Roanoke Doppler improvements by Nathan Fisher N0ZYC 10/18/2012 • Added an externally-accessible 1/4 amp fuse to the control box this is an obvious construction detail for many of us but not everyone is going to immediately think of it. Even with the diode in series with power, something can still short before that or come loose in the enclosure. • Increased the compass rose to 16 LEDs This is a popular mod and I don't think I need to explain it again. The additional resolution IS useful. But I don't think you'd gain anything by going beyond that to 32 LEDs. • Added low signal hold modification. Another common mod described in numerous places. • Changed all connections to the control box to be jacks in this way there are no cords running away from the control box when fully disassembled. RS232 for the display, 6-pin mic for the antenna control box, mini barrel for power, and 1/8" earphone jacks for the audio in/out. Keeping things modular instead of having the control box flailing wires all over the place when being transported will protect your gear from accidental damage, and make cable repairs easier. • Used the wideband modified antenna switching design explained elsewhere. PIN diodes are available on the CHEAP on ebay. One seller has a lot of 30 up at all times for $10 shipped, and they're the bigger old-style motorola black blocks with flat leads, not the smaller fragile glass diodes that are easily confused with a 1n4148 etc. This requires modification of the antenna array, antenna switching circuit, and a few changes inside the doppler control box, check online for instructions. • Designed the antenna control board and antenna array separately This allows you to design one control box and numerous antenna arrays. My control box has BNC bulkhead connectors numbered 1-4 to mate with BNC-terminated coaxes from the four antennas. Cuts your work way down when making a new antenna array for a different band, and helps conserve PIN diodes and other components in the process. • Moved the compass rose and overload/low signal LEDs to an external box I see a number of people talking about it but no one shows any examples. I used a 25 pin RS232 connector on my control box and some computer ribbon cable to run to the external display. This allows the control box to be placed in a more convenient location and for the display unit to sit on the top of the dashboard without seriously affecting the driver's vision. There will be 4 unused wires you can use for other things if you wish, possibly a speaker or other indicators. Recommend using velcro or some other method to hold it in place on the dash, and a light hood to make the LEDs easier to see during the day. • Placed an external speaker jack on the control box, and use an external speaker allows for slightly easier construction of the control box, and better placement of the speaker. A cutoff external jack could be used to make the external speaker optional as desired, but I chose to eliminate the internal speaker. • Changed the compass rose center LED to be powered off the main power this allows for higher current to the compass LEDs. This required inserting a resistor where the center LED was to prevent them from drawing too much current on the driver IC, but no additional wires are required to the external display, because one of the two other LEDs has a ground, and the other has a +5, so a 330ohm resistor in series with the center LED powers off those existing lines. The brighter compass rose is helpful in the daytime. • Eliminated the need for a fuse on the +5 leading to the antenna control box This was surprisingly simple to do, just get a 5v mini relay from radio shack and place its coil in series with the line going out to the antenna control box, and power a cheap piezo alarm off the raw +12 on the N.O. relay contacts. If your +5 finds itself shorted to ground the relay will engage and the alarm will sound, and nothing will be damaged or blown. Although the relay only sees 2.5-3.5v or so, that little relay has no difficulty engaging. The design is also self-resetting. This is of course only needed if you have done the widebanded antenna mod. • Designed the antenna feedpoints using copperclad board "etched" with dremel This turned out to be an easy way to fabricate a board to directly mount the antenna feeds, resistor, PIN diode, and coax to, while being both mechanically strong and keeping lead lengths very short. Be careful to mind what the other side of the copper board will be touching if you mount the antenna connector directly to the copperclad board. RG-174/u also works very well for the coax, or you can purchase 6' or 12' lengths of BNC-ended RG58 from radio shack and cut it in half, to avoid having to build your own BNC-ended cables. Beware, they use low quality coax in those jumpers and it's poorly shielded. • Keep a quality attenuator on-hand to help knock down powerful signals The doppler actually performs very well even with very strong signals, but using a stepped attenuator helps you to