|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
67.60.98.130
In Reply to: RE: P4 power supply using the BELLESON regulator and mouse problem posted by rickmcinnis@dogwoodfabrics.com on December 30, 2010 at 08:54:16
Rick,
Very much congratulations. Glad to hear it made such a significant improvement for you.
I'll write more on the power stuff later, but on the mouse, it should come back ok when you use the un-modified cable. If it does not, of course you can try turning one of the rear-panel ones again to confirm that it's not a mouse or motherboard USB header issue. I regularly pull my mouse when I have a new version of cPlay or just a recording or two to transfer and plug a USB stick there. I do have to boot with the stick in or it's not recognized... and when I'm done, I just put the mouse dongle back in & reboot & it's there. Also, you don't have the USB Mouse support disabled in the BIOS, do you?
Again, congrats!
Greg in Mississippi
there is also a setting f
Everything matters!
Follow Ups:
I went to FRYS and got an "unmolested" USB/MB connector - but only the standard mouse works, not the "remote" mouse.
Last time this happened I had to leave MINLOGON to get it to work. Once it worked I returned to MINLOGON.
What is weird is that my supply is definitely putting out the required 5 volts. I checked with a mouse plugged is (out of the computer circut completely)and there is no change in voltage (supply is up to the task). I had the same problem with the GRANITE supply - never could get it to work but when I went to the second ATX supply I had no problem using a 5 volts rail to power the MB/USB connector.
Does one need to have the ground from the supply/USB adaptor connected to one of the grounds of the MB? I have not tried this yet but it is the only thing I can imagine that would be the difference. I think I am the only one to have this problem. It is not confined to this MB, I could not get the GRANITE to work with the previous recommended MB.
I really am surprised how much better the linear supply sounds with P4. Of course, that is easy to implement. It will be far more exciting if I can the thing to turn on with the linear P24!.
Have you figured out what the mysterious (since he has disappeared) Jack Wong was doing with the memory supply? Was he powering the memory outside of the MB? I could never quite figure it out. And his post was not exactly clear about that. Just curious.
Thanks for the advice.
Bye,
Rick McInnis
I assume you are powering the mouse with a linear supply (I assume in conjunction with the HDD) - did you have to do something to ground it to the MB?
Interesting what's happening on your mouse. I had to go look how I'd done it... heck, I'd set that up almost 2 years ago! I did mine just as described in the cMP documentation for using the Granite Digital SMPS's to power the USB header... cut the black & red wires, tie-in a 4-pin Molex plug to those wires going to USB recepticles, and plug the power supply, GD or my linear, into that. The black & red wires that go to the plug that fits on the motherboard header are not connected. But now that I type this, I realized that I didn't check if it was tied to ground... I checked, I hadn't done it on purpose, but I did tie the shield on the cable to the ground from the Molex plug and that gets tied to ground through the USB recepticle mounting screws (Hmmmm there's probably some opportunity for better grounding paths there).
To answer your question, yes, the ground on my 5v linear supply for the USB is tied to the motherboard ground. So it wouldn't hurt for you to do this & might fix your USB mouse problem.
Oh, and of course, the likely reason it worked with the spare ATX supply is that it was grounded through the P4 connector.
BTW, a quick experiment to try... plug your spare ATX into the AC without it being plugged into the P4 & the USB header... I'm curious if you'll hear a difference just due to having it plugged into your AC.
On the missing, lamented JackWong & memory power, yes, he was supplying the 1.35v/160ma for the memory from a single, large-capacity (11000 mAh) NiMH battery. He also supplied the 0.65v/120ma for the graphic core this way with a variable resistor to drop the voltage from the battery. I figure the memory supply is worth doing someday (since he said it made the larger impact), but likely pretty far down the road... and I'd like to do it with a linear supply, not a battery to avoid the battery housekeeping issues.
I hope this helps with your mouse issues!
Later!
Greg in Minneapolis
Everything matters!
I have not, as yet, tied the supply to the MB's ground, but that has to be the answer.
I will let it be known when I get it to work.
One funny thing though about using the ATX ground for the USB - while this was what I was using my BIOS was very unstable. All would be fine for days and then the default BIOS would return. When one attempted to re-set the BIOS you were given the RED SCREEN saying that I was doing something the BIOS did not like. Usually a simple shorting of the pins would not do the trick. The battery had to be removed for a couple of hours.
