Attn Col. Bob Scott re the Milviz T-38

[ramprat 8]

Hi Bob,I heard you were a T-38 instructor during part of your tenure as an USAF pilot. I would love to hear your opinion of the Milviz T-38A as I have found you to be a reliable source of accurate information on this forum.Thank you in advance,Bill in Colorado

[Bob Scott 7,314]

Hi Bob,I heard you were a T-38 instructor during part of your tenure as an USAF pilot. I would love to hear your opinion of the Milviz T-38A as I have found you to be a reliable source of accurate information on this forum.Thank you in advance,Bill in Colorado

Yep, I've got somewhere around a thousand hours logged in the "White Rocket." Since I upgraded to a 30-inch monitor on the yoke side of my flight deck, the 24" monitor on the stick-n-rudder side is unplugged (used to have two 1920 x 1200 monitors in clone mode on left/right sides)...still have to figure out how to re-rig things with monitors of different resolutions before I can try out the T-38.I have high hopes with this one...I think the old FSD version from FS2002 was the last attempt at this plane, and it didn't even get close.

[dhenriques 7]

Yep, I've got somewhere around a thousand hours logged in the "White Rocket." Since I upgraded to a 30-inch monitor on the yoke side of my flight deck, the 24" monitor on the stick-n-rudder side is unplugged (used to have two 1920 x 1200 monitors in clone mode on left/right sides)...still have to figure out how to re-rig things with monitors of different resolutions before I can try out the T-38.I have high hopes with this one...I think the old FSD version from FS2002 was the last attempt at this plane, and it didn't even get close.

Bob;If and when you get the Milviz 38 up and flying, I'll be interested in your opinion on it. You can back channel me regular email if you like. It's my full name at gmail .comDudley Henriques

[Bob Scott 7,314]

Bob;If and when you get the Milviz 38 up and flying, I'll be interested in your opinion on it. You can back channel me regular email if you like. It's my full name at gmail .comDudley Henriques

Wilco...probably sometime in the next couple of days.Cheers

[Bob Scott 7,314]

