Uuuuaaaah..! I have started building now..!

[CyBerkut]

Damn, you do nice work. If I weren't so lazy, I'd even say you are an inspiration!

[Triggerhappy69]

Thanx mate..! That was probably the best compliment I've ever heard..

 

 

 

 

WIP on the rest of the pit.. Please tell me where I have gone wrong here? I know it's something, but cant find it

[CyBerkut]

What is your video display setup going to be?

[Schwinn]

Please tell me where I have gone wrong here?

 

You are setting the bar to high for the rest of us? LOL =) Seriously it's going to be one good looking rig!!!

[HitchHikingFlatlander]

Yeah ^what he said lmao!

[Triggerhappy69]

Pitbuilding Update - 12 July 2009 - Under the weather while the sun is shining?

 

What is your video display setup going to be?

 

Good question..! I have a couple of projectors arround here somewhere that I want to test? Mostly since I am planning on building the window frames? And my dream is to be able to project a 160 degree field of vision on three projectors. And in this way not having to rely on trackIR or any other headtracking device..

But from what I have seen while testing projectors with Flaming Cliffs, they are not good enough.! At least not the ones I have..!

So I guess my answer would be: I have no clue yet..!?:cry:

 

I have been knocked out by some infection in a mucle attachment in the back of my head this week. So my progress has been slow..

 

However me and AlexRC had a very nice work session (was it 5 hours on MSN AlexRC?). Me with a bunch of question, a multimeter and the MS Sidewinder FFB2 controller card on my desk. And AlexRC with he's infinate knowlege of electronics and a good portion of patience with my asking stupid questions! Not only was it really nice to "spend some time" with a fellow simmer (and one I must say I admire quite a lot to!)!

 

But we (when I write "we" from now on, please know that it's really AlexRC I mean. The rest is just me wanting to take credit for he's work.. Self image issues, remember..? LoL) managed to get some interesting results from our labour.

 

The picture shows the PCB inside the MS SWFFB2 stick:

 

The 4-wire plug at the far right side is the wires to the X and Y-Axis FFB motors.

The four small 8-leg microchips are H-Bridges that recieve a PWM signal from the larger microchips situated approximately center of the PCB, but never mind them for now!

 

The two upper H-Bridges feed the X-Axis motor, and the two low... you get ir right?!

 

Now the H-Bridge chips have eight legs. Four of them (the four on the right hand side of the chip) are bridged, and feed one of the the motors power cables.

The four on the right hand side (legs 1 through 4) recieve either a small PWM signal, or a 24,5 Volt DC current.

 

I'll come back to the details of H-Bridges next time, but for now let's just say that they work like this:

a tiny pwm signal is fed to the H-bridge Chip together with a HUGE current that the motors will require to run. And with a little magic and some logic, the tiny pwm signal is magnified to an identical looking but much bigger PWM signal that the FFB motors just love..!

 

Our goal is to either extract the original tiny PWM signal and feed this into a MUCH BIGGER H-Bridge. So I can satisfy my need to over do everything!:joystick:

 

Also with motors sucking power from a 6AMP power supply, and with lot's more torque. I suspect that the Trim function will feel more rigid? Naturally I am paralell to this building a new joystick gimbal in 20mm Aluminum and ballbearings all over to hold the bigger motors. Nothing to extravagant, just a beefed up version of the plastic one in MS Sidewinder FFB2.. With hall effect sencors.

 

Whats holding me back you say? Well I am still waiting for the H-bridge cips to arrive from Germany.

 

And then I have to start looking for new motors.. But that can wait until tomorrow..!

[Triggerhappy69]

Oh yeah..! I almost forgot..!?

 

I am still working on the drawings for the panels, glareshield and window frames..

 

I have started from scratch from the other end this time. Deciding the inner bulkheads that hold the panels first, and then moving to the outer skin after these are in place..

 

Here are some VERY WIP previews (I just mirrored left and right side so you could see the general idea):

 

The line of sight has been raised 50mm to emulate the real life Ka-50 pilots field of view.

 

 

The "bumps" on each side (the right side one will be smaller like in RL) is mainly for stabilising the sidewalls if I for some reason choose to not build a full enclosure.. Without them the whole pit would be like a cardboard box with no lid..! Real flimsy.!

 

 

just to give you a rough idea of the proporsions I work from..? the grey lines are scetched from a sideways photograph of a Ka-50 sitting on the ground, and then scaled to fit. I used the pilots line of sight and he's elbow as reference points. Since these are important in my pit.. I mean, they do after all sit on ME..!? LMAO..! I should definatly go to bed now..!

 

Cya L8R M8's

[Triggerhappy69]

You are setting the bar to high for the rest of us? LOL =) Seriously it's going to be one good looking rig!!!

 

Bar..? Is the Bar open now..? WAIT FOR ME..! :pilotfly:

[Schwinn]

For the bulkheads if you don't have a cnc router...I just bought a regular router table today for 80.00 Cad and inverted the router. Wow what a difference and a time saver too. The bulkheads I am making come out smooth with very little need for sanding.

 

PS I hate you for being so good at Solidworks lol. I can't use it very well =)

[Alex_rcpilot]

I have been knocked out by some infection in a mucle attachment in the back of my head this week. So my progress has been slow..

 

However me and AlexRC had a very nice work session (was it 5 hours on MSN AlexRC?). Me with a bunch of question, a multimeter and the MS Sidewinder FFB2 controller card on my desk....

 

Dude, tell us about the mysterious infection you got after you recover. And you're welcome about the help. It's kinda fun to work out problems from tens of thousands of miles away. :lol:

 

I figured I may help you cut some corners by drawing a couple of schematics to explain our work.

 

 

H-bridge fundamental:

 

The direction of DC motor's rotation can be controlled with the direction of current. An H-bridge enables this control by activating(asserting) appropriate switching components located on its arms across the bridge.

 

The table below explains how it works. Without proper protection, an inadvertent simultaneous assertion with both switching components on either side will cause a short circuit and fry the bridge.

 

 

Half bridges:

 

By having trigger's hands full examining the PCB, we found out that each of the 4 SO-8 chips is a "half bridge" chip, as illustrated below:

The 4 output pins are internally connected together to fit in the standard SO-8 package, they don't contribute much in carrying more current since power comes from pins #1 and #3.

 

 

Digging deeper:

 

By overwhelming trigger with the multimeter in hand probing throughout the board, we were able to picture the layout of the original motor driving circuitry on the MS FFB stick as shown below, but let's not hurry down at this moment, coz the diagram also shows a portion of the L6202 datasheet.

 

We're trying to migrate the Microsoft FFB control board to an electro-mechanical system composed of two motors with two L6202 full H-bridge chips. What the L6202 requires is a power supply (which is 24V in this case), and two control signals that tell the chip how to turn the motor.

 

The diagrams I copied from the datasheet explains what combinations the chip needs, and what magnitude they are of.

I didn't spend much time to speculate what model those SOT23-5 / SOT23* transistors could be, that would require more testing data from trigger. Currently we already have enough information to draw the table with "InputA | InputB | Motor" combinations.

 

And I think those transistors somehow estabilishe a protection mechanism to prevent cross-conduction damage from inadvertent false assertion. That's why I put "STOP" after the wrong combination instead of "screwed".

 

If we cross-reference the two tables within this diagram, we may find many similarities, and that means by stealling inputs A & B from the Microsoft FFB circuit board, we will have the required feed for our own FFB motors which are more powerful.

 

However, a resister ladder will be required to confine the input voltage well within 7V. The following values yield approximately 5.1V. Depending on the fact that IN1 / IN2 from the L6202 only require very weak current (-10~30uA), a resistor ladder of 20K+15K may still guarantee good singal transmission.

 

------------------------------------------------------------------------------------------------------------------

*SOT23-5 and SOT23: standard surface mount packages with either 5 pins or 3 pins as shown in the photographs above.

 

 

*Edit: Trigger was worried about possible confusions we might result in by explaining PWM in this thread. So I put it up here as an appendix.

 

Pulse Width Modulation is actually a combination of two distinctive states alternating so rapidly that they produce something "in between". Like when the motor turns clockwise as explained in the diagrams above, it spins at full power producing maximum torque(100%). And when it stops there's no torque at all(0%).

 

If we need half torque output, we may basically spin it at full speed for a duration of time(Ton), and let it lose for a while(Toff) after that. By repeating this ON and OFF operation at a very high frequency, say 10KHz(Ton + Toff = 100 microseconds), the motor output appears very smooth. By adjusting the duty, which is equal to Ton / (Ton + Toff), we may control how much torque is output from the motor.

 

A small side-note: a motor is an inductive load, which means when PWM is implemented, though the voltage across the motor is alternating rapidly between 0 and maximum, the current doesn't swing as abruptly.

Edited July 13, 2009 by Alex_rcpilot

[Sokol1_br]

Naturally I am paralell to this building a new joystick gimbal in 20mm Aluminum and ballbearings all over to hold the bigger motors.

]

 

Trigger,

 

I see that you is going to make a entire new stick - with M$ eletronics.

 

Maybe this have any usefull:

 

http://buggies.builtforfun.co.uk/Sim/force...back-yoke-1.php

 

S!

 

Sokol1_Br

[HitchHikingFlatlander]

 

 

 

 

 

Holy smokes your building a robot too!:thumbup:

[Swetrekki]

"So I can satisfy my need to over do everything!"

 

If its worth dooing its worth over dooing ;)

keep at it, in the end that thing will probably actually take to the skys :)

 

and thanks Alex, a very good explanation!

// Trekki

[CAT_101st]

HAHHAHAHAHA ROFL Hitch you are funny. Geezz (wiping tears from eyes)

[leafer]

Trigger, you are one loco cornfed and I meant that in a loving way. :D That pit is going to be sweet! Can't wait to see your two seater Apache pit. :D

[Triggerhappy69]

Pitbuilding Update - 18 July 2009 - Force Feedback amplifying, the story continues!

 

All I can say is WOW..!

 

This whole experiment has proven to be one of the most exiting things ever..! Alex..? I can't begin to thank you for sharing your knowlege with me like you have so far.

 

Two years ago I had no clue about electronics, and now I find the learning curve is getting steeper and steeper every day, and it's a real challenge to keep it up!

 

Me and Alex have had hours and hours on video conferance now trying to get it all together. And it's been a share joy everey second of it.. However.. I am not quite ready to post the How-to yet.. I know.. It sucks.. But before I make someone go bananas inside theyr MSSWFFB2 joystick I want to be sure this actually works..:music_whistling:

 

We have managed to isolate where the PWM signal is coming fro, on the MSSWFFB2 PCB.

 

When measuring the "voltage" (I use my rabbit ears because I have learnt that allthough it is shown as voltage getting stronger and weaker in the multimeter, It's actually a constant 12V current that is switched rapidly on and of with varying pulse time.. Hence the term "Pulse Width Modulation"! The multimeter shows you this as let's say a 9,7V voltage because it is showing you an average voltage over a certain time. Not nearly as short a time as a PWM pulse though! And that's why we read this 12V PWM signal as a 9,6V current.. Phuuh)

 

Enough with the digressions.. I measured the voltage between pin 13 and ground and pin 14 and ground, while moving the stick front to back in the Joystick's Y-Axis.

 

Stick in Center position: Pin 13 is 0V - Pin 14 is 0V

Stick in forwards position: Pin 13 is 12V - Pin 14 is 0V

Stick in Back position: Pin 13 is 0V - Pin 14 is 12V

 

So what does this tell us?

In simple terms it tells me that when I move the stick forwards from 50-100% on the Y-Axis the LM339 Chip is sending out a PWM signal thought Pin13 to the the H-bridge, so that the motor starts pulling the stick back to center. And the same goes for pin 14 when I move the stick for 50-0% on the Y-Axis.

 

So now we know where to pull the PWM signal for the Y-axis from, and feed to your own H-Bridge..

 

I have made a simple schematics showing ONE H-Bridge only (we need two identical like this.) No promises on the schematic being 100% correct though, because I haven't had a chance to run it by Alex yet..

 

 

I ordered the resistors, capasitors and all the other parts yesterday, so you'll have to wait a couple more days before I post a new update regarding this..

 

Some thoughts, and where I am now..

 

I have been thinking.. MS Sidewinder FFB2 USB is a OLD Joystick.. FFB joysticks really have'nt had theyr rennisance before now with DCS. And as such the technology is unprogressed for 15 YEARS.! However FFB Racing wheels have been steadily developed, and I hope improved as well..!? Would it be possible to use to PCB's from a High-end FFB wheel of newer date. And designate one axis to each card. And by this get higher resolution, and even bypass the whole "amplifying" issue for other morons like me with itchy fingers and a qurious mind?

 

... just a thought...

 

Today I am cutting out the hole for my ABRIS. I'll be using a 10,4 Inch Touchscreen for this. The "touch" part of the screen might come in handy if I'll ever bother flying something other than the Ka-50 (with I stronly doubt I will)..

 

I'll post some pics tonight.. Got my Camera back now.. :thumbup:

 

Cya L8r guys

Edited July 19, 2009 by Triggerhappy69
Had the wrong name for the microchip.. Fixed now!

[CyBerkut]

Some thoughts, and where I am now..

 

I have been thinking.. MS Sidewinder FFB2 USB is a OLD Joystick.. FFB joysticks really have'nt had theyr rennisance before now with DCS. And as such the technology is unprogressed for 15 YEARS.! However FFB Racing wheels have been steadily developed, and I hope improved as well..!? Would it be possible to use to PCB's from a High-end FFB wheel of newer date. And designate one axis to each card. And by this get higher resolution, and even bypass the whole "amplifying" issue for other morons like me with itchy fingers and a qurious mind?

 

... just a thought...

 

Semi-educated guessing here... I would think that you could use two FFB steering wheel PCB's to handle the X & Y axis signals... but it may very well require some work arounds that negate the potential advantage of increased resolution. I haven't followed the racing wheel controllers closely, but I seem to recall seeing some of them turning well in excess of 360 degrees. If you are using PCB's from a wheel that turns that many degrees, you may be looking at some serious gear ratios to get it down to the limited movement used by a cyclic (or joystick). It may be feasible to make the conversion electronically instead. On the other hand, the gearing down (or gearing up, depending on how you look at it) might be useful to get more torque.

 

Trigger, I know that you are no stranger to gears, and I'm not saying they are necessarily a bad thing to add in, but it does introduce the possibility of mechanical slippage... which would reduce precision, and could feel "sloppier".

 

I paused from writing this to attempt to quickly look up how many degrees of movement the Saitek R660GT had. I thought I had read somewhere that it was 270 degrees, but I can't find anything definitive with just a quick search. You would think manufacturers would list that in their specs. Ugh.

 

Any ways, so far I have no experience with FFB wheels, so take the above for what it is worth. I still have plans on the back burner to eventually meld a FFB sterring wheel (probably a Saitek R660GT) to my old Thrustmaster RCS pedals to make a custom FFB rudder control. That would merely be a case of mechanical hacking (I hope) with the device simply looking like a Saitek R660GT to the computer.

 

My electronics training was U.S. Navy circa 1975 (and not the latest and greatest stuff at that time, since they went with the time tested, tried and true approach to equipment selection). What you and Alex are doing is great, and looks promising. :) If I tried to figure that stuff out, I'd probably have a anneurism and stroke out... :book: :huh: :cry:

[Triggerhappy69]

Pitbuilding update - 22 july 2009 - Installing the ABRIS LCD screen

 

I had a Lilliput 10,4 inch touchscreen that I was planning on controlling my CnC Mill with. But since the ABRIS came before yhe CnC Mill was finished, it was re-allocated to thre cockpit.

 

These last weeks have been frustrating, and fun at the same time! Frustrating because I have'nt had time to do much "building". And as such haven't seen a lot of progress either... And fun because now I KNOW that I will get the HOCAS I have been dreaming of..

 

I thought I had busted the PCB for the MSSWFFB2 joystick the other day. All off a sudden the motors were dead, and I had no clue why.. Thoughts circled arround me short circuiting the microchip while testing. I am really not buildt for these 0,02mm size pads.. I can hardly even see them without a magnifying glass.. No really.. I use a magnifying glass just like Sherlock Holmes. And Alex is laughing at me calling me old..?! LMAO

 

And also finishing the drawings for the pit is taking FOREVER..! It's so hard fiinding the right balance between height and length and figuring out how to make it not fall apart after construcing the d*mn thing.. But I'm getting closer every day..

 

THE ABRIS

 

This was my starting point (more or less). I had cut the hole for the LCD screen with a hacksaw. A standard blade is fine for cutting aluminum. The fines tooth metal blades tend to get theyr teeth packed with aluminum that melts while sawing..

 

Remember to mark ALL the holes at this stage. You really don't want to have pieces of metal falling into open pcb's and short circuiting them because you forgot to drill holes for the gauges..

 

 

The inside of my touchscreen.. Basicly just two parts here. The screen and the PCB.

 

 

I used spacers from a computer cabinet to attach the screen to the panel. The spacers leave a 1mm gap between the aluminum panel and the screen that I'll fill with a thin layer of balck foam.

 

On top are 40mm spacers to fix the pcb on.

 

 

 

 

 

It does look cool right?

 

 

 

 

 

The PCB is fixed to a sheet of 2mm foam plastic sheet I had in one of my drawers.. It's absolutly non-conductive..

 

 

And it has a nice and white underside to hide the PCB behind

 

 

 

 

 

M-m-M...

 

Edited July 22, 2009 by Triggerhappy69

[Schwinn]

Damn man looks great. I have the 22 gauge steel just waiting and after seeing you do this I definitely have to do the same!!

