|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
178.219.42.74
In Reply to: RE: cMP - the open source high-end Memory Player posted by cics on December 30, 2007 at 05:42:01
I have followed to cics's advice.
My new configuration:
Intel Core i3-530
Scythe Kabuto CPU-Cooler w/o fan
Gigabyte GA-H55M-UD2H rev. 1.0 (rev. 1.0 differs from rev. 1.3 only chip LAN)
Corsair CM3X160C9DHX 1GB (memory remained from an upgrade of the desktop computer).
My BIOS settings:
It is received minimum downclocking of the processor, the graphics core and memory which allows stable cold start of system.
All devices on tab Integrated Peripherals except Onboard 1394 can be Disabled. PCI Latency Tools defines all devices.To receive dedicated interrupt (IRQ) for esi@Juli it is necessary to install it into the far from the processor pci slot, and also to disable all usb ports except two (one to leave for the wireless mouse, another to leave for connection external hdd, flash usb etc. - F_USB2 header on the system board) and SATA controller as is shown in figures.(Excuse me but I have Russian Windows XP).
It is necessary to disable 1394 controller too.
It is possible to leave only three services.
And others screenshots:
The temperature after long listening (temperature in the room of 26 degrees of Celsius):
Also I confirm that the voltage -12V is not required for system board start.
The delay of Power OK is not required for system board start.
Thus it is necessary only three powers (+3.3V, +5V and +12V) for system board operation.
Edits: 09/12/10 09/15/10 09/19/10Follow Ups:
I have Gigabyte GA-H55M-UD2H rev. 1.3.
From your screen shot I can see that you are using BIOS vers. F11.
The F11 version (20/08/2010) I D/L from Gigabyte site is slightly different for my revision motherboard.
It has no Spread Spectrum option, no Advanced Chipset Features, no Boot-Up Num Lock.
I hope such options will be implemented soon.There used to be a hidden menu in the bios, was there?!
ADDENDUM: although I cannot try it right now, the hidden menu is Ctrl + F1
I answered it myself! :-)
Edits: 10/22/10 10/22/10
Your post covers much of the optimizations. That will make it much easier for novices like myself. Your willingness to share is most appreciated!
My first build will be with the 32nm tech'. So I don't have the experience of building and optimizing a machine for CMP 2 at all.
Can you share any differences in implementing minlogon with the new platform that vary from the old platform procedures?
Thanks for your time and consideration
Julien
"There's someone in my head, but it's not me"
Hi Julien! There are no differences in implementing minlogon with the new platform. I use procedure described here http://www.audioasylum.com/forums/pcaudio/messages/4/40869.html.
Pay attention that there is no section PnP/PCI Configuration in BIOS settings of system board Gigabyte GA-H55M-UD2H and consequently there is no possibility to instal PCI Latency Timer (CLK) - [128] (as on an old platform). So it is necessary to install program PCI LatencyTool, to start it and to install value New Latency for an audiocard: 128 (as is shown in a figure):
All the others optimisations remained practically without changes (except those moments about which I wrote earlier).Good luck!
Edits: 10/17/10
Hi Mihaylov,
thanks for all your help.
Have you tried using 256mb of RAM? Did you see any difference in sound quality and/or power consumption? Why have recommended RAM changed from 256mb to 1gig?
I'm curious because I'm currently using a Peaktech linear psu (3A/15V) powering the P4 and the P24 (combined with a picoPSU).
Do you think 3A would be enough to power such a system?
From your power consumption details, 3A looks really borderline...
Thanks
Etienne
I think 256 MB too little. A minimum 512. I use that memory which remained with me from last upgrades of a desktop (1 GB). Unfortunately I did not try 512 MB (it is not easy to buy such memory).
"Do you think 3A would be enough to power such a system?" -
Yes I think that psu 3A/12V is quite enough: power consumption of my cMP2 (P4 and P24) now is 20W. So the margin of power will be 36W-20W=16W. It is quite good.
dear mihaylov,
thanks for all your outstanding contribution on this forum !
you have been keen on promoting the i3-530 CPU and the H55-3USB board several times, and made us understand that your experience is itīs superiority against previous XP based setups.
could you clarify if the XP version was completed with minlogon and all the recommended mods ?
is your current setup (win7 I guess) in any way software modified ?
how much does win7 allow to reduce and/or slim-down ?
windows does an embedded win7 version for industry applications....has that crossed your mind ?
we are eager to get some more information from you pioneers :-)
kind regards
Hi play-mate!
I used and I use now only Windows XP SP2 for my cMP2 Advanced (i.e. completely optimised)!
