How To...Modify Microsoft's Sidewinder Force Feedback Joystick For Precision Manual Flying

By Erich H. Leonhard

The euphoria wore off pretty fast! Kicked myself for selling off the 3D Pro! Ten months of fretting and fuming. So what was my gripe? I really hated two things. One was the throttle wheel. Every time I reached across with my left hand to adjust it, I felt I was crossing myself, which you needed to do anyhow if you were trying to pass a tricky test and needed precise power trim--and if, for reasons of ergonomics, I used my right thumb, it only got dislocated trying to hold the stick and make a fine adjustment! Try small power adjustments while making a delicate touchdown in the Bell 206. And you daren't let go of the stick for a moment! Which brings me to the second, and much more serious problem, and this was that the stick just flopped all over the place when force feedback was selected but centralized OK with force feedback off. (If I wanted this, I could have stuck with the SW 3D Pro!) This problem was with any default FS98 or FS2000 aircraft. The only time force feedback showed some signs of life was when the stick was moved violently in the pitch or bank axes, when it showed some signs of (equally violently!) self-centering. And, of course, it was hyperactive when taxiing on every runway with the default Cessnas, all of which appear to be cobble-stoned. The runways, I mean!

To be entirely fair to the stick however, a lot of bad flight modeling is to blame for bad overall results. I have some third party aircraft (e.g. VIP Classic Wings for FS98), which are modeled very well in terms of force feedback. Surprisingly, 90% of the aircraft I had selected for download were very acceptable flight models too (though one or two had frame-gobbling panels). But this was no excuse, for Microsoft default Cessnas, especially! This doesn't absolve the joystick, however, which appears to bring in its own foibles. The net result of all this being thorough compounding of the problem overall.

The best way of gauging the extent of the problem is to try flying the ATP Lessons, or worse, take the ATP or Instrument Rating Check rides. Not only did I seem to get even the elementary things wrong--one feels terribly small on front of the Lady Examiner as well, kind as I think she's trying to be! Once, twice, OK...but five times? To me she sounded almost as nice as Rod Machado! But that didn't help. She had a job to do and I wasn't helping. I'm an absolute pushover for passing tests, and must record here my mounting frustration every time I flunked. It was indeed difficult to stay within +/- 500' in some phases, let alone the +/- 50' I had psyched myself up for! Holding heading wasn't so bad since the tolerance is larger, at +/- 10 deg.

Make no mistake, I have the utmost regard for the designers, engineers and programmers at Microsoft. Let me mention, that they have provided thousands and thousands of avid flightsimmers around the world, me included, a hobby matched by no other. For every one-time outlay you are assured of hundreds, if not thousands, of hours of enthralling simulation, stimulation and education, be you a real-world professional pilot, a weekend Cessna or microlight flier, or just a 'for-the-sheer-love-of it' simmer. Right from the FS4 on two floppies, running on an ox-cart of a computer, through the years, to the mighty FS2000, operated on a Pentium III at 650 MHz I've got them all, jealously guarded in my own little Fort Knox.

So what does all the above lead up to? Only to the fact that I count myself a member of the MS FS family and have earned, I think, the right to constructively comment on something that is not up to scratch, and for which no fix has been provided! Not just criticize, but to provide what has proved, for me, an excellent fix. Maybe it will for you, too.

Back to the subject at hand. I must mention a few things here--some questions possibly passing through your own mind. Did he have a bum stick? Did he scour the Internet, particularly FlightSim.Com, for information or solutions? Did he use the latest MS patches/updates (including DirectX)? Did he change the 'Force Feedback' and 'Return-To-Center' Forces? Can he fly? Is he mad?

The collective answer to all the above is YES! I didn't find much from MS, except DirectX 7a and the SWGameEN download (same as I received on the Driver CD I got with the stick.) But where I hit an indirect bonanza was reading two articles, viz. Setting Control Sensitivity In FS2000 and the other on Flying The Big Iron. With Peter's recommended sensitivity settings things became much, much more manageable. (Actually, it never even occurred to me to keep any sensitivity at 0%, or any null at 100% Thanks, Peter!) But this was not the end of it. I was still left with a floppy stick I had to carefully keep centered where there was no center. I have always, like Peter recommends, flown hands-off as far as possible, using only pitch trim (my preference, trimmer set in the step mode, not continuous) to fine tune. (And this definitely goes for real life too. Heaving the control column back and forth is only for emergencies--and looks good in movies!) There was also an irritating mechanical dead zone exactly where micro stick movements were required back and forth, important on the elevator axis if you want to maintain altitude within +/- 50', or left and right to accurately maintain heading. I'm not referring to null, but to a loose feel. As if, in real life, the aircraft's control cables were slack. To prevent this dangerous situation in real aircraft, of course, the control cables are tensioned regularly as part of periodic maintenance, tensions being checked using a tension gauge. Now that's "really as real as it gets!"