Here are the comments I just posted in the (users only) Milviz forums on their new T-38 after about 4 hours of test flying and tweaking.They're off to a pretty good start with this one. There are some minor issues that need addressed, but I think it's worth $40, assuming the non-user-addressable issues do get fixed.=============================================================================Gents; I'm a retired USAF pilot with ~ 1,000 hours of IP time in the T-38, and a freeware flight sim developer with a fair amount of experience with gauge code and FDE. Here are some of my critique items on the Milviz T-38, and associated suggestions.1. Crossfeed is inoperative...switch configuration does not change fuel feed behavior at all. Also, "Crossfeed" is mis-spelled on the panel as "Crosspeed" Suggestion: needs panel programming and bitmap fix.2. Master caution light cannot be pressed to extinguish/re-arm the light. It does not show a clickspot pointer anywhere over the caution light. Needs panel gauge programming fix.3. Throttle/RPM response from low power settings (i.e. below 80%) is FAR too sluggish. Suggestion: Needs FDE fix.4. Flight control hydraulic system pressure drops out of limits (to ~2500 PSI) with throttles at idle. Suggestion: needs panel programming fix--both hyd systems should hold 3000 PSI at all RPM settings from idle to MAX A/B.5. Gear warning horn - sounds after gear retraction with power at MIL during touch and go landings. Horn should not sound with either engine RPM at/above 96% per your own docs...suspect the gauge is looking at raw throttle position or RPM and not indicated RPM as indicated on the gauge.6. Rudder overcontrol--the T-38 is equipped with a rudder travel limiter that limits deflection of the rudder to 7 deg with the gear up. Application of full rudder with a clean config in this sim demonstrates massively excessive rudder authority. Suggestion: this can be implemented in the panel gauges by intercepting the rudder axis input and limiting what is actually input to the sim, especially easy in FSUIPC. Barring that, if you have to choose between full rudder or limited rudder full-time without implementing the limiter, I'd recommend going with limited travel, as most people will overcontrol the rudder. I've implemented a compromise fix here by setting the rudder effectiveness to 0.257. Insufficient stabilator authority. The T-38 is equipped with a flap-stabilator interconnect that changes stab angle when the flaps are moved to prevent excessive pitch changes and adjust travel limits. I've found myself reaching the mechanical limits of stick travel at times, for example during landing flares. Suggestion: a quick compromise to fix this is to increase the elevator effectiveness to somewhere between 1.5 - 2.08. AoA index is not properly calibrated at 60% flap setting. Calibration at the 60% flap setting is important, as it is used for single engine approaches. I found that the gauge is not well-calibrated at this setting. Flaps at 60% produce nearly the same amount of lift (but much less drag) as flaps at 100%--in fact the approach speeds in both configs are the same. The problem here is that FS does not scale the lift properly at the intermediate setting. Suggestion: I added a dummy flaps.1 section that corrects the lift deficiency at 60% flaps and produces appropriate AoA readings:Code:[flaps.1]type = 1span-outboard = 0.5extending-time = 9.0flaps-position.0 = 0flaps-position.1 = 44flaps-position.2 = 0maneuvering_flaps=0lift_scalar = 0.409drag_scalar = 0.000001pitch_scalar = 1.09. Insufficient brake effectiveness. The T-38's brakes can hold the aircraft with power set at MIL--in fact formation takeoffs are done by having both acft set MIL power with brakes set, and lead then visually commands brake release from there. The toe_brakes_scalar is set to 0.5 in the aircraft.cfg, which is way too low. Suggest a range from 1.0 - 1.2510. Supersonic flight characteristics. Needs some work here. The T-38 cannot accelerate through mach in level flight at MIL power, but this model does. Suggestion: this can be fixed by increasing mach drag in FDE. Also, as the T-38 accelerates through the transonic regime at ~M0.98 into supersonic flight, the shock wave on the wing causes partial aerodynamic blanking of the stab, resulting in greatly reduced stab authority. This can be simulated much the same as the rudder limiter, with elevator inputs captured and scaled down before being applied to the sim while the acft is in trans/supersonic flight.Last, as the T-38 passes through M1.0, the shock wave attaching to the pitot boom causes the altimeter indication to drop ~300-400 feet in indicated altitude (momentarily...over maybe 1-2 sec). This can easily be implemented in panel gauge code.11. Afterburner nozzle indications. Any setting of A/B in the model results in a full open nozzle. The nozzle should vary from ~80% to 100% open (at MAX) while thrust is in the A/B range.12. Afterburner sound effects. Too loud, needs to be dialed down to about half its present volume.13. Flap motor sound effects. The T-38 uses electric flap motors that make a barely perceptible whir. The hydraulic motor effects are the wrong sound and too loud as well.14. Ignitor sounds at engine start. Again, these would be barely perceptible, if at all. Waaaaaaaaay too loud.15. Afterburner detent. I highly recommend adding a distinctive click when reaching a throttle setting at the A/B threshold. Most users will not have a joystick throttle with a mechanical detent...this would help setting MIL power without constantly overshooting into A/B range.16. AoA rumble and wing rock--works OK configured in the pattern, but I find myself getting rumble and wing rock at inappropriate times, such as in the break or in a closed pullup. The AoA readings that the effect is based on appear excessive in clean config with speeds 240KIAS+. A couple comments on your quick-start docs--you fail to mention setting flaps to 60% after engine start in the quick-start, so the eager new guy following those instructions will be attempting a no-flap takeoff first time out...not 'zackly setting him up for success there. Also, pressing SHFT-3 brings up the GPS/Autopilot panel...it should be SHFT-2 to connect the air cart.

[Milviz 919]

I appreciate these, They will deffo be implemented (along with some others) in the planned service pack. 2 weeks.Thanks again.

[CoolP 283]

Wow, nice co-work there, gentlemen. Impressive, and a nice read too.

I'm a retired USAF pilot with ~ 1,000 hours of IP time in the T-38, and a freeware flight sim developer with a fair amount of experience with gauge code and FDE.

Where can we find some of your stuff?

