| EQUIPMENT FOR FLEXNET NODES |
Subject: Equipment for a PC/FlexNet-Digi |
| GETTING FLEXNET NODES TO TALK TO THENET NODES |
| From:
"SANY Ski Reporting Service" To: <info@northeastflexnet.org> Date: Mon, 21 May 2001 10:30:11 -0400 Subject: A question? Hi there. I am setting up a FlexNet digi in Cortland NY and I am trying to find some information on how to gateway FlexNet to our regular TheNet node stack. Do you have any URL's on how to do this. I would appreciate it. I have searched the web and can't seem to find out how to do this. Chris Perrine - KB2FAF Operations Director Ski Areas Of New York, Inc. Chris: TheNet changes: Set parms/modes on neighboring TheNet X1J4 node to the TN<-Flex column in NEDA parms table. see http://www.packetnetwork.org/neda/ (This link has been as reported broken 6/2003 by KA2MYD - Thank you!) Of real importance is to set Parm /23 1 Mode /17 0 which allows FlexNet to digi through TheNet At this point, we set the mnemonic for each TheNet backbone node facing FlexNet to FLEXxx or FLXxxx or FXxxx so TheNet users have a hint how to get into the FlexNet network. For example: If the node is facing FlexNet in Syracuse, which converted to FlexNet, I would make it FLXSYR.. For multiple TheNET nodes facing Syracuse, name them FX1SYR, FX2SYR, FX3SYR... Hence, the backbone port now becomes a visible node with the name "Flex something"... Also good to add more info to assist users on the connect text for FXxxx TheNet nodes. This all can be done remotely with the cooperation of the node SysOp... On the FlexNet node: Add a link statement to each neighboring TheNet node. Format of entry is "Link (from) (to) " Ie "L (port number) (thenetcallsign-ssid) @" or "L 1 W2ABC-3 @" (no quotes !) the @ indicates link to non Flex Neighbor. Any TheNet node you link to will show throughout the FlexNet network. You can step further through TheNet, TNC by TNC, by "L (callsign-ssid) (callsign-ssid) @" .... Hence if node stack W2ABC has a port W2ABC-6 facing another TheNet node K2XYZ-2 you can add the following link statements: "L W2ABC-3 W2ABC-6 @" "L W2ABC-6 K2XYZ-2 @" Following this format you step through a TheNet node(s) to reach FlexNet network. The last step to the FlexNet digi would not have the @ ! Any TheNet TNC you step through needs Parm /23 1 Mode /17 0 set... The "LO"cal text on the FlexNet node can be quite extensive at first to assist users during the change over phase... Those IMPORTANT little DETAILS: Nothing works until the TXDelay is set for a port ! User ports and ports that link to non FlexNet neighbors need an SSID set !!! Any SSID's set to a port MUST be within the SSID range originally defined by MY command. Links to non FlexNet neighbors must have an "@" ! Links to FlexNet neighbors must NOT have an @ ! Links to neighboring FlexNet must point to the FIRST SSID in the range of the neighbor. Links to FlexNet neighbors with SSID beginning with -0 must be entered with no ssid !! Links to "callsign-0" are a special case, used to link to a none FlexNet neighbor without an SSID or SSID-0, such as an HF BBS. When installing a new FlexNet site, set the links on neighboring FlexNet pointing to new site first, then the links at new site pointing to neighbor site! (if you attempt setting the link on the near end first, you will not be able to connect to the far end to enter the link data.) A maximum of 20 link entries can be installed at any FlexNet digi (node)... The SSID on a FlexNet port can be removed by setting it to 16 ! The "U" flag for "User TXDelay test" is extremely harsh, will drop most users, don't use it!! The "Y" flag on port 15 gives automatic sysop privilege to local keyboard. DON'T put a "Y" other ports ! Other lessons we have learned: Better to be extremely verbose at first with the various information texts to assist users.. Link as far as you can, so users can find their way to familiar major nodes... But only link over good solid paths ! (once multiple paths exist, FlexNet will choose the best, and marginal paths can actually help the network as backup links) The 20 limit on link entries may be a minor problem during the change over to FlexNet phase, but disappears as soon as a neighbors convert to FlexNet... Converting every other site to FlexNet builds out FlexNet quickly, but leaves some TheNET nodes isolated with very small nodes lists.. Backbone only node sites (one without a user port, if you have any) may initially remain TheNET, as they were and will remain transparent to users.. Try to establish two paths to neighboring FlexNet network, or at least a fast solid path... A single link over a crowded user (ZOO Channel) frequency is the quickest means to a disaster. Hope this helps !! 73 John WB2CIK |
