mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Hardware

Reply
 
Thread Tools
Old 2016-04-20, 13:32   #12
Mark Rose
 
Mark Rose's Avatar
 
"/X\(‘-‘)/X\"
Jan 2013

37·79 Posts
Default

Quote:
Originally Posted by bgbeuning View Post
Those use 18 gauge wire which I would recommend limiting to under 10 A current.
(Wikipedia says 14 A heats AWG 18 wire to 90 C.)
300 W at 12 V is 25 A so the PSU wire needs AWG 12 or 14.
And it would be having 12.5 A on the first Y cable.
The above is the worst case. The load is probably spread over multiple wires.

If none of the wires feel hot to the touch, it should be safe.
A "loop" current meter will let you check each wire.
There are three 12 V wires, one in the ATX-24 connector, and two in the ATX-4 connector. Each circuit in the ATX-24 connector is supposed to be rated for 6 amps, and each in the ATX-4 connector 8 amps.

http://www.playtool.com/pages/psucon...onnectors.html

Four motherboards on a single power supply is probably very close to a limit on one or more of the circuits.
Mark Rose is offline   Reply With Quote
Old 2016-04-22, 05:25   #13
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

33·283 Posts
Default

Quote:
Originally Posted by Mark Rose View Post
There are three 12 V wires, one in the ATX-24 connector, and two in the ATX-4 connector. Each circuit in the ATX-24 connector is supposed to be rated for 6 amps, and each in the ATX-4 connector 8 amps.

http://www.playtool.com/pages/psucon...onnectors.html

Four motherboards on a single power supply is probably very close to a limit on one or more of the circuits.
Good reference link.

There are two 12V wires in ATX-24 connector and two wires in the ATX-4. Do we have any idea how well balanced most motherboards are in drawing from those 4 wires? I don't own a current meter.

My Skylakes are drawing 57W at the wall. PSU is 92% efficient so each board is consuming 53W. If all of that is 12V, that's 4.5A. My thought is to put one ATX-4 splitter on each plug of the EPS12V4+4 connector. This will halve the load on each of those lines. Since those are rated for 8A they will definitely not be a problem. The only question is the ATX-24 cable where the 12V lines are rated for just 6A. Do we think the motherboard will draw more than 1.5A from either of those 12V lines?
Prime95 is offline   Reply With Quote
Old 2016-04-22, 10:42   #14
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT/BST)

592410 Posts
Default

My dad's old Athlon II x4 was supposed to need a ATX-8 connector in the motherboard but would run with just a ATX-4 connector plugged in. When he got to the point where he needed to replace the power supply he discovered that the plastic on the connector/motherboard had melted. This led to a new pc.
I would be careful with this sort of thing.
http://www.playtool.com/pages/psucon...onnectors.html was an interesting read.

The Athlon II 620 x4 was a power hungry system though with a TDP of 95W while yours are lower.
It looks to me that based upon the link above you should be fine with the 20+4 splitter as you are not using graphics cards which need 12V power. It sounds like the two lines are usually connected so the total is more concern than the individual. Will your pc boot with just 20 pins? If so you should be fine.
henryzz is offline   Reply With Quote
Old 2016-04-22, 11:47   #15
bgbeuning
 
Dec 2014

111111112 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Do we have any idea how well balanced most motherboards are in drawing from those 4 wires? I don't own a current meter.

Do we think the motherboard will draw more than 1.5A from either of those 12V lines?
My loop current meter only seems to work on AC power.
bgbeuning is offline   Reply With Quote
Old 2016-04-22, 13:35   #16
Xyzzy
 
Xyzzy's Avatar
 
Aug 2002

5·1,663 Posts
Default

Quote:
Originally Posted by bgbeuning View Post
My loop current meter only seems to work on AC power.
https://en.wikipedia.org/wiki/Current_clamp
Xyzzy is offline   Reply With Quote
Old 2016-04-22, 14:40   #17
bgbeuning
 
Dec 2014

3×5×17 Posts
Default

Quote:
Originally Posted by Mark Rose View Post
The link for ATX-24 shows the +5 V (5 wires) and +3.3 V (4 wires) rails can have twice the current as the +12 V (2 wires) rail. But since the +5 and +3.3 have twice as many wires, that is OK.

If the 60 W is split fairly evenly over all 13 wires (11 in ATX-24 and 2 in ATX-4),
then the Y cables seem pretty safe. But like George said, the evenly part is the big question.
bgbeuning is offline   Reply With Quote
Old 2016-04-22, 19:02   #18
chalsall
If I May
 
chalsall's Avatar
 
"Chris Halsall"
Sep 2002
Barbados

132·59 Posts
Default

Quote:
Originally Posted by bgbeuning View Post
My loop current meter only seems to work on AC power.
Further to Mike's link... Yes, loop current meters only work with AC.

This is a fundamental truth based on physics; the "flux" generated by alternating current allows the measurement.

It is possible to measure the voltage of both AC and DC circuits by placing probes appropriately, and can be "out of circuit" (read: "parallel").

It is, however, only with AC circuits that it is possible to measure the current externally.

To measure the current of a DC circuit one must be "in circuit" (read: "in series").

(There are actually more complicated ways of measuring DC current "out of circuit", but they're non-trivial, and not available retail.)

Last fiddled with by chalsall on 2016-04-22 at 19:10 Reason: English is such a subtle language....
chalsall is offline   Reply With Quote
Old 2016-04-22, 19:58   #19
sdbardwick
 
sdbardwick's Avatar
 
Aug 2002
North San Diego County

68910 Posts
Default

Quote:
Originally Posted by chalsall View Post
Further to Mike's link... Yes, loop current meters only work with AC.
Some clamp current meters can work with DC. See from Mike's link https://en.wikipedia.org/wiki/Current_clamp#Hall_effect
and Fluke model 365 (around $250 USD) for example.
sdbardwick is offline   Reply With Quote
Old 2016-04-22, 21:21   #20
chalsall
If I May
 
chalsall's Avatar
 
"Chris Halsall"
Sep 2002
Barbados

132×59 Posts
Default

Quote:
Originally Posted by sdbardwick View Post
Some clamp current meters can work with DC. See from Mike's link https://en.wikipedia.org/wiki/Current_clamp#Hall_effect
and Fluke model 365 (around $250 USD) for example.
I am aware of that.

Are you aware that such measuring devices are highly inaccurate?
chalsall is offline   Reply With Quote
Old 2016-04-22, 23:00   #21
sdbardwick
 
sdbardwick's Avatar
 
Aug 2002
North San Diego County

13·53 Posts
Default

Quote:
Originally Posted by chalsall View Post
I am aware of that.

Are you aware that such measuring devices are highly inaccurate?
Depends on the context. Accurate enough for field work (diagnostics), but you wouldn't use one as an engineering tool. They are much better than they used to be - cheap DSP helps a great deal.
IIRC, non-specialized Fluke clamps (Hall or otherwise) are good for about 2% full range accuracy. Resolution is not great @ around 0.1A
sdbardwick is offline   Reply With Quote
Old 2016-04-23, 21:41   #22
Madpoo
Serpentine Vermin Jar
 
Madpoo's Avatar
 
Jul 2014

3,313 Posts
Default

Quote:
Originally Posted by chalsall View Post
To measure the current of a DC circuit one must be "in circuit" (read: "in series").

(There are actually more complicated ways of measuring DC current "out of circuit", but they're non-trivial, and not available retail.)
I'm going to agree with chalsall on measuring DC current. You really need to measure in series (directly for low amp or use a shunt for larger amperages).

Anything else is really no better than an educated guess. It might be close enough, so if that's good enough for whatever you need, then knock yourself out, but I wouldn't use it for anything where accuracy matters.
Madpoo is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Call for GPU Workers to help at the "LL Wavefront" chalsall GPU Computing 24 2015-07-11 17:48
Aouessare-El Haddouchi-Essaaidi "test": "if Mp has no factor, it is prime!" wildrabbitt Miscellaneous Math 11 2015-03-06 08:17
Speeding up double checking when first test returns "prime" Unregistered PrimeNet 16 2006-02-28 02:00
Oh noes! The "42nd Mersenne prime" isn't prime! ixfd64 Lounge 7 2005-04-03 19:27
Would Minimizing "iterations between results file" may reveal "is not prime" earlier? nitai1999 Software 7 2004-08-26 18:12

All times are UTC. The time now is 04:59.


Thu Oct 28 04:59:16 UTC 2021 up 96 days, 23:28, 0 users, load averages: 2.32, 2.02, 2.00

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.