Batteries
in [Mandrake]
|
From: Moe Trin on 29 Sep 2007 16:52 On Fri, 28 Sep 2007, in the Usenet newsgroup alt.os.linux.mandrake, in article <IFaLi.28395$nO3.6542(a)edtnps90>, Unruh wrote: >Christopher Hunter <chrisehunter(a)NOSPAMblueyonder.co.uk> writes: >>Unruh wrote: >>> Actually I doubt that you could push Gigabit signals down a coax. Already discussed. For longer runs, you tend to prefer a solid outer conductor (inside is silver plated) coax with a foamed polyethylene dielectric over a silver plated solid semi-conductor (skin effects), but that runs into physical size problems with increasing frequency. Above the size limitation, you tend to use waveguide. >>> I also find it astonishing that you can push Gigabit down a twisted >>> pair, but it works. Not quite twisted pairs, but the original British 10 cm (3 GHz) radars on ships (Flower class corvettes ~1941) were using open pairs as a transmission line - only for a very short distance (few feet if I recall correctly). You may also remember the "twin lead" flat cable about a half inch wide used for television receivers. Open lead replaced the plastic web with ceramic spacers some distance apart. >>The input to my 22 GHz spectrum analyser front end is a BNC! Oh, REALLY? What idiot manufacturer would use a connector rated to 12.4 GHZ at nearly twice it's cut-off frequency? 22 GHz is even beyond an SMA connector (18 GHz). Tektronix had a design (model 491?) where they used something similar to connect to an external _waveguide_mixer_ device, but the connector (which I believe was SMA) was actually carrying a sub-harmonic of the effective LO, and the resulting IF - no where _NEAR_ the "observed" RF, and pretty insensitive at that. >I have 5, 10 and 18 GHz signals around here going down cheap coax with >little loss. I once made measurements of the loss of RG-58A/U (the coax commonly used for "thin-net" a.k.a. 10Base2) at 5.2 GHz. With a BNC connector (UG-260/U which probably wasn't helping matters) the loss averaged one decibel per foot. A lot of that was probably mismatch and leakage loss due to the single layer of braid. Single layer of braid type coax is usable to 1 GHz, but gets far to leaky above that. A double braid, such as RG-142/U is usable to 12 GHz, but only in short lengths. >Very short I suspect. Ie, try doing a 50m run and seeing how much signal >gets through. 50 Meters of RG58 at 1 GHz - about 30 dB. But 50 meters of 3 1/8 inch copper hard line would be more like 0.55 dB. Quite feasible, and nearly as good as WR-975 waveguide (0.26 dB) which is nominally 5 x 10 inch in cross-section. On the other hand, no one is going to be using coax for a 50 meter run above ~4 GHz without extremely interesting reasons. I did a 5.2 GHz installation where they wanted a receiving antenna on the nose and tail of a DHC-5 Buffalo, with the receiver and antenna switch located just behind the copilot. Front antenna cable was 3/8 inch "Alumifoam" (aluminum semi-rigid coax from Times Wire and Cable), while the 70 foot / 21 meter run to the aft antenna was 7/8 inch "Alumifoam" with a _measured_ loss of 4.8 dB. The 7/8 inch coax has a cut-off (the frequency where it starts acting as an extremely lossy waveguide) around 5.4 GHz, so that was the largest and lowest loss cable available. Waveguide - even semi-rigid helical guide - simply wasn't acceptable. Coax wasn't ideal, but it did the job. Old guy
From: Moe Trin on 29 Sep 2007 16:52 On Sat, 29 Sep 2007, in the Usenet newsgroup alt.os.linux.mandrake, in article <DTgLi.1889$1d2.404(a)trndny05>, Adam wrote: Moe Trin wrote: >> Where might the air be coming in? Gaps around the drive[s]? > >I'd guess from the vent holes in the right side of the case. This is a >vertical tower, ATX form factor. That doesn't sound like the ideal solution. >I just discovered that I already have hddtemp installed: What's this??? Another 5 kilo bag of salt??? >The temp for hda (all Linux except swap) is staying constant at 29C, >but hdb (Linux swap, W2K, system-independent data) is fluctuating >second-by-second between 32C and 50C, mostly around 41-43C. Bad connection. Got an incandescent light handy? Hold the glass bulb, and turn on the lamp for one second. Gets warm pretty quick, but takes a LOT longer to cool down. Same idea, except that bulb is probably doing