keep an ear on whether you're getting closer or farther away, and keeps the "multipath rasp" at a consistent level for easier interpretation. You'll need to construct the doppler control box and antenna control box RF-tight also if you're serious about tracking powerful transmitters. My use of plastic boxes could be a limitation down the road, but is much easier to construct. • LED display on the control box to indicate which antenna is currently selected This would have been tremendously useful when I was troubleshooting new antenna designs. Tap signals off U5 pins 14,11,9,7 , run them through 100k resistors to the base of 2n3904's, ground collectors, emitters to LEDs, to a single common 330ohm to +5. Also serves as an indicator as to whether scan is on or off, since all four LEDS will appear on when scanning • Increment switch to change to next antenna when not scanning (cut trace between U8-3 and U9-1, use a momentary toggle to connect U9 to U3 normally, or to U8-8 through 1k resistor mom) Requires tapping the switch four times for each increment since the scan clock is 4x the antenna rotation clock. Use a DPDT momentary tac switch if you can find one, otherwise use a small bat toggle. • Light hood over the external compass rose, to make it easer to read in daylight Planned modifications: • Place relay on audio input/output, so that audio is passed through when control box is turned off, since when the control box is powered off, there will be no audio on the external speaker. • Add a set of antenna switch indicators on the control box to identify antenna connection problems when scanning. (open/short on the center conductor) Must be low current due to drive power from doppler controller (use transistor amplification?) so as not to burden the pin diode drive power from the line driver IC • Shield the control box from RF from nearby radios. I think the short warning piezo is resonating RF, even when the control box is powered down. • Convert the antenna switch system to 8 position for higher accuracy and possibly more stable readings when mobile. Convert antenna combiner from 5 to 9 pin (mic jack) and the regular 4 antenna should still be usable with 1/5, 2/6, 3/7, 4/8 pin diodes (total of 8 pin diodes even in the 4 pin switcher) WIll require modification to the control box as well of course. • Toggle on control box to switch audio between the radio and the output of the switched capacitor filter • Add output tap on the switched capacitor filter output as well as the clock, for scope input/trigger hookup Notes: • I don't recommend using a quick connector for the control wire It's too easy for that to come loose or have waterproofing problems. You can have an RS232-C on the control box end, but run the wire right into the antenna control box and waterproof/strain-relief it. I have recently had better luck using a mic connector, the 6 and 9 pin variety plugs are available at radio shack but the panel jacks will require ordering someplace online or from salvage. • YES you can make an antenna array for 10 meters. No part value changes are necessary, the inductors, capacitors, and resistors in the frequency-indifferent design will work fine. My antenna array is 6' x 6' and is made from 1/4" and 3/16" aluminum channel tubing, as well as 5" x 5" squares of 1/4" plywood with sheet tin on top, for the corners. A standard mirror mount bolt with feedthrough goes right through that and the copperclad board is on the underside, with the bottom nut going right through it. At 6' separation, FM deviation does suffer a bit, but is still very usable. Pick up a set of 4' fiberglass CB whips for the verticals, and rely on the channel tubing for your ground plane. Vertical length variation doesn't seem to affect things much, but vertical separation appears to be fairly critical - 2m set doesn't work on 10m. As long as you're using the fiberglass whips, receive sensitivity remains excellent on 10m. Visually averaging the position becomes more important with 10m due to increased multipath at the lower frequency. Despite several sources suggesting a required change in inductor/ capacitor values in the antenna feedpoints and combiner, I have found 8uH and .001uF work equally well for 2 and 10m. • A dremel can be used for fabricating both the antenna control box's board and the four boards at the feedpoints on the antennas. Make sure the four boards are all as close to identical as possible, and that the control board's layout is symmetrical to all four BNC jacks. Keep coax lengths inside the control box and also on the feeds to the antenna feedpoints IDENTICAL in length, or it will distort your bearings. Pay attention when using dual copper clad boards where you have feedthroughs such as for antenna jacks, the other side of the board conducts also and may require cutting an isolation out on the reverse side in places. Ramsey uses the Roanoke Doppler design with several of the mods already implemented: http://www.ramseyelectronics.com/cgi-bin/commerce.exe?preadd=action&key=DDF1