B737 Throttle Quadrant - Automated Thrust Lever Movement

/

Autothrottle arming switch is a solenoid operated switch clearly identified on the main Instrument Panel (MCP).  The switch is linked to the IAS/MACH speed window (adjacent) and to two A/T disconnect buttons located either side of the throttle lever handles. cp flight pro mcp

In this final post dealing with the conversion of the 737-500 throttle quadrant. I will discuss the automation and movement of the throttle thrust levers and touch on some problems that occurred.  I will also briefly discuss the installation and use of potentiometers.  Part of this post will be repetitive as I briefly discussed automation in an earlier post.

Avoiding Confusion - Automation

To avoid confusion, automation refers only to the movement of the two throttle thrust levers in relation to the %N1 output.  These N1 limits and targets are provided by the Flight Management Computer (FMC) and normally are used by the Autopilot Flight Director System (AFDS) and the Auto Throttle (A/T) to maintain airspeed and/or thrust setting.  

Automation and Movement - Interface Cards

Automation is the use of CMD A or CMD B (autopilot) to control the %N1 outputs from the Autothrottle (logic), and motorization is the moving of the throttle levers in unison with %N1 output.  To achieve this seamlessly, two interface and one controller card are used.

Alpha Quadrant Cards (2):  Each  motor controller card has the automation logic programmed directly to the card (via propriety software).  One card controls Auto Pilot CMD A while the other card controls Autopilot CMD B.

Phidget Advanced Servo Card (2):  This card acts as an interface and bridge between the Alpha Quadrant cards and the flight simulator platform used.

The card does not provide movement for the throttle thrust levers; this is controlled by a Phidget Motor Controller card.

Leo Bodnar BUO 836 A Joystick Controller Card:  This card will register in Windows the movement of levers, buttons and switches on the TQ.  Calibration of this card is done first in Windows, then in Flight Simulator (FSX/P3D), FSUIPC or the avionics suite used; for example, ProSim737.

The interface cards are mounted forward of the MIP within the Throttle Interface Module (TIM) and are connected to the throttle unit by custom VGA straight-through cables and to the computer by a single USB cable.

Main Controller Cards

The controller card I have used is not a Phidget card but a specialist card often used in robotics (Alpha Quadrant card).  The software to program the card has been independently developed by a software engineer and does not utilize Phidgets.

The technology used in the controller card is very similar to that utilized by NASA to control their robotic landers used in the space industry.  The technology is also used to control robots used in the car industry and in other mass production streams.  One of the benefits of the card is that it utilizes a software chip (firmware) that can be easily upgraded, reprogrammed, or replaced.  

The Alpha Quadrant cards provide the logic to operate the throttle automation (the movement of the thrust levers) and act as a bridge between the two cards and the avionics suite.

Being able to program each card allows replication of real aircraft logic and systems.  Whenever possible, these systems and their logic have been faithfully reproduced.

oem throttle. toga switches clearly seen

CMD A/B Autopilot - Two Independent Systems

Most throttle units only use one motor controller card which controls either CMD A or CMD B; whichever autopilot you select is controlled by the same card.  

In the real aircraft to provide for redundancy, each auto pilot system is separate.  This redundancy has been duplicated by using two Alpha Quadrant controller cards, rather than a single card.  Each controller card has been independently programmed and wired to operate on a separate system.  Therefore, although only one CMD is operational at any one time, a completely separate second system is available if CMD A or B is selected on the MCP.

Synchronized or Independent Lever Motorization

Synchronization refers to whether the two throttle thrust levers, based upon separate engine %N1 outputs, move in unison with each other (together) or move independently.

In earlier Boeing aircraft, such as the 707, 727 and 737 classics, the levers were roughly synchronized; however, the Next Generation has as a computer-operated fuel control system which can minutely adjust the %N1 of each engine.  This advanced fuel management can be observed in a real aircraft whereby each throttle lever creeps forward or aft independent of the other lever.

Programming flight simulator to read separate %N1 outputs for each engine, and then extrapolating the data to allow two motors to move the throttle levers independently is possible; however, the outputs are often inaccurate (for varying reasons).  This inaccuracy can often be observed on reproduction throttle units that exhibit a gap between lever one and lever two when automating %N1 outputs.  

It was decided to maintain the older system and have both levers synchronized.  Although this is not replicating the Next Generation system, it does make calibration easier.  If in the future incremental thrust lever movement is required, it’s a matter of adding another 12 Volt motor to the front of the throttle bulkhead to power the second thrust lever.  

Autothrottle activation will advance both thrust levers in unison to the fmc calculated %N1 output

Be aware that although both thrust levers are synchronized, the throttle handles may still show a slight difference in position in relation to each other.  This is caused by the varying tension that needs to be maintained on the fan belt connecting the 12 Volt motor to the mechanical system beneath the thrust levers.

Another aspect to note is that the position of the thrust levers during automation is arbitrary and is a visual representation of the %N1 output; it may or may not reflect the exact position on the throttle arc that the thrust lever would be placed if moved manually (by hand with Autothrottle turned off).  

Although the throttle is automated, manual override (moving the thrust levers by hand) is possible at any time, provided the override is within the constraints of the aircraft logic (programmed into the Alpha Quadrant card), and that provided by the flight avionics (ProSim737).  

Power Requirements and Mechanics

To provide the power to move the throttle thrust levers, a 12 Volt motor previously used to power electric automobile windows, is mounted forward of the throttle bulkhead (see image at bottom of post).  Connected to the motor's pulley is a fan belt that connects to the main pulley located beneath the thrust levers.  To enable the thrust levers to move in unison, a slip clutch, which is part of the main pulley assembly, is used.  

Captain-side TO/GA button is clearly seen below lever handles.  The button at the end of the handle is the Autothrottle disconnect button

ProSim737 Limitations - TO/GA and Auto Throttle Override

Unfortunately, concerning automation the ProSim737 is deficient in two areas: TO/GA and A/T Override.

(A)  TO/GA

In the real aircraft, the flight crew advances the thrust levers to power 40%N1 (or to whatever the airline policy dictates), allows the engines to spool, and then pushes the TO/GA button/s.  Pressing TO/GA causes the throttle to go on-line and to be controlled by the AFDS logic.  The throttle levers then advance automatically to whatever %N1 the logic deems appropriate based on takeoff calculations.

As at the time of writing. If you're using ProSim737, this will NOT occur.  Rather, you will observe the thrust levers retard before they advance (assuming you have moved the thrust levers to %40 N1).  The reason for this is nothing to do with how the throttle is calibrated, FSUIPC or anything else.  ProSim737 software controls the %N1 outputs for the automation of the thrust levers and the developer of the software has not fine-tuned the calibration in the software to take into account real-world avionics logic.  This thread located on the ProSim737 forum provides additional information. 

I have not tested Sim Avionics, but have been told this issue doesn't occur in their avionics suite.

There are two workarounds:  Engage TO/GA from idle (hardly realistic) or push the thrust levers to around 80% N1, allow the engines to spool, then push TO/GA.  Anything less that around 80% N1 will cause the thrust levers to retard before advancing.

Latest ProSim737 release (V133)

The latest version of ProSim737 (V-133) has provided improvement to the above issue.  Throttles can now be advanced to ~60% N1 and TOGA engaged without the throttle levers retarding.  This is possible ONLY if you calibrate the throttle levers within ProSim and allow ProSim to control the throttle output logic.  if you calibrate within FSUIPC then the same issue will apply.

According to ProSim developers, this issue is probably related to the calibration of the ProSim servo output. When you press TO/GA, the current N1 is taken and calculated back to a throttle percentage. This throttle percentage, when combined with the servo calibration data from ProSim results in a servo output. The servo calibration at the moment only has 2 calibration points, which are 0 and 100%. This results in a linear behavior between the two points, while depending on the construction of the throttle, the relationship might be non-linear. This would require a multi point calibration which is hard to do at the moment, because a throttle does not have exact readouts of the current position, so it will be hard to calibrate a 50% point.

This may need improvement in the code to auto calibrate the throttle system.

It's hoped that future release of ProSim737 will rectify this issue.

(B) Autothrottle Manual Override

In the real aircraft, manual override is available to a flight crew and the thrust levers can be retarded with the Autothrottle engaged.  When the flight crew release pressure on the thrust levers the Autothrottle will take control again and return the thrust levers to the appropriate position on the throttle arc dependent upon the speed indicated in the speed window of the MCP.

ProSim737 will not temporarily disconnect (manual override) the Autothrottle.  

At the time of writing, there is no workaround to solve this.

Potentiometers

There are many types of potentiometers; the two types most commonly used in flight simulation are the linear and rotary types. Linear potentiometers are inexpensive, often have a +- percentage variance, are compact, have a minimal throw depending upon the size of the device, and are not immune to contaminates building up on their carbon track. 

The last point is worth mentioning, as it is wrongly assumed that a potentiometer will remain correctly calibrated for the life of the unit.  General wear and tear, dust, and other debris will accumulate on the potentiometer; any of which may cause calibration and accuracy problems.  Keeping the potentiometers free of dust is important.

Rotary potentiometers (which may have a string attached) are very accurate, are in a sealed case and have very minimum chance of contamination. They are also made too exacting standards, are larger in size, and are expensive.

To read further about potentiometers

QAMP secured to base of throttle unit.  Thumb screws are visible on each corner of the plate.  A possible add on modification to reduce the risk of dust contamination to the potentiometers is a plastic cover that fits over the plate (a lunch box)

Calibration of Potentiometers

The main method of calibrating the position of the thrust levers is by calibrating the potentiometer in Windows, then in FSX followed by fine tuning in FSUIPC (if needed).  

At the moment I am using linear potentiometers; therefore, at some stage cleaning or replacement may be required.   The 737 throttle quadrant is not cavernous and only certain sized potentiometers will fit into the unit; this combined with other parts and wiring means that the potentiometers are often inaccessible without removing other components.  

