| |
ZL3AI > APRDIG 27.07.03 12:50l 203 Lines 9135 Bytes #999 (0) @ WW
BID : 1681-ZL3AI
Read: GUEST
Subj: TAPR Digest, Jul 24, 3/4
Path: DB0FHN<DB0RGB<DB0MRW<OK0PKL<OK0PCC<OK0NAG<OK0PPL<DB0RES<ON0BEL<ZL2BAU<
ZL3VML
Sent: 030726/1341Z @:ZL3VML.#80.NZL.OC #:5039 [Chch-NZ] FBB7.00i $:1681-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: Inverters and RFI
From: "James Lux" <aprssig@luxfamily.com>
Date: Thu, 24 Jul 2003 11:56:26
X-Message-Number: 17
Interesting discussion going on here about powering laptops etc from the
car. A few comments:
1) You desperately need significant isolation from the car's electrical
system: load dump transients, alternator whine, etc. National Semiconductor
has an old ap-note describing the automotive electrical environment which
should be required reading. You're pretty brave if you're running a
significant load that is sensitive with just a regulator. For small loads
at low voltages, series inductors, big capacitors, and a regulator may work
just fine.
2) There are lots of ways to make DC/AC or DC/DC inverters, and as noted by
others, cost doesn't much indicate RFI potential or anything to do with
EMI/EMC. There are fully isolated DC/DC converters and ones that share a
common connection from input to output, according to the eventual use.
Modular converters may or may not have any particular EMI/EMC specs..
you'll have to look at the catalog. Pay particular attention to the current
ripple!
3) Judging from the high pitched whine on my inexpensive DC/AC inverter, it
looks like it is doing the boost converter to a high DC bus voltage and
then making 60 Hz from that. Makes sense, because there are inexpensive
capacitors to serve for energy storage at 170 or 340V, and inexpensive
switch devices for the output. This design would be a lot cheaper than the
older scheme of inverting at 60 Hz directly, because the iron and copper in
the transfomer would be a signficant cost. The real question is whether
the output is a square wave or is a PWM sine wave with a bit of filtering.
Practically speaking, the square wave will work just fine into a switching
converter for a laptop, since the input stage of that is almost certainly a
bridge rectifier feeding a capacitor. However, if the power supply is
expecting 170V out of the capacitor, it might not get it. The DC/AC
converter is going to produce 115V RMS, and for a square wave, that means
the peak voltage is going to be 115V, not 1.4 times that. To the power
supply, it's going to look like it's being fed with a 80V sine wave. Peak
currents on the polarity changes might also be an issue, as they are with
any capacitor input filter.
Interestingly, if it's a consumer device, and it has switching rates
greater than 9kHz inside, then it probably has to meet Part 15, although,
since it is entirely battery powered (after a fashion) it might not. Of
course, meeting Part 15 radiated and conducted emissions doesn't mean that
it doesn't interfere. (Take a look at all the FCC filings for Broadband
over Power Line (BPL))
4) One might want to look at UPS's... They are designed to feed switching
power supply loads (as witness their radically different VA and Watt
ratings), but, in general, aren't designed for 100% duty cycle continuous
use (i.e. if you know the battery is going to die in 9 minutes, just make
sure junction temperatures don't get too high in that amount of time.. mass
is your friend). UPS's definitely are designed for EMI/EMC rules (CE and
Part 15)
I guess the real answer is to take a converter and hook it up to an
oscilloscope. We'd be interested in line current, fed from some reasonable
source impedance (say, 0.1 ohm or thereabouts) and load voltage and current
into whatever sorts of loads are convenient (a lightbulb would serve as a
nice resistive load, some sort of switching wall-wart would be another, and
a transformer type wall wart would the be the third). If anyone wants to
send me data (scope photos, data files, etc.), I'll be happy to collect it
and post it on a web site. send it to emiemc@luxfamily.com.
Jim, W6RMK
----------------------------------------------------------------------
Subject: Re: Trucks with 7 and more Wides
From: "Daron J. Wilson" <daron@wilson.org>
Date: Thu, 24 Jul 2003 10:15:26 -0700
X-Message-Number: 18
>>Simple things like stripping the wrong path information off and
>>replacing it or simply rejecting the packet with more than WIDE3-3
>>should be part of the network management.
>
>Not simple...
Yes, actually, very simple. If I plug my token ring computer network
card into the Ethernet network, it just don't run. Simple. There are
established protocols that set standards that I MUST adhere to in order
to run data on the network. My failure to do so and I can't get any
data going. I believe that is much wiser than letting me on the network
with incorrect and disruptive parameters.
<soapbox>
I spent many years with packet, enjoyed developing it, grew it into
nodes, etc. I got behind and just dropped into APRS a year ago.
Frankly, there is a fair amount of misinformation and a large lack of
correct information online for APRS settings. Different regions demand
different settings, etc. It reminds me of the published RS-232
'standard' which is implemented differently by almost everyone that uses
it. Honestly, I think it better to control things at the network level
than at the user level.
</soapbox>
>If we simply start rejecting packets, how is anyone going to
>know they are
>doing anything wrong? Yes, many of the header paths we see don't
>make much
>sense. Then, how do we program a digi to reject a path greater
>than
>WIDE3-3?
Again, If I plug into an Ethernet network and don't have an IP address
and subnet programmed CORRECTLY, I may not be able to go anywhere. I
have to reference (somewhere) the correct parameters.
73's N7HQR
----------------------------------------------------------------------
Subject: RE: Question on I-gate
From: "AE5PL Lists" <HamLists@ametx.com>
Date: Thu, 24 Jul 2003 13:39:16 -0500
X-Message-Number: 19
Nothing. The packets carry the path set into the IGate.
73,
Pete Loveall AE5PL
pete@ae5pl.net
>-----Original Message-----
>From: Dave Thege [mailto:dthege@conpoint.com]
>Posted At: Thursday, July 24, 2003 11:36 AM
>Subject: [aprssig] Question on I-gate
>
>I posted this a couple of days ago and got no response. What
>prevents a local digipeater from repeating traffic a home
>station gates to rf from the internet.? Our digester does a
----------------------------------------------------------------------
Subject: Re: Trucks with 7 and more Wides
From: "Gregg G. Wonderly" <gregg@skymaster.cytetech.com>
Date: Thu, 24 Jul 2003 14:49:53 -0500
X-Message-Number: 20
>Trying to understand why it would be ok for someone in Seattle to produce a
>WIDE7-7 that is converted to WIDE4-3, but it's not ok for a distant packet
>to arrive as WIDE7-2 and be dumped. If you were going to strictly enforce
>APRS in Seattle shouldn't it be the other way around?
I think that since the packet has already used 5 hops, that dropping it on
the floor keeps it from bouncing around any longer, right?
One of the paramount questions here is what is APRS for. It was designed
for a different purpose than many people are using it for. Some people use
it for advertising their existence and something that makes them feel good.
Some people use it for messaging around the world via the APRS-IS. Some
people use it correctly. Others would probably use it correctly if they
knew how. Still others are abusing it in ways that do amount to harmful
interference.
All of these things are a reflection of the flexibility of AX.25 routing,
and to the lack of control that such flexibiility will provide. It seems
like it is time to start thinking about more explicit routing control
capabilities for cross country routing. AX.25 doesn't have enough
extensibility to make that easily happen...
-----
gregg@cytetech.com (Cyte Technologies Inc)
----------------------------------------------------------------------
Subject: Re: Trucks with 7 and more Wides
From: Bob Bruninga <bruninga@usna.edu>
Date: Thu, 24 Jul 2003 16:10:49 -0400 (EDT)
X-Message-Number: 21
On Thu, 24 Jul 2003, Gregg G. Wonderly wrote:
>All of these things are a reflection of the flexibility of AX.25 routing, and
>to the lack of control that such flexibiility will provide. It seems like it
>is time to start thinking about more explicit routing control capabilities for
>cross country routing. AX.25 doesn't have enough extensibility to make that
>easily happen...
Well, it could happen easily if IGates just had a DIGIGATE function. That
is, if they see a packet on the APRS-IS that has their DIGIGATE call in
the original DIGIpath, then they will take it off of the internet and send
it back to RF just like a digi would.
THe only problem is the potential for some person 3000 miles from you to
originate packets and FLOOD your local IGate (and RF LAN)...
But I think this is somewhat solvable with filters at the local level if
it gets to be a problem. Then the APRS-INTERNET system is just like a
digipeater.. And nothing in APRS has to change. AX.25 would continue to
work very well... Its only software at the IGate...
de WB4APR, Bob
----------------------------------------------------------------------
Read previous mail | Read next mail
| |
