully loaded. The system isn't going to last long at these temps.

Why? How do you know?
Lots of laptops run hotter than that.

MikeC wrote:

maybe HP snuck in a change w/o telling anyone

HP indeed changed the PSU and let us know by changing the part number of the Microserver and the specification of its PSU. But the old version is still available in some shops. I don't know which version you got. I got mine early after the release so I have the 200W PSU.

I may have got a bad egg, I've posted on another forum about my silencing project, but everyone seems to be totally cool with the noise, I'm starting to think now maybe it's just the PSU I have that's noisy, quite a lot of people responded by saying the PSU doesn't make any noise and a handful of people have commented that it's too noisy.

With regards to the first point, yet the PSU fan may rev up to cool more. However the PSS makes noise on idle for me too.

DanaG · **Posted:** Sun Jul 10, 2011 7:51 pm

I just got mine about two weeks ago, and the power supply fan is really loud -- and sometimes, it's actually painful.
viewtopic.php?p=544669#p544669

I made a recording (attached to post above) -- if you stick it in Audacity and do Analyze -> Plot Spectrum, you get nasty spikes at 1899 Hz and 20142 Hz.

Does anyone else get similar tones under high load, or is my PSU bad?

EDIT: The 2kHz spike may merely be an artifact of the 44100kHz sample rate. The 1900Hz one is real, though... and not even constant amplitude.

HFat · **Posted:** Mon Jul 11, 2011 3:48 am

Sure the fan can get noisy under high load but, going by your description, I'd say your PSU has a problem.

I have one in an office where there are five fans which spin regularly: the two in the Microserver, one in a display and two in a couple of printers. Nobody complains about the noise of the Microserver. Most people can't hear it, even when the printers are the display are shut down and even if they're alone in the office. It takes a fairly low noise floor to hear it unless you're next to it. And the PSU fan is not the dominant noise anyway (that would be the drive seeks and the case fans).

Maybe the new PSUs are all like yours but, unless the noise specifications of the Microserver have changed between the two versions, I believe you could argue that it's defective and ask for an older PSU (or an older Microserver) as a replacement. Have you measured the level of the noise and compared that to HP's specifications?

TigerUK · **Joined:** Fri Jul 01, 2011 5:33 pm **Posts:** 29

DanaG wrote:

I just got mine about two weeks ago, and the power supply fan is really loud -- and sometimes, it's actually painful.
viewtopic.php?p=544669#p544669

I made a recording (attached to post above) -- if you stick it in Audacity and do Analyze -> Plot Spectrum, you get nasty spikes at 1899 Hz and 20142 Hz.

Does anyone else get similar tones under high load, or is my PSU bad?

Seems to be some varying degrees of quality components used my the micro-server, mine is loud too, by loud i mean it sounds like an ordinary PC, which IMO is too much for a network computer that's always on, i dont want to notice it at all.

I have my microserver in my room, and sometimes I'm reading a book on my kindle, and I don't want to have that distracting fan noise running in the background.

I don't know why the psu makes so much noise, anyways as i posted up earlier^^ just get a passive psu and sell the stock one on ebay to recoup your costs. my temps went down around 4 degrees celcius, as the passive brick psu is kept outside the case.

I took out all noise makers, the case fan, the psu, and even the stock hard drive made a lot of noise for me, it had a high pitched noise, that would have probably given my tinnitus, took it out and put in 1x500GB WG green caviar, 2X2TB WD green caviar.

Sure there is a bit of sound when one of the larger capacity drives are being written on, but that's just for storage backup, and is rarely used. the 500GB WD green is silent as it can get, i have to put my ear right next to the case in order to hear anything from it.

mjzglr · **Joined:** Tue Jul 19, 2011 3:24 pm **Posts:** 1

What do you think the power draw differences would be between the 200W and 150W model?

I'm curious to see what the efficiency spectrum looks like for low load 25-50W on each model? Do you guys have any of these numbers or a good guess?

MikeC · **Posted:** Wed Jul 27, 2011 8:54 am

mjzglr wrote:

What do you think the power draw differences would be between the 200W and 150W model?

I'm curious to see what the efficiency spectrum looks like for low load 25-50W on each model? Do you guys have any of these numbers or a good guess?

Probably next to nothing. I'd guess that sample variance between two 150W models would make as much difference as between a 150W and a 200W.

cjmccarthy72 · **Joined:** Wed Sep 14, 2011 2:25 am **Posts:** 1

I also have issues with the noise of the HP PSU and fan. I would put it on a par with the standard low power office desktops I find myself using in different workplaces. So not terrible but certainly not the silent machines the reviews have been talking about. It is certainly audible 5 metres away when I reading with all appliances off at night. The PSU is nosier than the 120mm fan and the hard drives. So perhaps I should ask HP for a new PSU. Does anyone have a passive solution working?

HFat · **Posted:** Sat Sep 24, 2011 4:39 am

FYI, I should have an opportunity to compare the latest update of the Microserver (with a slightly better CPU) with the original version next week. If there's anything in particular you want me to look for or to test, do ask.

DanaG · **Posted:** Sat Sep 24, 2011 8:49 am

Oh yeah, I had also asked about the PSU in the other Microserver thread, and they suggested that there must be something wrong with it. Not long after that, I had HP replace the PSU, and now it's merely annoying instead of "oh god my ears".
I'd be curious to see a comparison, with your equipment, between the old and new PSUs.

I also did see the news of the update, but 200MHz difference is not worth fretting over.
I'm curious whether the chipset and the BIOS changed at all, though. (I do have various gripes about the BIOS in the thing.)

I think I probably should get a PicoPSU for the thing. Does anyone know offhand if the wide-input ones will fit? I intend to hack an old HP 19-volt, 90-watt brick into it.

HFat · **Posted:** Sat Sep 24, 2011 9:19 am

I'm afraid I don't have equipment sensitive enough to quantify the noise of the original Microserver. But I can give you my impressions.

DanaG wrote:

I also did see the news of the update, but 200MHz difference is not worth fretting over.
I'm curious whether the chipset and the BIOS changed at all, though. (I do have various gripes about the BIOS in the thing.)

Exactly. The eSATA port is one thing to watch in particular. That, the PSU and the fan controller of course.

HFat · **Posted:** Tue Oct 11, 2011 9:41 am

As it turns out, I didn't have the right cable to test eSATA hotplugging.

Otherwise, while the chipset seems to be different (according to lspci), both versions are basically the same computer with the same BIOS.

Comparing a new N40L with an orginial Microserver (1 year of usage), there's no big difference in noise.
The new PSU seems a bit less noisy at idle while the new case fan seems slighty more noisy (but what's most obvious is a difference in tone which might be giving me the wrong impression regarding the physical sound level).
Especially with the new unit, the PSU noise is not going to matter if you don't mod the case fan. Except at some frequencies you normally wouldn't hear unless you're next to the unit, the case fan's noise overwhelms that of the PSU.

edit: according to my UPS, idle power consumption is about 6W higher but I don't trust it. It measured a significantly lower idle power consumption than SPCR with the original unit. Then again I'm not sure I got the same version as SPCR and I think we have different main voltages.

DanaG · **Posted:** Sat Oct 29, 2011 7:54 pm

Interesting... I had to get my N36L motherboard warranty-replaced (bricked BIOS), and the part I received was announced as a substitution: the N40L motherboard.

I don't remember what "lspci -nn" showed before, but it seems to be essentially the same chipset. Can somebody can tell me what's different?