[Bob Scott 7,314]

Wow, nice co-work there, gentlemen. Impressive, and a nice read too.Where can we find some of your stuff?

Right here in the AVSIM library. Do a library search using my name.

[CoolP 283]

TinMouse II Boeing 737-200 and the fuel gauge for the Mustang, that's nice work, Bob! I've been using this stuff but didn't watch the names closely it seems.Didn't know about the JELAIR - Maersk Air, Boeing 737-2L9 though. Worth a look. :(

[Guest bstolle]

6. Rudder overcontrol--the T-38 is equipped with a rudder travel limiter that limits deflection of the rudder to 7 deg with the gear up. Application of full rudder with a clean config in this sim demonstrates massively excessive rudder authority. Suggestion: this can be implemented in the panel gauges by intercepting the rudder axis input and limiting what is actually input to the sim, especially easy in FSUIPC. Barring that, if you have to choose between full rudder or limited rudder full-time without implementing the limiter, I'd recommend going with limited travel, as most people will overcontrol the rudder. I've implemented a compromise fix here by setting the rudder effectiveness to 0.257. Insufficient stabilator authority. The T-38 is equipped with a flap-stabilator interconnect that changes stab angle when the flaps are moved to prevent excessive pitch changes and adjust travel limits. I've found myself reaching the mechanical limits of stick travel at times, for example during landing flares. Suggestion: a quick compromise to fix this is to increase the elevator effectiveness to somewhere between 1.5 - 2.09. Insufficient brake effectiveness. The T-38's brakes can hold the aircraft with power set at MIL--in fact formation takeoffs are done by having both acft set MIL power with brakes set, and lead then visually commands brake release from there. The toe_brakes_scalar is set to 0.5 in the aircraft.cfg, which is way too low. Suggest a range from 1.0 - 1.25The T-38 cannot accelerate through mach in level flight at MIL power, but this model does. Suggestion: this can be fixed by increasing mach drag in FDE. Last, as the T-38 passes through M1.0, the shock wave attaching to the pitot boom causes the altimeter indication to drop ~300-400 feet in indicated altitude (momentarily...over maybe 1-2 sec). This can easily be implemented in panel gauge code.

The FD design took half a year in very close cooperation with two T-38 pilots, so I do think that making changes after only 4hrs of flying is a bit early. Especially #7. Doubling the elevator authority???? According to all pilots in our team it's almost too easy to pull into heavy buffet and into a stall. #6. According to our T-38 pilots you don't need rudder in the T-38 but once you get into the higher alpha range you wouldn't have any rudder authority at all by using limited travel. The high alpha rudder roll needs to be included as this is in the training syllabus for T-38 pilots. Furthermore many people (including me) aren't using FSUIPC so this is not an option.#9. The manual says that you might need to pump the brakes to achieve mil thrust, this is not possible in FSX but if you increase the brake effectiveness to the value you mentioned, the landing run will be way too shortI agree with the supersonic drag problem. This problem occured with the last version of the FDs and has been corrected already. (BTW, mach drag wasn't the culprit)Concering the altimeter dip for 1-2 sec...this is a 35USD sim and not a multi million dollar sim. The question is where to stop with the development as you could go on for ages.

[WarpD 824]

You're a developer? Without a 'gold tag'?

[Guest bstolle]

You're a developer? Without a 'gold tag'?

Freelance FD design, but more free than payware.

[N1G 88]

Bob,That was quite impressive, and I and many I am sure that are using the talon will agree a big thank you is in order Sir!! I know Milviz is very dedicated and sure they appreciate your input. I just purchase the talon and am really enjoying this bird but I will enjoy it more knowing that Milviz will implement what is possible of course within the limitations of FSX. Aso your gauge programming experience is truly a plus. How many t-38 pilots actually use FSX? I am sure it is rare so we are very luck here as many are really enjoying this bird. Appreciate your input and time and looking forward to your further reviews as time moves forward. I have really enjoyed flying this but boy is she a hot one!RegardsBob

