PoC or GTFO, Volume 2

by Manul Laphroaig

  N0CALL-9>N1CALL-9,WIDE1-1,WIDE2-2::N1CALL-9 :This is a test for APRS messages{1

  Internet Gateways

  Gateways between the Internet and APRS radios are known as Internet Gateways or iGates. Typically iGates are used to forward APRS beacons heard over radio to some website, but there are a lot more interesting things we could do with them.

  Tricks with iGates

  Some iGates support transmitting data from the Internet out to radio, effectively bridging the local RF spectrum to the APRS-IS network.

  There is no official way to list iGates, so our best bet is connecting to the backbone servers they report to, passively listening for frames and beacons that announce their presence. We would also like to distinguish iGates that are capable of transmitting from those that only receive. When we find some such iGates, they allow us to perform some gnarly tricks!

  We can send an APRS message from an Internet-only host in Asia to an individual driving in Pittsburgh with only a radio receiver and a TNC. Hide locations of control sites by first proxying your packets through the Internet iGates, only to target your local RF nodes through a separate, sacrificial iGate bridge.

  The system is only limited by APRS-IS rules in terms of traffic congestion control. Because all RF nodes receive from and transmit to the same frequency, overlapping transmissions can and will reduce the ratio of successfully decoded packets for everyone else. Therefore, be neighborly!

  Traffic caps are enforced by the iGate operator’s configuration. Commonly a given node, as identified by its callsign and SSID, will only be able to use the Internet-RF bridge for transmitting a fixed number of packets each minute. This is to prevent accidental jamming of the RF channel.

  Packet Validation and RF Digipeating

  Some architectural limitations of APRS need to be considered carefully. First, most iGates in the APRS-IS network will only digipeat packets to the RF side if the station is located within a fixed radius of so many kilometers. Second, we might not get to know if a given area has an iGate capable of bridging RF, or transmitting to RF. We can’t simple wait for a response, as APRS is a response-less protocol. Third, packets marked RFONLY in their path won’t reach APRS-IS. Packets marked TCPIP won’t reach RF nodes. iGates forcing or restricting either will be dead-ends if we aim to bridge over APRS-IS. Finally, user-defined packets are ignored by most of the APRS-IS infrastructure. For example, aprsc ignores them. Third-party packets are allowed, with caveats.

  Bypassing Validation

  There are a few ways to bypass the restrictions imposed on bridging RF in iGates that require geographical proximity.

  You can try to spoof your location by sending a beacon positioned at fake coordinates near the iGate. You can then send your actual data packets, remembering to regularly send a position beacon to the iGate to remain in the last-heard list.

  You could limit use of user-defined packets to RF side, operating a a rogue iGate that does not ignore them, instead transforming them to third-party or steganographic standard packets, delivered to APRS-IS. User-defined packets are not displayed by most equipment. This also applies to unused or obscure DTIs.

  To avoid potential roadblocks, the following considerations may help. If trying to reach the RF side, do not use—and verify that the iGate/APRS-IS nodes don not use–TCPIP in the path. If trying to reach the Internet side, do not use RFONLY in the path. To avoid packet drops from rate limiting, throttle your packets, sending just one every few minutes.

  Albeit completely illegal on the actual air, as an experiment in a controlled environment, automatically generated callsigns can be rotated to avoid being detected or banned from the system.30 Finally, client version strings, as used during registration with APRS-IS nodes, could be rotated and mimic real clients.

  Looking up standard TCP/IP pivoting techniques may help for accessing the APRS-IS network, but first and foremost, remember to be neighborly.

  International Space Station (ISS) and APRS

  Space, the final frontier! It suffices to say that a digipeater installed onboard the ISS makes APRS into the tool of choice for legal ruckus communications on a worldwide scale. So as long as the TNC of the ISS’ radio validates your packets, you can deliver your covert messages in a fully decentralized fashion!31

  Whether commercial TNCs out there relay packets with unused DTIs is a question left to the reader as an exercise.

  Parting words: legal status of subterfuge in radio communications

  Amateur radio laws generally prohibit steganography and also encryption, with a few narrow exceptions.32 For example, the US Electronic Code of Federal Regulations §97.309 states, “RTTY and data emissions using unspecified digital codes must not be transmitted for the purpose of obscuring the meaning of any communication.”3334 Governments do monitor the airwaves where they care about them the most, and having your antennas, expensive equipment, or house ransacked sucks. Also keep in mind that amateur radio is self-policing; if you mess up and create a nuisance that affects everyone else, your future experiences with that small, tight-knit, but global community may be seriously soured.

  So be neighborly, have fun, and stay safe!