To allow speedier access to the potentiometers, a Quick Assess Mounting Plate was designed.

quick access mounting plate (QAMP). four linear potentiometers are mounted to the plate. Two grub screws secure the plate to the throttle chassis

Quick Access Mounting Plate (QAMP)

The potentiometers are mounted directly onto a custom-made aluminum plate that is attached to the inside of the throttle unit by solid thumb screws.   To access the plate, the side inspection cover of the throttle is removed (a few screws) followed by turning the thumb screws on the access plate.  This releases the plate.

A similar plate has been designed and constructed for use with the stand-by potentiometer that controls the flaps.  A more detailed picture of the QAMP can be seen here..

Below is a video showing the movement of the thrust levers with the autothrottle engaged recorded during a test flight. 

 
 

Teething Issues with the Throttle Conversion

It was envisaged that more problems would have surfaced than have occurred.  The major issues are outlined below:

(A) Trim Wheels

An early problem encountered was that the trim wheels when engaging generated considerable noise.  After checking through the system, it was discovered that the two-speed rotation of the trim wheels were causing the two nuts that hold each of the trim wheels in place to become loose.  This in turn caused the trim wheels to wobble  slightly generating undue noise.  

Solution:

Tighten the two nuts at the end of the rod that holds the two trim wheels in place.

(B) Flaps 5 Not Engaging

The problem with the flaps 5 micro-button has been discussed in an earlier post.  To summarize, when you moved the flaps lever to flaps 5 the correct flaps were not selected on the aircraft or registered on the PoKeys 55 interface card.  Several hours were spent checking connections, micro-buttons, wiring and the custom VGA cables that connect the flaps section of the quadrant to the appropiate interface module; the problem could not be discovered.  

Solution:

One of the two Belkin powered hubs located within the IMM had been replaced with another powered unit.  It appears the problem was that the replacement hub had too low a voltage, as a replacement with a higher voltage solved the problem.

(C) Throttle Thrust Levers Not Synchronizing (A/T on)

The two thrust levers of the quadrant did not synchronize when the Auto Throttle (A/T) was engaged; one lever would always be ahead or behind of the other.  At other times they would split apart (do the splits) when A/T was engaged.  

Solution:

The problem was easily solved by altering the tension on the slip clutch nut.  When the nut was  tightened, both levers moved together as one unit.  The secret was finding the appropriate torque.

(D) Throttle Thrust Levers Difficult To Move in Manual Mode (A/T Off)

The ability to move the thrust levers in manual mode (Autothrottle turned off) was not fluid and the levers occasionally snagged or were sticky when trying to move them.  

The fan belt is barely visible linking the pulley of the motor to the main pulley inside the quadrant

This is caused by the fan belt not moving smoothly through the groove of the pulley wheel.   The Autothrottle when engaged overrides any stickiness due to the power and torque of the Auto Throttle motor.

Solution:

Unfortunately, there isn’t a lot you can do to rectify this issue as it’s a by-product of using a mechanical system in which the fan belt is central to the consistent operation of the unit.  

The conundrum is that if you tighten the fan belt too much you will be unable to manually move the thrust levers as they will be exceptionally stiff and difficult to move (as you are pushing against the tension of the fan belt); however, if you loosen the fan belt too much, although the levers will move fluidly by hand, the fan belt may not have enough tension to move the levers when Auto throttle is engaged.  It’s a matter of compromise; selecting an appropriate in-between tension to allow acceptable manual and Autothrottle operation.

A more reliable method is to use a small gearbox, a simple slip clutch and a coupler to connect to the spur gear.  Another option is to use an electrical system.

Further thought needs to be done in this area before a decision is made to replace the fan belt system.  If a new system is incorporated, the change-out will be documented in a future post.

Conclusion

Despite some of the shortcomings to this conversion, in particular the mechanical fan belt system, the throttle unit shows a marked improvement on the earlier 300 series conversion. 

Since the project began there has been three throttle conversions, and wth each conversion has built upon knowledge learnt from earlier conversions.  Initially there was the 737-300 conversion in 2012, which was converted in a rudimentary way and only operated in manual mode.  This was followed by the conversion of the 737-500 throttle in 2016.  This throttle was then rebuilt and upgraded in 2017.

Further Information

  • A summary of the articles that address the conversion of the 737-500 series throttle quadrant conversion, and the rebuild and update can be found in Flight Controls/Throttle Quadrant.

Acronyms and Glossary

  • AFDS - Autopilot Flight Director system

  • A/T – Autothrottle

  • CMD A/B - Autopilot on/off for system A or system B

  • Flight Avionics Software - Sim Avionics, ProSim737 or similar

  • FMC - Flight Management Computer

  • MCP - Main Control Panel

  • QAMP – Quick Access Mounting Plate

  • Throttle Arc – The arc of the thrust levers from the end of the blocks to fully forward.  The term refers to the curved piece of aluminum that the throttle levers are moved along

  • TO/GA - Takeoff Go-around switch

  • %N1 -  Very simply explained, %N1 is throttle demand and as N1 (and N2) spin at absurdly high speeds, it is easier to simply reference a percentage and display that to the crew. It's much easier for our brains to interpret a value on a scale of 0-100% rather than tens of thousands of RPM 

Replacement OEM 737-500 Throttle & Center Pedestal - Conversion to NG Style

/

737-300 throttle quadrant with old style paddle-style stab trim levers

The last few months have seen quite a bit of activity regarding the throttle quadrant and center pedestal, which has culminated in me selling my former 737-300 series throttle quadrant and pedestal and replacing it with an another unit from a late series 737-500 aircraft.

