totm june 2019 Custom hardware / simpit repository. For people who take KSP a little too far.

[Monsterlunch]

Just started a thread about my ongoing project!

 

 

[richfiles]

Oh! OH Nice!

I contacted the supplier of the futura stamp set, cause the numbers just DID NOT add up. At all... They said the stamps in the small set were 50x80x.... 370mm... 
That seems... not right... (for us weirdos using inches... that's 14.56 inches)!

So I contacted them, and they confirmed my suspicions... Somehow, their product page had an extra 0 added to the end of every dimension listed. Weird. 
They also shared the actual text dimensions of the small stamps as being about 3x4mm... In other words, these stamps are THE perfect size for me! 

Now, If only I could find a US seller. The product is from Korea, yet I've only found it sold in Australia!? The ebay stores don't even offer US shipping, and the only places I found selling it will charge me $20 for international shipping. Joy. I can certainly handle that, but it'll also cost me at LEAST $30-40 for the numbers alone, as each set only has ONE of each 0-9 digit, and wastes a ton of stamps on symbols and punctuation. As in there's MORE symbols than numbers... Ugh... What were they thinking! It's obvious they expect you to buy extra sets for doing numbers larger than one digit... But why so many duplicate symbols int he process! It's a waste! They should have had a symbol set, and number set with duplicate numbers. The letter set at least gives you two letters for the most common letters... 

To add insult to injury... No < or > symbols at all... I so wanted ">9000" (or at least ">9999") for my high altitude reading! 
I suppose I could buy the alpha set, and turn the "v" sideways.

So long as they don't bleed my wallet completely dry, I should be able to get a set of futura stamps. Wish I needed them for more than a SINGLE project... 

Truth be told... I need a MINIMUM of three number sets and one letter set... and that's if I cheat and do my proposed top meter reading as ">9999" by flipping a "6" upside-down and using the "v" sideways. I could skip the letter set altogether and use numbers only if I switch that to "+9999" (there is a "+" symbol). If I stop it at "+9000" or "+9900", then I'm down to needing only 3 number sets. Numbers like "1000", "2000", etc force me to buy 3 sets, no mater what, as I absolutely need three "0"s.

Even buying 3 sets, I pay nearly as much in shipping as I am on the sets!  It's hard sell, just to go with futura over whatever that font style on your average office number stamper is. The thing about that, I I can spend $10 on the basic office number stamper, or spend 5x that for the ABSOLUTE MINIMUM to do the numbers in futura. Up to 8x to have some variety, text, and make the stamping go easier... That's a REALLY tough sell... 

Edited June 10, 2016 by richfiles

[Pvt. KASA]

Hi there!

i just wanted to say hello and "present" my plans for my ksp cave:

I am sleeping in a raised bed with 1,5m space under it.

So I had the Idea to build some sort of ksp capsule below my bed.

i will the lie parallel to the floor with raised legs, the console above my head. (Just like the launch situation with Apollo, Gemini and Mercury.

I want to have it fun but also a little bit realistic.

It should be a mix of Mercury, Gemini and Apollo.

I think it's ok if it's primarily designed for only two ksp companies crafts:

one for space station docking and orbit operation and one for a Mun landing.

At this point I'm figuring out the positions of the switches, buttons, gauges and lights.

I also got some parts to try some things.

I want to be able to do a complete session/mission without leaving the capsule once so I think I'll need at least a small mfd.

I will control the inputs with a unknown number of arduinos.

(got at least 1 arduino micro for keyboard emulation experiments.)

Thats it for my first post, hope there weren't to much grammar mistakes, I'm only in 8th class so we had englisch for (only) 4 years yet.

It's written on my phone so the autocorrect helps me 

Thanks for Reading

KASA (name wasn't available anymore)

edit: Two more Informations:

There should be one window forward like the Gemini ones and a periscope in heat shield direction to observe the ground before Mun touchdown

I'll Send some drawings tomorrow maybe

Edited June 14, 2016 by Pvt. KASA
Added infos

[Pvt. KASA]

Ok, Took a while but I now made the first Tapemeter Prototype:

The blue white is the timing belt to drive it (I ordered it but not yet delivered)

I also put the wheels in a Lego construction, but didn't made pictures...