  —Vogelfrei

  9:10 The Galaksija Home Computer

  by Voja Antonić

  This article on the Galaksija computer first appeared in the January 1984 special edition of Dejan Ristanović’ Yugoslavian science magazine, also called Galaksija. We reprint it in English as a salute to fine neighbors such as Mr. Antonić, to all those who build strange and lovely contraptions in their basement laboratories and then share them with the world. —PML

  Do It Yourself Guide for the Galaksija Computer

  A serious but pleasant work awaits us, which will be rewarded with the unusual satisfaction of having created an intelligent device. Do not feel discouraged if you don’t have a lot of experience. That is a sign that you have a self-critical spirit which is, in this business, much more important than experience. Take a moment’s pause to examine every minute detail; if it’s well done, the Galaksija will surely work on the first try!

  Important Decisions

  Before we start working, we need to make a few important decisions. First, do we want this system to be final or will we leave space for potential future expansions such as a printer, more memory or a music box? If we don’t want these expansions, we save one additional multi-pin connector and one integrated circuit. (74LS32, for which we instead use just a short circuit marked with dashes on the mounting diagram.) If you are unsure, we advise that you do mount these two parts, although it’s never too late for that afterwards, either.

  Mounting: Layout of Galaksija components

  The second decision is whether to use a raw or RF modulated video signal. Raw video signals don’t require an additional RF modulator and give a stable, higher quality image, but they can’t be used with just any TV, requiring either a special display or a black and white TV modified with a raw display input. This modification does not require any additional investment, but it does require certain prior knowledge and experience in working with TV receivers. Next, a TV like that must be transistor based (vacuum tube ones are not suitable), and it has to have a mains transformer (and not a so called “hot chassis”). Usually, both of these requirements are satisfied on smaller, portable, black-and-white TVs that have a 12V battery connection. We’ll go through some of the details for adding a proper display port to such a TV further in the text. But, if we do install an RF modulator, we are freed from all these complications and we’ll be able to connect the computer to the antenna port of any TV.

  We will also have to decide which ICs to socket and which will be soldered directly to the board. You should definitely use sockets for the EEPROMs (2716 and 2732), but for the rest, the choice is yours. The advantage of using sockets is that there’s less risk of damaging an IC and it’s a lot easier to diagnose a problem by swapping ICs because desoldering ICs is a very delicate job. Unfortunately, if the sockets aren’t of the bes
t quality, they can cause problems with bad contacts. To be very reliable, a socket must be of high quality, and that can sometimes make it more expensive than the IC it holds!

  Because of high quality and affordable price of professionally made PCBs, making them yourself isn’t worth the time.

  Connections to the outside world:

  Inputs and outputs on the back of the Galaksija

  Connector pin numbers and descriptions.

  Double sided PCB layout: Expansion connector in a form of a printed circuit board.

  Keyboard mask: The final layout depends on the space bar type, so you should wait for keyboard parts to arrive before making this part. Those who ordered the keyboard in the first round don’t have to worry, the parts will fit perfectly.

  The heart of Galaksija computer: Z80A microprocessor and 2732 EEPROM with BASIC interpreter.

  1. In front of us we have laboriously gathered all the parts which will, in a few hours, grow into a Galaksija computer. At the bottom we easily recognize buttons and caps of keys with printed labels, to the right we see 1/8W resistors, with capacitors to their left and integrated circuits in the middle. Make note of the MOS and CMOS ICs.

  2. Because the PCB is single layer, we will need a lot of jumpers. They are easy to make from a single core copper wire that you can easily source from popular blue-white telephone twisted wire pair. The fact that they are of standard length (5, 10, 20, 30 and 40mm) makes things easier, so you can easily make a tool for their precise bending. (Take note of wire gauge when making the tool.)

  3. We start building the computer by placing the first jumper. Some jumpers pass beneath the ICs; this won’t create problems if the jumpers are neatly bent and rest flat on the PCB. (This view is from the component side and not, as it may first seem, from the trace side.)

  4. When we turn the board over to solder the first jumper, it’s obvious why we start soldering the lowest components first. If we had, for example, started with keys, other components would fall out when turning the board. If you haven’t soldered before, it’s good to first experiment a bit on another board. The tip of the soldering iron should be prepped with a file, cleaned and tinned. Put solder on one side and hot soldering iron tip on another side of the pin. Be careful not to leave too much solder on the pad, because however odd it might sound, this would make a bad soldering joint.

  5. All jumpers are in place and soldered. Count them carefully: there should be exactly 119. If you are missing some, consult the mounting diagram. Pay close attention to the 74LS32 IC; as we said at the beginning, we can substitute it with a jumper (dashed line on mounting diagram) if we don’t want future system expansion connectors. That would then make 120 jumpers.

  6. The next phase is soldering the resistors, which are very similar to 10 mm jumpers.

  7. When mounting ICs, take care to use the correct orientation, because even hardened professionals sometimes mount the ICs backward. Some are marked with a semicircle as on the mounting diagram, while others have a dot over pin number 1. It should be pointed out that the inscription on the IC isn’t always printed so it starts from first pin. Since the PCB has a silk screen marking component orientation, there should be no problems.

  8. The ICs are mounted, but not all of them. We leave out MOS and CMOS ICs CD 4017, CD 4040, 6116, 2716, 2732 and Z80A. It’s best to leave them for the end, but there is no reason not to solder their sockets. Now is the time, before soldering, to check once again that the ICs are all in the right places and correctly oriented. We aren’t repeating this to be pedantic: every bit of impatience and negligence when soldering can cost a lot when first turning on the unit.

  9. Soldering the ICs requires some precision, as distances between pins are only 2.54mm, and they sometimes have a trace going between them. If, a solder bridge is accidentally created between two pins, the simplest way to remove it is by applying more fresh solder on the same place and them removing it all with the tip of the soldering iron.

  10. Next by height are capacitors. Let’s then solder them, too. It is advisable to use disc capacitors as they are smaller and cheaper, but if they are hard to procure, use whichever you have. Capacitance values and voltages aren’t critical. We will skip soldering C5 as, with a suitable quartz crystal, it probably won’t be needed. We’ll say more about that when we come to powering on the unit.

  11. We also have two NPN low power transistors on the left and right sides of the PCB. A little bit of caution and we won’t make a mistake when soldering these; looking at the transistor from below, we can see that its pins form an isosceles right triangle. The holes for transistor pins on the PCB have the same layout. There’s a place for a small diode at the upper left corner of the PCB. Usually, a diode will have a ring marking a cathode side of its cylindrical housing.

  12. We have reached the keyboard mask! Whether you have cut your own out of FR4 or aluminum, which we wouldn’t wish upon our worst enemy, or you ordered it directly with keys, it is essential: without it every key would move around and caps will scrape over each other. The mask is self standing, so it doesn’t get connected to the PCB in any way.

  13. First, place a couple of keys at the corners of the keyboard mask without their caps, then solder them in so the mask is stable. Take care that the keys aren’t backward: you can see that on the mounting diagram, the pins are toward us. Jumpers won’t pose any problems because they are placed right between the keys. After that, it’s easy, as all fifty-five keys are the same.

  14. Since we are nearing the end, we’ll solder or socket the remaining MOS and CMOS ICs. Be careful, as these ICs are very sensitive to static electricity. You should study the “Dangerous Paths” section of this article first.

  15. Click — click — click! Put the caps on all the keys and the whole thing is starting to look serious. It’s almost taunting us to start programming, but we’ll need to have a little patience.

  16. Notice that the ENTER keycap is twice as wide as the rest. That one is mounted on two keys. Taking a closer look at the traces on the PCB, you’ll see that the contacts of those two keys are connected in parallel. Therefore, only one of the keys has an actual function, the other is just there for mechanical reasons.

  17. The choice of jacks we’ll leave up to you. You can use whichever you have, as long as they have at least three pins. As far as we can tell, the standard 5-pin DIN plugs are perfectly usable and easy to get, as they are made by Ei. They are cheap and, who would have guessed — very reliable. Since they all have five pins we suggest the same layout as on the mounting diagram. A good feature of this layout is that we won’t cause any short circuits by swapping the jacks by accident.

  18. Since it’s not very easy to find a multi-pin connector in our country, we have designed the PCB so it’s possible to mount several different types of connectors, if they have the standard 2.54mm spacing. As optimal solution, we have decided to add one more, small, double-sided PCB that is designed in such a way so that a 44-pin edge connector can be used with it, because this connector type is the easiest to find at an affordable price.

  19. Of course, now we will make a final check of the whole PCB by shining a strong light through it and carefully examining every trace. Minuscule solder bridges are very common. Take a look at the circled part of the image; we’ve found a bridge which shorts together two traces!

  20. Our labor has been rewarded by the beautiful sight of nice and tidy PCB, a device which will repay all the labor and patience in multitude. Galaksija will work for you much better than many electronic devices in this era of electronics, exhibiting one characteristic we haven’t seen before. It will communicate with us in such a way that we’ll start to think of it as part of a family. And really, it’s no wonder that many people consider their computers their friends, too!

  Dangerous Paths

  If you already have a few working projects behind you, you probably won’t follow every piece of our advice. But there are some rules you should never break because those certainly c
an lead to permanent damage to components.

  Short circuit between positive and negative power supply traces of the computer will damage the 7805 voltage regulator. Some manufacturers build this IC with over-current protection built-in, but it’s better not to even test it. Similarly, accidentally swapping the polarity anywhere between power supply and the computer would probably prove fatal to all ICs.

  Almost all ICs in the Galaksija computer have a working voltage of + 5V, with tolerances of ± 0.25V. ICs will survive over-voltage of up to 7V, but anything higher is dangerous.

  Short circuiting any pin of a 74LS-series TTL IC to a positive rail will lead to permanent IC damage. Short circuits to ground are harmless and we can use this to experiment. You should still take care that not too many pins of any one IC are grounded at the same time.

  In case of bad image synchronization on the screen, we’ll have to experiment with different values for resistors R12, R13, R9 and R10. Having R12 or R13 less than 330 Ohm poses no problem, as well as having R10 less than 40 Ohm.

  Connecting the raw, unmodulated display output to a TV receiver with a hot chassis poses danger not only to ICs but to your own life. A later section describes these modifications.