Code:

00.0 Host bridge [0600]: Advanced Micro Devices [AMD] RS880 Host Bridge [1022:9601]
01.0 PCI bridge [0604]: Hewlett-Packard Company Device [103c:9602]
04.0 PCI bridge [0604]: Advanced Micro Devices [AMD] RS780 PCI to PCI bridge (PCIE port 0) [1022:9604]
06.0 PCI bridge [0604]: Advanced Micro Devices [AMD] RS780 PCI to PCI bridge (PCIE port 2) [1022:9606]
11.0 SATA controller [0106]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 SATA Controller [AHCI mode] [1002:4391] (rev 40)
12.0 USB Controller [0c03]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 USB OHCI0 Controller [1002:4397]
12.2 USB Controller [0c03]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 USB EHCI Controller [1002:4396]
13.0 USB Controller [0c03]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 USB OHCI0 Controller [1002:4397]
13.2 USB Controller [0c03]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 USB EHCI Controller [1002:4396]
14.0 SMBus [0c05]: ATI Technologies Inc SBx00 SMBus Controller [1002:4385] (rev 42)
14.1 IDE interface [0101]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 IDE Controller [1002:439c] (rev 40)
14.3 ISA bridge [0601]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 LPC host controller [1002:439d] (rev 40)
14.4 PCI bridge [0604]: ATI Technologies Inc SBx00 PCI to PCI Bridge [1002:4384] (rev 40)
16.0 USB Controller [0c03]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 USB OHCI0 Controller [1002:4397]
16.2 USB Controller [0c03]: ATI Technologies Inc SB7x0/SB8x0/SB9x0 USB EHCI Controller [1002:4396]
18.0 Host bridge [0600]: Advanced Micro Devices [AMD] Family 10h Processor HyperTransport Configuration [1022:1200]
18.1 Host bridge [0600]: Advanced Micro Devices [AMD] Family 10h Processor Address Map [1022:1201]
18.2 Host bridge [0600]: Advanced Micro Devices [AMD] Family 10h Processor DRAM Controller [1022:1202]
18.3 Host bridge [0600]: Advanced Micro Devices [AMD] Family 10h Processor Miscellaneous Control [1022:1203]
18.4 Host bridge [0600]: Advanced Micro Devices [AMD] Family 10h Processor Link Control [1022:1204]
05.0 VGA compatible controller [0300]: ATI Technologies Inc M880G [Mobility Radeon HD 4200] [1002:9712]
00.0 Ethernet controller [0200]: Broadcom Corporation NetXtreme BCM5723 Gigabit Ethernet PCIe [14e4:165b] (rev 10)

I did a couple of things to make the thing more tolerable:

Got a PicoPSU. The 120W wide-input one fits fine in-place on the motherboard. It just needs a 4-pin "P4" power cable extension, and a bit of electrical tape to protect the top edge. I'm using a 19V laptop brick with it.
Turned around the rear fan, but left it on the inside of the case; reattached the metal fan guard to suit.
Got a 120mm "washable fan filter" (from FrozenCPU), to keep out dust. I used the original torx fan screws (much better than Philips!) to attach it. It seems to work fine on top of the bulging metal grill.
Stuffed a laptop-sized chunk of foam (tear-out from an HP laptop service box) in the optical drive bay, to force the air to go between the hard drives.

Now I'm just left with a quiet, but annoying, clicking/grinding sound from the fan. I'd assume that's the bearings. That's much better than the "oh god my ears" I got from the PSU fan.

HFat · **Posted:** Sat Oct 29, 2011 11:28 pm

DanaG wrote:

RS880

It used to say RS780 instead.

Strange that some of you seem to have noisy PSUs. Many of us have relatively quiet PSUs. Maybe the electricity you feed it is a factor...

Monkeh16 · **Posted:** Sun Oct 30, 2011 8:47 am

HFat wrote:

DanaG wrote:

RS880

It used to say RS780 instead.

Nah it didn't.

There's no difference except the CPU.

HFat · **Posted:** Mon Oct 31, 2011 9:21 am

Why would I say it did if it didn't? Some people...
http://www.google.com/search?q=microser ... +bridge%22

Monkeh16 · **Posted:** Mon Oct 31, 2011 10:10 am

If you'll look closer they all say 'RS780 Host Bridge Alternate', which was the incorrect name given to 1022:9601. If you check the lspci output above, you'll see 1022:9601 as RS880.

Hint: The name is stored in a regularly incorrect and regularly updated text file.