Here are the comments I just posted in the (users only) Milviz forums on their new T-38 after about 4 hours of test flying and tweaking.They're off to a pretty good start with this one. There are some minor issues that need addressed, but I think it's worth $40, assuming the non-user-addressable issues do get fixed.=============================================================================Gents; I'm a retired USAF pilot with ~ 1,000 hours of IP time in the T-38, and a freeware flight sim developer with a fair amount of experience with gauge code and FDE. Here are some of my critique items on the Milviz T-38, and associated suggestions.1. Crossfeed is inoperative...switch configuration does not change fuel feed behavior at all. Also, "Crossfeed" is mis-spelled on the panel as "Crosspeed" Suggestion: needs panel programming and bitmap fix.2. Master caution light cannot be pressed to extinguish/re-arm the light. It does not show a clickspot pointer anywhere over the caution light. Needs panel gauge programming fix.3. Throttle/RPM response from low power settings (i.e. below 80%) is FAR too sluggish. Suggestion: Needs FDE fix.4. Flight control hydraulic system pressure drops out of limits (to ~2500 PSI) with throttles at idle. Suggestion: needs panel programming fix--both hyd systems should hold 3000 PSI at all RPM settings from idle to MAX A/B.5. Gear warning horn - sounds after gear retraction with power at MIL during touch and go landings. Horn should not sound with either engine RPM at/above 96% per your own docs...suspect the gauge is looking at raw throttle position or RPM and not indicated RPM as indicated on the gauge.6. Rudder overcontrol--the T-38 is equipped with a rudder travel limiter that limits deflection of the rudder to 7 deg with the gear up. Application of full rudder with a clean config in this sim demonstrates massively excessive rudder authority. Suggestion: this can be implemented in the panel gauges by intercepting the rudder axis input and limiting what is actually input to the sim, especially easy in FSUIPC. Barring that, if you have to choose between full rudder or limited rudder full-time without implementing the limiter, I'd recommend going with limited travel, as most people will overcontrol the rudder. I've implemented a compromise fix here by setting the rudder effectiveness to 0.257. Insufficient stabilator authority. The T-38 is equipped with a flap-stabilator interconnect that changes stab angle when the flaps are moved to prevent excessive pitch changes and adjust travel limits. I've found myself reaching the mechanical limits of stick travel at times, for example during landing flares. Suggestion: a quick compromise to fix this is to increase the elevator effectiveness to somewhere between 1.5 - 2.08. AoA index is not properly calibrated at 60% flap setting. Calibration at the 60% flap setting is important, as it is used for single engine approaches. I found that the gauge is not well-calibrated at this setting. Flaps at 60% produce nearly the same amount of lift (but much less drag) as flaps at 100%--in fact the approach speeds in both configs are the same. The problem here is that FS does not scale the lift properly at the intermediate setting. Suggestion: I added a dummy flaps.1 section that corrects the lift deficiency at 60% flaps and produces appropriate AoA readings:Code:[flaps.1]type = 1span-outboard = 0.5extending-time = 9.0flaps-position.0 = 0flaps-position.1 = 44flaps-position.2 = 0maneuvering_flaps=0lift_scalar = 0.409drag_scalar = 0.000001pitch_scalar = 1.09. Insufficient brake effectiveness. The T-38's brakes can hold the aircraft with power set at MIL--in fact formation takeoffs are done by having both acft set MIL power with brakes set, and lead then visually commands brake release from there. The toe_brakes_scalar is set to 0.5 in the aircraft.cfg, which is way too low. Suggest a range from 1.0 - 1.2510. Supersonic flight characteristics. Needs some work here. The T-38 cannot accelerate through mach in level flight at MIL power, but this model does. Suggestion: this can be fixed by increasing mach drag in FDE. Also, as the T-38 accelerates through the transonic regime at ~M0.98 into supersonic flight, the shock wave on the wing causes partial aerodynamic blanking of the stab, resulting in greatly reduced stab authority. This can be simulated much the same as the rudder limiter, with elevator inputs captured and scaled down before being applied to the sim while the acft is in trans/supersonic flight.Last, as the T-38 passes through M1.0, the shock wave attaching to the pitot boom causes the altimeter indication to drop ~300-400 feet in indicated altitude (momentarily...over maybe 1-2 sec). This can easily be implemented in panel gauge code.11. Afterburner nozzle indications. Any setting of A/B in the model results in a full open nozzle. The nozzle should vary from ~80% to 100% open (at MAX) while thrust is in the A/B range.12. Afterburner sound effects. Too loud, needs to be dialed down to about half its present volume.13. Flap motor sound effects. The T-38 uses electric flap motors that make a barely perceptible whir. The hydraulic motor effects are the wrong sound and too loud as well.14. Ignitor sounds at engine start. Again, these would be barely perceptible, if at all. Waaaaaaaaay too loud.15. Afterburner detent. I highly recommend adding a distinctive click when reaching a throttle setting at the A/B threshold. Most users will not have a joystick throttle with a mechanical detent...this would help setting MIL power without constantly overshooting into A/B range.16. AoA rumble and wing rock--works OK configured in the pattern, but I find myself getting rumble and wing rock at inappropriate times, such as in the break or in a closed pullup. The AoA readings that the effect is based on appear excessive in clean config with speeds 240KIAS+. A couple comments on your quick-start docs--you fail to mention setting flaps to 60% after engine start in the quick-start, so the eager new guy following those instructions will be attempting a no-flap takeoff first time out...not 'zackly setting him up for success there. Also, pressing SHFT-3 brings up the GPS/Autopilot panel...it should be SHFT-2 to connect the air cart.