I also ordered some stepper Motors.

 

Im trying to figure out a way to paint the tape at the moment.

 

KASA

Edited June 18, 2016 by Pvt. KASA

[richfiles]

Very nice work!!

Make sure your wheels and belt both have matched teeth. You'll need that to maintain accuracy of the tape position, or it'll slip out over time. If you have a controller with enough processing omph, you can use something like a scavenged optical mouse as a mini scanner to scan the markings on the back of the tape as a position guide. That takes some more powerful processing power though.

Oh yeah, and you might want to look at a VCR for inspiration for your drive. The supply and take up reels do not spin at the same speed, except for a single moment when both reels have the same amount of tape on them. VCRs handle this with a gear that engages one side or the other, and a slip clutch

Edited June 18, 2016 by richfiles

[Pvt. KASA]

I just wanted to have a small hole with a UV Photocell at one end to drive there at startup to calibrate it.

I thought the Problem with the differing diameters of the wheels that hold the tape would be solved by simply pulling the tape with a spring...

The error caused by it schould be within 5mm, or what do you think? If it is, it`ll be ok.

BTW when does this "needs to be approved by a Mod" thing stop?

What is it for?

Thanks, KASA

Edited June 18, 2016 by Pvt. KASA

[richfiles]

Having an index point will help, but rolling from the spools is not that precise. It's also not an issue of "error" but of speed. If both spools are fed at the same speed (by belt), then the supply reel will unspool faster than the take reel until they are the same size. The amount of unspool has to do with the circumference of the spool being unwound, so depending on how thick your tape is, that could be a few centimeters difference. I used a small 5 foot/150 cm tape and it had a difference of 4.5 cm from beginning to end of the spool's diameter (5 cm to 9.5 cm). It's more than 5mm. One possible mechanism would be to find a windup clock and use the mainspring on one of the two reels, and then have a stepper on the other reel. The spring will take up the slack, and you can then have one motor wind and unwind the whole tape, with the spring pulling it taught. The other benefit of this, is you can attach an encoder to the stepper for accurate position feedback, if you really want to (though step counting is fine too, as long as it won't unwind), and you can calibrate for the differing spool diameter in software. One catch is that you might need to keep the stepper powered to keep the spring from taking put he spool and back spinning the stepper, if the spring is strong enough. In general, you don't need much spring... just enough to keep it tight. Another alternative, is a worm screw. I have found DC motors that use worm screws. Add an encoder to the reel, drive it with the DC motor with a worm screw, and it will never unwind. An index hole can still mark the "home" position not he tape, and the encoder will tell you exactly how far the spool has spun. Again, adjust for the growing diameter of the spool in software.

If using that method, I recommend that the "un-wound" spring configuration be orbital altitude (aka >10000). You are unlikely to launch anything so quickly that the motor can unwind the reel faster than the spring can take it... Though I've seen some INSANE 5 digit landing speed before at steep angles that would spin an altimeter from 10000 to 0 in half a second... If anything survived long enough to reach radar altimeter altitude without popping first!  Still matters for that last "Oh Shi..." moment on atmosphereless bodies. 

Also, since you only have 3 posts, that's why there the mod approval thing. it'll go away after you've made a few posts.

Another trick is if you have a pair of motors (one on each reel), making your circuit short the terminals of the supply reel motor will turn it into a generator (this is best with a permanent magnet DC motor). That means the supply reel will create mechanical resistance, tanks to the shorted motor. When the altimeter wants to reverse direction, the circuit unshorts the previous motor and drives it, and shorts the other motor instead. Since the motors only ever turn one direction, you can handle it with only 2 bits on a controller. Direction and speed. Speed can be a PWM. Direction sends opposite actions to the motors (and would use a VERY simple driver circuit).

You can still attach an encoder to one of the spools (you only need one) for position feedback, and have your index hole/mark on the tape to re-center the home position.

