I-Appliance BBS
The Official Source for Internet Appliance Upgrades and Mods
Amazon Honor System Click Here to Pay Learn More
BBS Main List | Sign In | Sign Up | Search | Help | Linux-Hacker.netReply to Thread | Printer |

Home / Other I-Appliances / Virgin Webplayer
Possible NIC problem found. May also play a role in CPU heat.

New MessagePossible NIC problem found. May also play a role in CPU heat. (modified 0 times) sheepdog
Profile
In the thread discussing the power supply being too weak, I decided to hook up a volt meter to a USB port.

After watching it for a while, it was pretty stable between 5.28 and 5.3, which is a little on the high side. I noticed on large files the voltage would climb, and also occasionally spike.

Turns out, at 5.5 volts my Nic shut off.

A quick check on my own system, showed USB to be perfect at 5.0-5.03 volts constant.

So I decided to see how it could possibly be getting 5.5 volts. Since the power supply is only 5 volts.
Another quick check revealed that the power supply is kicking out 5.40-5.42 volts with no load! I even verified this on a second WP power supply.

So seeing as how its feeding too much juice out the power supply, I would assume most of the system is being over-juiced as well. Which would explain why they run so damn hot. Then again, that could just be the chip (probably a little of both).


I am not sure if this helps anyone, but it did explain one problem at least. Next step will be to find a way to drop it down to where it should be.

Anyone got any ideas for a cheap and easy way. Or does anyone feel its even really necessary for all but the NIC.

07-04-2002 02:02:19

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) ChocoNutDancer
Profile | Email
could this explain why I couldn't hook up a regular IDE cdrom drive that
was externally powered?
07-04-2002 07:45:11

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) mbaha
Profile
What nic are you using? Also has any one found a good replacement for the stock power supply? This voltage problem seems like it would solve a lot of problems we are having. In full size computers I have had a ton of problems due to cheap power supplies. I was getting ready to give up on webplayers but I think I will give it one more go.

Thanks
Mbaha


hack your own you will love it more
07-04-2002 08:28:38

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) im_eguy
Profile
A few of us are wiring up AT power supplies. I think even a cheap AT would be more stable and provide more power than the stock webplayer PSU. I am heading out to Radio Shack now to see if I can get a connector the same size as the one on the webplayer PSU.
07-04-2002 08:33:57

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) sheepdog
Profile
I am using a Netgear FA101 USB Nic.

On the Cd-Rom it is possible with it being that far out of spec.

I know on Motherboard monitor 10% is cause for alarm, and almost any fluctuation is bad in either direction.

We are running over voltage, (in the danger zone it seams) and its not even stable when its there. not a good combination.

On a Cd_r drive, you could pretty much count on burn problems I would imagine..

I am just amazed they could put something like that out the.

it seems to me like a time bomb for the electronics.

On a desktop, this would be completely unacceptable.

07-04-2002 10:39:59

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) im_eguy
Profile
I imagine it was a cost issue (I am sure these PSU's are cheap). Since the thing was not intended to have a HD, CD-ROM or even a NIC, why would Boundless or Virgin care as long as it's cheap and "good enough"?
07-04-2002 12:05:54

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) sheepdog
Profile
I can understand it from a cost perspective, but you would think they would use something that was actually operate within spec.

.4 volts out of spec is an awefull lot.

I am honestly surprised no one has fried anything yet. Including cpu, screen, motherboard, etc...

I over volted ram (to overclock) by .15 volts, and it died early. It takes no more than that to kill a cpu as well.

If your going to run that far out of spec, you should at least ensure it goes no higher.

07-04-2002 17:45:53

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) mbcook
Profile
that's pretty odd. Have you tried anything like a few simple zener dioeds attached between the USB port and USB nic just to see if that keeps it from dieing?
--Michael
07-04-2002 17:54:42

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) sheepdog
Profile
I am more worried about the system as a whole. USB Nics are cheaper than WP's.

As for the diodes. I am not that good with electronics. I can re-wire, and check voltage and simple stuff, but thats about it.

My main goal was to make people aware of it, I plan on going to an AT PSU soon.

I am sure there are others on here who could easily figure out a way to fix this, and let us know, but at least we now have an idea of where and what the problem may be.

07-04-2002 19:28:04

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) boosted
Profile | Email
I remember someone once told me I can solder a zener diode of the proper voltage (5v in this case) in parallel to the load, and it would keep the voltage constant. Of course we have to use a zener diode of the correct wattage too. Do we have a circuit guru that can confirm this?

-boosted

07-05-2002 10:02:32

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) ST225
Profile
Finding a zener that can handle that much power is going to be hard. The easy thing to do is take advantage of the .2v drop of a normal diode. Using two standard power diodes in series with the ps should drop the voltage to the right level.
07-05-2002 11:08:40

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) boosted
Profile | Email
Would the diode need to be 25watt also? Also, from experience, I have seen power supplies that put out more voltage with no load and dropping to the correct voltage with load. So perhaps it is putting out 5v when it is plugged in?

-boosted

07-05-2002 11:25:37

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) im_eguy
Profile
What is the input voltage for the SODIMM and the CPU? I doubt they are 5V anyway (meaning the voltage will be dropped by the MD anyway). I think the only problem with the 5V would be USB (might be another reason virgin didn't care).
Also, I don't know if you want to drop the voltage since it will then drop below normal once you remove USB devices. The problem is that the PSU is not reguilated well
07-05-2002 11:58:58

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) zmoz
Profile
I remember I read once on a thread that I can't find again that someone used some kind of power supply that gave the webplayer exactly 5 volts, but the webplayer wouldn't work until it was uped to 5. somthing...
07-05-2002 13:00:18

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) Lucubrate
Profile
Woah, hold up yall..

Switching power supplies must have a minimum load to stay within specs for voltage. Measuring the voltage of an unloaded supply is pointless.

It would probably be even worse to put a PC power supply in service on a webplayer without making sure you meet the minimum load.

07-05-2002 19:12:07

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) jpe_dty
Profile
The webplayer power adaptor is putting out 5.3v under load (both the ones I tried) which is more than 5% out-of-spec. I am considering attempting to "repair" it, however I am not absolutely certain that this thing has a real voltage regulator in it (from what I have seen of its low-voltage section.) It looks like adjusting R58 might do it, though. Either that or blow the thing to bits!
07-05-2002 19:29:00

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) ChocoNutDancer
Profile | Email
I found the proper power plug connector for the WP.
it's a 5.5mm O.D. x 2.5mm I.D. power plug. Radio
Shack part 274-1573. also known as a Radio Shack
power plug "N".

I played around yesterday with a powered USB Hub
($10 at compgeeks.com). at first I thought it
was a solution but with time the network would
fail again. I suspect that as the power
supply gets hot the problems increase.

tomorrow I'll connect a regular AT PS to the WP
and see what happens.

07-05-2002 22:12:04

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 1 times) sheepdog
Profile
Lucubrate, I realize my measurement of 5.4 with no load should be pointless, but if the USB is putting out a minimum of 5.4, then obviously the PSU is putting that out weather it is under load or not.

Only reason I mentioned that the PSU even puts out that much was I was trying to determine if it was a correct reading and why was it so high.

If the PSU was only putting out 5 volts under load, then the USB would certainly not be running at 5.4

Not to mention the fact that it only went UP from there.

Seems to me, that if they were going to regulate it anyhow, why exclude the USB ports, especially since it appeared they were planning on using them.

Also, an At power supply has more than enough power. The only question is if the system requires a little more, however we have already seen evidence that the system runs better with a computer PSU.

I may have my psu running my WP tonight and will see how it runs.

ChocoNutDancer I noticed that it did go up and fluctuate more as it warmed up. Large single file transfers made the voltage rise as well.

07-05-2002 22:18:43

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) boosted
Profile | Email
I found a 5.1v 5watt zener diode at partsexpress.com. perhaps 5 of them in paralle will make 25watts? and if it does require more than 5v, then 5.1v is perfect. Compgeeks have a 6v 1a power adater, perhaps that can be used with some resisters to lower it to 5v. they are priced at $.25 each.

-boosted

07-05-2002 22:46:52

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) Lucubrate
Profile
I told myself I wasn't going to bust into my WP again, but now I just have to :)

I measure 5.25 volts under load. But, if you look at the board you will see a power regulation section that supplies most of the chips on the board with 3.3 volts so I don't think you have to worry about the chips getting fried. There are a few ICs that use the 5 volt rail, but most are just logic ICs and should be able to easily take 5.5 volts.

It appears the 5 volt rail does go unmodified right to the USB power. My first thought is that a self powered hub will make the USB supply voltage a moot issue.

The only WP that I have left is one with a mini-pci etherenet card, so I don't use USB devices on it. As such, I can't comment as to the stability of USB on this one. The others I had were the ethernet-less version, and I did use a USB NIC without any real problems. The NIC would occaisonally stop TX/RX, but it would resume in a few seconds. The boards I had were the ones with the IC near the USB section, but I think we already determined there was no correlation there.

And sheepdog, my point about the PC power supply was not capacity, but that you need a minnimum load on it as well. Depending on the power supply, 25 watts might not be enough to bring the 5 volts into regulation. In fact my WP only uses about 2 amps. All I am saying is be carefull. And good luck.

07-06-2002 08:43:44

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) jpe_dty
Profile
Ok, I replaced R58 on the bottom of the power supply board (measured as 3.33Kohm) with a 10Kohm micro potentiometer. If you are foolish enough to attempt working on a live switching power supply yourself, I would recommend using a 4.7Kohm one instead. Anyway, if I used this pot. to adjust the unloaded voltage to about 5.12V, the voltage hovered around 4.98-5.02V with the webplayer plugged in and any variety of usb devices.
07-06-2002 11:53:38

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) jpe_dty
Profile
The webplayer psu really doesn't work very well. When adjusted to start the webplayer at about 5.02V it slowly drifts up to about 5.18V as it warms up. It may not be worth the effort "fixing" it.
07-07-2002 05:36:59

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) sheepdog
Profile
If mine would stay down that low, I would be happy.

However both of mine run a bit higher. Odd since both came from different sources.

Oh well, my AT PSU is done once I get the connector.

I am glad to hear there is a regulator on some of the system at least.

Thanks Lucubrate.

07-08-2002 17:33:17

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) im_eguy
Profile
The connector is at Radio Hack for $2. It's a type "N" (was posted already I think)
07-09-2002 13:14:32

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) im_eguy
Profile
BTW, for all ppl from the [H] forums, I am using a Xircom and I have the same problem as the people do with the 3COM NIC. However, with my AT PSU, the problems have disappeared...
I think STRESS has a more stable power siupply than myself, sheepdog and some of the others do.
07-09-2002 13:27:23

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) sheepdog
Profile
Don't include me yet. Sometime during modding, my AT PSU, took a poo. :(

I have to rip it apart and see whats up again...

If it ain't one thing...

Figures, works fine untill you put Windows on ;)

07-09-2002 19:56:46

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) im_eguy
Profile
Shouldn't be a big deal though.. you can find AT PSU's for $5 :)
I won't lose any sleep if mine dies!
07-10-2002 07:57:57

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) sheepdog
Profile
I would not worry either except for the fact that I built a whole new housig for it.

Turns out the PSU is fie, I did not ground it completely where it needed it, unfortunately, the WP refuses to run from it. All I get is light codes from the front LED.

07-11-2002 03:25:56

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) im_eguy
Profile
Most of the work you did is on the case. You can likely find another PSU and put it in that case. Just check that it works first :)
07-11-2002 08:00:16

New MessageRE:Possible NIC problem found. May also play a role in CPU heat. (modified 0 times) jpe_dty
Profile
FYI I purchased three replacement power adaptors at: http://store.yahoo.com/absolutelyspecial/acad125v5a.html
for $18 each. It runs at about 5.10 to 5.18V, which is not great but better than the 5.5v the webplayer power supply puts out, and it is a direct replacement (it even looks the same.)
07-19-2002 17:23:30

Reply to Thread | Printer |
All times are PSTPowered by UltraBoard v1.62



Copyright © 2000, Netmake Inc. All Rights Reserved.
See Terms and Conditions for more information.




i-opener opener laptop notebook computer help drivers dll free windows dos repair fix linux mac macintosh 2000 95 98 nt pc configure hardware software sound video netscape explorer network networking lan wan software cmos fat bios printer card mouse modem ide scsi cd rom controllers scanner tape hard drive cgi scripts source code mp3