[srburger 0]

Bob,<<Here are the comments I just posted in the (users only) Milviz forums on their new T-38 after about 4 hours of test flying and tweaking.>>Fascinating read, and much appreciated! Thanks for taking the time to do this!Best,Ken

[Bob Scott 7,314]

The FD design took half a year in very close cooperation with two T-38 pilots, so I do think that making changes after only 4hrs of flying is a bit early. Especially #7. Doubling the elevator authority???? According to all pilots in our team it's almost too easy to pull into heavy buffet and into a stall.

That was a quick-and-dirty fix because I was seeing heavy stick forces and actually hitting the stops during the flare. It is indeed too easy to pull into heavy buffet and wing rock, but I suspect that has more to do with the CL vs AOA curve in the FDE. The AoA goes up far too fast as you slow down...flying final at just 5 KIAS slow produces a red chevron and AoA at/above 0.8...that's too sensitive. At liftoff I'm rotating right into that rumble...again, never ever saw that...believe that the lift curves need some work and some recalibration of the AoA and associated effects is also necessary.

#6. According to our T-38 pilots you don't need rudder in the T-38 but once you get into the higher alpha range you wouldn't have any rudder authority at all by using limited travel. The high alpha rudder roll needs to be included as this is in the training syllabus for T-38 pilots. Furthermore many people (including me) aren't using FSUIPC so this is not an option.

I used rudder all the time--in extended trail, for example, it's a very effective way of turning the jet when you're loaded up to 4-6 G. Anyway, with rudder effectiveness dialed down to 0.25 I still have LOTS of rudder authority at high AoA.

#9. The manual says that you might need to pump the brakes to achieve mil thrust, this is not possible in FSX but if you increase the brake effectiveness to the value you mentioned, the landing run will be way too short

If you aerobrake to ~100KIAS and then use the wheel brakes, the landing roll with braking effectiveness at 1.0 looks pretty good to me. If you apply wheel brakes much over 120KIAS in the real jet, you're asking for a blown tire anyway...so braking effectiveness is kind of a moot point--a better approach would be to simulate the blown tire. But overpowering locked brakes at just 80-90% RPM is a much bigger break with reality IMHO.

Concering the altimeter dip for 1-2 sec...this is a 35USD sim and not a multi million dollar sim. The question is where to stop with the development as you could go on for ages.

C'mon, that one's trivially easy to do...that's a five-minute coding job in the altimeter gauge. It's a unique characteristic of the T-38 that tens of thousands of AF pilots who first broke the sound barrier in this jet observed, and one of those cool details that any guy who's been on a boom ride might see and go "wow."Besides, I think I read somewhere "In a world of compromise, we don't." :(