So I suppose you were in a way correct, but so was I: Nothing has changed.

I would be hugely surprised if HP used two different chipsets for one product without any obvious distinction between hardware revisions being made, as the BIOS would not be compatible.

HFat · **Posted:** Mon Oct 31, 2011 10:21 am

I was hoping they had a fix for the eSATA hotpluging issue I hadn't had an opportunity to test. If it's the same BIOS and not just a BIOS with the same user-configurable options, I guess I shouldn't waste my time...

If nothing else the PSU is also not the one which came with the original version. Is yours noisy by the way?

Monkeh16 · **Posted:** Mon Oct 31, 2011 10:23 am

HFat wrote:

I was hoping they had a fix for the eSATA hotpluging issue I hadn't had an opportunity to test issue with a slightly different chipset. If it's the same BIOS and not just ones with the same user-configurable options, I guess I shouldn't waste my time testing that.

If nothing else the PSU is also not the one which came with the original version. Is yours noisy? None of mine are.

You can just flash a modified BIOS to enable AHCI on all ports. I did.

I'd love to tell you if the PSU is noisy, but it lives in my cupboard, along with three other boxes, two eSATA HDD enclosures, a heavily modified AT PSU with a dodgy fan, and a 24 port switch with a pair of 40mm fans. So I can't really hear it.

ukjim · **Joined:** Mon Nov 14, 2011 2:24 pm **Posts:** 1

Hi great info on this thread and site!

I have just purchased the N40l if I understand correctly the bios is the same as the N36l?

I should be able to use one of the modded bioses to enable full speed sata ports on this motherboard?

thanks for any info.

regards

HFat · **Posted:** Tue Nov 15, 2011 9:20 am

I have not tested the modded BIOS nor do I intend to.
But I thank Monkeh16 for letting me know they exist and that they allow you to fix the issues with the SATA ports other than the 4 main ones. I was curious but the 4 main ones are all I need really (another device on the network can pick up from there) and I don't want to do something unsupported and potentially non-reversible to a production server under warranty...

The 4 main SATA ports if they are not full speed (I don't know) are fast enough for most purposes with the supported BIOS and I doubt an alternative BIOS would make them faster.

Monkeh16 · **Posted:** Tue Nov 15, 2011 3:46 pm

It has nothing to do with speed, it has to do with having a usable driver for the two extra ports.

The flash is safe and entirely reversible (if it fails so badly you can't reflash it, how the hell are they going to find out you've flashed it?).

HFat · **Posted:** Tue Nov 15, 2011 3:56 pm

Monkeh16 wrote:

(if it fails so badly you can't reflash it, how the hell are they going to find out you've flashed it?).

I don't think they'd care. But I don't think I'm supposed to do that and I don't think it would be honest to claim a warranty afterwards. Maybe I got the wrong impression...
I'm not above being dishonest when the situation warrants it but there's no justification in this case.

Monkeh16 · **Posted:** Tue Nov 15, 2011 7:49 pm

HFat wrote:

Monkeh16 wrote:

(if it fails so badly you can't reflash it, how the hell are they going to find out you've flashed it?).

I don't think they'd care. But I don't think I'm supposed to do that and I don't think it would be honest to claim a warranty afterwards. Maybe I got the wrong impression...
I'm not above being dishonest when the situation warrants it but there's no justification in this case.

If the flash didn't cause the failure, there's nothing dishonest about claiming on the warranty.

DanaG · **Posted:** Tue Nov 15, 2011 10:21 pm

I only managed to brick mine by messing up the batch file, and then apparently mistyping the flash command manually.
When you get it right (as the batch file does), you'll see only a very small number of blocks actually marked 'p' for 'programmed' -- everything else is unchanged.

There are also a bunch of other interesting-looking hidden settings that you can unhide with AMIBCP... but I didn't dare mess with the ones where I had no idea what they did.

One thing I wish HP would let you choose: leave IPMI card enabled as boot device, but also enable the Radeon.

HP MicroServer

Who is online