Does indicate that grounding (no surprise) is very important for stability. Wish we could get Dave Davenport interested in this but I have tried and he thinks we are all insane to use computers. Well, maybe not insane, he does not think it a good direction. YET, he loves the PS AUDIO Perfect Wave CD system which is very much like one of those dreaded computers!
He is an expert on grounding. Wish he was interested.
You are correct that there is probably much to be gained by some kind of grounding "tricks".
Thanks again for the advice.
Bye,
Rick McInnis
Keep working on Davenport... it'd be great to have his insight.
BTW, are you using his I2S-interface? Very curious how that works for people.
Greg in Mississippi
Everything matters!
That was really driving me crazy.
Of course, it is/was obvious.
Before I was using a separate GRANITE supply for the HDD and I guess that explains why I was never able to get this to work then.
It will be interesting to see if this eliminates the BIOS instability.
Thanks, again,
Welcome... and same to you!
Your BIOS instability does not surprise me. I noticed something similar on stability when I went to the fully-linear supply from the Pico-PSU-based linear-hybrid one... I could run a step or two lower in voltage.
Also on the grounding, when I was tracing out the USB power connection last night, I noticed that the ground pin on the USB cable side of the Molex plug had pulled out (I HATE those connectors... I bet there's a good sonic upgrade available in replacing them!) and was not making contact. I pushed it back in and made sure my wire dressing and component placement (especially HDD) was right... and when I re-booted the system, the sonic 'robustness' through the lower midrange & sense of dynamics & 'bounciness' that had been missing for a couple of weeks and that I assumed was due to Alan Maher CBT's & AC filter chokes breaking in came back. They were (and still are) breaking in... but it was not just that.
I definitely need to setup better ground routing on my 'dirty' connections... I did it very haphazardly back then.
Very glad that the grounding fixed your USB mouse issue.
Later!
Greg in Mississippi
Everything matters!
I have configured my pc to use cMP as per the recommendations on the site. Implied Minlogon as well. But since last evening, none of the USB ports work. If I connect my USB hdd, it gets powered but does not appear in "My Computer", nor does the wireless mouse work. What has gone wrong???
Read, please, more attentively "Step 4 – disable unused USB hubs & controllers" here http://www.cicsmemoryplayer.com/index.php?n=CMP.07Optimisations. Probably you disabled the necessary usb ports.
http://cmp2-mihaylov.narod.ru/
No. All USB ports are enabled. When I connect my USB hard disk it powers up, but it does not get recognised in windows. So does the wireless mouse. They worked fine just a day earlier...
When I connect my USB hard disk, it powers up but it does not get recognised in windows.
If all the USB ports are definitely enabled, you could try re-enabling the Logical Disk Management service (assuming of course you have disabled it) and taking a look at what it reports. If LDM sees the drive but Windows in general doesn't, it's a known Windows bug (that will certainly, definitely be fixed in SP106).
Change the drive letter [Action > All Tasks > Change Drive Letters] and try again. I get this all the time when hooking backup drives in and out of various boxes.
If that doesn't work (and to help retrieve your mouse), go into Device Manager and see if the devices are recognised there. At least until you retrieve the thing, I'd be minded to use a wired mouse.
You should see the drive under Disk drives , the mouse under Mice and other pointing devices and both under Universal Serial Bus controllers .
They worked fine just a day earlier...
That's probably why you didn't post a day earlier . . . :)
BTW, have you checked that the HDD works fine on a different computer?? Especially as it's bus-powered, I'd be minded to check. Bus-powered drives can be moody.
HTH
Dave
Yes the hdd works perfectly fine on my other computers. I will do as u said & post back...
Just connected a USB pendrive to the port ( flash drive). It performs normally, but the USB hdd does not. The USB root hub has reserved only 200ma for each port. Seems too low I guess. Can this be the cause??
Sounds like you're providing several new wrinkles that will benefit us all.First, thanks for sharing the info on the new Belleson 2A regulators. I hadn't seen them yet. They look to be THE current contender for DIY supply regulators... I'd even look at their 2A regulator if I were doing another linear/hybrid supply for the main 12v rail. I strongly suspect they are a better choice than the AMB Sigma... and based on your comment about the high-current 5v Salas, it may not be suitable (BTW, can you point to any web references or discussions on this aspect of the Salas? I'd like to see what can be done to make the voltage level more consistent).
You know, as I ponder it more, I think I need to put an order in for a couple of 12v & a 5v 2A regulator from them.
On voltages, earlier this summer, I had been in communication with a lurking-inmate from Oz who had a fully linear supply built. He had pots to adjust the voltages and found (similar to other postings here) that a slightly lower voltage was best sonically. He didn't get around to reporting on the 3.3v rail, but found the sweet voltages for his setup were 4.6v & 11.8v for the ATX connection.
BTW, he used the John Swenson choke-filtered supply for his basic raw DC supplies and was very happy with the results. Very interested in what you're doing for the choke-filtering for your 5v raw supply if you're willing to share it.
Those are the voltages I settled on based on his suggestions.
And I'd agree that using an LT1086-3.3v is likely your best bang-for-the-bux for the 3.3v rail. Sorry, tho, I haven't taken the time to put in my LT3080-based regulator on that rail, so I don't have a definitive opinion on that.
As for heatsinks on the rectifiers for the 5v raw supply, I totally agree with that. Those devices are the hot-spots of my fully-linear supply and putting heatsinks on them was one of the first things I did when trying to resolve my early issues that I thought were related to overheating (but were not). I'd also suggest heatsinks for the other regulators too, especially if you have a fairly high input-to-output voltage differential on them. All of my regulators (all LT1083 or LT3080 based at this time) run from moderately-warm to hot, but I do have fairly high differentials on all of them.
On powering the HDD, if you have a 2.5" SATA drive, they only need the 5v rail. I believe that 3.5" drives need both the 5v & 12v, but I cannot confirm that... it may have changed with the newer 'Green' drives. But I strongly suspect that the best sound quality will be had by using an entirely separate supply (or pair of them) for the HDD instead of sharing it with the USB 5v supply.... the drive(s) seem to be electrical-noise hot-spot(s) in our cMPs.
Finally, my experience with the sonic benefits parallel yours. 'Dense' is the adjective that comes to mind... lot of details, lot of textures, all presented in an easy-to-listen-to, but musical fashion. But I bet your Belleson-based supplies with the choke-filtering on the 5v raw supply might be the best sounding implementation out there currently.
Thanks... And Happy New Year too!
Greg in Mississippi
P.S. I have a few tweak comments too... will put them into the EVS Ground Enhancer tweak sub-thread you, Theob, & Playmate discussed awhile back.
Everything matters!
Edits: 12/31/10 12/31/10
Using the HAMMOND 100mH 5A (195T5)choke between two caps - nothing clever.
Then on to the DANIELs PCB. I am able to use the PCB on the heatsink I had laying around which is absurdly oversized but better safe than sorry.
I am stealing Swenson's rectifier filter and using his recommended rectifiers for the P24 supplies. For the other supplies I used FAIRCHILD "Stealth" rectifiers.
BELLESON has led me to believe that the 2A reg will be able to do the job with a large heatsink. I am using the latest recommended MB which I hope will allow me to get away with this regulator. I can always plug in the ATX for outside of cMP duty like adding music files.
I am using heatsinks for the others, just not big ones. This is what I am using: http://www.mouser.com/ProductDetail/Ohmite/RA-T2X-64E/?qs=sGAEpiMZZMttgyDkZ5WiumlCfl50RTwzl4D6wMWH1QQ%3d.
For the JULI@ and the P4 (what I have done so far) they are warm. No need for any more. To make them more efficient I am not using anything between the heatsink and the transistor. I am using that trick thermal compond the computer gamers use for their CPUs. One does need to make sure there is no way the heatsink comes in contact with anything else, though. You can touch them without getting shocked as long as you are not grounded! From what has been said about the P24 12 & 3.3V requirements I cannot imagine these not being sufficient. Nice thing about these sinks one can mount the regulator at the connector - the one for the JULI@ is on the board, for the P4 I am using about two inches of wire before the connector. The JULI@ regulator is not using the DANIELs PCB.
For P24 this will not be possible, obviously!
One can certainly make a shunt reg have a constant voltage - you have to add some kind of feedback and an error amplifier of some kind. This is what Allen Wright does and I feel certain Hynes does this also. Hynes uses lots of SMD devices so it hard to see what is actually happening. It has been awhile since I had the ones I have in front of me to look. Nonetheless, there is no other way.
For audio circuits the SALAS type reg is probably superior. There really is no need (that I can see) for voltages being absolute, though it is a nice bonus. One can worry about just what mischief an error amp and feedback loop can do - they still lag behind the actual event so you get the see-sawing. I would think a slow variation is better than the quick (but never quite quick enough) reactions of an error amplifier.
Not to say I hear anything wrong with the Hynes, and plenty of other people are keeping the man very busy, so it must not be much of a problem. (Excuse my going in circles)
My HDD is not a 2.5" so I guess I need to have both rails. I think you are right about separate supplies for USB and HDD. It is so easy with the DANIEL's PCB and fixed voltage versions of the LT1086. No need to push the limits of the regulator.
Now, if you could just tell me where I am going wrong with the "mouse" supply!
Thanks,
Rick McInnis
Thanks for the info on your choke setup. What size caps? How'd it look in the sim?
On the rectifier heatsinks, I am using very oversized rectifiers... IXYS DSEP30-06A, a 30A 600v FRED. My heatsinks on them are pretty small in size... but do make a diff. The other difference is that I've setup all my supplies as 2 diodes full-waves instead of a 4 diode full-wave bridges. I've also used the MSR860 diodes that Peter Daniel uses in his NOS DAC, but prefer the IXYS ones.
I need to try the Swenson filters. See his comments here that suggest it'll work on just about any PSU setup:
http://www.diyhifi.org/forums/viewtopic.php?f=15&t=1957&start=15
I'm running about 2.2A on the GA-G31M-S2L motherboard. I need to check with Belleson if their 2A reg will work with that much current with a substantial heatsink. My P4 is at just under .5A & I use a smaller heatsink than what you referenced. All of my supplies have about 5V-6V of input-to-output differential.
I checked the Hynes shunt I have here for a Juli@ upgrade & it does appear to have an error amp.
I'll have to build up the high-current Salas & see how tight they can be set. I was assuming the 2 LED reference would keep the output pretty stable.
Again, thanks for all.
Greg in Mississippi
Everything matters!
fluctuations in line voltage.
If yours is super stable it will not make much a difference.
I'll readily admit I did not feel like fooling around with the SALAS - as usual, I bought the parts and then thought that the BELLESON's make it easy! I will do that!!!
I might eventually use a SALAS on JULI@, I do not think super stable (divorced from the line) voltage is all that important there.
I am using the i3 board so I am hoping that it does indeed use less power.
(You need to make this upgrade, it really makes a difference) Board is on sale at NEWEGG, who I despise at the moment.
Can't remember if I said, that the reason I am using the choke with the % volts supply is that I am going to have to minimize the in/out voltage differential and figured I should give the regulator as little work as possible to do.
On the other regulators I have used more dropout voltage, I suspect similat to your choices.
I am still enjoying what the EVS version of the Purvine idea does for my speakers, I have only used them on the TAD drivers, the 4001 especially known for aggressiveness, and found they made them sound much more refined.
Theo is far more adventurous than me with his DIY efforts. I tried the battery on my woofers but did not hear much of anything but it might take time to hear the difference below 500 hz (?). I have not tried the battery in place of the EVS devices. I will.
I have not tried Mr. Maher's devices. I am intrigued. They strike me as even more mysterious than loops of wire and batteries hanging off of drivers!
If the battery works as well as Theo reports, and as you know I think Theo has very perceptive hearing, Mr. Schultz might be back to modifications soon. The battery is easy to implement. What Ric was/is doing is time-consuming. I know I would not want to make one in the same fashion as his. I was glad to send him the money. They are well made.
I have been involved in a very silly thread about the battery at AA TWEAKS - a guy who makes cables is going on and on about how the "pool of electrons" idea is preposterous and CANNOT work. A guy making cables saying this. Says: where's the proof? A guy making cables wants proof? I asked him about cables being PROVED to work. He says they have not (duh!) but that doesn't stop him. He has not tried the tweak. I think he has been on an extended NEW YEAR'S "imbibement" or maybe he is lacking some other form of diversion, who knows?
Bye,
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: