That Darn Drake


The folowing is excerpted from the Jan/Feb 2000 issue of The AMSAT Journal:

Still have that Drake 2880 you bought last year sitting on the shelf, unmodified? Here is a step by step guide of how to get it done and be one step closer on mode S for P3D.

The first thing you will need is a computer. You should visit at least two of the many sites that discuss the modifications to this downconverter. I will limit this discussion to the excellent data (and pictures) presented in two of those sites. If you have a color printer, go ahead and print out the pages for future reference. The first one is one of the original postings by Masa Arai, JN1GKZ:
http://www.ne.jp/asahi/m-arai/gkz/

The second one is a further discussion of various mods by David Bowman, G0MRF:
http://www.qsl.net/g3pho/drake.htm

The second thing you will need is the right toolbox. You will need a very good, small Phillips-head screwdriver: the screws are very soft and easy to strip. You will need a magnifying glass (yes you will), a set of small needle nose pliers, solder wick, small diameter solder, and a small soldering iron (25 W or less). Also useful would be a Pana-Vise and a pair of tweezers.

Ready!
Now that you have researched the on-line material, it is time for a little soul-searching. How much experience and/or skill do you have with VERY SMALL electronics? I don't want to unnecessarily worry you, but this Drake 2880 is definitely not your father's Heathkit. Let me offer several options, based on my own experiences, before we go any further, and you can decide where you feel comfortable.

Level 1, Are You Crazy? The unit will work just fine with an IF of 120 MHz, but with very low gain, about 12 dB, and a noise figure of about 8 dB. If you 1) use a receiver capable of 120 MHz SSB, 2) have a high gain antenna, such as the popular Myers/Conifer 24 dB "barbecue grill" dish, and 3) can provide 12 Vdc up the coax, then you can leave the screws on the cover and stop reading here. From what we know today, P3D is still within your grasp.

Level 2, OK I'll Do The Easy Stuff: Take the screws out of the cover, take the screws out of the interior shield, and remove the two red coils and the two chip capacitors as shown in the N1GKZ article. The chip caps provide you a chance to practice your surgery skill with a magnifying glass and tweezers. You now have a unit with 15.5 dB of gain and a NF of 7.8 dB.

Level 3, That Was Easy, What's Next? Now it is time to replace the crystal and move the IF up into the 2 meter ham band. Carefully de-solder both the N-connector and the F-connector. The N-connector has a holding nut, but the F-connector screws out and solder wick is recommended. Be very careful of the latter as it is easy to damage the center pin. Next, remove the remaining screws on the circuit board, including the two on the voltage regulators and the small one in the middle by the vertical shield. Remove the circuit board and carefully de-solder and remove the old crystal. You will likely have to de-solder and suck out the holes (solder sucker bulb is very handy) before the replacement crystal will insert. Make sure the new crystal is flush and cut the leads off very short after soldering. You now have downconverter that puts the IF at 144 MHz for 2400 MHz input. You, of course, could pick another IF, like 140 MHz (my preference for easy mental conversion) or even somewhere in the 70 cm band. A 70 cm IF would be convenient if you plan to use 2 meters for uplink or to drive an L-band converter. I personally think the 2 meter IF is preferable, allowing mode U/S with a common dual-band rig (FT-847, IC-821H, etc.). I anticipate modes V/U and U/V to give way to modes U/S and L/S very quickly.

Level 4, Tweak That Baby: G0MRF has two mods that will bring you up to about 22 dB of gain and a 5.5 - 6.0 NF: adding a pair of 22 pF capacitors to the IF stage and physically extending the image stripline filter. You can find suitable 22 pF ceramic disc capacitors in Radio Shack 272-806 (all I could find on Sunday afternoon). Cut the capacitor leads short, about 4 mm, and bend them halfway at 90 degrees to facilitate soldering in place (you will want the magnifying glass for this operation). For the stripline filter, cut a couple of capacitor or resister leads to about 3 mm and solder to the ends of the board striplines, extending them about 1.5 mm. Note the two striplines are not the same length. Compare your final "extensions" to the one in the fine picture on the reference site above. This is a far as I was willing to go, but all went well and the patient survived. I am indebted to the above referenced authors for making their knowledge and experience available. I had this unit tested at AMSAT-UK and it came in at 22.5 dB gain @ 5.6 dB NF: I'm happy!

Level 5, Oh, I Forgot Power: The Drake unit will accept 12 Vdc up the coax (into the F-connector). If your rig has that capability or you are inclined to add it externally, that is the easy way to go. The JN1GKZ site has a simple mod for routing external power up to the unit, but it requires drilling a hole in the case. I cringe at the thought of drilling a hole in an obviously well-made weatherproof enclosure, so I opted to bypass this mod and power the unit from my rig (IC-820H).

Level 6, Where Only The Brave Follow: G0MRF has the ultimate mod, replacing the stock Drake GaAsFET with a higher performance device, an MGA88578. This mod requires extensive revision to the RF amplifier section of the board, but results in an impressive 32 dB of gain at almost 2 dB NF. Now you are talking real low-noise performance. The JN1GKZ site also has a RF amplifier mod, but it appears to offer less for the effort. For my money, a low noise MMDS preamp (California Amplifier model no. 30947, $60 plus shipping) may be a safer bet to get to this kinds of performance level. If you want to try it, though, you can see the K5PK photos here.

Set!
As I noted above, I took my unit up to Level 4. The next step is to calibrate the IF for the desired frequency. There is a (very small) variable capacitor on the board that can be adjusted to set the IF frequency right on the mark. Be sure and use a non-metallic screwdriver (or wide toothpick) to adjust the frequency. I made up a signal generator using a high power switching diode, 1N4005, and five 47 K ½ Watt resistors. These six devices are all soldered in parallel, with the cathode end of the diode connected to the coax center conductor and the anode end connected to the shield. With a short piece of wire loosely coupled (one loop) around the "signal generator" and inserted into the Drake N-connector, I transmitted a 1 Watt CW signal at 29.9999 (the upper limit on my HF rig) into the signal generator in order to receive the signal at 143.9992 on my two meter rig (30 x 80 = 2,400). Tuning was smooth and easy-but very little adjustments make a very fast frequency shift. I also noted the Drake drifts about 3 kHz in the first five minute warm-up period.

Go!
But what about testing it on a satellite? Fortunately, OSCAR-11 has a beacon at 2401.500 MHz. I quickly cobbled together a 10-turn helix with a ring reflector (see http://members.aol.com/k5oe/trio_ant.htm), fixed it at 45 degrees elevation, and waited for the next pass. On the first attempt I picked up the beacon at about 20 degrees elevation at 2401.530 (145.530) MHz and followed it up through 57 degrees elevation and back down to 20 degrees where it was now at about 2401.470 (145.470) MHz. Signal levels were S3 for about half the six minutes I copied the beacon, against an S1 background: success by any measure.

Reflections:
I used a casually constructed, untested and unproven antenna. Still, I was able to copy the beacon with ease. If you consider OSCAR-11 is 1/2 Watt at 2,000 km and P3D will be 50 Watts at 40,000 km, you could still hear P3D at apogee--being only about 6 dB down from the above comparison signal. And this does not even take into account the high gain antenna on P3D. Just imagine how good it would be with a high-gain antenna at the ground station. 73 es listen for me on mode U/S.

(C) 2000, Gerald R. Brown, K5OE