No such luck on a sim plane! So, was my stick defective? I borrowed my buddy, Mike's, and guess what? Same problem. Two bum sticks? Hardly likely. I tried it on another machine as well and--you guessed it! About this time I made another horrible discovery. Microsoft has introduced the SWFF-2. According to them, "Drastically redesigned to...200 different forces...etc.". I read this in utter disbelief, truly dismayed! Had my Version 1, by implication, flunked in its most crucial role, as an aircraft joystick? I could not believe a new design being needed to give puerile games like 'Sheikh!' a more realistic feel of a Plasma Gunutron or a Thoron Kerboom! Mere conjecture, but was the SWFF-2 being aimed specifically at more realistic aircraft feel and control where it had fared poorly in the earlier version? Make no mistake I'm all for progress, but how about a fix for the guys who invested so recently on something professing to deliver so much? (Wonder if Microsoft has a buy-back scheme! Is the moon made of green cheese?)

So, what did I do? I am not going out to buy the SSWF-2. My problem had to be solved, and it looked like I wasn't going to get any help on this one. A bit of mental blood and sweat...and now, definitely, less tears! And I'd like to share these ideas with you, since you too may be having a similar problem and feel the same way I do.

Let's look at the centering problem only in this part. We'll try and rid ourselves of the problems according to their IQ (irritation quotient). A few rough and ready ideas proved not wholly satisfactory. The best of these involved stuffing a piece of 1" thick sponge with a keyhole slot cut in it between the joystick flange underside (the one near your little finger), and the top of the base. This, among other similar attempts using springs and car gear-lever rubber boots, only partially solved the problem. The centering was not consistent though the flopping about stopped. And they were all visible, and looked pretty messy! However, many of you not desirous of the invasive techniques that follow might opt for these simpler expedients.

What follows is not for everyone, and I strongly recommend against it for the guy who thinks a screwdriver is part of a bar menu, or a file is a little yellow box for stuffing data in. But if, like me, you're the desperate type, or have a modicum of mechanical ability and/or are willing to learn, please read on! Because I've done this especially for you!

Please be warned! Opening the bottom cover on this product will render any warranty null and void. If you are planning to sell this joystick, and have been lucky enough to find a buyer, then do not proceed further.

Even if you do consider modifying, read this article end to end first to get an overall picture of all that is involved.

Please also be warned that you undertake any work internally at your own risk. My personal experience has resulted in excellent performance on my computer, with both FS98 and FS2000 particularly.

Not a single hole is drilled, or any material removed from the joystick. The modification can be removed in a matter of minutes, should you for any reason find it unsatisfactory, and absolute originality will be restored. (But I'll be very hurt!)

So here's where ModFlight starts! We've taken off, done all the usual things then instructed George (our faithful autopilot) to hold the fort, while we do some unusual emergency stuff that we must complete before the end of the flight. OK? So, let's start by unscrewing the control column from the cockpit floor...I warned you!

DIAGNOSIS

One look at the inside and my hat's off to Microsoft. Superb internal workmanship! Drat! Why must it be that the 90% of something that could have been great be so completely marred by the mere 10%! Surely 90% is passing marks! Why can't we just live with it? We do with lots of other things in our life. Why can't the manufacturer provide a fix? Why indeed! OK, back to the problem at hand.

Vital Precautions!

