| |
ZL3AI > APRDIG 02.04.04 18:22l 241 Lines 9264 Bytes #999 (0) @ WW
BID : 3084-ZL3AI
Read: GUEST
Subj: TAPR Digest, Mar 30, 4/6
Path: DB0FHN<DB0RGB<OK0PPL<DB0RES<ON0AR<ZL2BAU<ZL2BAU<ZL3VML
Sent: 040402/1219Z @:ZL3VML.#80.NZL.OC #:21820 [Chch-NZ] FBB7.00i $:3084-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: RE: Kenwood TM-D700A as a DIGI?
From: "Scott Miller" <scott@opentrac.org>
Date: Tue, 30 Mar 2004 15:55:21 -0800
X-Message-Number: 20
Or wait 6 to 8 months and check http://www.omnipacket.com. =] Hopefully
it'll end up being cheaper than the KPC-3+, anyway.
Scott
N1VG
----- Original Message -----
From: "Christensen, Eric" <CHRISTENSENE@MAIL.ECU.EDU>
To: "TAPR APRS Special Interest Group" <aprssig@lists.tapr.org>
Sent: Tuesday, March 30, 2004 3:40 PM
Subject: [aprssig] RE: Kenwood TM-D700A as a DIGI?
> Dave,
> I have used my D700A for a digipeater on a special event. Just include
> WIDE, RELAY, and TRACE in the UIDIGIPEATER part under the APRS settings.
> BUT, for a lot less, you can get you a TNC2 clone and get a real
digipeater
> firmware (UIDigi?) for it and hook it up to a radio. You will probably
> spend half as much as what you paid for the D700 and it will support n-N
> protocols and lots of other things to help keep the network clean.
>
> It is a lot better choice.
>
> Eric KF4OTN
>
> -----Original Message-----
> From: Dave Land [mailto:kd5fx@despammed.com]
> Sent: Tuesday, March 30, 2004 12:28 PM
> To: TAPR APRS Special Interest Group
> Subject: [aprssig] Kenwood TM-D700A as a DIGI?
>
>
> Has anyone used the TM-D700A as a DIGI? If so, how well did it work? What
> settings did you use? I've got an VHF antenna at 250ft and looking for a
> good, fairly inexpensive but easy setup DIGI for APRS.
> Thanks & 73,
> Dave
> KD5FX
>
> ---
> You are currently subscribed to aprssig as: christensene@mail.ecu.edu
> To unsubscribe send a blank email to leave-aprssig-22306P@lists.tapr.org
> Questions regarding the SIG go to the SIG administrator: wa1lou@tapr.org
>
> ---
> You are currently subscribed to aprssig as: scott@opentrac.org
> To unsubscribe send a blank email to leave-aprssig-22306P@lists.tapr.org
> Questions regarding the SIG go to the SIG administrator: wa1lou@tapr.org
>
>
----------------------------------------------------------------------
Subject: Re: New tracker design suggestions
From: David VanHorn <dvanhorn@cedar.net>
Date: Tue, 30 Mar 2004 18:56:32 -0500
X-Message-Number: 21
>
>It is going to be the setup charges that kill you in manufacturer, in
>particular the pick and place tapes for SMD. Not really that much? I think a
>few hundred dollars, but with small projects that can be your profit margin.
>Plus, there is always the risk something wrong.
The "easy-Bake" SMD production method is somewhat labor intensive, but very
cost effective for small quantities.
>A thought that has often crossed my mind, but I didn't try it yet, is make
>"convertible parts" in my PCB program. For example, a 1/8 watt through hole
>resistor that also has 0805 pads on the inside of it. That way, when your
>designing it, it is a standard resistor on the schematic, but the PCB part
>gets laid down as a 0805 part, with two standard through hole pads flanking
>it. Same thing with 0.1" thruhole caps, maybe with a 0603 on the inside of
>that.
Might be better to route two PCBs, otherwise the SMD version isn't gaining any
of the size benefits.
One approach that's easy to do with standard tools, is Thruhole ICs, and SMD Rs
and Cs on the bottom of the board. 1206 parts fit nicely under a DIP, using 8
mil track and space, and the right pad sizes.
----------------------------------------------------------------------
Subject: Re: New tracker design suggestions
From: "Scott Miller" <scott@opentrac.org>
Date: Tue, 30 Mar 2004 16:09:07 -0800
X-Message-Number: 22
> It is going to be the setup charges that kill you in manufacturer, in
> particular the pick and place tapes for SMD. Not really that much? I think
a
> few hundred dollars, but with small projects that can be your profit
margin.
> Plus, there is always the risk something wrong.
Yeah... a lot of constraints on my design are imposed by manufacturing
restrictions, like the panel size used by the board house, the enclosure
sizes available, and so on. There's no way I can afford a custom
injection-molded enclosure, but there are semi-custom solutions where you
can get end plates and things customized to go in standard products.
