These are the instructions for building a switching power supply for the laptop PC to run it from the 28 Volt airplane for the $200.00 color moving map (if you don't already have your own GPS) discussed in the tent forum at Oshkosh 1998. You need this if you fly an airplane with a 28 volt battery and stay up longer than your laptop battery lasts.

I have used an old Toshiba Tecra 700CT as the design example, because it is representative of what you might pick up in a used computer store for cheap. It works off the automobile cigar socket, when the engine is running, but not with the engine shut off. I think it wants a minimum voltage around 13.3 volts, and that doesn't run it well. This thing says, at its power input connector, that it wants "DC IN 15V" and that the center should be positive. The connector you need is available at a Radio Shack. Its the really big one, 6.3mm outer diameter, 3.0mm inner diameter. 274-1572 at $1.79 for two.

I think this will work in a 14 volt airplane, because the battery chemistry is quite similar. However, if you have a 28 volt airplane, the computer would get aloft faster than the airplane if you connected the two. We fix this with this switching power converter - 28 Vdc to 15 Vdc.

While you are at the Radio Shack, pick up a project box, 2.4x3.8x0.9, RSU 11340007, $ 3.79..You might have to order this delivered to your house through Radio Shack Unlimited for $ 5.19, pay now, wait a week for delivery. This is also available from PacTec at 1-215-365-8400 as 61914-510-039 (with EMI shielding). I recommend shielded project boxes. A fuse holder would be a good idea - 270-362 $ 1.59, or 270-1238 $ 1.59, your preference, and don't forget fuses 270-1068 $ 1.99.

Now, I don't like wires dangling from underneath the dash on the airplane, so I made the power cord from the converter to the PC to be removable. Unfortunately, there is no socket that mates with the Radio Shack plug (these guys work for the FAA?), so I picked up a matching set 274-1573 $ 1.69 and 274-1583 $ 1.99. These allow for a socket on the box, and a plug for the cord (other end of the cord is the bigger plug to run to the computer). Too bad they don't match, but, no matter, it won't fit backwards. Let's keep the center is positive convention, though, just in case some FAA Airworthiness Inspector breaks into the airplane and tries to force fit.

We design now everything between the computer power connector and the airplane master bus.

Requirements: Looking at the portable power module (black plastic box with a cord to a 120 Vac plug, and a cord to the computer connector, (came with the computer)) it says it makes 2.7A, so that is the design spec for the airplane switching power converter.

Call 408-432-1900, or write Linear Technology Corporation at 1630 McCarthy Blvd in Milpitas CA 95035-7487. Tell them you are an avionics design engineer, because you are, and ask for kit DEMO DC133A "LT1620/LTC1435 Synchronous High Efficiency Low Drop Out Battery Charger". Don't plug this in, or the power converter will get to altitude faster than the airplane. There needs to be changes to the circuitry (else the smoke will get out).

First is the two switching transistors - they will not survive life in a 28 volt airplane! Call Vishay, also known as Temic, also known as Siliconix, at 408-988-8000 or write to 2201 Laurelwood Road, Santa Clara CA 95058-0951, and get a few Si4450DY-T1 transistors. You need two, but these are really small, so you might want extras. These replace the ones on the kit board, and will survive transients to 60 volts. They are not available through distributors. Replace the two Si4412DY transistors on the kit board with the Si4450DY transistors.

Second is tiny resistor ectomies. The resistors you will buy replace the ones on the kit board that set the correct voltage and current limit. You need one of each, but these are really really small, so you might want extras. Call Mouser at 1-800-346-6873 or write to 958 N. Main Street, Mansfield, Tx, 76063. I ordered 6 of each size, 48 resistors, $ 7.99 delivered, and used the bigger ones for 1181, 1372, 1582, but needed the smaller one for 3001.

ME290-1.18k, 292-1.18k

ME290-3.0k, 292-3.0k

ME290-13.7k 292-13.7k

ME290-15.8k 292-15.8k

Remove the AB switch, and R3 and R4. Using a switch is not going to be reliable, and your computer is at stake. The last thing we need is an intermit contact at the voltage set level. I put the 1181 resistor from the lower pad of R3 over to the B pad where the switch was. This was JP1A and JP1B and the selected R3 or R4 on the schematic. All are now gone, and the 1181 resistor permanently wires in an R value of 1.18k ohms where that unreliable switch and wrong value choices were.

Remove R2 and replace with the 1372 resistor.

Remove R6 and replace with the 3001 resistor.

Remove the stakes at E1 and E3 and E4 and E5. Scrape the green insulator coating off of the copper at the corner near where the escutcheon says R5 on the top copper near where E4 was. Solder the 1582 resistor between the through hole right next to E4 and the bare copper. This makes Rprog to select Iprog, the maximum current limit. This is a handy circuit, just in case an FAA Airworthiness Inspector breaks into the airplane, turns on the master, and sticks a paper clip into the power socket of this power converter, you know, to check for safety - see can he set it on fire by applying a prolonged dead short at the output.

Time to assemble in the box. If you got this far, it is time to cut off the unused part of the circuit board, so it will fit in the box and leave room for the fuse holder and power connector on the front panel. I also put an led indicator, with current limit resistors (3.3k each side [in case one ever shorts out] across the output connector). I usually carry a Radio Shack 22-169 with me, but, this would be good for a quick check while flying. Anyway, I cut the board across the top to eliminate the part that said "Synchronous High Efficiency Low Dropout Battery Charger Demo Board DC133A Linear Technology (408) 432-1900 LT Linear Technology". I put the cut through the words "Battery Charger" and "Linear Technology". No sense in letting the FAA see that part, and we need the room. Make sure you don't leave copper around to short together a path where L1 could talk to Q1 C1 C2 (that would be bad)!

Now, CAREFULLY drill two holes in the board, one just above the E1 pad (you will cut into it slightly, and the other on the other side of the board. Drill to fit the SMALLER standpipes on the plastic box cover. Now the board will rest on the PCB mount ridges of the box, held in place by the standpipes, through which the cover screws hold the two pieces of the box together.

Airplane power bus goes through fuse, then to the pad where stake E1 was. Airplane ground goes to stake E2. Don't use the Kapton wire that McSweeny insists upon! Computer plug center goes to stake E6, with outer ring to stake E7. Check your resistors with an ohm meter, and your transistors with a magnifying glass BEFORE applying power.

For better temperature operation extremes, or different frequency of operation, you might want to replace components U1 with LTC1435ACS or LTC1435IS, and U2 with a different LTBC (LT1620 flavor). C11 controls the switching frequency, in case you need to move it.

Try to stay out of near 50 KHz (where Ryan put the stormscope (I have no idea if 3M or BFG changed that, or where strikefinder is)), 100 KHz (Loran-C), 200-400 KHz (ADF) and 535-1600 KHz (AM radio). Jamming any is not good (and you want to be able to use the Peshak NDB/AM/FM clear channel radio position sensor instead of your GPS position sensor (when the FAA turns on the GPS jammers located at the VORs all over the country for GPS Connelrad)0, and a portable loran (like the Ray Jefferson PL-99) is a great idea for backup. That shielded case is a good idea to keep the electrical noise from getting out of the project box, and use of chokes or RFI filters on the input and output lines is approved. Anyway, I didn't do the shielding or RFI stuff (I'll probably wish I had) with Pharley#1.

So, anyway, Pharley#1 bench test time. It idles at a variable switching frequency around 588 Hz at 15.03 Vdc no load (except the led). When I place a load of 24.2 ohms on Pharley#1, he goes to 197 KHz at 15.00 Vdc, at a 157 degree duty cycle, and draws 0.35 Amps from a 28 Vdc bench power supply. That is 9.8 watts in, and 9.3 watts out! Wow. It's time to put it in the airplane for further testing. Preliminary conclusions are I might want to add a pf to C11 to slow it down a tad.

Flight test with all instruments showed no problems. Didn't seem to bother the Loran-C, storm scope was quiet as a mouse (no weather), the ADF clearly heard the noise but only at just below 200 KHz. If I ever take Pharley#1 out of the airplane, I'll add a little capacitance to C11 to lower the switching frequency just slightly, refit with the shielded project box, and add RFI filters at the input and output. However, by pulling the plug out from the computer, the noise goes away, because the switcher goes to 588 Hz and stops throwing magnetic field all over the place.

The switcher powered the computer for the two hour test regime, with no loss in battery. Switcher didn't even get warm, producing 3 amps at 15 volts (45 watts). When we stopped for fuel, the computer kept running on its own internal battery, what with the master turned off, and the switcher charged that right back up on the short flight home.

Now, for those of you that don't like to roll your own projects, RST has offered to produce this product for you. Call Jim Weir at (530) 272-2203 or write to RST Engineering, Grass Valley, CA, 95945, or just go visit http://www.rst-engr.com

Now, in this day and age, nothing as complicated as a switching power supply, what with all the careful designing and layout needed for loop gain, phase delay, and all that stuff, never gets done right without assistance from qualified experts in this narrow field of endeavor. Particularly, in switching power supply development, until you get all the coefficients set just right, there can be some interesting experiences during bench test. "Well, Keith, we think you should adjust this resistor to this value to solve that start up droop problem you described." "OK, I'll try that." !BLAM! "Well, that was spectacular, you get the halon bottle, I'll go for the hose, you there, call the fire department." You gotta watch stray capacitances, and ground plane impedances can be a real problem, too.

Grateful appreciations and kudos for patience go to the Linear Technology Applications Engineering folk, Dave Dinsmore in Dallas at 972-733-3071, and sales office folk, Andrew Pinkman in Austin at 512-343-3679.

Detailed instructions are available on request.