Never handle electronic boards unless you're electrically grounded or earthed. Static electricity tends to be far from static. It leaps around and pounces on unsuspecting CMOS IC's. For grounding, even Rover's chain is fine, one end round your wrist, the other attached to the back of your computer, which, if the mains cord is attached, is earthed. The Main Circuit Board (MCB) is held by two lonesome black Philips screws. Carefully lift off the board a couple of inches, tilt it and (once again, carefully!) unplug the five little plugs from the circuit board. Hold the plugs themselves and ease them off the board--don't yank the wires. Once the board is free, carefully roll up the cable, at the end of which is the joystick plug, and put it safely in an anti-static bag. (Look after those 11 black Philips screws as well.) A savvy buddy, (nerd, geek or even alien), is of great value, if you feel while reading this you're not so hot yourself. Oh, you needn't worry where the little plugs came out from, in case you are. They each have a different number of pins and go in only one way. So now you already know how to put them back! And for goodness sake handle the board as if you were carrying a dozen eggs in a tissue paper bag. Don't touch any of the electronic components left with the stick after you remove the main circuit board. These are infrared optical devices and don't need your fingerprints or dust to enable them to function properly.

The mechanical works is now easy to study. Two small DC motors are used to "drive" the joystick's pitch and bank axes respectively. (These really should have been stepper motors that are able to make movements up to fractions of a degree, making things far more subtle and realistic.) The movement is transmitted by gears. This is blunder #1. Gears are marvelous positive transmitters of motion--other drives may slip, but gears, never. But simple gears have one major disadvantage--backlash, or, to simplify, loss of motion while reversing direction, due to the necessary small gaps between the running teeth. Logitech recognized this problem, so their Force Feedback Joystick design uses push-pull cables to transmit motion to and from joystick and motors. Beautiful--no slip, because of the cable tension, and no backlash. We, however, are stuck with this problem. (Eliminating this 100% is my next ground challenge project!) One more area of loss of motion is where the stick motion is transmitted to the gear sector (i.e. the part gear). A plastic pin (with a funny hole in the middle) sticks up through the gear sector and you can easily see the free movement between this pin and its bore. Blunder #2! Score two minus points! Serious ones if you're into any sort of precision manual flying, such as the ATP Lessons and Checkride, and the Instrument Rating Test.

I mention the above two points right up front because even after doing the modification detailed below there will remain a very small amount of free play, which can be considered negligible if you stick to Peter's sensitivity settings. These settings tend to mask the problem--unfortunately, partially only. (There is one other minor fix to remove 50% of this, but that's the subject of another part).

It took me a couple of hours, sitting absolutely motionless, scowling at the works to figure out the scenery and to come to a solution that would remove the flop. Meantime, the better half and a couple of look-alikes tiptoed around, I hope, solicitously. When my daughter touched my shoulder and I actually moved, she seemed positively relieved. If I sat like that any longer, my wife might have started dusting me! (My wife, bless her, is a compulsive duster!)

I'll pass over my own early trials. I won't bore you with the details, particularly of why I selected the materials I selected. You, dear prospective modifier, get just the cream.

So, what can you expect at the end of a couple of hours of aerobatics inside the joystick? You'll have a stick that consistently centralizes, whether force feedback is on or off, regardless of the vagaries of the stick, or of funny downloaded or default aircraft characteristics. You'll have much less strain with any kind of precision manual flying. After a nice long relaxed cross-country, when you punch the autopilot off, you find that there is less of a relationship between the normally docile 737 and a suddenly bucking bronco. Even if your hands are off the controls when AP is put off, there should be no immediate violent attitude changes. Most important I have found that the modification does not in any way offset the action of force feedback--rather, it unobtrusively enhances it. It acts and feels like the airflow over perfectly balanced control surfaces. You can now use the trim effectively for practically hands-off manual flying. All because you have a defined self-centering reference point, but not to the extent of mistaking it for a detent. (More about detents later.)

In spite of what I'm saying, if you find you are satisfied with your present performance, then this modification is not for you. But if, on the other hand, you have identified with my kind of problems, read on. Naturally, to obtain the above results specified, you have to do the things specified, in the order specified, and with the precision specified.

MODIFICATION DESCRIPTION

Mechanically, the modification consists of two small plates, one for each of the two axes mentioned above. On each of these plates are fitted bolts, nuts and various washers, as also four levers you will fashion from the 3 mm wire. Two of your screws will be specially shaped to tightly fit into the plastic pivot pins we talked about earlier (where the second play problem was). The two plates will be clamped by two screws each (the ones exactly 46 mm apart running parallel to the pivot pin travel) that clamp the steel plates to the plastic joystick base. This is the reason we don't need to drill any new holes with this mod, as these screws will hold them in place. What about the springs? These are the little guys who, when stretched between your wire levers on the two plates, keep each axis of the joystick centralized. So now, even unpowered, the stick self-centralizes smartly. (What a difference some spring makes! Pure mechanical Viagra!)

Hey, cheer up! I'm going to walk you every step of the way. You can do it! Just follow my directions accurately for a couple of hours and you'll be OK. I promise.

And hey, all you mechanical whiz kids out there, please forgive all this fine, irrelevant, (for you), detail given below. You've probably worked out a much better fix already. The concept, however you do it, remains.

TOOLS AND MATERIALS REQUIRED

TOOLS (beg or borrow...middle option may delay project, depending on sentence!)

6" round nose pliers, 6" file (second cut), rat tail file, small metal saw (fine blade), hand or power drill, 4 mm and 5 mm drill bits, small bench or hand vise, 6" Philips screwdriver, small metal cutting snips, 6" ruler, pair of dividers, set square, center punch and hammer, cutting pliers, 2 nos. 5 mm set spanners, fine-tip felt pen.

MATERIALS (not counting cost of chartered chopper, brains, and Tax...around $2!)

150 mm x 23 mm strip brass or MS, 1 mm thick, 8 nos. 5 mm x 30 mm long MS screws, 10 nos. 5 mm hex head nuts, 12 nos. 5 mm flat washers, 4 nos. 5 mm spring washers, 8" length of 3 mm galvanized iron or MS wire, 2 nos. tension springs (4.5-5 mm ID made with 0.4/0.5 mm diameter spring wire, 2" long...eyes not required. We'll make eyes!)

DETAILED FABRICATION INSTRUCTIONS

a) Cut two pieces from your 23 mm MS strip, each one 58 mm long. Flatten and square up with the file, if necessary.

b) Mark off a small rectangle 5 mm x 15 mm, starting at one corner of one of the plates, the long side corresponding with the long side of the plate.

Step b) is not necessary for the other plate.

c) Use the snips to make two cuts to remove this rectangle. Clean up and tap flat.

d) Draw the short center line on each piece dividing the plate into 2 equal parts of 29 mm each. When drawing the centerline for the slotted plate, keep the slot on the right and close to you. (Please be accurate to within half a millimeter or less).

e) Mark two points 5 mm from each end of and along the short center line, and center punch each mark (ready for drilling).

f) Repeat Step e) for the other plate.

g) Drill two 5 mm holes at the punch marks, in each plate. Clean up burrs.

h) Place the plate lengthwise from left to right. From the center line of each plate make marks at 10 mm and 18 mm to the left and right of the center line. Draw short through vertical lines at each of the marks exactly at 90 degrees to the long sides (use the set square).

i) Repeat Step h) for the other plate

j) On each of the plates, fit two bolts, with nuts and flat washers, and fully tighten. Heads must be toward the unmarked side of the plate.

Now prepare to learn a new step...The Joggle! Sounds like fun! (Joggling simply means getting two components or surfaces in line by appropriate bending).

k) With the marked side facing you clamp the plate in a vise accurately at the 10 mm mark, with plate gripped dead vertical. Carefully bend along the clamped edge, away from you, to an angle of roughly 30 degrees from the vertical. (Keep your thumb(s) as close as possible to clamped point for a clean, sharp bend).

l) Repeat Step k) at the other 10 mm mark on the plate. Check both bends are equal. (Visually will do for now).

m) Repeat Steps j), k) and l) for the other plate.

n) Compare the two assemblies. They should be identical in all respects. Make small adjustments to achieve this, if necessary.

Congratulations--You've done the half joggle!

o) Take one of the half-joggled plates and accurately clamp in the vise, at the 18 mm mark, this time with the unmarked side facing you, dead vertical. Bend along the clamped line, away from you, to an angle of 30 degrees from the vertical.

p) Repeat Step o) at the other 18 mm mark on the same plate.

q) Place the now fully joggled plate on a flat surface, screws up. The outside ends should sit absolutely flat, the bolt heads should be just clear of the base surface, and your work must look perfectly symmetrical. Make small adjustments to your bends until you achieve this. (Judicious use of hammer and pliers.)

r) Repeat Steps o), p), and q) for the other plate.

s) Compare the two assemblies. They should be identical in all respects. Make small adjustments to your bends, if necessary, to achieve this.

Congratulations again! You're a full fledged joggler now!

We now need to drill the two fixing holes in each of the plates exactly 46 mm apart. Please do it my way, crazy as it seems! (Seriously, there's a good reason...Nike!)

You'll need a thin piece of stiff card (old greeting card will do) with which to make 'templates' (temporary guide plates). Mark and cut a strip 23 mm wide and 130 mm long. Square cut the ends.

t) Place your plate assembly on the card at one end, and with a fine pencil, mark the other end of the plate assembly. Cut out this template. Check it neatly matches your plate. Repeat with the other plate. Notice something? The little 5 mm x 15 mm rectangle needs to be cut out too, from the second template. Draw the short centerlines on each template. (Keep all this aside for the moment).

u) Pick up your joystick, upend it, place the stick part between your knees and rest the base on your thighs, (unclassic position for inverted flight!).

v) Identify the aileron and elevator plastic pivot pins (the ones that come through the sector gears). They're easily identified because each has a cross between a slot and a hole (a 'slole' what else!) in its center. With your fine-tipped felt pen draw a line by eye, at 90 degrees to the line of travel of each of the pins.

(There may be grease abounding around here. Clean it off first, then use a damp alcohol-dipped cloth, or your pen won't mark. (Don't use too much alcohol or SWFF could flunk the breathalyzer!)

w) Now take each of the axes to its extreme ends of travel and draw a line on the static metal part in line with the 'slole' centerline. (Reason? After our mod we should get full travel on both axes, and these are our checkpoints) then mark each axis' dead center.

x) Let's start with the aileron pivot pin. Close to and running parallel with the travel of this pin are two shiny Philips screws exactly 46 mm apart. Set your dividers to 46 mm, locate these two screws, and remove them.

y) Place your 'slotted' template in position (now you know why we cut out that little rectangle...it only goes neatly one way) with its centerline aligned with the axis center mark), parallel to the pin travel and spaced about 5-6 mm from the white plastic sliding strip. Once correctly positioned, press down simultaneously with both thumbs to transfer the hole impressions to the card. Remove the card and carefully mark the hole centers (by eye should be OK but try to keep things as accurate as possible). Don't worry if you see these marks out of line. Because they are! That's why we took the template route.

z) Stick the template to the base of the plate assembly, keeping the hole marks impressions visible and the little cutaway rectangles aligned. Set aside to dry. (Use envelope glue.)

Blast! I've overshot the alphabet!

i) Repeat steps x), y) and z), for the other plate, after removing the 46 mm spaced screws near the elevator end. Make relationship marks here since this plate is symmetrical.

ii) Pick up the aileron plate assembly and punch the hole centers from the template onto the metal. Tear off the template and remove all traces of glue. (Use water, Dear Henry!)

iii) Drill both holes with a 4 mm drill from the bolt head side. Tricky! Clean up the burrs. Check hole alignment by placing the plate assembly over the joystick body holes. If you've been diligent the holes should fall dead center. If not you'll need to elongate the plate holes, using the rat-tail file.

iv) Repeat steps ii) and iii) for the elevator plate assembly. Check hole match at elevator pin end. If not you'll need to elongate the plate holes appropriately.

(When both sets of holes are correct, please return the rat-tail file to the concerned rat.)

2) LEVERS:

a) Cut four pieces of the 3 mm wire, each 78 mm long. (This is just a bit longer than required. You may have to trim off during assembly later).

b) Use the round nose pliers to form an accurate eye with an inside diameter of 5 mm. Start the bend at the extreme end of the wire, and almost close up the eye fully, leaving a gap of 1 mm. It helps in fine tuning the size. Use one of the 5 mm screws as a gauge, and carefully bend for an accurate, but free fit on the screw thread area. What you end up with should be a 'b' (or 'd') with a very long leg. (Do a few practice loops first, just to get the hang of it.)

c) Finish the other three levers exactly as at b) above. All four levers must lie down flat. Tap flat to ensure (mustn't be like a four legged table with uneven legs.)

3) ASSEMBLING LEVERS TO PLATE ASSEMBLY

a) Let's assemble the aileron plate first. Hold the plate assembly with the screws facing upward and cutaway rectangle away from you. Keep all nuts and washers at hand. Slip on one flat washer on the far screw (on top of the nut), then one lever such that its arm falls on the right of the near screw, (let's call it the stop screw), the end pointing at you.

b) Slip on another flat washer over the first lever, then slip the second lever over this, but with the arm on left of stop screw, end pointing at you.

c) Slip on one more flat washer over the second lever, then one spring washer, then another flat washer, then run on one nut, then another spring washer, then another nut. Don't tighten anything yet.

d) Better double check a), b) and c)! Nuttiest levers I've seen! e) Put a little oil or grease on all the sliding areas. (For lubrication, not cooking!) Wipe off any excess.

f) Finger tighten the nut nearest the levers, to squash the spring washer just sufficiently to allow the levers to move with a little drag. They should neither flop about nor be too tight. (A look at the springs should give you some idea of just how much drag to give. The springs have to work against this drag easily).

g) Without allowing the bottom nut to move, fully tighten down the top nut. (Now you know why we have two 5 mm spanners.) The levers drag setting should not change. This is a fiddly operation and you may have to repeat a couple of times to get it just right. Helps if you have three hands!

h) Repeat steps a) to g) above for the elevator plate assembly. Should be easier now with all that practice.

4) MORE JOGGLE-IT?

Noticed something funny about the levers? They're way outta line, height wise. So what do we do? You're right! We need to get them level by joggling the upper one to the same height as the lower one.

a) With round nosed pliers, grip the upper lever at a point between the two screws and bend downward towards the plate about 10-15 degrees. Now grip same lever just level with the outside edge of the plate. Bend upwards about 10-15 degrees. The idea is to adjust both bends so that both levers are exactly at the same height, separated by the stop screw.

b) More bends! The lever arms must also run parallel from the base of the eye to the tips. Look at the left side lever (tip toward you). Turn it to touch the stop screw. It should be 90 degrees to the edge of the plate. You may have to bend it a few degrees just a little away from the eye on the straight portion so that it stops exactly at 90 degrees to the edge of the plate. Repeat this for the right side lever that must be bent until it is exactly 90 degrees to the edge of the plate. When you get this, you'll also find the lever arms running parallel to each other. Exactly what we want! Everything symmetrical. (For the moment, just trust your eyes for judging the 90 degrees. C'mon, you're a flightsimmer!)

c) While your hands are still set, better finish the other plate and lever assembly. Just repeat a) and b) above.

d) All the lever ends must be trimmed to a length of 53 mm, measured from the eye center to the tip. Don't dismantle anything. Just measure along the long leg of the lever from an imaginary eye center. That's accurate enough.

Hats off to you...you've hung in this far! I'm pleased to inform you we're more than halfway there on our ModFlight. Cruising happily at 280, engines purring at 70% N1. George doing a great job maintaining everything. Just time enough for a quick coffee, a bit of friendly banter and a quick scan of the coffee-bearer. OK, screw those eyeballs back. Next sector coming up fast...Focus and patience 100% for next 40 minutes.

5) PIVOT PIN EXTENSIONS:

Those plastic pivot pins are going to help us. The levers are going to press against extensions we will be adding to these pins, to centralize the stick in each of the two axes. Microsoft kindly left us ready made holes, funny shaped ones, but holes. We need to shape about 8 mm length of the remaining two screws to the exact shape of the funny-shaped pivot pin holes, and press these into the holes. The size must be such that you have to force them in with your thumb, and there is no shake inside the hole. We are not going to use any adhesive, as I promised we would be able to remove everything, in case you didn't like it. Let's stick to that. Remember, no shake, or we will introduce exactly what we are trying to eliminate--play! You'll need a lot of patience, plus a glass, can or barrel of your favorite beverage. Anything to keep you happy and focused and patient. (You'll need to double this in case you've grabbed yourself a savvy buddy!)

a) Cut off the heads of your remaining two screws. File finish and check a nut runs on. (This nut acts as an 'extractor' later.)

b) Clamp 8 mm of the screw in the vise horizontally, with little less than half of the thread protruding. File an 8 mm long flat portion squarely just removing almost all the thread. Remove the screw, and turn and clamp at 180 degrees and repeat the above. Check the width of the flat part by trying to insert it into the pivot pin hole. It won't go vertically yet. You'll have to check with the corners.

c) Repeat b)...sip...repeat b)...sip...until you have the tip end tight. You can hold it in your hand while fine trimming. Work slowly and carefully. If you lose patience and go undersize, you'll have to start over. One trick is to just get the first 2 mm of the tip perfect, then trim file the remainder to match, in slow steps. I repeat, aim for a tight fit.

d) Next, attend to the curved surfaces of the screw (it's your fault if you get distracted this easily). File to reduce the thread, while maintaining the curvature. This helps keep the fit tight. Work equally on both sides. After the initial shaping, try as before to insert the first 2 mm to fit tight, then shape up the remainder to match. Aim for a tight fit, but not so tight you're going to crack the plastic, or need dynamite to separate the parts! If while fitting, you push the extension screw in and you need to pull it out for trimming, run on a nut and keep screwing it downwards until the screw is 'extracted'. (Don't do this too many times.)

e) Repeat a), b), and c) to prepare the second screw.

f) Finally, thumb press both screws into place. Check for shake inside the holes.

6) TRIAL FITTING PLATE ASSEMBLIES TO JOYSTICK:

a) Secure the elevator plate assembly first, using the two shiny screws you removed earlier. Keep everything as central as possible. The lever ends will foul, so you'll have to bend them about 45 degrees about 12-13 mm from the tips forming a 'Y', to prevent this. Both levers must be free to touch the stop screw. One of the levers when moved outwards will touch a white plastic fitting. Use a fine tipped felt pen to mark the center of where it touches.

b) Move the 45 degree bend with the round nose pliers exactly to the above mark. Move the joystick until the pivot pin extension screw is at the extreme end of travel unrestricted by the lever. Correct the bend point until you get this right. You may also need to shift the plate assembly slightly to get its center line aligned with the center of travel of the axis. It's been made so that you can get some minor adjustment if required.

c) Note that the other lever is OK. It doesn't have any problem. Just bend exactly to match the first lever at 12-13 mm from the tip.

d) Once you have found a final position use the felt pen to trace around the securing screws, as we will have to remove the plate assembly in order to trim and prepare the levers for the springs. We will then be able to fit back in that identical position later.

The above screws are screwing into plastic. Be particularly careful if you have 17" biceps! Use butterfingers to 'catch' the first thread, screw in to just 'touch up' the screws to the base, then tweak, maybe one tenth of a turn, beyond that. That's it!

e) Repeat steps a) to d) for the aileron plate assembly.

So, what have we now? Two identical plate and lever assemblies. The arms of the levers run parallel, forming a 'Y', the bends about 12-13 mm from the tips. When the joystick is moved to left and right extremes the levers should not restrict full axis movement. Slight tweaking may be necessary, but be careful not to allow the pliers to slip and damage anything. If you're not sure remove the whole assembly and do the job outside, then refit and recheck. (All this is practically not half as difficult as it sounds!)

And where are we on ModFlight? Well, we've just told George to commence our let down from FL300 to FL100...slow to 240. Lots of things for us to do. Some tricky bends and twists. Real accurate stuff. For the next 30 minutes or so.

7) FITTING THE SPRINGS:

a) The springs have to be stretched between the 'Y' arms, about 2 mm below the tips. The spring eyes need to be retained there by cutting a shallow groove about 1 mm deep and parallel, not to the angled arm, but to the horizontal plate edge. The idea is that the groove must be deep enough and angled correctly so that the spring eye cannot jump out, and not so deep as to weaken the lever wire.

b) Measure the distance between the levers from groove to groove with the levers gently clamped and stick centered. Now cut off a length of spring 10 mm less than this. Bend the end loops (i.e. 2 turns on each end) 90 degrees outwards. The eyes must be level with the outside curvature of the spring. Make only one spring at a time, and check the action on one axis. When that particular length gives you the desired return action, not too stiff, not too sloppy, you can use the same length for the other axis.

c) You may find that the spring fouls the motor clamp. If so, bend the 'Y' arms a little wider to give a clearance of 1/2 mm or so.