a lot more power. >I don't like that 50C, if it's accurate, but it doesn't stay there, >goes right back down within 2-3 seconds to around 43C. I don't know where the sensor might be, but the 50C is usually the ambient temperature limit. Inside the drive, it may well be higher by a number of degrees. >I tried looking for the sensor chip yesterday, but couldn't find it >easily. I'll have to look again with better light and mirrors (some >chips are hidden behind the power supply), and also a better idea of >exactly what I'm looking for. Hard to say - could be a small surface-mount package about 1/4 inch square. >I was surprised at how few chips were actually on the motherboard. A bit different, even from designs from the late 1990s. >While it was printing, I used printerdrake (thru MCC) to try to >optimize settings, and it suddenly stopped printing, and won't even >display anything on the front panel, although something starts whirring >every time I plug in the power. That's going through the self test. It SOUNDS as if you've changed the printer interface. >> I don't have access to a 5 or 5M, but do recall that they had one, >> vaguely on the top on the display. > >Nope. Time to RTFM. For both. I'm not the printer person - try a.o.l.mandriva. >NIC card installed, crossover cable connected, still nothing displayed >on front panel of printer. /sbin/ifconfig only shows 'eth0' (DSL >router) and 'lo'. Just an additional screen ("Intel Boot Agent Version >2.6") displayed during system power-up, after detecting IDE drives but >before displaying system info. Haven't set up a JetDirect card in years, but recall that it initially came up on an RFC1918 address range - something in the 192.168.x.y perhaps, and you had to take it from there. The second NIC isn't showing up because you haven't configured it - Mandriva has some whizzy GUI that does that, but I work directly with the files in /etc/sysconfig/network-scripts/. If the default address that the printer comes up on doesn't conflict with the address range shown for your eth0 interface, I'd set eth1 to a static setup using the next higher/lower address from that used by the printer. >> I was under the impression (for unknown reason) that you had more than >> one system. > >Nope, this one's not antiquated enough to justify buying another. It's >a Celeron 1.7 GHz, low-end desktop as of early 2004. Age has nothing to do with it. It's how much system you need, and how you've arranged things. There are five systems on my home LAN, and two additional stand-alone (company systems from my wife's and my employer), and two more systems connected but not running... for the two of us. >> The 1500 is a large beast meant to handle the heaviest loads running >> unattended. For a single system in general use at home, there are >> UPS available that cost about the same as that battery. > >Sounds like overkill, then. I don't think I really /need/ a UPS, as >long as I have surge protectors on the power and phone lines. Oh well. Your power never goes out? Our power distribution is primarily underground, but we still loose power occasionally. Even using a journaled file system, it's nice to be able to allow the systems to shut down cleanly in the event of a power failure. Old guy
From: Moe Trin on 29 Sep 2007 16:53 On Sat, 29 Sep 2007, in the Usenet newsgroup alt.os.linux.mandrake, in article <LTgLi.1890$1d2.1097(a)trndny05>, Adam wrote: >Moe Trin wrote: >> -rw-rw-r-- 1 gferg ldp 41481 May 23 13:03 UPS-HOWTO > >How'd a version of that end up on my HD? :-) Yeah, but is it _that_ version? That's what's current on sunsite at the moment - and the Revision History indicates the changes are about battery information. The previous version was two years old. Old guy
From: Christopher Hunter on 29 Sep 2007 19:50 Moe Trin wrote: > Oh, REALLY? What idiot manufacturer would use a connector rated to > 12.4 GHZ at nearly twice it's cut-off frequency? Hewlett Packard on the 22GHz front end for their 141 series analysers. Works well enough with good quality double-braided coax. I can't be bothered to measure the losses involved - I'm only interested in /relative/ levels to look at harmonic outputs compared with the fundamental. Most of my RF work is done at much lower frequency - particularly around 100 MHz at present. C.
