Transmit-Receive Frequency Offset
This should be zero. Synthesized digital radios have no problem accomplishing this; however, vacuum tube rigs may struggle. Surprisingly, the less expensive HW-series transceivers and the SB’s with the vacuum tube based LMO (VFO), in my experience, have little shift between transmit and receive. Later SB-series transceivers with the solid-state VFO may have an offset. If this offset is > 100 Hz, you’ll notice it during SSB conversations (“leapfrogging” as you chase the fellow you’re talking to), and you’ll want to fix that for digital communications. Happily, the solid state LMO includes a FSK (frequency shift keying) terminal where a small trimmer variable resistor of (say, 25K or 47K) from “FSK” to chassis ground will easily adjust the frequency a few hundred Hertz. To make this automatically switch in and out for transmit/receive, add a small transistor switch (2N3904) between this resistor and chassis ground and switch this by driving the base with a 1 Mohm ½ watt resistor to +300VDC available on one of the sections of relay RL2. Choose the connection in order to move the transmitted and received frequencies closer together, then adjust the variable resistor until they are within 25 Hz or so. The problem is solved. A friend on a SSB contact can help you figure out when you sound “right”. Otherwise, you can measure the frequency of a second (solid-state) transceiver’s transmissions to which you’ve tuned, and then compare the frequency of your vacuum tube rig (when both are sending the same audio tone). You can do this with two rigs both connected to dummy loads; send an audio tone into the reference rig and tune the receiver of your test rig until the tones match. Then measure both transmitters’ frequencies.
Latency
You need to switch from transmit to receive in under 250 mSec, and preferably around 100 mSec, to keep up with the handshaking responses. The RMS server to which you connect will be firing responses back quickly, and if you miss them your throughput and connection suffer badly. Your Heathkit, when controlled by the Signalink, will usually achieve this (assuming VOX delay and Signalink delay are all minimized), but if you make a small modification to the T/R receiver control circuitry within your Heathkit you’ll have latencies below 100 mSec. V14A audio preamplifier is turned off by a large negative voltage to grid during transmit (through relay RL2), charging capacitor C322 (primarily) to that negative voltage. When the relay moves to “receive” position, a substantial delay occurs because that voltage is leisurely bled to ground through 3.3 Mohm R336. Heathkit purposefully created this delay, to avoid clicks and thumps in the speaker while the relay repositioned, as extremely low latency wasn’t needed then. You can dramatically speed up the transition by wiring a 1N4007 diode in series with a 100K resistor and putting this assembly in parallel with R336 3.3 Mohm resistor, with the cathode of the diode oriented toward V14A (or away from the relay contacts). An alternative solution would be to change the 0.02 microfarad C322 to .001 microfarads. Your latency, as measured by WINMOR, will drop below 100 mSec with this very simple upgrade.
Signalink Connections
Because Hams typically made up their own mic and speaker connections to Heathkits, Tigertronics provides an unterminated shielded multi-twisted pair cable (RG45 to fit the Signalink) similar to CAT6 computer cable. You’ll need to make your own connections to the microphone and speaker wiring. If you can find shielded CAT6 cable, you can use that. Connect the foil shield to the chassis ground of the transceiver.
RFI Prevention
All of these digital phase or frequency-shift signals over audio are fairly sensitive to radio frequency interference, including the Signalink and your computer running the RMS EXPRESS software. If you have trouble connecting to WINLINK servers, and your frequency is correct, this is the most likely cause, and it is very insidious. Use the best grounding RFI prevention techniques you can to insure success. There are two additional items that may make a huge improvement, and you should do these pre-emptively. These two pre-emptive actions are: adding an effective “RF Isolator” to the coax line out of your transceiver, and adding ferrite snap-on cores around ALL signal lines. There are multiple manufacturers of RF isolators.[7, 8]. Ferrite snap-on RF chokes can also be found in various sizes and from many suppliers. Representative chokes can be found here:[9]
Speed Considerations
The least expensive method of getting digital connection to the WINLINK system is certainly via the WINMOR software-based TNC (terminal node controller), which can use a sound-card type interface such as the Signalink. For most purposes, this provides adequate throughput, somewhere north of PACTOR II. With the email overhead, even a simple message will take a minute or two to transmit. If you really need speed, you can get faster performance by moving to a hardware-based TNC, using PACTOR protocol III (or if outside of the United States, using PACTOR protocol IV). PACTOR is a patented protocol; an example is the P4-DRAGON 7800[10], which is widely available commercially, as it is used by many sea-going sailors. Used PACTOR II or III capable SCS modems are sometimes available at greatly reduced prices. Your transmit-receive latency may not be quite up to what solid state rigs can perform and are normal for PACTOR. Your preferred RMS server station sysop may be willing to adjust upwards their delay (default: 30 mSec); there is an option for them in their Trimode software settings that allows up to about 80 mSec of delay.
Solid EMP Protection
Once you get this all going, you’ll want to add some sort of metallic Faraday protection around your laptop (or other computer used to generate, send, and receive your email) and the Signalink, as well as any power supply for the laptop. Pass the AC power wires from the computer through a small perforation through the Faraday enclosure and put ferrite chokes on the AC wires both just inside and just outside the perforation. Pass the wires from the Signalink to the Ham transceiver through minimally-sized holes through the Faraday enclosure, and add another shield around them all the way to the transceiver (using aluminum foil or copper braid) and connect that shield to the Faraday enclosure. Apply ferrite chokes on the Signalink wires just inside and outside the Faraday perforation, just like the computer AC wires. The Heathkit transceiver needs an appropriately-chosen gas-discharge surge arrest or connecting the coax center conductor to shield (350 volts is reasonable for just the transceiver) as EMP protection, and then it probably does not have to be within the Faraday enclosure. Provide three-wire MOV surge protection for the AC lines feeding the Ham rig and the computer. If you are unable to support a vacuum tube transceiver and attempt to use a modern solid state transceiver, significant additions will need to be made to the transmitter output and receiver input circuitry. The antenna feedline is likely the source of your greatest incoming surge, because the antenna is specifically designed to absorb electromagnetic energy.
Conclusion
Being able to maintain access to world-wide email abilities, including attachments, completely independent of the wired Internet is a significant step toward guaranteeing communications redundancy for yourself. Thousands of users have already succeeded, and tens of thousands of messages are handled every month, all over the world, in boats on many different waters and on land in many situations. Many emergency coordinators are also turning to this option. This is a great capability for a prepared individual to attain!
References
[7]T4+ isolator from RadioWorks: http://www.radioworks.com/ct-4.html
[8]http://www.mfjenterprises.com/Product.php?productid=MFJ-2912 FJ-2912 wall mount balun RF isolator.
[9]Ferrite snap on chokes that accept up to a 0.250” diameter cable. https://www.digikey.com/product-detail/en/laird-signal-integrity-products/28A2025-0A0/240-2074-ND/242802