Note on detents: Ever got yourself tramlined in a car? A detent in a yoke or stick is a neutral position tramline. Very unnatural and very irritating. If your springs are too stiff when stretched across the 'Y' arms you will get that groovy feeling (OK somewhere else, not here). Also you will be nullifying too much of the force feedback action. On the other hand, too slack springs will not allow the stick to center accurately, and this defeats the very purpose of this whole exercise. Steer a middle course! Start with longer springs than necessary and shorten gradually (lop off one eye, then make a new one) until you get just the right return action. Obviously you'll have to fit back the plate and lever assembly after each adjustment to check stick action. Get one axis right first then just duplicate the spring length for the other axis.

8) PLAY CORRECTION:

There's one more vital setting to be done. Please ensure maximum accuracy here.

a) Fit the aileron plate and lever assembly in position.

b) Hold the aileron axis in the dead central position.

c) Check that both levers positively touch the stop screw on both sides, and the extension screw on both sides, simultaneously. There must be zero gap, otherwise there will be lost motion, and this axis will literally remain a flop. The tiniest gap here translates to a 1 mm free flop at the top of the stick.

d) If you can see any gap at any of these four points with the stick held in the dead center position you can bend to fine tune the appropriate lever. Wiggling the stick a very small amount helps you check for small gaps. Even a thin piece of paper should not be able to pass these four points in dead center position.

e) Repeat steps a) to d) for the elevator axis.

f) Finally, check the action of both axes simultaneously. The stick should consistently return to a now very definite center. There will be a very small free play--we know why its there.

This check must be done every time the plate and lever assembly of either axis is removed and refitted. Takes a couple of seconds, once you've got the hang!

ModFlight's slowing to 180. We're cleared to 4500' and a straight-in ILS approach. All the heavy stuff is behind us, but still no time to relax. Gear down (3 greens), flaps 20, touch of throttle to hold 180! Hang in there, George!

9) REFITTING MAIN CIRCUIT BOARD:

Time to get hold of the MCB from where you carefully stashed it in its antistatic bag. Rover (poor chap!) will think it's 'lamp-post' time when you get hold of his chain. You're going to need this while handling and fitting back the MCB. Hope you have all 11 black Philips screws. (I've lost 2...Didn't have a nice chap like me to tell me what to do!)

a) Place stick in unclassic inverted flight position. (Forgotten? See 1) u) ).

b) Remove MCB from its storage. Get oriented. Semicircular slot in board goes into black plastic male protrusion and long transparent plastic piece goes into its opening.

c) Identify the five little plugs and relate them to the MCB pins. Each one has a different number of pins and goes only one way...remember? OK. Plug them in carefully and recheck.

d) Dress (arrange) the wires neatly so that they will not foul any moving parts, place the board in position and secure. Move the joystick to all its extremes and check no wires foul moving parts. Reorient any wires that do.

e) Position the joystick plug wire grommet in its slot.

f) Carefully orient the cover, make sure the grommet is properly clamped and fix back all the screws. Not too tight...you're screwing into plastic. Still got the two rubber feet? Press 'em back.

Back to ModFlight! Screw back control column to cockpit floor!

500 feet...Say thanks and punch George (poor guy) out. Oh boy! We're manual. Flaps 40. Touch of down trim. Slowing to 140. 400 feet. Aah...feels good. Right on the beam. Terrific stick feel and response. 50 feet...40...30, back a tad, 20 feet. Just over the threshold. Power to idle. 10 feet. Check. Back, back, back...let 'er sink. Aah! The rest just falls into place!

Congratulations! We're at the end of ModFlight. One part of it was real and the other fantasy. I enjoyed both parts...hope you did too!

Back to reality. Plug in the 15 pin 'D' plug and joystick power pack plug. Bring on FS2000 and set joystick 'Forces' and 'Return to Center' sliders to default (or to your taste). Set Control 'Sensitivities', as given in Peter's "How To Fly The Big Iron", and away you go...this time for real!

So! Did I pass? Yes! Not in one shot...but yes! I think the Lady let me off a bit. Definitely fluffed a few things. Must do it again.

You know, I've been reading a lot about yokes and pedals. I just wonder...if it's possible to convert a 3 axis joystick to become a full fledged...? Hey, got to rush off...just got a couple of ideas!

Meantime, Happy precision flying!

Erich H. Leonhard

ehleonhard@yahoo.com