Entries from November 2012 ↓

PSK Interface

In it’s most basic configuration a computer to radio interface for PSK and other digital modes consists of the following:

1. Radio audio output is connected to sound card mike input.

2. Sound card audio output is connected to radio mike input.

3. Com1 activated PTT, goes to PTT mike input.

This basic approach, while conceptually accurate, is doomed to failure. The signal levels coming in to the sound card will most likely overdrive the card and generated IMD on received signals. The sound levels going to the mike input of the radio will most likely overdrive the mike input causing IMD on the transmitted signal. Then the fact that there is no galvanic isolation between radio and computer will most likely introduce hum that will swamp all the other signals anyway. All these problems have simple solutions and there are more than a few simple solutions. Unfortunately there are even a greater quantity of not so simple solutions. Complicated circuits are not needed. The following solutions were chosen for their simplicity and effectiveness.

1. Transformer isolation between radio audio and sound card input with fixed resistance signal attenuation.

2. Transformer isolation between sound card and radio mike input with adjustable signal attenuation.

3. Push to talk is tone generated by the transmitted audio signal.

Transformer isolation prevents ground loops that cause hum and interference. There are other ways to eliminate ground loops but the transformer method is inexpensive in this case and works every time.

Note that the sound card is stereo. There are two channels in each audio connection. Use stereo 3.5mm connectors and utilize both channels by using coupling capacitors to combine both channels. Some of the PSK software available may only use one channel. By combining both channels, you don’t have to worry about which channel is the active channel.

If your radio is equipped with a constant volume audio output, use it for the audio input. The constant volume level allows a fixed attenuator to reduce the volume to levels that the sound card can handle. Once set, it needs no more adjustment as long as the volume from the radio remains at a constant level. Note that this assumes the constant volume output is independent of the speaker and speaker outputs are those which are effected by the radio’s volume control. You do not want the volume control to effect the level of audio going to the sound card because overdriving the sound card input will cause IMD on the received signal. Yes, indeed, those helpful folk advising you of IMD on your signal might actually be causing that IMD themselves by overdriving their sound card.

A potentiometer adjustment is necessary for the microphone input to the radio because of the high sensitivity of that input. The input level should be adjusted to 20 mv for a TS-950. Other transceivers would require levels in that same range.

We decided to reserve the comports for typical com port applications. Besides, a MacBook does not have comports. We had no desire to complicate the project with comport to USB converters. We also wanted a plug and play device which could be used with an Apple MacBook. The MacBook has USB ports but a very simple sound interface providing one input and one output, only.

Deriving the push to talk signal from the audio out to the transceiver mike input is much like using VOX but without actually having to use the VOX feature of the transceiver. Some interfaces go to great lengths providing all sorts of op-amp circuits to implement the audio derived PTT function. All that is required is a simple rectifier circuit working off the sound signal from the computer. The simple rectifier circuit provides enough voltage to turn on a transistor switch which provides the PTT function. No external power is required.

This entire circuit is built into a cast aluminum box resulting in a fully shielded interface. Two cables from the box go to the computer to connect the audio input and output from the sound card. A specially made cable composed of three individually shielded cables connect to the audio out, microphone in, and PTT terminals of the 13 pin ACC2 connector of the transceiver, a TS-950 in this case.

The tone generated PTT did not work. I only got .37 vdc to the base of a 2N2222. I checked for continuity with an ohmmeter from collector to emitter and did not find any with .37 volts on the base. Perhaps I should have tried a different transistor or maybe a FET. The original circuit called for a 2N3904. By accident I found that the VOX in the TS-950 worked just fine with the audio from the interface so I ended up using that. Doing so risks transmitting computer sounds but the MacBook is pretty quiet as it is, not as noisy as when run under XP and I hardly ever run XP on the Mac.

Signals in and out are very clean and neither signal needed any adjusting. The component values from the original circuit worked perfectly.


Still having trouble with the vox. It seems to be a problem with level. Not a strong enough signal to be reliable. FLDIGI has a strange soundcard volume adjust so we switched to DIGIPAN which seems to work right now. Had to opt to use XP because DIGIPAN does not run under OS X. That is okay because we want to use TIGHTVNC which also needs XP. Besides DIGIPAN copies multiple signals at the same time but only PSK. Where as FLDIGI only copies one signal at a time but can do many different modes. That is Okay, not interested in multiple modes but am interested in multiple signals.

I have ended up with a very basic and simple interface that uses two miniature transistor output transformers. Their primaries are around 1k ohms with secondaries around 8 ohms. The constant volume output of the accessory connector goes to one 8 ohm winding through a 1.5k resistor. The secondary of that transformer has one leg in series with a 40k resistor and is shunted with a 2.2k resistor before being connected to the cable that feeds the mic input to the sound card. Or the line input can also be used. Both channels are used and connected through .1mf capacitors. That setup works well to give me a decent waterfall display. Volume levels can be adjusted in software but I found that setting the levels to maximum worka just fine.

That takes care of the receive side. On the transmit side we take the output from the sound card and couple both channels directly into the 8 ohm transformer winding through .imf capacitors. The secondary has one leg going to ground. The other leg also goes to ground through a 4.7k resistor. The hot side of the 4.7k resistor is connected to a 100k resistor which goes to the mic input of the transceiver. That mic input is found at the accessory connector of the transceiver.

We use VOX to key the transmitter. I can understand folk using the serial port for PTT for transceivers and transmitters not equipped with a decent VOX feature, but when you do have a decent VOX feature it would be silly not to use it.

The most expensive part of this project is the cast aluminum box. The hardest part of the project is making the cable that goes from the transceiver to the interface box.

Bootable USB Memory Stick

There is nothing magical about making memory storage devices bootable. We have been doing it ever since magnetic storage media became available but the largest amount of memory available on the old floppy disks was 1.2 meg. Today we are dealing with memory sticks of 4, 8, 16, and 32 giglebytes.

The term bootable refers to loading an operating system to ram memory, from there the operating system can load other applications and operating systems depending on what the user wants to do. Just about any operating system will do but DOS based systems are readily available along with utility software to make boot disks (sticks).

Since we will be dealing with a 15gig memory stick and wish to have all the memory included in one single partition we have elected to use the windows 98 system files and the HP utility to create a bootable memory stick. Both the HP utility and the win98 system files may be downloaded from the internet. Since the download links might change by the time this is read, search on HP memory stick and win98 system files for the software which can be run under XP. The process formats the memory stick with FAT32 file system and installs the win98 system files to the boot sector.

Once created the memory stick can be tested by setting the computer bios to boot off USB, inserting the memory stick into a USB port, and doing a cold re-boot of the computer. The screen should come up to a DOS prompt and announce the win98 operating system. Since there is no application software, this is a perfectly useless system but it is bootable. NOTE: Booting off USB does not prevent the computer from booting off other bootable devices. At least not for the BIOSes used in the systems we use. Our computers are set up to boot in the following order; USB, CD, Floppy(A), Hsrddrive. Thus if there is no USB stick, or CD, or floppy disk inserted, then the boot is off the hard disk and up comes whatever the hard disk boots to.

In our case we wanted to create a bootable version of Linux Puppy 5.4. We can now load the puppy system from CD and go to the universal installer and install the puppy system to the memory stick.

The installer will offer to reformat the memory stick to a Linux file system and claim to make it bootable. We tried this for the better part of a day and had no luck. None of the MBR manipulations provided by the installer worked. The puppy files were written to the memory stick but it was not made bootable. Possibly we could have solved the problem by using the grub loader but it was getting late and we were more interested in using the computer instead of having the computer use us.

Another installation option is to simply install to a FAT32 partition and we tried that and it worked. We ended up with a Linux puppy system on a 15 gig memory stick which was bootable.

We wanted a poetable linux system designed for ham radio applications that could be run on any computer we own and not require messing with its hard drive. Actually this method will run on a diskless computer. Hard drives, CD drives, and floppy drives are not needed.

Memory Storage Computers

Mass storage has always been on hard drives and it still is but I have started using them differently.

The conventional use of hard drives is to install them permanently to the machine where they can be powered up, recognized by the system and ready for use. A computer may have several drives but it only needs one boot partition to operate and if that partition is large enough, it might be all that is required to make the system usable for most applications. That boot partition is normally on a hard drive but it could just as well be on a USB memory stick if the stick has a large enough capacity. An inexpensive 8gig memory stick can run a variety of operating systems and applications.

Perhaps the most desirable way to implement solid state memory storage is with a solid state hard drive. We can get away from mechanical storage medium and its problems and achieve better reliability and lower power consumption at the same time. Unfortunately solid state drives are still fairly expensive. You pay twice as much for less than half the capacity of a mechanical hard drive.

We can also increase the life and reliability of mechanical hard drives in the way we use them. Normally, if they are installed in a system, they are on-line all the time even when they are not in actual use by the computer. This can be remedied by using external USB drives which obtain power as well as data access through the USB port. For those applications that require multiple drives, an inexpensive USB hub with its own power source is more than adequate to power multiple devices.

The current preferred configuration in use here is a solid state drive for the main boot partition and multiple 500gig mechanical USB drives for data storage. Experiments with various operating systems are done using USB sticks with the 32gig sticks preferred.

This extends the life of the drives because they are only powered up when needed. The 500gig capacity is chosen for price and reliability. Experience has shown that economical 500gig drives just seem to last longer than larger capacity drives.

The external USB drive solution is better than housing bare hard drives in individual enclosures because USB drives come with their own enclosure and do not require a separate power supply.

OCF (windom) Antenna

I finally got tired of twisting knobs on the antenna tuner. Time to consider a no-tune all band antenna. The OCF is a half wavelength, 80 meter dipole. One leg is 45 feet, the other leg is 90 feet. Feed point impedance is roughly 300 ohms. A 4:1 balun is located at the feed point, followed by a 15 turn 3 inch diameter coil of 50 ohm coax to prevent feed line radiation.

In this particular installation the feed point is located on a tower at the 30 foot level. The 90 foot leg is routed to an insulator at the top of a 60 foot mast and then down to roof top level. The 45 foot leg is routed to a tree at the 15 foot level.

The long leg is trimmed to provide a minimum SWR on as many bands as possible. An SWR of less than 2:1 was obtained on 80, 40, 20, 15, 17, 12, and 10 meters.

Other than the multiple band capability there is nothing special about this antenna. It is simply a half wave dipole fed at a point that has a fairly constant impedance over multiple bands. That impedance is passively matched to 50 ohms to ensure a maximum output from the transmitter and a maximum transfer of energy into the antenna.

The very conditions that some claim are only possible with a ‘resonant’ antenna. So, I guess this antenna could be considered ‘resonant’. Resonant on multiple bands even though the feed point impedance is not 50 or 75 ohms.

The fact that part of the antenna is at 60 feet increases it’s effectiveness but it is still just a half wave 80 meter inverted vee with apex at 60 feet.

Limited Edition

Sometimes you see ads for equipment claiming ‘only a few were built’. The small quantity made implies they are rare and possibly valuable. Real collectors items.

It may also be that they were part of an unsuccessful test market. Perhaps the manufacturer discovered the rig would end up overpriced, underperforming or both and end up being rejected in the marketplace. There is always a down to earth reason why equipment does not make it into mass production.

Limited Editions may have application in literature and art, not in ham radio. Unless you are stocking a museum or collecting junk, limited edition radios are not the golden gems sellers would have you believe.