> A thought that has often crossed my mind, but I didn't try it yet, is make
> "convertible parts" in my PCB program. For example, a 1/8 watt through
hole
> resistor that also has 0805 pads on the inside of it. That way, when your
Sounds like a layout nightmare. I've settled on 0805 size parts wherever
possible. They're compact and still fairly easy to handle for
hand-soldering. This board's going to be tight as it is, without using lots
of thru-hole parts. My OpenTracker's got so many holes, the board house
charges me extra for the drilling. It's not a big deal for production, but
it makes prototyping a pain.
> That would give folks the option of doing the kit as SMD or as standard
thru
> hole. And the pad would do double duty as a via as well, which you almost
I'm not even sure I'll make it available as a kit. It'd be a pain to
package everything properly, too easy for parts to get lost, too much
trouble to help troubleshoot someone else's poor soldering job, and besides
that the MCU has to be programmed with the bootloader code first. For small
quantities, I can get the boards professionally assembled overseas and not
add more than 15% to the price.
> Oh, one other thing, SMD wise. I was talking with one of my assembly
houses
> the other day, and he was telling me if I used resistor arrays it would
lower
> my costs. Apparently he chargers a placement cost per part, so going from
8
> resistors to one 8 up array would lower my cost quite a bit.
Yeah, depends on how they calculate the cost, I guess. Resistor arrays can
also be less convenient for routing, and might mean adding more vias.
Scott
N1VG
----------------------------------------------------------------------
Subject: RE: Kenwood TM-D700A as a DIGI?
From: David VanHorn <dvanhorn@cedar.net>
Date: Tue, 30 Mar 2004 19:15:58 -0500
X-Message-Number: 23
At 03:55 PM 3/30/2004 -0800, Scott Miller wrote:
>Or wait 6 to 8 months and check http://www.omnipacket.com. =] Hopefully
>it'll end up being cheaper than the KPC-3+, anyway.
Over the air updates? That's impressive, assuming it can handle packet loss and
duplication gracefully.
----------------------------------------------------------------------
Subject: Re: New tracker design suggestions
From: "Scott Miller" <scott@opentrac.org>
Date: Tue, 30 Mar 2004 16:17:02 -0800
X-Message-Number: 24
> The "easy-Bake" SMD production method is somewhat labor intensive, but
very cost effective for small quantities.
You mean doing reflow with a toaster oven, like this:
http://www.seattlerobotics.org/encoder/200006/oven_art.htm ?
Haven't tried it, for the simple reason that the solder paste is $42 for 35
grams, has to be shipped overnight and refrigerated, and then has a shelf
life of weeks. It wouldn't be so bad if I could keep a tube around for a
few months.
> One approach that's easy to do with standard tools, is Thruhole ICs, and
SMD Rs and Cs on the bottom of the board. 1206 parts fit nicely under a DIP,
using 8 mil track and space, and the right pad sizes.
The bigger problem for me is that most of the parts I want to use aren't
available in thru-hole at all. Only one of the critical chips in my design
is available in a DIP package, and that's a big 42-pin SDIP that'd take up a
good portion of the board. The modem, FRAM/RTC, and USB interface are only
available in SMT packages.
I figured it was finally time to bite the bullet and start working with the
stuff. As long as you've got a good soldering iron, fine-tipped tweezers, a
flux pen, and good eyesight, it's really not so bad.
Scott
N1VG
----------------------------------------------------------------------
Subject: RE: Kenwood TM-D700A as a DIGI?
From: "Scott Miller" <scott@opentrac.org>
Date: Tue, 30 Mar 2004 16:23:54 -0800
X-Message-Number: 25
> Over the air updates? That's impressive, assuming it can handle packet
loss and duplication gracefully.
I've got a good idea of how I want to do this. It'd be done with UI frames,
using a distinct PID to avoid confusing other devices. You'd basically just
set up a workstation to start dribbling out frames over the course of maybe
a day or two, and after they've all been sent, the system would ask the
digi(s) what blocks it's missing. It'd go into another slow transmit cycle,
and ask again when that's done. When it's satisfied that all targeted
devices pass checksum, it'd issue a write command to update the firmware.
Of course, a ping operation at the start of the transfer might be a good
idea to make sure you're actually getting through to the device...
60,000 bytes at 120 bytes/sec = 500 seconds of transfer time, over 48 hours
is maybe 15 seconds of airtime per hour, including overhead. Not too bad,
especially if you're able to reach multiple devices at the same time.
Firmware updates with WIDE7-7 would be discouraged, though. =]
And of course, it'll have a security mechanism to keep others from messing
with your digi...
Scott
N1VG
----------------------------------------------------------------------
Read previous mail | Read next mail
| |