[CAT_101st]

looks great

as always

[londo-cat]

Hey there TriggerHappy - I'm having a heck of a time getting dimensions for the right consoles for the BS. I'm just looking for the base dimensions (length, width, height) but so far no luck. Any help would be appreciated!

 

~ Londo-Cat

[Triggerhappy69]

Hey there TriggerHappy - I'm having a heck of a time getting dimensions for the right consoles for the BS. I'm just looking for the base dimensions (length, width, height) but so far no luck. Any help would be appreciated!

 

~ Londo-Cat

 

You are lycky mate..! ;) That is the one part I consider myself done with as of today..! I'll make some drawings and post them here tomorrow

[ruprecht]

Damn man, that is looking fine.

[Triggerhappy69]

I seem to recall seeing some of them turning well in excess of 360 degrees.

 

What you and Alex are doing is great, and looks promising. :) If I tried to figure that stuff out, I'd probably have a anneurism and stroke out... :book: :huh: :cry:

 

You are right about the wheels turn up to 900 degrees on some..

But on my FFB pedals I use the PCB from a wheel that turn 200 degrees, and just by using a hall effect sencor turning inside a magnetic field I cut that down to 30 degrees.? Am i mistaken when I say that the turn rate of a wheel is more dependant on the potmeter than the electronics? Hmm.. Dude..! Now you got me all confused here..:helpsmilie:

 

 

The thing is that the resolution on today's joysticks basicly stink..! The old Microsoft Sidewinder 2 USB (Red edition) has a 10bit resolution (1024 points of measuring on each axis) if I remember right?

 

However the new Thrustmaster T.16000M has 16000 points of measuring on each axis. AND it has hall effect sencors to.. Ruling out spiking and/or mechanical glitches that potmeters suffer from.. Makng it MUCH more precise (in theory at least) when trimming!

 

I assume that racing wheels have had a simular evolution, and they are quite cheap to buy used on ebay..

 

But will DCS accept this many FFB devices. And export the correct DirectX values to the correct card? That's another issue?

 

As far as I could read ED haven't even done anything with the problem of FFB stick users not being able to use curves in the new patch?:cry:

[CyBerkut]

You are right about the wheels turn up to 900 degrees on some..

But on my FFB pedals I use the PCB from a wheel that turn 200 degrees, and just by using a hall effect sencor turning inside a magnetic field I cut that down to 30 degrees.? Am i mistaken when I say that the turn rate of a wheel is more dependant on the potmeter than the electronics? Hmm.. Dude..! Now you got me all confused here..:helpsmilie:

 

Actually, I think you have it right. I wasn't thinking about your wise penchant for switching to hall sensors. Presumably, the circuit boards take the same voltage range of input for positional information as usual. That ought to allow you to do your hall sensor conversion. That being said... I would think that switching to hall sensors (necessitating the removal of the input pots) would largely negate the improvement on resolution that you were seeking by going with 2 FFB steering wheel PCBs. For potentiometer input based steering wheels, better resolution on newer models probably largely comes from using potentiometers that have more turns for a given resistance range.

 

But then, your hall sensor conversion may very well give you better resolution in and of itself (which could be true whether you are doing a MS Sidewinder FFB 2, or using 2 steering wheel PCBs...).

 

Of course, if the newer FFB steering wheels have finer control signal outputs to the motors... then you could still realize some improvement.

 

Has your head exploded yet? :D

 

The thing is that the resolution on today's joysticks basicly stink..! The old Microsoft Sidewinder 2 USB (Red edition) has a 10bit resolution (1024 points of measuring on each axis) if I remember right?

 

However the new Thrustmaster T.16000M has 16000 points of measuring on each axis. AND it has hall effect sencors to.. Ruling out spiking and/or mechanical glitches that potmeters suffer from.. Makng it MUCH more precise (in theory at least) when trimming!

 

I assume that racing wheels have had a simular evolution, and they are quite cheap to buy used on ebay..

 

But will DCS accept this many FFB devices. And export the correct DirectX values to the correct card? That's another issue?

 

As far as I could read ED haven't even done anything with the problem of FFB stick users not being able to use curves in the new patch?:cry:

 

I only have the single Saitek Evo FFB joystick at this point, so I can't test multiple FFB inputs simultaneously. I know that setting one of its axis to be the rudder input worked fine, so the software seems capable of keeping the axis signals straight.

 

In seem to recall that you did some satisfactory testing with multiple FFB devices already...

 

As for the patch and FFB curve use, I wasn't looking for that specifically when I first browsed the posting of the readme file. I'll have to give that a closer perusal when I can give it more time. :)