| FLEXNET ON EMERGENCY POWER |
| -----Original
Message----- From: Dave Packard Sent: Sunday, May 20, 2001 1:18 PM To: Bob Anderson; John Papson Cc: John Kushneir (K2JFK); John Driscoll N2MKH; Jim Wzorek Subject: flex query Hi guys. K2JFK and I were discussing emergency power for our packet nodes, when John brought up the question of how one could supply enough power to a computer in a FlexNet node during a power outage. We understand the advantages Flex has over NetROM in recovering from a power failure, but much of the interest in packet up this way relates to emergency communications. An off-the-cuff estimate of the power requirements of a 386/486 computer sounded like a lot of current drain for a back-up battery. Most of our nodes are on deep cycle battery backup for short interruptions and manual start gasoline generators for any outage that lasts about 12 hours or more. Have you had any experience along these lines? Comments? Regards, Dave K1YHR. Hello All: Yes, computer does draw more power, but overall improvement of FlexNet outweighs this concern. FlexNet's major benefit being it's ability to automatically route around node site and link failures. One has to ask oneself, how often does a momentary power blink extend beyond an hour or two ? What we have found is that there is never enough battery for a real emergency, such as a hurricane or ice storm. We also found that 15 minutes of backup carries us for months on end.. The "up time" in the "ST"atus dump shows how long a node has been continuously running. Backup time can be maximized by careful hardware selection. Trick is to find computers that draw a minimal amount of power. 1) First off a keyboard draws power and is not needed... 2) Use a PC with a highly integrated motherboard, one with large IC's. Also look for one WITHOUT a big heatsink on the processor IC.. Later 386 boards with about 5 big IC's and plastic 386SX processor. 386SX20's work fine for moderate size node stacks. Use of 16550 UART serial cards out weighs need for a super fast PC ! 3) Finally, load the computer with the minimum amount of memory, 1 meg for a small node, 2 megs for a large multiport site. Use newer memory such as SIMM's, modern memory draws less per megabit than older ! 4) Check the label on the hard drive, some draw a lot more power then others ! (always use IDE hard drives, old MFM and RLL drives are power hungry and plain unreliable !) It is possible to replace the hard drive with a "flash card" and small adapter board. A Flashcard has an identical electrical interface as an IDE hard drive. TAPR has a kit for the Flashcard to IDE adapters. 5) It's possible to remove fan, saving more, requires testing before doing.. One can glue heatsinks to all the chips, and let PC convection cool. 6) The floppy can also be disconnected as it also is not needed. Leave it in the machine, just unplug both the power and data cables ! That way it is available if needed for an update or something.. 7) Or it is possible to run FlexNet from a Floppy drive, forgetting about a hard drive all together. (I prefer a hard drive over a floppy, HD's are sealed, floppy's collect dust and dirt.) 8) If you want to run a DXCluster, BBS, or whatever at a remote site, that can be in a big fast computer linked to main node computer. Hence the node computer can be a slower lower power unit... Battery backup only the node computer, and forget about the frills running in the big power hungry computer in time of emergency! I've measured several typical 386SX20 machines, and they all draw under 50 watts, with fan, floppy and keyboard enabled.. (when one adds up how much power all the TNC's and radios draw when passing traffic, it will most likely greatly exceed the PC power drain !) A small UPS with a marine deep cycle gel battery will keep the PC running for hours! Don't forget that most UPS's are designed to