From: Adam on 30 Sep 2007 23:53
Moe Trin wrote: >>> Where might the air be coming in? Gaps around the drive[s]? >> I'd guess from the vent holes in the right side of the case. This is a >> vertical tower, ATX form factor. > > That doesn't sound like the ideal solution. It doesn't? Exhaust fan on left side of case pulls air in through right side of case and over HDs. >> I just discovered that I already have hddtemp installed: > > What's this??? Another 5 kilo bag of salt??? Sorry, that one went right over my head. System temperatures now seems to be pretty steady at: M/B Temp: +40�C (low = +15�C, high = +40�C) sensor = thermistor CPU Temp: +28�C (low = +15�C, high = +45�C) sensor = diode /dev/hda: SAMSUNG SP0411N: 25�C /dev/hdb: MAXTOR 6L020J1: 41�C Maxtor HD's spec for operating temperature is 5-55C. >> I tried looking for the sensor chip yesterday, but couldn't find it >> easily. I'll have to look again with better light and mirrors (some >> chips are hidden behind the power supply), and also a better idea of >> exactly what I'm looking for. After a LOT of work, I found it. It's by the top of the CPU heatsink, half under it. IT87??F (partially covered by plastic frame that CPU fan/heatsink attaches to), probably an ITE 8705F, which is a Super I/O chip. (And yes, I put thermal compound on the CPU before reattaching the heatsink.) Looks like everything's now in spec except maybe the motherboard is occasionally one degree over according to 'sensor'. >> Time to RTFM. Printer manual described location and use of power switch, which solved that problem. <blush> BTW this printer (HP Color LaserJet 5 if anyone's forgotten) seems to be drawing power even when it's switched off, but the manual doesn't say how much. I'm tempted to go out to the electric meter to see just how much. Is it worth unplugging the thing when (a) system is on but not using printer, (b) system powered down, and/or (c) when system will be off for several days? > Haven't set up a JetDirect card in years, but recall that it initially > came up on an RFC1918 address range - something in the 192.168.x.y > perhaps, and you had to take it from there. The second NIC isn't > showing up because you haven't configured it - Mandriva has some > whizzy GUI that does that, but I work directly with the files in > /etc/sysconfig/network-scripts/. If the default address that the > printer comes up on doesn't conflict with the address range shown > for your eth0 interface, I'd set eth1 to a static setup using the > next higher/lower address from that used by the printer. This is where I'm stuck. NIC card is recognized as eth10, and I used ifconfig to give it the address 192.0.0.192 which I'm told is the default for JetDirect. [root(a)ozymandias ~]# ifconfig eth10 eth10 Link encap:Ethernet HWaddr 00:03:47:42:12:A1 inet addr:192.0.0.192 Bcast:192.0.0.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:76 errors:0 dropped:0 overruns:0 frame:0 TX packets:166 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:36360 (35.5 KiB) TX bytes:56518 (55.1 KiB) Interrupt:11 [root(a)ozymandias ~]# cat /etc/sysconfig/network-scripts/ifcfg-eth10 DEVICE=eth10 BOOTPROTO=dhcp ONBOOT=yes METRIC=10 MII_NOT_SUPPORTED=no USERCTL=no RESOLV_MODS=no IPV6INIT=no IPV6TO4INIT=no DHCP_CLIENT=dhclient NEEDHOSTNAME=yes PEERDNS=yes PEERYP=yes PEERNTPD=no When I try to use MCC to change the printer's connection type to "network printer," I gave it 192.0.0.192, and it accepts it, but won't print a test page. I think that's not the right address, but I have no idea how to find out what's correct. >> this one's not antiquated enough to justify buying another. > > Age has nothing to do with it. It's how much system you need, and how > you've arranged things. You're right. I guess what I mean is that so far, anything I want to do, I can do with my current system. >> I don't think I really /need/ a UPS, as >> long as I have surge protectors on the power and phone lines. Oh well. > > Your power never goes out? Maybe once every two or three years. Not often enough to be a big worry. A UPS would be nice, but I don't think it's a priority for me. Only real problem over the years has been two dialup modems that had their hook relays frozen in "off hook" presumably by power surge on phone line. >>> -rw-rw-r-- 1 gferg ldp 41481 May 23 13:03 UPS-HOWTO >> How'd a version of that end up on my HD? :-) > > Yeah, but is it _that_ version? Nope, never said it was. Newer version has been downloaded and will be read. Adam |