Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Scotty! We need Warp power now! [can't deduce 5v draw!]

27 views
Skip to first unread message

Louis Ohland

unread,
Mar 18, 2023, 6:52:33 AM3/18/23
to
Had a few questions on 95 PSU load. At first, I thought just monitoring
the AC amps in would give us wattage. Not quite so.

There are two "levels" of power, one being the triple device power
sockets. +5v and +12v. SIMMple to deduce, read drive datasheet, done.
Just leave all taps disconnected, SIMMplify reading planar power draw. I
ass-u-me the three taps are overload protected separately from the main
overload. Dunno. But it makes sense that a device could be pulling more
power than it should, yet not enough to pop the breaker for overall current.

The second "level" is the 21 pin planar power plug. We got +5v, +12v,
and -5v.

It seems possumble that we measure the wattage into an unconnected PSU,
using either a paperclip on the 8595 PSU, or the test button on the 95A
PSU. We get a "no-load" wattage, just for the PSU.

Now we can plug it into the planar, and test with one set of SIMMs, then
two, then three, then four. -MAYBE- the SIMM power draw is not lost in
the overall efficiency of the PSU.

The best weg would be to measure all +5v at once with a clamp. The
planar pretty much exclusively uses +5v, so we aren't missing much by
letting -5v go unmeasured.

Grant Taylor

unread,
Mar 18, 2023, 12:08:22 PM3/18/23
to
On 3/18/23 4:52 AM, Louis Ohland wrote:
> The best weg would be to measure all +5v at once with a clamp. The
> planar pretty much exclusively uses +5v, so we aren't missing much by
> letting -5v go unmeasured.

I would think that it would be possible, if not annoying, to measure the
actual voltage & current draw (thus wattage) of each rail if one has
enough meters and proper connectors to make and place a pig tail between
the PSU and the motherboard.

Though I suppose that someone would need to really want to do that to go
to the trouble to do so.



--
Grant. . . .
unix || die

Louis Ohland

unread,
Mar 18, 2023, 12:43:50 PM3/18/23
to
Uh, the 21 pin AMP connector is unobtanium.

Grant Taylor

unread,
Mar 18, 2023, 1:26:13 PM3/18/23
to
On 3/18/23 10:44 AM, Louis Ohland wrote:
> Uh, the 21 pin AMP connector is unobtanium.

I was thinking sacrificial / donor parts from a dead system plainer and
power supply.

I hear that it's possible to do amazing things with 3D printers these days.

Louis Ohland

unread,
Mar 18, 2023, 2:46:15 PM3/18/23
to
Oddly, I -MIGHT- have foundt both wire and board connectors. Sendt links
to MAJ Tom

moussa

unread,
Mar 19, 2023, 10:08:05 PM3/19/23
to
Louis,

Computing power consumption is dynamic, you need to stress it to get an
average overall consumption, isolating the active at the main with a
clip on amp meter still the best way to do it.

Shifting bits around do cost power.

--
Moussa

"People alike with a similar circumstances, tend to find each
others across space and time, given enough time in life, no
matter distance, language, race, colour and believes" (c) MEK
*** Do Not Copy, Duplicate or Use without my Permission ***

0 new messages