Bendix Air Data Computer - Part 2: Master Ken Explains How It Works

Master Ken explains the amazing jet fighter analog computer made out of gears, that we opened up in this episode: • Bendix Air Data Comput...
Bendix MG-1 Restoration Playlist: • Bendix Central Air Dat...
Ken's in-depth blog article: www.righto.com/...
Plenty of aerospace videos, including the ones about a later version of this computer on the "Le labo de Michel" channel:
• LDM #302: Bendix Centr...
Stuff that supports the channel:
Amazon links for the tools I use in the lab (supports the channel if you buy from there)
www.curiousmar...
Channel merch on Teespring: teespring.com/...
Support us on Patreon: / curiousmarc
Our lovely sponsors
PCBWay: fast turn PCBs, www.pcbway.com
Electro-Rent: www.electroren...
Keysight: test instruments: www.keysight.com
Samtec: connectors: www.samtec.com
R&D Microwaves: rdmicrowaves.com
Get more technical info on the companion site: www.curiousmar...
Contact info: www.youtube.co...

[@CuriousMarc]

Sorry had to reupload due to some rendering bug in the middle of the video. Also corrected my tongue slip of True Ground Speed instead of True Air Speed. No way to re-upload cleanly on KZbin unfortunately, 15,000 views get erased and previous comments disappear in the internet black hole. So please re-post your sassy comment!

[@erkinalp]

Well, you could have posted a correction notice.

[@fgaviator]

15,000 views do not get lost: we all just re-watch the re-upload... 😀

[@vicmac3513]

KZbin behaved very weirdly because I was right at the rendering bug when you deleted the video.

[@reneschmitz4845]

Will rewatch, re-like....

[@bengelman2600]

I watched it again for that reason.

[@akersmc]

Pilot here, in English "pitot tube" is pronounced the way you have been pronouncing it. France played a significant role in the history and development of aviation so there are a bunch of french words in aviation. Aileron, empennage, fuselage, monocoque, canard, longeron, avionics, to name a few.

[@8BitNaptime]

mayday...

[@akersmc]

@@8BitNaptime m'aidez

[@Damien.D]

@@akersmc and the least severe "panne panne".

[@MarcelHuguenin]

@@akersmc hmm… never thought about that 👍🏻

[@patrickshaw8595]

Nacelle. "See that Grumman Bearcat? It looks like they strapped guns wings and a pilot on a DC-6 engine nacelle !"

[@zh84]

You call it "backed up since Christmas with editing videos": I call it "having four months' worth of videos in the pipeline to look forward to"!

[@ErinJayEldridge]

A note on indicated airspeed: even though it's not the true airspeed, it's surprisingly useful because the lift from the wing is effected by many of the same factors. For a light aircraft staying below about 5km, you almost always are flying to a indicated airspeed, and only want to know true airspeed to guess at your groundspeed (which nowadays you mostly measure more-directly anyway).

[@kh40yr]

"Antikythera Mechanism V2.0, I wonder when they dig this up 2000 years from now, they will be just as amazed. Love the miniature ball bearings on each countershaft.

[@yyzkevin416]

7:18 "Don't turn the screws!"

[@kennethbeal]

Thank you! Nice name. :) Great ending! Showed me why they're called "synchros" -- they synchronize two devices that are at a distance. Cartoons growing up, "synchronize your watches!" :)

[@KeritechElectronics]

Insane complexity indeed. Almost beats the Antikythera mechanism, haha!

[@BGTech1]

Incredible piece of engineering and craftsmanship! Thanks for this.

[@gryffuscze]

Strangely enough, what amazed me the most and still gives me hard time to accept is seeing Master Ken standing next to a computer full of gears and SPRINGS..... WITHOUT GLASSES 0:40 😮😮🤣 Cheers and thanks to you guys, you are the best.

[@Zadster]

I imagine most of the mechanical hardware here was probably developed around the time of WW2, way back in the 1940s, for bomb-aiming computers etc. Mindboggling!

[@FrozenHaxor]

What amazes me most is that a few transducers and a calculator grade microprocessor would replace all this nowadays, amazing effort on the manufacturer's part!

[@sashimanu]

More like three microprocessors running overcomplicated code in lockstep with outputs going into a voting comparator for redundancy and fail-safety

[@guidoscalise]

@@sashimanu can you define what “overcomplicated” means?

[@FrozenHaxor]

@@sashimanu This old school mechanical computer has none of that.

[@lwilton]

@@guidoscalise It's probably written in ADA, requires at least a 32 bit CPU with at least 8 MB of ram, and I'd be real surprised if the source code was under 100K lines. I would not in the slightest be surprised if the source code was over 400K lines.

[@mfree80286]

@@FrozenHaxor And, it's realtime. Unless you devote a LOT of time into developing an RTOS or dedicating cores to a given task, you'll have some possibly significant lag in a coded solution.

[@awesome3165]

Haha Ken. "Don't turn the screws!!!"

[@tomteiter7192]

Totally off topic, but seeing the little tektronix 222 sitting on top of the instrument stack sent me right to eBay again... It's SO cute! I want one!

[@RoryMacdonald-pfff]

I’m curious what the servicing of such a system would be like - particularly the lubrication of the various pivots/axles. Fascinating unit - the use of cams and gears to conduct calculations and display results … this is the kind of stuff that would have brought maths at school to life.

[@dziban303]

This video was in my recommendations. Ken Shirriff is an amazing guy, his is one of the accounts I miss the most from Twitter.

[@dbenson3114]

6:53 this is my favorite part of the machine

[@thesteelrodent1796]

rewatching the reup to get that viewcount (re)up 😉

[@whiskeytuesday]

I mean who wouldn't want a chance to watch this video for the first time for a second time. The masters at work!

[@Allan_aka_RocKITEman]

_"At last we meet for the first time for the last time!"_ -- SPACEBALLS [1987] 😉

[@dirty364]

Blows my mind 🤯 how intelligent some people are! What a piece of art! Id like to think I’m somewhat tech savvy and mechanically inclined but when I see what people were building in the 50s and watch stuff like this I feel dumb lol

[@HiwasseeRiver]

Aww - Takes me back to my class in LaPlace Transforms - in LaPlace space the transfer functions look very simple (inputs and outputs). Side story - back in the pre computer days I remember a flow rate display in an oil refinery control room - the only catch was it was the square of the real flow rate. You had to take the square root (in your head or look it up) to get the real flow rate - but the operators worked with the display square - crazy

[@craigcooknf]

Absolutely fascinating. The mechanical computers in aerospace, yesterday and today are amazing!

[@SkyhawkSteve]

Any chance of a video to explain magnetic amplifiers? I've got vague recollections that the core saturation was adjusted to change the coupling of signal from primary to secondary... but that could be very wrong. For applications that required robust parts, a magnetic amp seems like a clear winner over vacuum tubes.

[@hydranmenace]

My wife wanted to know what we should watch before bed. She was not expecting this.

[@AndyMarsh]

Such a cool machine! Company I used to work for once made a Syncro Trainer to teach studends how they worked. It was before my time and I always wondered what syncros were used for, now I know!

[@jnbfrancisco]

I used a synchro trainer when I was an instructor in the USAF at Chanute AFB Illinois in 1974 to 1977.

[@waynestock300]

Much better. Fascinating as always and master Ken never disappoints! ❤

[@mikefromwa]

A truly incredible electromechanical device. The amount of work that went into creating it must have been staggering.

[@selimkucukates30]

Thank you very much for this wonderful work you have done.

[@ropersonline]

Commenting on the connectors for a 2nd time allows me to go into greater detail: Turns out the innards sport a different set of connectors (male DD-50, DC-37 & DA-15 D-subs, see 0:39) than the faceplate, which has round connectors (see 0:13). So there must be a set of matching and perfectly aligned female D-sub connectors on the inside of the faceplate, an example of which arrangement in the form of a DC-37 (not the one on the faceplate but further down on the side) was seen in part 1 at 2m26s. Interesting that they used internal connectors at all, prolly for ease of assembly & maintenance. Pls show us the inner faceplate.

[@SubTroppo]

My comment in the original was about Bendix's change from mainly electro-mechanical to electronic in this form of instrumentation. When was that?

[@CuriousMarc]

It’s very progressive. Some part of the calculation gets replaced by an op amp or two, while the rest stays on gears and synchros. eventually everything gets the analog electronic treatment, then the digital enters. Michel (from the channel “Le labo de Michel”) has a later model of this where you can see that happening.