The current consumption of the CPU (P4, +12V):
standby - 25 mA
start up - 275 mA (max) (first jump - 400 mA)
idle and playback w/o upsampling - 210 mA
playback with upsampling (96 kHz, Tiny, SoX VHQ) - 220 mA
loading of files in memory before playback - 270 mA (max)
The current consumption of the socket P24:
+5V
start up - 2.9 A (max)
idle and playback (all modes) - 2.54 A
+3.3V
all modes - 0.70 A
+12V
start up - 0.26 A (max)
idle and playback (all modes) - 0.21 A
Thus the computer consumes about 20 watt in a playback mode (w/o hdd, lcd monitor and sound card).
Thanks. This is very usefull.
Could you please give me the power consumption of internal LCD screen.
I am following your ATX linear PSU design. I need to work out exactly +5V and +12V module consumption, powering a 1TB WD Caviar Green, a BD-ROM and internal LCD screen.
Bibo,
I may be able to help out here somewhat. When I was considering trying the LT3080 on these voltage rails, I wanted to make sure the chip (which is limited to 1.1A) could handle the current requirements.
I measured the draw of the +12v supply to the internal monitor and that of the +5v to the USB ports (for the monitor's touchscreen connection + my cordless mouse).
I am sorry that I don't remember the exact measurements, but both of these were under 750Ma. My recollection is that the +12v monitor was in the range of 500Ma-700Ma and that the +5v USB was in the range of 250Ma-500Ma.
If these work, great. If you need more precise measurements (and noone else comes up with it), I can measure again.
Greg in Mississippi
Everything matters!
Greg,
Thanks.
I will stick to the idea of internal powering HDD and USB devices only - no BD-ROM.
Unfortunately I do not remember a current consumption of the LCD monitor too. To measure anew it is difficult. Sorry. But I agree with GStew.
I nevertheless insistently recommend to change 1TB WD Caviar Green by 2.5 inch hdd. 3.5 inch hdd consume about 10 W against 1 W at 2.5 inch hdd.
I can easily understand isolating the HDD from the system/MB but once it is powered independently what difference could it possibly make whether it uses ten watts or one?
If one was to power with a battery I could understand this concern. Using AC makes it very hard to understand the motivation.
- 2.5 inch hdd allows to use less powerful PSU which to make qualitative easier, i.e. with smaller level of ripple, which is cheaper and which releases less heat.
- More powerful electric motor in 3.5 inch hdd emits stronger RF noise and an electrical noise into ground bus.
My fear was that you were saying you could hear a difference.
I am relieved!
Rick I was a late comer to the 2.5" hd's. I used 3 3.5" disks and because of the electrical reasons mentioned above by Mihaylov the 2.5" disks put a lower strain on any power supply and yes I heard more purity with 2.5" disks. Just think about the ssd's... they are a further reduced load on the power supply and they definitely sound different (but not necessarily better). They are too 'upper bass' shy sonically in my system. The 2.5" disks are just right for my system balance.
if one is powering the HDDs from the same supply as the computer I can easily see the advantage that would arise from the one requiring the least power BUT if the HDD has its own supply I cannot imagine it could possibly make a difference.
I used to use the originally recommended 2.5" HDD's (were they SAMSUNG? can't remember, when I changed to the WESTERN Caviar Black I heard no difference whatsoever, but that is simply my experience.
Maybe I am lucky but the WESTERN drive makes a little noise when starting but after that is silent. It is placed on a sorbothane pad within my think walled wooden box. Even if it did make noise it would not be heard from the listening sofa.
Just my experience.
Wish I was brave enough to try an SSD for software but I have read too many stories of difficulty with XP and SSD so I will wait.
I measured it to confirm I wasn't going over the current limits of the LT3080 I wanted to try a few weeks back.
As far as hearing a difference, I have here three HDD's (500Mb WD, 640Mb Samsung, 1Tb WD) and will be swapping among them to see if I can hear any differences between them... I also want to try home-made SATA data cables (Ala Peter Daniel... anyone got a source for SATA cables with snap-on ends? I know they are available... but I can't find them!), additional bypass caps at the HDD, screen, and USB connectors, and I now have copies of both XP SP1 & SP2 and need to setup a dual-boot setup so I can compare them.
So many things to try, so little time!
Greg in Mississippi
Later!
Everything matters!
Did you have to change anything in your linear PSU to adapt to the new motherboard?
Was it just a question of swapping new hardware and installing cMP2 again?
Thanks
Any changes in a power unit were not required and that is quite natural. This power unit has a sufficient margin of power on bus +3.3V and +12V, and the current consumption on the bus +5V of the new system board is less than of the previous.