Edited June 19, 2016 by richfiles

[Pvt. KASA]

Ok...

I think i´ll go with two stepper Motors. Is a Encoder needed or is it optional?

Do i loose so many steps?

Do you have a Idea how to code that diameter correction?

Thank You!

KASA

[richfiles]

Unfortunately, I am beyond new to the software side of... all of this. I'm a hardware guy. My best recommendation would be to experiment. The idea with software adjustment, is you step the motors in different amounts to get a net uniform movement of tape. You can still have your spring roller, but it's there to just keep the tape tight enough to not flap around when it moves. Might not even be necessary.

If the reels and tape move smoothly, and the steppers have no problem rotating, then you won't need an encoder. You could have an index hole at the top and the bottom of the tape though, so you can index it for both launching from the ground, and landing from above.

[Pvt. KASA]

Do you have a picture how VCRs do this?

I didnt understand it

[richfiles]

I'm thinking the easiest way to do it with your setup, would be the two stepper method. You honestly don't even need to know any fancy electronics or do any complicated mechanical stuff. At best, you might use a roller on a spring to keep constant tension. That's still not a bad idea.

As far as software controls go, what you need to do is figure out a stepping pattern for the two motors that keeps nearly all the slack of the tape. You can store the stepping pattern as a lookup table, and step based on reading opposite ends of the table and scanning across. Where the spool is smaller (has little tape), the stepper will have to make more steps. Where the spool is larger in diameter (has more tape rolled on it), the stepper will have to make fewer steps for the same movement. It's just a matter of balancing it out so one motor doesn't end up going faster than the other. You can also save memory on the device and do it mathematically. If you have any errors, the spring roller will take up that slack, but it'll be very little error if you figure out the formula right. I don't know that off the top of my head though.

Imagine the "O" is a spool.
When it's small (o), it needs to step more than once, to match the amount of tape spooled by the large spool (O) stepping just once.
When the two spools are half empty/full, they step equally.
When the first spool is nearly full, it becomes the larger spool, and must step less than the other spool, which now is stepping more, to keep up. Make sense?
You can handle the number of steps in software, but you'll likely need to actually do experiments... Build the hardware and test out what the difference in steps is. See how much slack happens if you step them the same, see how many steps it takes to take up the slack. You'll need to determine if the relationship is linear, or if it's non-linear (linear is easy to work with, but since we are dealing with circumference on a  spool with a changing diameter... Ugh... my brain doesn't have that answer for you off the top of my head). Trust me though, it can all be done right, and software can indeed do it.

o     O
..... .
.... ..
... ...
.. ....
. .....
O     o

It's still a good idea to have an index mark/hole at the two ends of the ribbon. If anything mis steps, or if the numbers end up off in any way, then you can re-home the tape at both the lowest and the highest altitude.

Unfortunately, you'll have to search for the VCR stuff... Everyone spends time describing the video head and how the tape works with that... I actually had a hard time finding anything remotely useful about the clutch and engagement gear used for managing the reels of the tape. It does occur to me that a VCR is NTO a good mechanism, as a tape meter (especially if it shows radar altimeter) is a BAD idea... Ground can drop nd come up at you rapidly as you fly over the surface... VCRs have to "change gears" to switch directions... Not suitable. That's why a single stepper and a wound sprung reel, or a dual stepper configuration is absolutely the best option. Gives you instant tape reversing response.

Now, the reason I suggested the option of going with DC motors instead of steppers is super simple... You take all your measurements off of an encoder, so you end up with an absolute feedback relating to where the tape is at. After that, it's only a matter of controlling direction and braking. You can handle that with just one PWM wire, and one digital wire. (or you can use multiples, and select the right ones in software, if you don't wanna build the very basic controller that would operate it).

I think the circuit below is close enough to an approximation of how DC motor control would work. I could have something off... Basically, D0 and D1 are always meant to be opposite of each other. If you use a transistor inverter, or a logic inverter, you can do that in hardware, but you can also do it in software too... But NEVER let them both turn on at once. Both off, one on at a time... but never both on. The PWM signal runs a MOSFET that provides power for the motors, and the pair of MOSFETs in line with that select which motor actually gets the power. The pair of MOSFETs that tie the other motor lead to ground are there to short the opposite motor to ground, thus connecting it to it's other permanently grounded lead. This turns the motor into a generator that is basically shorted, which means it acts like brakes. This prevents the reel from unspooling and losing slack. You could use cheap DC motors and belt drive the spools if you like. For position feedback, you'd add an encoder to one spool. You'll have to have a calibration table in the controller's memory, again, cause spools don't ever keep it simple and run a uniform circumference as they fill up.

-[Motor A]----o---[MOSFET]--o---[MOSFET]g---[PWM]
  |+    -     |         g   |         |
  |        [MOSFET]g   / `--+---[D0] [+]
  |           |     \ /     |
  o--Ground---o      X      |
  |           |     / \     |
  |        [MOSFET]g   \ ,--+---[D1]
  |-    +     |         g   |
-[Motor B]----o---[MOSFET]---

Hope that makes sense. At this point, you really have to just make something and experiment with the software. I have no good advice in regards to software, because I've only done a very limited amount myself. Personally, i keep building hardware,c use that's just what I know, and I haven't even LOOKED at the software side of things.

[richfiles]

Guys... I found something super cool! I'm gonna link you guys my web page "The Richfiles", so I don't kill ALL my bandwidth with hot linking. It's not that the host has ever actually complained about hot linking, but I made these pics SUPER high resolution, so I'll play nice! 

So, just click "Models" at the top of my navigation bar. The page is still being worked on... after only 20 years... No, seriously! 

What you'll find is the model of the space shuttle's cockpit I built when I was in grade school, using some VERY NICE prints of the instrument panels. I think I ordered two copies of the book these came from a few years ago... I got them somewhere. I dug this out of my stuff in storage though! VERY awesome blast from my past!


Brought to you by glorious cardboard, staples, and Scotch™ tape! 
A couple pieces have been lost over the years, such as the left instrument panel area and some of the structure of the right side. It also sags. Badly.  Out of frame, this is being squeezed together by stuff on the left and right. Whatever works! 

I used to know the pre-launch, launch, OMS retro burn and landing checklists, and all the switches involved...  Man, that was ages ago. Forgot it all since.

Anyway, I put high resolution pictures up on my website. Most are 2560 pixels on the long dimension. You'll need to specifically click the link to the high res page. If you wanna download them, feel free to just right click and save. 

Edited June 21, 2016 by richfiles

[Pvt. KASA]

Yeah, I understood the Basic Geometry about it.

Buuuut, Ive got another Idea to save one Motor:

How about shafts that only transfer rotation in one direction?

So both shafts where the tape is rolled up are connected to one stepper and if the stepper turns right, one shafts pulls and one shafts slips, and in the other direction, the other shaft pulls and the other just slips.

Then I would just need an algorithm to tell the diameter of the roll/how much I have to turn to get a given tape movement.

Would that work?

Edited June 21, 2016 by Pvt. KASA

[richfiles]

I'll start by saying, i think I found the ideal solution, but that's later. I wanna get the problem with ratchet and pawl type systems out of the way first... Don't be discouraged... I seriously think I found the ultimate solution for SIMPLE one motor tape meters!

First off, you idea is sorta kinda maybe a little similar, but not quite what a VCR does. If you use a ratcheting mechanism, then it can't work. You need to be able to actually fully disengage the other reel. When it's unspooling, it goes against the direction that would lock against the ratchet, not with the release direction of the pawl... Aka... It'll lock up. VCRs have a swinging lever with a gear at the end. There is another gear that rotates around the same pivot as the lever's fulcrum. When the gear rotates one way, the swinging gear swings in the same direction and engages one reel. When he gear reverses, the lever flips the other way and the swinging gear engages the opposite reel. There is a tiny delay as the gear flips from side to side. That delay would not really work well with a tape meter. Think how long it takes for a VCR to switch modes between fast forward and reverse... Now imagine trying to display measurements on the moving tape... Then add to that, the fact that you need to have mild braking to prevent the supply reel from unwinding if the take up reel suddenly stops or slows (You need to stop inertia from unwinding your "loose" spool).

If there's any delay in switching directions, it could get problematic with some types of data display? Hmm... I actually don't recall. Were you using it for your radar altimeter, or something else? I honestly have no idea what you're doing with it. A radar altimeter needs the capacity to do rapid changes in direction to register uneven terrain. Something that just displays speed, or some other thing like that will not often change directions all that rapidly.

In all honesty, if you use two separate reels, you're far better off using either two separate motors, or a wound spring and one motor. A wound spring keeps one reel tight, while the other reel is attached to the motor. You just wind it up or un wind it. One motor, nothing fancy. You can even wind a cable around a pulley attached to the reel and attach that to something else... A weight, a different sized wheel (so you can use a shorter spring, etc). As cheap as steppers are these days, it makes no sense to skimp on functionality, especially with complex mechanics... that quite frankly, are going to introduce more points of potential failure. How many VCRs out there eat tapes too! 

I know I've posted this before, but this is NASA's actual mechanism. Examining it more closely, it seems they use spools geared to each other, that are counter wound, with a spring linkage between the gear and the reel itself. This "absorbs" all the slack and makes up the difference in circumference. They are driving the drum with the sprockets to actually feed the tape and have absolute positioning for it. The reels are "generally" wound under power, but "float" on springs, keeping the tape tight. NASA chose to use a mechanism that's held under spring tension. Their contractors have the precision machining capabilities to make this mechanism work well.



Still, it just simply comes down to one simple fact, if you do not have the option to do precision machining, then 2 reels, 2 motors. If you have separate reels, the most reliable and mechanically simple mechanism possible will eliminate any fancy mechanisms and just use two motors. I may not be a software guy, but I know that adjusting for all the quirks of the mechanism in software will be way more reliable in the long term than relying on ratchets, pawls, disengaging gears, etc.


Ooooh! OHHH!!! 
THIS! This is a GOOD idea! Here's another NASA goodie, and quite frankly, one that I TOTALLY missed before!

So, your tape isn't SUPER long (it's basically a tailor's tape). It's just too long to be otherwise all that easy to work with.  This is a VERY interesting concept. Take note of the larger black drum. The tape is spooled around the drum and into a retainer in a slot. I think the tape at the top does the same, but is wound in the other direction on the small drum beside the first drum. if you ditch the second drum, and reel the tape side by side on the same drum, It lets one drum be both supply and take up reel! If you make the diameter kinda bigger, you won't have to worry as much about slack, and I think there are two simple ways to take up the slack. The easiest way is a simple roller and spring, like your original idea. You could also attach the drums together with a spring. The only catch is the tape has to wind up side by side with itself, and you need to extend it far enough to reach the drum. You also need to probably set your end rollers at a teeeensy tiny little angle, to get the tape to line up right with the offset side by side spools on the reel.

THIS. This is your simplest solution, and probably your absolute best bet for control. It only takes one motor, handles the spooling and unspooling, is instantly reversible, and isn't terribly complex! To make this, you need a drum wide enough to fit your tape side by side. Maybe glue some manner of ring in the middle, to keep the spools from overlapping (not even sure how necessary that is, but better to keep it simple. Then you just secure one end of the tape to one side, spool it up, and when the other end of the tape (when fed through your display and two rollers) reaches it's max length, where it just reaches the drum, secure it to the other side of the drum, from the other direction. You could secure it with a flat topped screw, or whatever works. Then you'd have the mechanism below.



The larger the drum diameter, the less slack will be introduced spooling from one side to the other, and a simple spring roller should be all you need to maintain tension. Heck, the whole DRUM itself could BE the tension roller. You just mount the drum and servo to a plate, mount it loose at the bottom, and have a spring pulling on it at the top. As slack tightens or loosens, the whole assembly lightly flexes forward or back. Super simple! the drum and servo, the top roller, the bottom roller... and that's it!