[@CuriousMarc]

I found Michel's video: kzbin.info/www/bejne/mKbSf4asa51sqJo . One of the next revisions of this, where you start to see a mix of op-amp analog electronics and electro-mechanics.

[@VladimirTolskiy]

How would a typical cam be connected to a differential in this computer? Cams produce a sweeping motion and differentials work with full revolutions.

[@Patrick_B687-3]

Time and gain this channel reminds me how smart I’m not. Ken and Mark are just incredible.

[@johnkaufmann4711]

That is sooooo cool . Thanks .

[@soniclab-cnc]

Master Ken at it again deciphering the impossible.

[@frankwales]

[Sassy comment re-post, as demanded :-P ] As an old software guy, I'd love to know how this kind of gear (sic) was debugged and verified.

[@TheIceGryphon]

It’s amazing to see part of one of my favorite planes. F-86 Sabre was amazing aircraft.

[@DanielGBenesScienceShows]

Very Large Scale Integration microchips don’t blow my mind (they should but they just don’t). THIS, however… THIS completely blows my struggling, broken, little, feeble mind.

[@deanmaclanders6115]

Can confirm...As someone who flew small aircraft, and spent several years working in a wind tunnel, in english Canada, it was always pronounced peet-oh

[@benjaminhanke79]

I hope someday Marc and Master Ken get their hands on a fluidic computer, a pneumatic logic device without moving parts.

[@mattikaki]

I have an old mechanical plane altimeter and it is astonishing how accurate it is. If I move it up and down in my hand it shows the difference.

[@andrewkieran8942]

“Any sufficiently advanced technology is indistinguishable from magic” -- Arthur C. Clarke

[@senilyDeluxe]

Watching Marc's channel, I use to say "Any sufficiently antiquated technology is indistinguishable from magic" - case in point with 1960s frequency counters that go up to over a Gigahertz even though the electronics couldn't handle more than a few dozen Megahertz - and went to using fancy math to do the 100MHz and GHz range. Or sensing Microamps by chopping the signal with a wild neon light opto coupler circuit.

[@michaelbrodsky]

Except that this belongs in a museum.

[@abrunosON]

Any sufficiently old technology is indistinguishable from aliens.

[@luthmhor]

Pressure is just energy per unit volume. The dynamic pressure component (kinetic energy) doesn’t actually exert pressure on the pipe walls in a flowing fluid, so it isn’t measured with regular pressure gauges. To measure it we use the pitot tube which takes and stops the fluid, thereby converting that kinetic energy to potential energy where it does exert a force and so can be measured. The easiest way is as shown in your diagram, just the difference in pressure head. I used to be a pipeline engineer, liquids only. Compressible fluids like air are much more problematic to constrain with assumptions like engineers love to do 😆.

[@chrisnizer5702]

Thanks for showing elevator music the proper love and respect it deserves. The air data computer is pretty cool too by the way! Thanks for another great vintage technology video my friends, Semper Fidelis.

[@frederickwise5238]

Watching this makes me feel even worse than I did back in 1962. I had no idea how complex an ADC was. Fresh out of the Navy, I went to work for NAA on the A3J Vigilante. One of my earlier tasks was installing an ADC. Not sure why my lead man for the training period didnt catch it but I hooked up two of the "air inputs" backwards and of course it wrecked the computer. I wasnt penalized but I felt really bad and stupid for all of the work I caused for the instrumentation repair lab and the expense to the company. I never made another mistake. Made "D" sure I knew what I was doing from then on. (Somebody else on 3rd shift moved the gear up lever when the airframe was powered up for some tests, the gear came up and the plane sat down.)

[@MegaGman61]

Amazing machine!

[@ptonpc]

Amazing what can be done with clever engineering.

[@henryD9363]

Minor note. At 3:23 the drawing implies that the air going past the static port is at velocity V. Actually, the static port is located at a place where the air velocity is close to zero. If you attach a little string at the static point you would see it's not moving in any particular direction. Just flopping around due to a bit of turbulence. The locations of a static port on an aircraft are chosen so there is essentially no airflow velocity at the surface. Hence, static port. If there was a velocity at that point, then the Bernoulli effect would cause a suction. In other words, a negative pressure relative to the static value.

[@mikebarushok5361]

Yes, exactly. Usually well past the beginning of the taper of the tail part of the fuselage. But also normally pairs plumbed together from opposite sides to reduce the effects present when crabbing or in significant yaw maneuvers.