Mihalov,
Are you still using the PicoPSU-based power supply or did you go to a fully-linear one like your associate's?
Has your associate with the fully-linear supply switched to the 32mn tech boards? If so, did he have to do anything special to get his supply to start the new board?
TIA!
Greg in Mississippi
Everything matters!
Greg, in the near future I am going to replace picoPSU with the linear power unit which is giving out voltages +3.3 V and +5 V (as I already wrote the voltage -12V is not required for the system board start and +12 V already is available). It will be the toroidal power transformer of 60 VA with two secondary windings with 6 V on everyone and the printed-circuit board with diodes, capacitors and regulators on heat sinks.
In particular the possibility concerning easy changeover picoPSU by a linear power unit has induced me on platform change.
My friend has not passed to a new platform yet but any changes in his linear power unit is not necessary as I already wrote .
I have replaced picoPSU with the linear PSU as I wanted. But the new system board is not started without delay of signal Power OK (grey wire) (with Pico PSU is OK!). Therefore it was necessary to use a delay circuit.
Thanks for the circuit.
Could you please tell me how you implemented the monitor on/off on the volume knob.
Excuse me but the circuit was not saved unfortunately, and it is difficult to restore it in accuracy since it was changed during customization.
Here is the circuit for monitor off/on by sequential push the volume knob on the front panel.
Edits: 11/01/10
Thanks but there is a problem with your image.
Now it's OK.
Mihalov,Thanks for the info. That sounds like it will be a great-sounding power supply. What regulators do you plan to use?
One thing I have been meaning to ask of someone who has the new motherboard... these new Gigabyte boards have the dual-bios setup like the GA-G31M-UD2H board I've tried. One thing that it did that made the startup a little more complicated was that it started up, then stopped, then started again after about 10 seconds (you could see it do this by the flashing on and off of the LED's on the board and on the PicoPSU).
Does the new H55M board do this too?
TIA!
Greg in Mississippi
Everything matters!
Edits: 09/18/10
I am going to use my favourite regulator LT10xx-adjustable version (LT1084CP) for +5V and LM1085IT-3.3 for +3.3V.
My new system board starts also as well as previous, i.e. pressing the button on the front panel, the pause about 4-5 seconds, LED lights up over the button (an accessing to a hard disk), approximately in 3-4 seconds the screen with POST turn on.
Just to be clear...
Your "revised" plan has:
P4 module:
(+12V) regulator LT1083CP
P24 module:
(+3.3V) - regulator LT1085IT-3.3
(+5V) - regulator LT1084CP
(+12V) - regulator LM1085IT-12
-12V not necessary
Toroidal power transformer 60 VA with ONLY two secondary windings at 6V for +3.3V and +5V.
(Originally you had toroidal power transformer 100 VA with four secondary windings at 5-6V for +3.3V, at 6-8V for +5V, at 12-15V for 12V and 12-15V for -12V.)
Peripherical module:
(+5V at 3A) - regulator LM1084IT-5.0
(+12V at 2A) - regulator LT1083CP
Could you please confirm it.
Eventually, as I am not so technical, let me know exactly where the differences are between the two.
Thanks
Look here http://www.audioasylum.com/forums/pcaudio/messages/7/72756.html and here http://www.audioasylum.com/forums/pcaudio/messages/7/79532.html please.
(I used now the alternative decision of PSU is the so-called hybrid supply, which basis PicoPSU 120, providing all necessary voltages for socket P24 (P20) from entrance voltage +12 V).
This is the photo of my old platform (but the psu remained the same so far).
Edits: 09/22/10 09/22/10
Thanks for your detailed reply.
Sorry I got confused between your system and your friend's uncompromising ATX linear supply.
This is the one I am interested in.
Does the PDF file of the full linear PS at cMP's site still stand or do you have any revision for it?
I do not know yet. It is possible to make the new revision without the voltage -12 V. Nevertheless it is desirable to have practical acknowledgement of working capacity of such version. Though I do not doubt it.
Has forgotten to add that the sound on the new platform much better than on the previous.
I had had my JULI@ in the slot nearest the CPU.
I had noticed the sharing of the interrupt request and had turned off the other devices that were "sharing".
Nonetheless, I moved it to where you said it should be (had to re-install the drivers) and I hope I am not imagining things but it seems to sound better. I was also able finally to get the INTERNAL GRAPHICS CLOCK to 400 without a BIOS failure. But, I cannot imagine that making that much difference, but how would I know?
More than a slight improvement.
Thanks, again,
Rick McInnis
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: