This is the electronic version of the Stratosphere Newsletter. Occasionally final editing is done to the actual layout of the newsletter and spelling checks, and other corrections may not make it into this edition. The content is complete (except for graphics).


Another EOSS Launch

At the July meeting, I was glad to see that everyone thought having a newsletter was a good idea. I was however, a little disappointed that no one stepped forward to act as editor. So, what the heck, I'll give it a try for a while(i.e. not forever). At the meeting, I said that I thought the newsletter should be bi-monthly, but for now at least, or until someone else steps forward, I'm only planning to publish it quarterly.

I suggested the name "Stratosphere" and it seems to have the endorsement of the group. If anyone else has a better idea, bring it up at the next meeting.

Speaking of meetings, I was very happy that we could finally agree on a set meeting night, the second Tuesday of each month at 7 PM. I know there is at least one of you that cannot be at the meeting on Tuesdays. I am sorry but Tuesday was clearly the group's choice. I would also like to thank Dave Clingerman for continuing to offer us a meeting place. If we clean up after each meeting, or until we grow too big for the Pines Rec hall, it should serve well for our purpose.

And speaking of growing, we continue to sign up new members. Welcome aboard to the new members and dive in! There is plenty of work (and fun) for everyone.

We have some real go-getters within our group. One who is doing an outstanding effort is Bob Schellhorn, W6ORE. While many of us were too busy thinking about a microprocessor/controller for our balloon projects, Bob went out and built it! If you made it to the DRC hamfest, you got to see it in action. There's a description of its capabilities inside this issue. I'm sure you will agree, it's a piece of fine work. Thanks Bob!

At the next EOSS meeting, final proposals for our next project will be presented. If you have an idea for a project, this is the time to put it on the table. I'm hoping that we can get two more off this calendar year, but time is slipping away, so please hurry. I have a help sheet to assist you in planning your proposal, ask for it and one will be yours.

Whatever our next project is, remember that we need to put more emphasis on involving students and young people. If you have contacts or ideas that can assist us in this effort, please speak up.

I'll clam up for a while, hope you enjoy this issue of the "Stratosphere", another fine launch for EOSS.

Return to Contents

EOSS Balloon Controller

This controller is comprised of a Z80 based microcontroller which has 32K of ROM, 32K of RAM, two PIO's (parallel I/O chips), a Real Time Clock chip, a DTMF decoder chip, an Exar tone encoder chip used for generating packets, a National voltage to frequency converter for translating the minute voltages from the pressure sensor at high altitudes into values the CPU can handle, and a National 8 input, 8 bit analog to digital converter used to measure different temperatures. The CPU is kept at 1 MHz clock speed to hold down any interference to receivers on board (see figure 1, page 5).

Because the next launch is scheduled less than two months away, there just is not enough time to put into operation all of the features that a controller like this one could be capable. However, it looks like this launch will include the following features:

  • A two meter beacon transmitter which identifies the balloon in CW including the present altitude in thousands of feet.
  • Following the CW ID, a packet message is transmitted telling the status of the balloon which includes:
    • Present altitude in thousands of feet.
    • Present temperature of the payload interior.
    • Present temperature outside the payload.
    • Local time and date.
  • A movable TV camera positional from the ground to any vertical angle from straight up to straight down.
  • A balloon release device commandable from the ground which allows for the payload to be detached before it is damaged by the exploding balloon.
  • A command receiver to accept commands from designated ground control stations.
  • A confirmation packet message is sent following any commands to indicate if the command was completed or no good. This also is a time stamped log of any and all commands used.
  • A General information packet message sent on command which tells all stations the nature of the mission and where to send any confirmations of receiving balloon signals.

In addition, a terminal can be plugged into the payload prior to launch time to allow for updating parameters. These include:

  • Setting the present time and date.
  • Changing to a different CW speed.
  • Entering a different Call for the ID.
  • Entering a selected password used in commanding.
  • Changing to a different general informational message.
  • Check out other functions such as the release device.

A note about the altimeter. The sensor is a Motorola device which gives an output voltage from 5.0 to 0 volts for a given air pressure and is very linear. As you know, the difference in pressure per thousand foot is much greater at lower altitudes but at near 100,000 feet, the difference in the output of the sensor for each 1000 feet will be only a few millivolts. Since this is too big a job for our 8 bit A/D converter, a voltage to frequency converter was used.

The sensor is connected directly to the LM331 with its output running about 4.0 MHz. This output is then divided by 256 and run into a bit on one of the I/O ports. The program detects when the signal changes state and starts a counter which steps until the signal changes again. This number is then used with a lookup table to determine the present altitude.

The temperature sensors used are National LM335's. We tried using LM34's and LM35's which read temperature directly in fahrenheit and centigrade, but since the temperature range we will encounter will go to about -50 degrees fahrenheit, it was difficult to use sensors that go minus. The LM335s give a linear output of 10 millivolts per degree kelvin, which never goes minus. We convert the output read by the A/D converter, to degrees fahrenheit via a lookup table which is inserted into the packet status message to read out from -59 to 179 degrees.

Return to Contents


Numerous times in the past couple of years I have been asked if I thought a "DISH" antenna would help with the reception of Amateur Television (ATV) any better whether it be from the Denver ATV Repeater or when tracking one of the Edge of Space Sciences (EOSS) balloon payloads (spacecraft). My response to these queries has been, not that I'm a New Yorker, answering a question with a question. My usual questions:

  • Are you "line of site" to the signal?
  • What sort of winds do you encounter throughout the year?
  • Are you contemplating portable operations with this "Dish"?
  • What are you presently using to receive the signal?

Here, a couple of facts need to be brought to light before further discussion or consideration is continued. The "Dish", as we lovingly refer to the parabolic reflector, is on the order of abut 55% efficient and therefore has, for most of us, a low frequency cutoff due to size. Since we have this rather low efficiency rating there is a wavelength-to-size ration that if exceeded, renders the "Dish" no more effective than if four stacked Yagiis has been used. This ratio is about 6 wavelengths. A "Dish" that is 6 wavelengths in diameter can produce 20-22 dBi gain. Some of the "long yagiis" of today can produce about 17 dBi gain and stacking four of these will yield about 20-22 dBi gain. That is why I use this as a reference. And, if you'd like to have just a silly 3 dB more gain, you've got to double this size, effort and capital output to obtain it.

Let's take a look at the VHF/UHF bands and see where we might consider using a "Dish." Some of us still consider the 6 meter band VHF, so we'll start there. Six meters is 19.7 feet, so a "Dish" would have to be 118 feet in diameter in order to meet the 6 wavelength criteria and achieve the gain of four stacked long yagiis. How many of you have room for a structure of this size? This would be twice the size of either of the dishes on Table Mountain, near Boulder.

How about 2 meters? You'd still be dealing with a structure having a diameter of about 40 feet. Not in my back yard or probably yours either. Before the government scoffs up the rest of our 220 HMz band, you might want to use a dish there. How about 25 feet in diameter. Does that ease the pain? Still, probably not. The 70cm band is probably where you might start thinking that a 14 foot dish is possible, practical and even economical. But, consider the wind loading problem. To hold that devil in one place all winter long might take some doing, not to mention the requirement for considerable engineering. The newly acquired 33 cm band (another use it before we lose it band), can use a 7 foot dish and give us a goodly 20 dBi gain figure. Using our "L" band (1250-1300 MHz), will satisfy our criteria with under a five foot diameter parabola. We could go on up through the UHF and microwave bands but I think by now, you get the picture (no pun intended). Personally, I wouldn't consider using a dish on any frequency less than 1250 MHz. The answer to whether to use a dish or stay with yagiis is, as you see, rather equivocal. If you need 22 dBi gain to make your link, then it's a good deal. If not, save your money, time and effort. In the future EOSS will be flying some experiments that will be using "L" band and higher. You may want to be ready for them and if contemplating a dish, consider a polar mount rather than just hard mounting it looking at Lookout Mountain (home of the Denver ATV repeater).

Return to Contents

Balloon VOR Navigation Link Experiment

During the past two balloon flights, inflight tracking and final recovery of the payload depended strictly upon 2 meter DFing by the field T-hunt team. although recovery was successful in both cases, the in-flight tracking left some uncertainty regarding the balloon's actual path.

Last year, the flight that ended near Stratton, CO (EOSS-1) was a real white-knuckle operation, since all the DF bearings were taken from nearly due west of the bird, yielding essentially no range data. Because of beacon antenna damage on landing, the touchdown site was discovered only from the air.

This May, the bird flew due north along the hogback (low range of mountains just east of the Rockies). Lessons learned from the earlier flight resulted in a well dispersed T-hunter team, and recovery was swift and sure from a spot only 200 feet from a road. But even so, a post flight review of DF bearings called out during the flight showed that at best, the bird could be located only to about a 5 or so mile radius. In this case, bearing errors were attributed to reflections from the mountains, since all were taken generally from the east.

Payload recovery in both cases depended upon functionality of the single 2 meter beacon. Only one or two T-hunters have 10 meter DF capability, and bearings at that frequency are not as precise as on 2 meters; if their search were to extend into the night, final recovery would be slowed considerably. And the batteries have but a finite life. Could be ugly.

Loss of a payload is a hit which EOSS can ill afford to absorb, so a redundant source of relatively accurate location data would be a nice piece of insurance. There has been talk of carrying a GPS receiver aboard, but such Cadillacs are beyond our budget. A lower cost and reasonably accurate approach uses the same method used by aircraft for years, FOR or VHF Omnidirectional Range.

VOR stations are located all over the country, and there's one at most major airports. They are spaced such that a pilot is guaranteed to be within 40 miles of a station wherever he is over the lower 48 states. In operation, he simply tunes in the desired station and his magnetic bearing to or from that station appears on an instrument on his panel. This instrument, a VOR decoder, typically has a manually adjustable 350 degree bearing dial and a zero-center meter needle showing deviation between the dial and the actual bearing. The needle will center up at two bearings 180 degrees apart; a small TO/FROM flag will show whether that bearing is how to fly TO the station or FROM it. To find his location, the pilot can plot bearings from two different stations on a special aeronautical map, and the intersection is his fix. The FAA requires VOR bearings to be accurate to better than 2 degrees.

VOR stations operate from 112 to 118 MHz with 100 KHz channel spacings. Each station transmits continuously and has a 1 KHz CW ID to confirm proper tuning. Denver Stapleton's VOR operates at 117.0 MHz and ID's as DEN. Each VOR operates at abut 200 Watts and radiates a horizontally- polarized AM signal. VOR has been in commercial use since the 50's, so it doesn't require a lot of high-tech hardware.

Reception and decoding of VOR data is strictly passive, and the receiver doesn't have to be terribly hot. It does have to reject nearby FM broadcast signals up to 108 MHz, though, and must recover both a 30 Hz and 10 KHz subcarrier with little amplitude distortion. Phase error between the received 30 Hz RF and the demodulated audio shows up directly as a bearing error, so demodulator and AF amp response should go essentially to DC. Received signals can vary all over the map, depending on altitude and distance, so good dynamic range is important as well. Once demodulated audio is available, it is simply fed to the VOR decoder for bearing display.

The VOR signal comprises three components. One is a 1 KHz CW ID which carries no VOR bearing data. The second is a 30 Hz signal produced by a directional antenna rotated at 1800 RPM (30 rev/sec). The easiest was to visualize the way this signal is by imagining a beam rotation continuously on a vertical axis. A receiver some distance away will see the signal strength vary through one full cycle every revolution, peaking when the beam is pointed directly at the receiver. Another receiver at a different bearing will see the same signal, except that the peak will occur at a different time.

If both receivers know exactly the instant in time when the beams were pointed north, then they can figure their bearing simply by timing from that north reference to the peak of the signal. At 30 RPS, the beam moves 360 degrees times 30 persecond, or 10,800 degrees per second, or one degree every 92 microseconds.

Providing this essential north reference is the role of the third VOR signal component, the 10 KHz subcarrier. The frequency of this component is varied sinusoidally at the same 30 Hz rate of the beam rotation. The phase of the 30 Hz modulating signal is set to pass through zero precisely when the beam is headed north. The 10 KHz component is transmitted by a fixed omnidirectional antenna, so it's FM demodulating this signal. One now has a north reference for comparison to the 30 Hz AM component. One's bearing to the VOR station, then, turns out simply to be the phase lag in degrees of the 30 Hz AM part from the 30 Hz signal from the 10 KHz subcarrier demodulator.

Although one could homebrew some filters and PLLs to demodulate and display this phase angle, it's easier to use an aircraft VOR demodulator. Thanks to Dennis, W0EPG, of Denver Avionics, EOSS is the proud owner of a working used Narco VOA-9 decoder, along with a VOA-4 which needs work and full up Mk16 COMM/NAV transceiver. It runs from +12 VDC and requires a 0.5 Vrms VOR audio signal. It's pretty heavy, though, and extraction of the bearing signal requires manual tweaking of a knob until the needle centers. As such, it can't go on the payload.

All we really need is the VOR audio as received at the balloon. By downlinking the audio, decoding can be done on the ground. The payload thus must carry a 112-118 MHz AM receiver and horizontally polarized omni antenna. The 10 KHz subcarrier won't fit in a standard 5 KHz deviation FM channel, however, so we must use the 75 KHz deviation ATV audio channel. Telemetry data which used to occupy that signal will now be sent via packet as processed by the W6ORE controller. Since a fix requires bearing data from at least two different VOR stations, the receiver must be tunable in flight. Tuning commands uplinked to the Controller will be processed into either analog varactor tuning signals or sent to a frequency synthesizer. The former is simpler and lighter, but less precise. Whether it's good enough will be determined experimentally.

A Ramsey AR-1 receiver has been built up. Although it tuned up to about the same sensitivity as the Mk16, it suffers terrible overload from local FM broadcast signals. A helical resonator with a Q of about 800 was built up, and it definitely helps, but I'm still trying to reduce its insertion loss. A downside of the helical is that its tuning must track the LO; the plan is to use a 1N914 as a varactor. Further improvements in the mill include replacing the fixed gain NPN front end with an AGCed FET; the NE602 mixer has but 90 db dynamic range, and may be the overload culprit.

After we win the upcoming 10 GHz contest, I'll get back on this project and report further progress in this rag.

Return to Contents