Everyone who's been paying attention to the Suncom throttle thread knows about my little eBay score here, but now that it's actually arrived on my doorstep today, it's time to give this baby its own thread.



Yes, it's in the older A-10A configuration, but it's nothing a little modification can't fix. However, there is no way this thing can sit on my desk, it's just too tall compared to the Thrustmaster rendition. No, something like this deserves a whole cockpit to be built around it, don't you think?



Yep, all the basic engine management switches are there. Of particular note are that the DN position on the flaps switch and the MOTOR position on the two engine switches are both lever-locking, so you actually have to pull out the switch a bit before you can move it into those positions.

Also weirdly enough, the IGN momentary position has much more spring resistance on the left engine switch than the right engine one. I should look into that.

There's a weird round black button thing on the right engine grip...except it does nothing. Doesn't budge one bit. Gotta peek inside the throttle grip to see what that's all about.



That plastic plate on the side of the right throttle just pulls right off, as you can see. While an A-10 throttle obviously isn't going to have anything there, it looks like Mason Controls just made it very easy for me to plop on whatever antenna elevation rotary I see fit.

Also worth noting is that the F-15 and A-10 right throttle grips do indeed have their mounting plate screw holes in all the same places. Swapping out the two grips shouldn't be difficult if I feel the need to do so.



As expected, the two throttle arms do not have any angle sensors attached. Those are something I'll have to figure out a mechanical linkage setup for.

But what I didn't expect were these two rollers actuating all these microswitches, ten for each side. I'm not entirely sure what purpose all the extra ones serve to the A-10 flight control system, but a couple of them are actuated in just the right places to be linked to the idle/cutoff detent.

I've got a lot of rewiring and enclosure-building ahead of me with this, particularly to make it as functional or even moreso than the A-10C and F-15E throttles, but once it's all done, don't be surprised if that TM Warthog throttle finds its way onto the B/S/T section...

Amazing find. Love the looks of it.

I'm bookmarking this thread!
How much was it?

Just $325 with an extra $12.70 for shipping, astonishingly enough!

I wasn't even looking for A-10 stuff at the time, but rather F-15 throttle stuff to complement my right-hand grip. Then eBay showed that in the related items list, and, well, I couldn't help myself, especially as someone who knows how to use a soldering iron and wire up stuff to joystick controller boards...

Seriously, at that price, I can easily recoup most of the cost just by selling the TM Warthog throttle, though I'm not quite sure I want to do that until the USB conversion project's fully done.

I carefully unscrewed the right throttle grip and the slew control. Wanted to tighten up the slew control's click-down function a bit.



Well, this is a surprise. There ISN'T caulk and glue covering up everything in here like with the F-15 grip. Makes my job a heck of a lot easier, in a way.

But I also wasn't expecting that huge PCB in there. It looks like it has resistors, a capacitor, trim pots, and perhaps the latter two parts have something to do with how the slew control's wired up to it. Could it be a load cell amp? That'll make using it so much easier.

It also looks like the "button" may be a couple of those IR proximity sensors like force-feedback joysticks usually have. I think it's a dead man's switch.

The only issue is that the PCB is clearly in the way of where I'd mount the boat switch or any other switch between the speedbrake and china hat switches. I'd have to move or possibly even outright replace it.

The alternative option, of course, is to use my F-15 grip in its place, which doesn't have a gigantic PCB in there blocking things. I can always amp the slew control signal outboard if need be.

Meanwhile, I'm probably going to hit a local hardware store this weekend and see if there are screws that'll fit the mounting holes on the front and back neatly. Once I know that, I can start planning and building a proper enclosure to hold this throttle, one that'll hopefully expand to being part of a full-fledged cockpit later.

WOW, great find at that price!

It sure is the biz...

(edit: insufficient caffeine)

Knocked out the mic switch and the flare dispense/right engine start button without too much trouble, in preparation for switch replacements.

I gotta say, I wasn't expecting the two-way mic switch to look and feel like cheap plastic crap. It actually has a sort of gritty feel to its movement and doesn't move too smoothly, likely because of wear over the years. The Suncom throttle already has a much better-feeling mic switch at the top with how smooth it is, and anyone with a HOTAS Cougar loaded with IJ's aluminum speedbrake/dogfight switch housings already has something way, way better.

The speedbrake switch isn't too different in terms of switch type and construction, but at least it feels a little smoother when pulling back, if not Warthog throttle smooth. Could be a matter of wear, could be a matter of even the mil-spec stuff not being perfect.

The china hat switch, though...it feels SO NICE. Solid spring action, nice click in both directions, makes me kinda wish I had a boat switch cap that fits these switches so I could stick it on either the F-15 weapon mode switch or A-10 china hat switch, depending on which aircraft I wanna stick closer to. That would save me the hassle of finding another mil-spec equivalent.

I also counted the wires leading into the grips. The right-hand grip fortunately has enough to provide for the upgraded switches I intend to install, assuming everything has a common ground wire. That's mainly because of redundant sets of wires on the mic and speedbrake switches, along with an extra wire for the slew control push-down.

The left grip, on the other hand...the pinky switch was single-pole, and as such they only routed in the minimum of wires necessary. I'll have to change out that end entirely if I'm going to add an antenna elevation rotary, and I was hoping to avoid that in the first place because of the amount of disassembly that would entail.

I just need to source some hat switches and an antenna elevation rotary with center detent and a cap of the right size. I could e-mail TM/Guillemot support for some cheap Cougar/Warthog spare parts; mil-spec they aren't, but they'll suffice for now.

I also need proper controller boards; when my next paycheck hits, I'm ordering a couple of Atmel SAM4S XPLAINED Pro evaluation boards and Hempsticking 'em. This project only needs one; the other is for the Suncom F-15 HOTAS conversion project I'll cover in the other thread.

I just realized I've had this throttle for almost a month now and still haven't really done anything with it due to other expenses preventing me from procuring the parts I need to really get going with it. (Controller boards are secondary to just not having the hat switches I need to begin with alongside not having a momentary boat switch. OTTO Controls prices are no joke!)

I feel kinda bad about that, but I'd rather not get too rash with this throttle and break something I shouldn't.

What I did find when looking for A-10 pits was that the A-10C, at least a real mil-spec simulator that I saw a photo of, has a completely different throttle grip assembly, speedbrake switch shifted forward and everything. There's no way I'd have enough spacing to tack a boat switch onto the A-10A grip, especially with that PCB in the way.

I have little choice other than to use my F-15 grip on that side and make this some kinda weird hybrid throttle to get the level of functionality I want. Fine by me, considering that I'm planning on putting an antenna elevation rotary on the left throttle anyway and it'll be more of an F-15-ish throttle on an A-10 quadrant base in the end.

By the way, it looks like the seller of this particular throttle had another to spare, in case any of you want to do your own conversion project.

http://www.ebay.com/itm/Fairchild-Republ...t-/261635104138

Wow. I never looked at a real A-10 throttle that closely before. Apparently I've had a set of A-10A grips for years, just sitting in a box. *laughs*

NamelessPFG, you might want to look into creating a 3D model of the grip shells and have them printed. I can make you a deal on printing them.

g.

If I knew how to create 3D models of them in the first place, I'd be more than happy to.

One guy over at the Star Citizen forums already had a 3D-printed copy of the left throttle grip for his own DIY project around a surplus F-15 right throttle grip, but it has some intentional deviations from the real thing. Added space for a switch next to the usual pushbutton, no accommodation for an antenna elevation rotary.

He did it based on some Mason Controls throttle grip blueprints he found, but while said blueprints have an F-15 throttle with MSIP hat, they don't have anything for either the A-10A or A-10C. Not that it matters too much when they're more or less the exact same grips with different switch mount cutouts...

I grabbed it.... THANK YOU NamelessPFG, again!!!

I will model it in 3D... check out http://www.hempstick.org/The_Official_Hempstick_Site/Flight_Sim_3D_Models.html. I released these models in Creative Common, non-commercial International. So, will the A10 throttle be. I am slowing building up a library of them.

Congrats! I was hoping someone of your expertise would nab it, especially considering you know how to 3D-model stuff for printing already.

You might also be more comfortable with figuring out how to extensively dismantle these things, at least enough to get the throttle arms and pivots separated from the rest of the throttle quad base. That would make working on a few things much easier for me, but I'm a bit hesitant to do anything really rash with this throttle at the moment unless it's about rewiring things.

I just got the throttle about 20 minutes ago. And I think I figured out what that bank of rotary white wheels and microswitches are for.

It's a mechanical digital sensor!

It's quite simple, really. It's 10 bit digital for each throttle. There are 10 micro switches, hence 10 bits. The white plastic wheels are the activators (sort of like a CAM device). You can see that the white plastic has bumps, each bump defines when a corresponding micro switch is switched on/off. Each white plastic also has horizontal grooves on the lower circumference. These grooves are the thread that engage the corresponding 10 screws right next to the white plastic. The screws are the adjustment mechanism to change when a microswitch is turned on/off. When you turn each screw, it engages the grooves and adjust the angle of the white plastic, changing when the microswitch is turned on/off, sort of like rack and pinion, except that this time the rack is the grooves on the circumference.

Wow, that is some ingenious old time digital "sensor" constructed out of mechanical means!!! That's how they did it without ICs!

If you are going to directly interface with the throttles.... you would need a controller with at least 39 digital inputs plus 2 analog for the force sensor. Out of the 39 digital inputs, 20 of them for the two throttles cannot be matrix'd. Because the shadowing & ghosting of the the matrix will totally screw up your 10 bit reading.

I don't know what code the 10 bit generates yet. It's for sure not 2's complement binary (if it were, one of the wheel will have on/off/on/off/one/off.... but none of them does, and it's not wise to use it two's complement anyway). If I have to guess, it's likely 10 bit Gray Code, because that's what Gray Code is designed for -- to make sure only one switch transition happens at each possible position. However, take this with a huge grain of salt, it's just an educated guess. Also, since it's adjustable, it's possible to take the 10 bit and generate only a few positions, say, idle, cruise, full power, and nothing more. In any case, it's "up to" 1024.

I'm wondering what all the microswitches for the throttle arms are for when it comes to the underlying A-10 avionics. I'd only need one or two of them for the idle/cutoff detent transition.

Needless to say, for our purposes, I feel that it would be more practical to just use angle sensors connected to the linkages with bearings, which most likely went to the real jet's engine throttle linkages. Fine-tuning the actuation wheels for proper digital readouts throughout the entire throttle range sounds like a big pain.

Also, I've moved my throttle back and forth, and I've noticed considerable deviations in the actuation and release of those microswitches, enough that setting it up in a Gray Code sort of arrangement may very well be impossible unless you like hysteresis.

I would have to agree that the original mechanical digital output is too finicky. I will be surprised if the latest A10 model still uses this. Just throw two Hall Sensors there and be done with it.

NamelessPFG, let me know if you need some MLX90316 programmed. I have the PTC-04 programmer & I always have some MLX90316 & MLX90333 lying around.

Apparently, he has yet another of this A10 throttle for sale.

http://www.ebay.com/itm/Fairchild-Republ...=item3a9abb9a30

I was about to bring that up! Too bad this one's missing the slew control cap for some reason.

Where did he get all these A-10 throttle quadrants, anyway? Must've been one heck of a score, just like those folks with the seemingly endless amount of F-15 right throttle grips.

About those Melexis sensors, aren't they generally used in a two-axis manner, usually at the heart of a stick gimbal? I figure I could use pretty much any sort of Hall sensor or digital magnetoresistor for the throttle lever axes.

One of those might come in really handy for updating this Suncom SFS Stick I've got lying around, though. The stock slide pots are kinda crappy, and I don't think it would be too much trouble to adapt Suncom's saturn ring gimbal to use one of those sensors.

MLX90333 does both X&Y axes, the same used in the Warthog stick. MLX90316 is a rotary Hall sensor used in the 2x throttles of the Warthog (Warthog uses both of these chips in digital SPI mode instead of analog mode though). The biggest advantage of using MLX Hall chips vs other magnetometer chips is that MLX chips are integrated with Auto Gain Control (AGC, i.e. amplifier with auto gain control), and a DSP module on it. So, MLX condition their own signal, which they know better than anybody else how to do that. It's your signal, you do it and don't make me do it! Similar to what Milan said.... b/c of the AGC, it's very resistant to magnetic field variations, and you don't need to add a separate DSP chip to do it. Not only that, it comes out with with two ratiometric analog output (if so programmed), thus you can program it to whatever range you wish. Say, your board is a 3.3V, program it to output 0 to 3.3V. If your board is 5V, program it to output 0-5V (although the input is still 5V). And you can set the center point and the range of the response. I used it to construct the Drop In Hall sensor for Cougar to replace the Cougar throttle pot. Not only that, you don't need to fuzz with the location of the magnet. I have used cheap toy Neo magnet. Put it 0.5mm away, or 1cm away, work just fine, unlike the Allegro Hall chip with which you have to be very carefully locate your magnet; spending a whole afternoon twiddling the magnet with the Allegro chip is not uncommon.

Sure MLX chips are more expensive than the raw magnetometer chips (but only a couple of bucks), but you save on the DSP and trouble. MLX90316 is just great for constructing your own pot replacement with minimum fuzz. But, like I said, the only trouble is the ridiculously expensive programmer required to program the chips.

Whoops, got my Melexis sensors mixed up. Looks like the MLX90316 just needs a magnet rotating next to it...gotta figure out how I'd arrange that physically and link it to the throttle assembly.

If they're not as finicky as the Hall sensors I'm used to, that'll be much appreciated. Modding that Saitek X-52 Pro the way I did by moving the magnets was a pain! I had to keep checking for linearity and responses at both extremes, and it all could've been avoided if they just designed their Hall sensor/magnet configuration properly and shipped it that way from the factory.

I have to ask, though: why do they need such an expensive programmer to use? Can I calibrate them without needing such a programmer, much as TM has the firmware calibration utilities for the Warthog stick and throttle?

I figure it's kinda important to be able to program them with whatever calibrations suit my particular sensor/magnet/linkage arrangement, unless you're willing to make self-contained angle sensors out of them or it's not really that necessary for a consistent response once programmed.

Once programmed, it's extremely easy to use. And the programming is quite easy too (but a pain in the ass to setup the whole stack).



Because they could, I guess... it's probably the old way of doing business... you charge the developers for what it takes to develop the tools, instead of amortize the cost of development of the tools to the chips you sell. These days, very few manufacturers charge developers much for the tools, except a few like Melexis. However, they are changing a bit. The new MLX90363 does not need a programmer. But it's all digital SPI (and standard 4-wire SPI, instead of the non-standard nasty 3-wire SPI of MLX90333). But you do need to write the whole driver for it. I haven't finished this one yet. Milan does.

Warthog does the 3-wire SPI digital mode of the MLX90333, reading off raw X/Y/Z values. I guess, Warthog's firmware is the one doing the calibration, not doing to the MLX90333 (at least I would). I guess TM chose to use a special calibration program to do this.




Yes, you do need to know where your center should be, and your response curve to be programmed in. But you don't need to be very accurate about it. Just about there so you don't waste too much resolution, and then calibrate it in Windows. And the chip also needs 3 capacitors to run in analog mode. So you might need a PCB to host the chip and capacitors (it's quite simple, just copy the recommended circuit from the spec sheet and route the PCB to the way you like it).

The way I did DiH is that I simply drill a 1/8" D hole in the D shaft and super glue the magnet in. What's left is just a case to fix the PCB and the shaft. Then, I program the MLX90316's response curve to fit. No linkage needed at all.

Actually, I have been experimenting with a couple of self contained designs (same size as a regular pot; well, the old DiH was one of them). But I am waiting for the "Super Vat" to be released for my SLA 3D printer before I do much about it. Other methods for making the case was too painful... like the DiH, the epoxy potting method was just too much work and too much fuzz to make.