keep the monitor running to allow user to shut down. No monitor is needed for a remote site, extending backup by a factor of 4 or more ! A more expensive solution is to use a "Single Board Computer" (SBC) running off of a small 5 volt supply... These typically have built in DOS on a chip, and can run FlexNet out of a Flash ROM. A small one chip switching supply can convert 12 volt battery to 5 volts for the SBC. I have a couple that draw less than 5 watts! Finally, an older Laptop or Notebook PC can also serve remote site. You may be limited to a single serial port, but can drive four 1200 baud links by daisy chaining TNC's RS-232 at 9600 baud... Laptops have the best designed in power management. We have these now running at two sites, one with the display (was busted) and backlight totally removed. 73 John WB2CIK |
| FBB Settings for FLEXNET |
to: flxnet@ww subject: F6FBB 7.00G <--> FlexNet 3.3G Hello All: This is in response to a query by Jocelyn F6IQA on how to interface the F6FBB BBS with FlexNet. As described in his original bulletin, running under DOS, FlexNet with TFEMU loads and runs in the background. F6FBB BBS is then loaded and runs in the foreground. This arrangement does work perfectly, as we are running it at many sites here in the Northeastern USA... The following lines in three files are key to making F6FBB 7.00g work with PC FlexNet 3.3g.. ************************************************************** INIT.SRV file: <snip> # # DRSI and TFPC interrupt (Hexadecimal, default FF and FE) FF FD # <snip> *************************************************************** PORT.SYS file: #Ports TNCs 1 1 # #Com Interface Address (Hex) Baud 5 5 0 9600 # #TNC NbCh Com MultCh Pacln Maxfr NbFwd MxBloc M/P-Fwd Mode Freq 0 0 0 0 0 0 0 0 01/03 ---- FILE 1 10 5 1 230 2 4 10 00/20 UDLY FLEX # **************************************************************** INITTNC1.SYS file: M UISC I WA2PNU-4 O 2 Y 10 **************************************************************** Hope this helps at least one struggling SysOp! |
| A FlexNet Node in a PC Case - Self Contained! |
Greetings,
|
| BASIC BCT COMMANDS |
| Basic BCT commands: "BCT" from DOS prompt will bring up terminal program... "Alt-X" shuts BCT down, do so before leaving a node site... ":C" connects you to node as sysop (if a "Y" is set FlexNet port 15)... ":Q" forces a disconnect from your end... F1 thru F6 are 6 different streams, have to connect individually on each stream. Each stream can have a different callsign if wanted, set with ":MY <Callsign-SSID>" F7 F8 have no function F9 selects between panels F10 is a channel monitor Top Panel is sent text, Center Panel is Received Text, Bottom Panel is Channel Monitor... Home, End, PgUp and PgDn scroll back whatever panel your in (F9)... Ctrl Home, Ctrl End, Ctrl PgUp, and Ctrl PgDn change size of panels... |
| How To TCP-IP with FlexNet |
How-to
install MFNOS on a pc-FlexNet link Now to configure! Simply
connect NOS to pc-Flexnet as you normally would. I
personally happen to use the IPPD module for Flexnet and
use AXIP over ethernet between my MFNOS and Flexnet,
however a KISS serial interface may be treated in an
identical fashion. =>mycall n1uro 1 4 74e2 0 0 N1URO-5 AA1UE-8
FlexNet Connected and the AX.25 route: AA1UE-8 FlexNet Perm default
N1URO-1
|
| Mods to MFJ-1274C or MFJ-1270C for FlexNet 6PACK |
| For all those that
have purchased a MFJ-1274C or MFJ-1270C for the purpose
of running FlexNet's 6PACK type EPROM. As produced, these units are not the same as the previous MFJ-1270B / MFJ-1274B models, as an IC (U40) was added to do memory bank switching. They will not function correctly without the following simple modification: 1) Remove the IC U40. (CD74HCT374E) (Save IC for future use in case you want to use TNC for normal packet activity.) 2) Jumper Pin #16 to PIN#1 of the IC socket. (this grounds pin#16) A small piece of thin solid wire inserted in the socket for U40 will surfice for the jumper. After this modification, it will function as a standard TAPR-2 compatable TNC. (Same modification is needed for NETROM/TheNet EPROMS) 73 John WB2CIK |
| Home | FlexNet Coverage Maps |
| Links to other Sites | The Software & Firmware |
| NE FlexNet News | Node Photos |
| Screen Shots | FAQ's and How To's |
Click on icon to send me e-mail