Brief Recap

In late 2012, I decided to convert the 737-300 throttle to full automation.  A dilemma I faced was whether to keep the throttle unit as a 300 series throttle with the attached two-bay pedestal, or do a full conversion to make it similar to the Next Generation. 

After careful consideration, it was decided convert the throttle quadrant.so it appeared as close as possible to the Next Generation.

Stab Trim Switches

One of the biggest differences, apart from thrust lever handles, between early model throttle units and the Next Generation units is the stab trim cut out switches.  On the earlier 300 series units, the switches are paddle / lever style switches while the Next Generation uses toggles and T-Locks.  T-Locks are a safety feature and sit beneath the toggle switches and are spring loaded; the pilot must push down the T-Lock to activate the toggle.  

To convert the trim switches requires cutting out the old switches and fitting new reproduction Next Generation switches.  This is a major task requiring precision work.  Although reproduction switches can be made, the reproduction T-Locks don't operate as the real T-Locks should.  I did search for some genuine T-Locks and toggles, however, my search was fruitless as these parts appear to be reused by airlines (or recycled).

Replacement 500 Series Throttle Quadrant & Three-Bay Center Pedestal

A friend of mine informed me that a late model 737-500 throttle quadrant was for sale.  This unit was in better shape than my existing throttle, included the genuine Next Generation style stab trim switches complete with T-Locks, and also had a three-bay center pedestal.  It appears provenance was shining on me as the new throttle appeared for sale a day before the stab trim switches were about to be removed (with a metal cutter...)

The throttle and center pedestal were purchased (you only live once!) and the 300 series throttle sold to an enthusiast in Sweden.

Next Generation Conversion

To bring an earlier style throttle and center pedestal to appear similar to a Next Generation throttle quadrant requires, at a minimum:

  • Attachment of a Next Generation style throttle lever shroud to existing aluminium levers;

  • Removal of TOGA buttons and relocation to bring design in-line with a Next Generation (the buttons are identical, but the housing is different);

  • Possible replacement of the stab trim switches;

  • Painting of throttle housing and center pedestal from Boeing grey to Boeing white; and,

  • Painting of all throttle knobs from Boeing grey to Boeing white.

The biggest hurdle is usually replacing the trim stab switches, however, as these are already present on the new throttle, and are the Next Generation, considerable time and expense was saved in not having to replace them.

Main Differences - Next Generation & Classic

The Boeing airframe that most people associate with today begins with the 737-200 and ends with the 737 Next Generation.  In between we have the classics which refer to the 737-300, 400 & 500 series airframes. The 737 Next Generation series includes the 737-600, 700, 800 & 900 series airframes.

The main differences between a classic and Next Generation throttle quadrant are:

  • The stab trim switches are slightly different; the classics having two flat levers while the Next Generation has toggle-style buttons with T-locks;

  • The throttle thrust lever handles; the classics are bare aluminium and the Next Generation is white aluminium that is ergonomically-shaped.  The TO/GA buttons are also positioned in a different place on the Next Generation.  The knobs (handles) on the levers are also coloured white rather than off-grey;

  • The method that the throttle thrust levers move during automation.  The classics move both thrust levers together when auto throttle is engaged.  The Next Generation moves each lever individually in what often is termed the throttle dance (this is due to the computerised fuel saving measures incorporated in the Next Generation);

  • The spacing (increments) between each flap lever position is identical in the Next Generation, but is different in the earlier series throttles;

  • The center pedestal in the classics is either a two-bay pedestal (early 300 series and before), but more likely a three-bay pedestal.  The Next Generation always has a three-bay pedestal.  Base materials for the center pedestal are also different - aluminium verses a plastic composite material;

  • The speedbrake knob is very slightly more elongated on the Next Generation unit; and,

  • The telephone, circuit breakers and mike assembly differ in type and location

Next Generation Skirt - Thrust Levers

Boeing when they designed the Next Generation style throttle didn’t design everything from new; they added to existing technology.  All Next Generation throttles utilise thrust levers which are identical to those of earlier units.  

Boeing designed a shroud or skirt that attaches over the existing thrust levers encapsulating the older thrust levers and sandwiching them between two Next Generation pieces.  The assembly is made from aluminium and is painted white.

The TO/GA buttons are located in a different position on the Next Generation units, although the buttons used are identical.

To alter the position of the TO/GA buttons you must detach  the small aluminium box from the 300 series thrust levers, remove the TO/GA buttons, and then re-solder the buttons in the appropriate location on the new unit.

I did not make the Next Generation skirt for the thrust levers, but rather had fabricated, from design specifications, a reproduction skirt.  The skirt is produced from aluminium and replicates the dimensions of the Boeing part.

Time-line, Functionality and Conversion

The throttle is initially being converted in the United States.  The advanced work (automation) will be done by a good friend in California, and then I will follow on with more mundane tasks.

The replacement unit will feature several improvements which will allow: full motorized functionality, full speed brake capability, accurate trim tab movement, alternate trim wheel spin speeds, correct park brake release, trim wheel braking and several other features. 

I want the functionality of the throttle to be as close as possible to that found in the real aircraft; therefore, the methods used to ensure this functionality will be slightly different from the norm.

When the throttle is fully functional and tested, I'll publish a post providing further information and detailed photographs of the various functions.

It is hoped everything will be completed, and the throttle and pedestal installed by late May 2013.  The next month or so will be quite exciting.

Two-bay Pedestal Will Be Missed

I know I will miss the narrower two-bay center pedestal.  A major advantage that will be lost is the ease in climbing into and out of the flight deck; the two-bay provided more room between the pedestal and the seats.  At some stage, I probably will need to install J-Rails because the seats I'm using are fixed-claw feet Weber pilot seats; J-Rails will be needed to allow lateral seat movement.

BELOW:  Montage of several images showing main visual differences between 737-300 classic series throttle quadrants and the 737 Next Generation. The 737-300 throttle is my old throttle but, the Next Generation throttle quadrant belongs to a mate of mine.

Montage of several images showing main visual differences between 737-300 classic series throttles and the 737 NG style throttle units. The 300 series TQ is my old throttle unit but, the NG Throttle quadrant belongs to a mate of mine

Sticky Autothrottle Button - Repaired

/

oem autothrottle button. note the circular circlip

I noticed soon after the throttle quadrant arrived that the engine number one auto throttle button was a bit sticky.  Depressing the button, it would stay pressed in for a few seconds even though pressure had been released.  The autothrottle buttons are one-way buttons meaning that they are click buttons.  It’s probable that after many hours of service, sweat, dead skin cells and dirt has built up on the inner button behind the spring mechanism; a friend suggested that DNA analysis of the built up debris would probably provide a list of suspect pilots!

Whilst the button was still in place, I attempted to loosen the built up material using a can of pressurized electronic cleaner fluid.  The fluid, I hoped would dislodge any loose material before evaporating.  Unfortunately, this didn’t work in the long run, although once lubricated with the evaporate solvent, the button operated correctly for a short time.

oem autothrottle button removed

Circlip

The button is held in place within the throttle handle by a ½ inch circlip.  Beneath the circlip and button there is a spring mechanism that pushes the button out after being depressed.  Using a pair of circlip pliers, I very carefully removed the circlip making sure that the spring mechanism of the circlip didn’t propel my button out the window and into the garden! 

With the circlip removed, the inner portion of the throttle handle slides out revealing the button and attached wiring.  The button is a modular design (shaped to fit inside the throttle handle) and unfortunately cannot be disassembled further, Therefore, I reassembled the button and sprayed a small amount of silicone spray around the button, allowing the silicone solution to penetrate around the the edge of the button. 

The silicon lubricant (which is non conductive, so there is no issue with power shorting) seems to have solved the problem as the button no longer sticks, however, this is only an interim solution.  I'll search for a replacement button module.  Sometimes the most simple solution will fix your problem.

No doubt I can purchase a new replacement from Boeing for err $800.00....  I think not.  Eventually I'll find a disused button module in my travels.  

Throttle Teething & Calibration Issues - It Was Expected

/

oem 737-300 throttle quadrant

The throttle quadrant works well and I’m pleased with the result; however, as anticipated there are a few minor teething issues that require fixing.  There is a background humming noise, The engine one autothrottle switch is sticky, and there are some minor issues with the calibration of the throttle reversers and deployment of the speed brake. 

Background Humming Noise

When the phidget software is turned on with FSX there is an annoying background hum.  Initially, I thought this background hum to be the low frequency AC noise, but then realized that everything is DC – so there shouldn’t be any noise.  The cause of the hum is probably related to either of the following issues:

  • When the phidget software is turned on it’s activating power to the servo motor to deploy the speed brakes.  The servo motor is ready and waiting for a command, but as there is no command for movement and the servo motor has power running to it, it’s humming.  If this is the reason, then the installation of the Phidget 004 card will solve this issue. 

A Phidget 004 card has four relays which allow for three situations – on, off and always on.  When connected, the relays will tell the servo motor to switch off“until activated by movement of the speed brake.

  • The power to the throttle quadrant is from a 400 watt computer power source and a bench-top voltage reducing board).  I’ve been told that because all the power requirements are coming from a singular source, then this maybe a cause of noise.  The easiest method to solve this is to use two or three independent power sources.

Speed Brake Calibration - Auto Deployment of Handle

Calibration is always an issue when simulating a complex piece of machinery.  Calibration must take into account the various positions and operational requirements of the speed brake.  The speed brake must be recognised by the flight software in the following positions: off, armed and part/full detent.  It must also be configured to automatically activate (deploy) upon flare and touch down when the landing wheels touch the ground. 

The Boeing Operations Manual states: the thrust reverser can be deployed when either radio altimeter senses less than 10 feet altitude, or when the air/ground safety sensor is in ground mode.  Movement of the reverse thrust levers is mechanically restricted until forward thrust levers are in idle position.

Once touch down in achieved, the mechanical speed brake arm on the throttle quadrant will move automatically to the deployment position (full detent).  This is done by programming a squat switch.  A squat switch is standard on/off relay that tells the brake to either deploy or remain in the non deployment position.

Squat Switch & FSUIPC Programming

To program a squat switch I used a Phidget 0/0/4 card and programmed the F2Phidgets software to read squat switch in the interface.

To ensure that the speed brake was calibrated to FSX correctly I used FSUIPC.  One important aspect of the calibration is to ensure that the speed brake handle matches, more or less, the same movement of the virtual speed brake handle.  To check this you must open the throttle in FSX and observe the virtual movement of the handle while manipulating the real handle.

Using FSUIPC, open the Axis Assignment tab and move the speedbrake handle checking that the arm and detent positions are correct.  Select send to FSUIPC and tick (check) the spoilers in the drop down box.  Finally save the adjustments.

If you have not done so already, it's a good idea to have a FSUIPC profile set up to ensure that your changes are saved to specific aircraft.  For example my FSUIPC profile is called 737 Project.

Reversers

Once a Phidget 0/0/4 card is installed and the card relays calibrated appropriately to the speedbrake, it’s hoped that the calibration of engine 1 and engine 2 reversers through detent position 1 and 2 will be straightforward.

After consulting with others and solving these issues, I'll post an update to this thread (here).  Perhaps the information may benefit someone else doing a similar throttle retrofit.

Update

on 2014-12-23 12:50 by FLAPS 2 APPROACH

The hum is now gone.  The reason was straightforward; there was no commands being issued to the servo motor, so the motor was making a noise (this continual power to the motor eventually lead to its failure).  Once a command was directed to the motor, the noise disappeared.  To stop the motor from being turned on when there was not a command; a phidget relay was connected (on/off).  When the command is issued, the phidget relay opens and when the command is rescinded the relay closes and power to the motor is stopped. 

Reverse Thrust Levers

This also was relatively straightforward to solve. 

My problem lay in the fact that I was trying to program the reverse thrust levers to read three power settings (three detents) as in the real 737 aircraft.  The first detent/notch on the reverse lever opens the clam shells/buckets to redirect air, the second detent/notch provides 50% idle thrust and the third detent/notch applies full reverse thrust. 

The third position is rarely used as it can easily ingest foreign bodies into the engines from a contaminated runway causing engine damage - the exception being in very wet and/or snow conditions, when it is regularly used.

Unfortunately, programming the reverse thrust levers like this is not possible as flight simulator does not have the appropriate logic. 

Using FSUIPC to assign buttons & offsets

Once the problem was understood (with a little help from a real-world 737 pilot and another simmer), FSUIPC was used to program individual thrust to engine 1 and engine 2.  In FSUIPC, you assign a button to each of the reverse thrust levers, and then assign a flight simulation command (throttle set 1 and throttle set 2).  An offset parameter is required for each throttle which is -16384 / 0.  The repeat box must also be checked/ticked.

Real World Throttle Nuisances – they had me baffled for a time

One thing that baffled me was the changing of the FSUIPC button sequences as I activated the reverse thrust levers. As the level travelled through the first detent, each lever would indicate, in the FSUIPC software, the correct button number.  However, as the levers travelled through the second and third detent positions the button number would change and indicate button 4.  Button 4 is assigned to the speed brake lever.  Why was button 4 continually being activated? 

It was then realized that at the forward end of the spee brake handle, where the brake arm recesses into the throttle, there is a small button.  This button had been wired to button 4.  Button 4 is activated when you either raise the brake handle to arm/deploy the speed brake, or when a cam is activated by the reverse thrust levers (the levers travel over the cam as they pass detent two and three position). 

Once the issue was understood, it was a matter of changing the Button 4 assignment.

Speedbrake Calibration

This is far more difficult and challenging than I thought it would be.  I've spent many hours attempting to have the speedbrake operate in the correct manner, but have failed on all counts.  Without going into detail, sometimes it sort of works and at other times it fails.  The 0/0/4 phidget card has been installed and the two relays programmed (squat switch), however, for some reason problems continue. 

This is going to require some in-depth thinking and help from others.  It’s possible that a problem exists with the servo motor.

Continuing the above theme, I've just spent a few hours talking to a friend in the US discussing this issue. It seems that it's highly probable that the servo motor has been burnt out.

The intermittent behaviour and motor humming was most likely caused by the motor beginning to ove heat which eventually caused the motor to fail.  

In the meantime, everything else on the throttle quadrant appears to work correctly, which includes using the actual speedbrake lever to arm, partially open, or fully deploy the spoilers.  The only part that does not"yet work correctly is the automated movement of the speedbrake lever when the squat switch is activated on landing.

Powering, Wiring and Configuring the 737-300 Throttle Quadrant

/

oem 737-300 throttle quadrant. installation of phidget 1064 interface card is to the forward bulkhead

The picture shows the front of the throttle quadrant with the attached 0064 and 0066 phidget cards and the BUO 836X Leo Bodnar card.  I thought this to be the best location for attaching the cards rather than having them either sit loose or be mounted on a separate board.  The wiring and cards will not be visible when the quadrant is sealed against the front of the main instrument panel (MIP). However, if servicing is required, access to the cards and wiring can easily be achieved via the front of the MIP.

The Phidget cards are required to provide functionality to the trim indicator, motorizing of the trim wheels (via a servo motor), and to allow the deployment of the auto speed brake.

Different Voltages Required

The throttle quadrant requires different voltages to operate correctly.  Apart from the obvious USB power through the USB cable connected to the cards, external power is supplied via a standard style computer power source, rated to 400 watts.  To reduce the main power, which is 240 volts in Australia, to that required by the phidget cards and integrated back lighting (IBL); I installed a bench top power board kit.  This small kit comes unassembled in a box direct from China.  Assembling the kit and card isn’t difficult but it does taken considerable time to solder all the terminals in place.  The bench top kit allows the power from the computer power source box  to be reduced to: 3.3 V, 5 V, +12 V and -12V.  Each power selection is protected by a 5 amp in-line fuse.  In an attempt to try and maintain neatness I mounted this card directly to the power source box. 

Functions on the throttle quadrant that require power are:

  • Integrated back lighting (IBL – aircraft bulbs) – 5 volts

  • Main parking brake light – 12 volts

  • Fire suppression module backlights and handle lights – 5 volts

  • Speed brake servo.  Phidget controlled servo motor - 5 volts & 12 volts

  • Trim wheels (spin when electric trim is activated from yoke) Phidget controlled servo motor – 12 volts

  • Lighting on/off switch (TQ IBL only) – 5 volts

  • Hobbs meter (to indicate length of time TQ has been operational) – 24 volts (12V + 12V)

connecting wires to power distribution board

The other avionics that will be installed into the avionics bay are powered directly via USB (unless real aircraft modules are used)

I wasn’t exactly sure what the amperage draw was from the servo motor (that spins the trim wheels and activates the speed brake).  Therefore, to connect the external power through the bench-top power kit, I decided to use 10 amp wire. I have a sneaky suspicion that 10 amp rated wire is overkill for the task, but at least I know it won’t melt.

If you want to view more detailed images, please navigate to the image gallery and select construction

Phidget & Leo Bodnar Card Programming

power distribution board positioned above computer power supply

Most of the buttons and levers located on the throttle are assignable to standard flight simulator controls through the windows joystick controller (or Leo Bodnar card).  But, those throttle functions that are controlled by a phidget card, initially require mapping through a registered version of FSUIPC, so that they can be seen within the phidget's interface to allow assignment and configuration.  I used a FSUIPC profile to map the functions controlled by phidget cards, which were: the trim indicators, trim wheels and speed brake. 

I'll be the first to admit that my knowledge of phidgets is lacking; Until recently I couldn't spell the word.  With the help of a very kind person from northern California who is exceptionally knowledgeable on phidgets my worries were soon overturned - at least for the time being.  During a two hour telephone hook-up, the correct computer drivers and phidget libraries were installed on the computer and the attached phidget cards on the throttle quadrant were programmed to the required throttle quadrant fields with various FS variances and offsets (after they were mapped in FSUIPC). 

As with many software related products, there was a bit of troubleshooting and configuration that needed to be done, but nothing too drastically complicated.  It all seems quite easy when you know how.

The throttle now has full functionality with the exception of the automatic deployment of the speed brake on flare and touch down.  This requires an additional Phidget card (004 card) which has four relays that can be computer controlled.  The relay is needed to activate the squat switch to turn off the servo motor allowing the speed brake to deploy.  This additional Phidget card will be installed shortly.

It was quite amusing when we programmed the phidgets to the trim wheel movement.  I hadn't expected the movement and was leaning on the trim wheel while discussing the issue on the phone.  BANG WHIRL as the trim wheel began to spin at a high number of revolutions.  The movement and noise startled me and I almost fell from my perch!  The TQ shook madly as the trim wheel rotated (as it isn't yet screwed to a platform) - I can now understand how real world pilots spill their coffee!

Programming the Leo Bodnar card was straightforward; this card follows the standard for windows joystick controllers.  Essentially, you just follow the screen prompts and allocate button functions to whatever devices you choose.

One aspect that required careful attention is to check that the flight simulator controls are not duplicated in either the  phidgets, Leo Bodnar, yoke, or other joystick controller settings.  duplicate settings will cause problems.

Throttle Functionality Includes:

  • Independent forward and reverse thrust to engine 1/2 throttles

  • Speed brake arming

  • Speed brake flight deployment (spoilers)

  • Speed brake deployment on flare & touch down (requires another Phidget card)

  • Trim wheel rotation/revolution when trim applied

  • Trim wheel indicator functional and moving when electric trim is activated from yoke

  • Park brake and light

  • Cut off Levers (fuel idle & cutoff)

  • Flaps

  • TO/GA button functional (to go around)

  • A/T disengage functional (auto throttle)

  • All IBL backlighting functional

The stab trim switches I have had wired in such a way to stop the trim wheels from spinning.  Although the spinning trim wheels are accurate to the real aircraft, they can be annoyingly noisy, especially at night when others are trying to sleep.  To disengage the trim wheel motor from the spinning trim wheels,  I flick the stab trim switch.  To activate the them again, I reverse the process.

The horn cut out switch is currently not connected to throttle functionality, however, can be allocated to another flight simulator function if required.

Fire Suppression Panel (FSP)

A communication error with my friend, who was converting this panel to flight simulator use, means a little more work is required to add flight simulator functionality.  At the moment I have power running to the handles causing the lamps to be lit all the time, and some of the module buttons to be back lit.  To my knowledge, the handles should only light when the backlighting is switched on or when they are activated.  I still have the original Boeing circuit boards and solenoid switches, and although I haven't given the matter a lot of thought, I believe that it should be possible to connect a Phidget 004 card, which has relays, to allow activation of APU and fire handles via the original solenoid switches.  I'm not quite sure on how to activate the buttons and switches - perhaps FSUIPC offsets and phidget software.  Rome wasn't built in day, so more on this later.

B737-300 Throttle Quadrant & Center Pedestal - Arrived at Last

/

A big orange truck from TNT Express parked outside the house this afternoon and the driver began to offload a large wooden crate that weighed around 80 kilograms.  I could be only one thing – the Boeing throttle quadrant and avionics box (center pedistal) had finally arrived.   

Together, the driver and I manhandled the crate through the hallway of the house to the room in which construction of the simulator is taking place.  Removing a heavy piece of machinery from a wooden crate can be tricky, and the only method was to disassemble the box screw by screw – WOW what beauty!

Initial Thoughts

The throttle and avionics bay is a genuine aircraft part so there wasn’t much to not like; you can’t “immerse” yourself or get a more authentic experience than by using a real aircraft part.  The throttle originally was in use in a Boeing 737-300 with South West livery.  Unfortunately, the guy at the tear down yard didn’t document the tail number of the aircraft it was removed from.  It would have been nice to have a photo of the actual aircraft to place on the Blog.

The first aspect I noticed about the throttle was the build.  It’s a solid piece of engineering built to withstand the neglect of pilot use and now simulator use.  I don’t believe the throttle will ever be damaged from neglect my end – its’ solidly constructed.  The feel when you push the two power levers forward is - well – you just have to be here!  Manoeuvring the flap lever through the various indents is equally rewarding.  Knowing that the throttle was once used in a real aircraft by real pilots adds a completely new dimension to flight simulation.

Retrofitting and Connectivity

During the refurbishment of the throttle, I had decided to not bastardize the throttle to try and replicate the appearance a throttle from a Boeing NG.  Therefore, the throttle remains a 300 series throttle.  It has been repainted only where necessary and decals have been replaced only when they were unreadable.  The internal mechanism of the throttle has been completely striped, cleaned and serviced.  Parts, such as the huge cog wheels and unnecessary internal wiring have been discarded as these are not required for simulation use. 

To allow the throttle to connect correctly with flight simulator, three Phidget cards (0066 & 0064) & a Leo Bodnar card (BUO 836X) have been used.  The cards are connected directly to the front of the throttle casing and will not be visible once the throttle casing is connected to the centre stage of the main instrument panel (MIP). 

All the functions of the throttle operate with the exception of the stab trim switches, which can be linked to another FS function if required.  Trim wheels are functional with the use of a servo motor and the trim spins when electric trim is activated on the yoke.  Back lighting is integrated back lighting (IBL) using genuine Boeing 5 volt bulbs.

oem 737-300 throttle quadrant - initial thoughts: it’s built like a thunderbox

Current Status

At the moment I’ve only taken delivery and am in the process of connecting a Benchmark card to an external power source to allow power to reach the 5 volt lighting bulbs and servo motors.  I have little doubt that there will be teething issues with software as I configure everything for correct functionality, but I believe that this extra effort is worthwhile to be able to use a real throttle instead of a replica.

Center Pedestal

The avionics bay is a two-bay type.  Two-bay types were mainly used on the earlier Boeing classic series jets up to the 200 series, however, a number of 300 series aircraft used them as well as 400 series.  The bay was attached to the throttle when I bought it, so rather than dump it and replicate a NG three-bay; I’ve decided to use it to maintain authenticity.  I may at some stage in the future replace it with three-bay – I’ll see how things develop once I begin to populate the bay with avionics instruments.  One benefit of using a two-bay style is that once Weber seats are fitted to the flight deck there will be more room to squeeze past to get into the seat!

An interesting feature to the unit is the positioning of two oddly shaped aluminium pull downs.  At first, I had no idea what these were used for.  Then it dawned on me – they are retractable coffee cup holders.  What more can you ask for (laughing). 

oem 737-300 fire suppression panel

Fire Suppression Panel (FSP)

The fire suppression module was an afterthought.  A second hand unit was available and I decided to retrofit this with limited functionality to flight simulator.  At the moment IBL works, and when pulled, each fire handle does what it’s supposed to do.  At some stage in the future I may activate the fire bell.  But, at the moment it’s early days with regard to this.  Basically it’s a module that has to be installed into the avionics bay for aesthetics; a TQ without a fire suppression module looks a slightly naked.

More on the actual avionics bay at a later stage when I begin to populate the bay with instruments - much kmore interesting than looking at "naked bay"