It might look kinda complicated, but this can be made with a drill and hand tools. It's basically a plate with a few holes and a slot cut in it. Everything else can be strips of metal with holes drilled into it, and the metal can be bent as required with a a pair of pliers. You could probably just get some angle brackets from the local hardware store and bend them to the correct angles. Mount your rollers, pivots and slides, as required, and you've got it! The drum might look hard to come by, but honestly, you could probably make it from a piece of trim hardware (again from the hardware store). A wood cylinder can probably be cut from something that you would buy to make a stair railing or something. No reason you couldn't use wood there. The stepper motor is just that, nothing special. All you need are 5 holes... the shaft and 4 screws to secure it. I think this thing could be made quite easily!If you wanted tapes side by side, you could have one set with a little longer tape, and have the drum set behind the other. It's just a matter of spacing out your pieces so everything has clearance for everything else. And, if the width is a concern (say you have instrument mechanisms that have to sit beside it), you can use a hollow drum, and mount the stepper inside the drum, with the shaft attaching to one side of the drum (it would be shaped like a larger version of a bottle cap. Open on one side, and solid on the other.

If you want custom numbers, like me, then buy an office stamp with as many digits as you require, and get yourself either a measuring tape with only one side printed (use the back then), or look for some type of fiberglass reinforced plastic or vinyl strap. That's basically what those tailor's tapes usually are. I still need to ask the local screen printing shop what ink they recommend for a flexible surface. If I find out, I'll share their recommendation here.

If you have more room to soak up slack, then there may be one MORE option... (confused yet! ) If you mount two reels on a sliding mechanism like above, and use two large reels, you can probably gear the reels together and use one motor for both. The whole reel/motor assembly would pivot like above, and as long as the pivot amount can take up 100% of the slack in the reel circumferential differences, then I think you might just be okay. Remember though, large drums are your friend, as that minimizes the difference in circumference change as it spools in and out.

Edited June 21, 2016 by richfiles

[Pvt. KASA]

Bit that basically the Same I planned, isnt it?

Just two rolls (on one shaft) and a spring mechanism to take up the slack.

Ive got an 3D Printer, so I could print the Parts.

I think that precise enough.

I want to use one for a Radar altimeter and one for a sink/climb rate meter.

And I didn't understand why the ratchet Idea won't work.

Just to Rachet-like shafts that slip in opposite directions.

KASA

[richfiles]

The ratchets don't work, because when a reel wants to unspool, it'll be turning the wrong way, and lock the ratchet. Remember, the other half of a ratchet looks like it locks in the other direction. You can visualize this by considering a bicycle. You can pedal forward, and coast in the same direction, but it you reverse the tire, suddenly, the pedal locks and is forced backwards. Same with your reels. If you attached two bikes facing each other like the reels, and then linked the pedals with a chain... it'd lock up. It would not let the other side slip. You may as well not bother with ratchets then. Also consider, that both meters need to be able to rapidly change direction. those two meters are the ones that most need to change direction. ratchets need a small lead up to lock. Even a good ratchet needs a tiny amount to do so. If you're relying on the stepper motor counts to keep accuracy, this will quickly become impossible, as direction changes introduce backlash in the position count... It'll quickly loose accuracy. It's why I highly recommend an encoder. You can mostly ignore backlash with an encoder.

The only reason I went with the side by side drum configuration, is that a large diameter drum basically absorbs much of the error, since the tape does not end up forming as much difference in circumference. A small reel will have a larger error, due to the tape layering over itself as it spools up, increasing or decreasing the diameter (and thus the circumference) more drastically. A large reel will show less error because of this. You certainly don't have to do it that way. You could have a roller and spring, just like in your original configuration too, but making a floating mount makes the entire drum keep itself taught, and reduces the part count and the number of surfaces flexing the tape. It also simplifies the drive mechanism.

Also... totally wish I had a 3D printer! 

Edited July 1, 2016 by richfiles

[Antipaten]

You don't need an advanced mechanical solution to handle two spools of tape with step motors. All you need is a spring loaded wheel that take up the slack of the tape that is caused by rounding errors between the two step motors. Put two sensors (optical or microswitches or whatever) that detects maximum and minimum allowable tape slack and adjust the relative speed of the reels when you reach one of the limits. The simplest algorithm should be to just stop one of the reels for a few steps when the slack is too small or too big.

EDIT: You need indexing on the tape too in order to adjust for the changing tape feed per step, unless you add a third step motor and another slack roll.

Edited June 30, 2016 by Antipaten

[Stoney3K]

On 19/06/2016 at 2:00 PM, Pvt. KASA said:

Do you have a picture how VCRs do this?

I didnt understand it

VCRs and other tape decks don't control tape speed with the reel motors, the reel motors just maintain a constant level of tension to take up the tape if it moves in any direction.

The tape speed is controlled by a capstan and pinch roller which "squeeze" the tape and keep it moving at a constant speed.

BTW, when I look at those NASA tapes, it looks like they just used perforated 8mm cinema film as the base to print their indicators on. Interesting.

Edited July 4, 2016 by Stoney3K

[richfiles]

There are records that indicate that some of their meters (including the standard edgewise meters) had their graphics created on film, but there was an alternate style used as well, that used etching. The Apollo CM and LM used differing methods. Color areas were paint... which they had a lot of trouble with. The 8mm film idea is interesting. Sadly, film isn't as easy to come by these days, save for the occasional disposable camera, and the few places left doing film development. The NASA machines, rather than use a capstan and pinch roller, appear to use the cogged drum as the actual drive/position roller. You can see exactly how the gear train is set up because you can actually see it for the tape not he next layer in. The gears at the bottom front are fro another pair of reels and another cogged drum roller The spool gears are spun opposite of each other, and each spool is bound to the gear driving it with a mainspring. Pvt. KASA, if you can make those with your 3D printer, you're pretty much good to go.

The little tape meter is different. It has two reels, one smaller, the other larger. I suspect that one is sprung, and the other is not. The one that is not is directly driven, and the sprung drum lets the tape unwind.

It really is a shame we don't have a good way to selectively expose and develop camera film. I certainly don't have a darkroom, nor do I have the materials to do a controlled film exposure. That'd make great tapes! I'm fortunate to have actual tape meter hardware on hand, even if it was just a display for a manually controlled device. The fact that I already have perforations and a cogged drum is great, but it doesn't help me or anyone else with new creations.

For that reason, I think the second meter picture I showed is more relevant to custom builds. It does not appear to use cogged indexing system. I think one reel is sprung with a mainspring, and the other, I suspect, is a direct geared drive. Interestingly, the easiest way to mark out divisions on a setup like that... Is to just step the drive equally, and mark out your divisions on the tape (assuming you're making your own tape, maybe with stamps and a screen printing ink). By stepping the drive equally, the tape will move to the desired spot, you can stamp your next number, and then repeat. Over time, the numbers will gradually space a little wider apart as you take up more and more tape, but you'll always use the same number of steps between numbers. Depending on reel sizes, you might even be able to get half division numbers printed between your number steps, which would be nice for when you're getting closer to the ground! You get a mildly logarithmic tape out of it, with more detail on one end, and you can drive the motor easily, with no software corrections, and the same stepping between numbers across the entire tape. It also needs the fewest custom parts 3d printed/machined, with only a single sprung reel being required (as the reel attached to the stepper motor is fixed, connected directly to the motor).

Edited July 4, 2016 by richfiles

[Stoney3K]

On 4-7-2016 at 7:04 PM, richfiles said:

For that reason, I think the second meter picture I showed is more relevant to custom builds. It does not appear to use cogged indexing system. I think one reel is sprung with a mainspring, and the other, I suspect, is a direct geared drive. Interestingly, the easiest way to mark out divisions on a setup like that... Is to just step the drive equally, and mark out your divisions on the tape (assuming you're making your own tape, maybe with stamps and a screen printing ink). By stepping the drive equally, the tape will move to the desired spot, you can stamp your next number, and then repeat. Over time, the numbers will gradually space a little wider apart as you take up more and more tape, but you'll always use the same number of steps between numbers. Depending on reel sizes, you might even be able to get half division numbers printed between your number steps, which would be nice for when you're getting closer to the ground! You get a mildly logarithmic tape out of it, with more detail on one end, and you can drive the motor easily, with no software corrections, and the same stepping between numbers across the entire tape. It also needs the fewest custom parts 3d printed/machined, with only a single sprung reel being required (as the reel attached to the stepper motor is fixed, connected directly to the motor).