[@martindowney7915]

Wow amazing engineering.

[@TeslaTales59]

Always worth a 2nd watch!

[@yuglesstube]

Excellent! Thanks. What year is this unit and how did you get it?

[@wildcat64100]

Fascinating to see this type of equipment that bridged the gap from analog to digital and were probably necessary to force electronic advancements as these electronic-mechanical hybrids became increasingly unwieldy and expensive.

[@jamesbronson8713]

Amazing engineering.

[@videoviewer2008]

Good to have some videos "in the can". Thanks.

[@MarcelHuguenin]

no worries, you get another complete view and comment. I think this is yet another piece of amazing technology. Nowadays this would probably be of an astronomical cost to develop and build such a device. So how was that done back in the day?

[@Shady97342]

It'd be neat if you could get your hands on an IBM AP-101. Not only was it in the Space Shuttle but also in a number of jets including the B1 Lancer! It runs a language I just learned about called JOVIAL.

[@StupidLilAdv]

Glad it's back! I was watching when it was taken down. "I was using that!"

[@ShainAndrews]

Really is an interesting solution.

[@38911bytefree]

This thing is a marvel, a piece of art.

[@valentinrusu8844]

Soo cooool! Mind blowing....

[@mixolydian2010]

Mindbogglingly wonderful, thanks a lot.

[@usvalve]

Why is it that this is so much more fascinating than a microcontroller and yards of program printout? Yeah, when I put it that way I think maybe I know :-D It does remind me of the Antikythera. Now I want somebody to do an animation showing each piece of it working!

[@repatch43]

Analog computers are immensely fascinating. I saw a military course from I think the 50s? that described the basic functional elements of analog computers, ending with a description of how all these components were put together to form a targeting computer.

[@n7275]

Very cool.

[@bigwinterboy]

Awsome content and technology

[@pickle5051]

How can I get one of these. I would love to hear and see this in action

[@johnnymnemonic69]

Next level tech even today

[@rael5469]

What's the difference between a syncro and a resolver? The Angle of Attack vane on an Airbus A320 has three.....resolvers? ....or is it syncros? The A320 has three air data / inertial reference units and three angle of attack vanes.....each with three resolvers.

[@CuriousMarc]

They are cousins, the difference is explained in part 3: kzbin.info/www/bejne/boLZoIxqlN11oNk

[@tomsherwood4650]

I am surprised that you cannot get tech data for that device anywhere, it is obviously obsolete for military purposes. After working on similar stuff on vintage autopilots, I have a lot of respect for anyone can go thru that thing and make it work to specs and certify it.

[@iamdarkyoshi]

Can't remember if I commented on the original or not. Very cool piece of equipment.

[@robertobryk4989]

How does the coarse and fine synchro setup work? I don't understand how can the coarse one be useful there at all. Fine synchro alone could work, as long as the value never changed quickly enough/we always had power when the axles were moving: the total number of its turns would agree between both sides under these circumstances. If we want the coarse synchro to help deal with fine synchro desynchronization, we need it to be able to overpower the fine synchro. But then I'd expect its error to also overpower the fine synchro and thus to make the fine synchro useless. What am I missing?

[@mfree80286]

It may be to widen the "response" of rate change that the synchros can transmit electrically. Probably more to do with the second derivative, the rate of change of the measured rate. The synchros geared together would always be locked, but the coarses may aid in the acceleration or braking of the fines to keep the total reflected output changing closer to reality. At least, that's what my mind's applying to the problem, right or wrong.

[@largepimping]

Just Ken doing Ken Things (tm). No big deal.

[@nateerb3114]

Said it before, and I'll say it again: I wish Master Ken was my Dad. Keep up the hard work of making vids, Marc. How else are we going to learn out here?

[@cpm1003]

It's always surprised me that they're able to get static air pressure from anywhere on a plane, when it's travelling so fast. It seems like there would always be some pressure or vacuum, and it would be bouncing around like crazy.

[@mikebarushok5361]

It's not quite as simple as described. Generally there are pairs of static ports, one on each side, plumbed together to get the average. And either during development many locations are tested starting with scale models in wind tunnels and also full scale testing on prototype(s) or, extensive computer modeling can determine a nearly optimal set of locations. In any case there are errors and then airframe specific corrections. None of this is necessary for one specific type of airborne use. Hot air balloon altimetry requires nothing more than the open port being in relatively undisturbed air.

[@glenncronise7775]

In the early 70s I worked on a printed circuit board that would do the same thing with opamps and comparators. Board was about 9X5 inches. No micros.

[@bogywankenobi3959]

I maintained the test software for the F-16 ADC at Hill AFB for a while back in the late 80s. It was nothing as interesting as your Bendix ADC. And frankly it could be improved by replacing all the electronics with an off-the-shelf Arduino Nano.

[@mx0r]

I was just watching the blank section. You were like 30 seconds late with re-upload. 😅

[@donmoore7785]

This shows that you guys are not afraid of diving into pretty much anything, except a pile of garbage, to figure out how it works AND how to get it working. That is an impressive piece of equipment. Just imagine people looking around for what could be improved with microelectronics once they were invented - I assume that a good portion of this is done in electronics these days.

[@simonturnill]

Love your videos Marc, have been a subscriber since you were at 15k, but can you change the elevator music? It gives me a 12 hour ear worm every time I hear it...!!!! Haha!

[@TechGorilla1987]

Master Ken is the only person that I can think of that would leave Kim Peek in silence and awe...

[@alexdehotot2712]

I've got to say, i've always heard the correct, french, pronunciation of Pitot all round the world from many aviation people!

[@rael5469]

Used in the F-86 fighter of the Korean war ?

[@AIM54A]

Interesting.. That picture at 5:04 shows a B-52 doing Mach 1.5. or about twice its actual top speed.

[@tfhhjh3372]

The reality is that an analog computer has advantages over digital ones, but the disadvantages must be constantly calibrated. And as the measurements are repeated with the same parameters, you will never have the same results since, being continuous, there will inevitably be variations.

[@timmoles9259]

I had to slow the speed to .75 just to absorb it all, not counting pause and understanding. Thank you

[@wouterke9871]

Follow the link to Ken's website, all and more is there written out with close-up pics 👌

[@Allan_aka_RocKITEman]

@CuriousMarc >>> Great video...👍

[@pauldzim]

Holy crap, this is like Flintstones technology, amazing that jet airplanes in the 50s and 60s didn't crash more often!

[@MoisesCaster]

2:54 pito in portuguese Brazil means p3nis

[@Consequator]

Imagine having to design this thing, and then having to actually fit all that into an enclosed box as small as possible.

[@Derpy1969]

Ken looks like a young Christopher Lee. He just needs the voice.

[@alvarobravo177]

Re-like!

[@frankhollein7093]

How in the heck do you work on something like that, without screwing it up?

[@benjaminhanke79]

That was only possible because it was soldered shut, no screws involved.

[@ndm13]

Forgive my incredibly uninformed question, but why 115VAC? Is there a specific reason for AC over DC generation, and at that high of a voltage?

[@markevans2294]

There's a document, MIL-STD-704, defining the standards used for aviation power by the USAF. 115/200 VAC 400Hz 3 phase, 28VDC & 270VDC. The latter apparently only used on military aircraft whilst the former two are also commonplace on civil aircraft.

[@marvintpandroid2213]

Re-engagement

[@RJGamer-zb4lb]

go to the weapon systems lol 1:14

[@624Dudley]

I’ve been wondering, Marc: are you an aviator yourself?

[@CuriousMarc]

Glider pilot, but I haven’t flown in years.

[@624Dudley]

@@CuriousMarc Roger. I’ve had a lifetime of being owned by airplanes, but recently the high cost has made me throw in the towel.

[@nmccw3245]

As an aviator, I can assure you the French pronunciation of aviation terminology is correct.

[@davidf2281]

I think I said this in the last instalment but holy heck the sheer cost of this thing at the time hardly bears thinking about. Must be a couple of houses at least.

[@Line-Ways]

Thats even more complex than my brain

[@gertebert]

Oh that obnoctious 110VAC 400Hz. I once owned an aircraft tube amp tripler containing a 2CV2797 If I remember that number correctly. It was sort of a 4CX250. I rebuild the input to 430 MHz and used it for over a decade to make 400 Watts CW at 70 cm. I aslo had to rebuild the power cuircuit to run on EU 220 VAC 50Hz.

[@NUTTER8291]

MY HEAD HURTS !!! 😂