It does look like that method could even be very feasible to do with a photographic film, provided you can get an appropriate image source. Each 'step' of the drive gear would then be a different exposure which could be switched out manually. The actual display window of the meter could in that case serve as the film gate, and the letters to be printed could be printed on larger sheets of paper or transparency and reduced with a standard 16mm or 35mm camera, in a similar way that cel animation studios would do it.

In the 50s and 60s, it would have made a lot of sense for such a process because they didn't have laser printers that could print micron-sized dots. It would also mean they didn't have to worry as much about the decrease in spacing, which is actually linear and not logarithmic. NASA fabricators could draft up the indicator legends on large sheets of paper and reduce them to the size necessary for the meters, which would greatly increase precision.

You can get blank B&W 35mm film for cheap which could act as a nice indicator tape, and it's easy to develop with inexpensive chemicals. Would be a nice touch for historical purposes.

[Pvt. KASA]

Update:

I was out of Time the Last weeks but now we finally have hollydays 

At the Moment i'm trying to create the gears that fit the Timing belt which was just delivered.

Do you have a Idea where I can create STL gears for 3D-Printing?

I only found JsCAD and there arme too less Parameters so i just try to scale it up and down...

Not very efficient...

And I try to create the Layout for the Panel (Im drawing all Instruments in Paper and cut them out)

Maybe I will add an edit later to Show the the 2nd prototype if I manage it to Finish it today.

See u later

Edit: No Edit for today

I first have to recalibrate the 3d printer (I wanted to fine-tune it and messed it up )

Buuuuuut I was able to create a wheel stl and it fits perfectly.

(So only another wheel to go)

Tomorrow I will try to get the printer working, until then im working on the layout.

BTW do you know how i can get rid of these big distances between the lines?

KASA

Edited August 2, 2016 by Pvt. KASA

[Freshmeat]

If you need a CAD program, I swear by OpenSCAD (when I don't swear at it). Can do anything, if you are able to work your way through the parametric programming language.

[Pvt. KASA]

Thank you! @Freshmeat

I worked with blockscad all the time (I know it's a bit childish but I am a child )

But I didn't manage to do a gear in Cad...

Do you have a simple guide how to do this?

For what I know, you need the coordinates of the peaks of the teeth... And I don't know how to get them.

Im in class 8 and we didn't learn it in school so I thought it would be simpler to use an easy gear generator...

@richfiles I See you watching EevBlog alot 

BTW I got the 3D Printer working again...

bye

KASA

Edited August 2, 2016 by Pvt. KASA

[richfiles]

6 hours ago, Pvt. KASA said:

@richfiles I See you watching EevBlog alot 

BTW I got the 3D Printer working again...

bye

KASA

Guilty as charged! I love those Mailbag Mondays and Teardown Tuesdays! Besides those types of videos, I swear I've learned more from YouTube than I ever learned in college! 

You'll also find me watching a lot of Louis Rossmann's videos. Man's got the mouth of a sailor, but he knows component level repair! STRONG supporter of proposed Right to Repair bills. Big companies like Apple and others have strongly lobbied against passing such laws, which would guarantee people's right to modify and repair the devices they buy. Those companies would rather sell you a new $2000 gadget, and send your old stuff, killed by a defective 5 cent part, to the recycler.

So awesome you got the 3D printer up and running! I'm so jealous, though! I STILL don't have such a wonderfully amazing contraption! 
I can't offer you any recommendations for gear making... or 3D modeling... or 3D printing... I have no experience there. I wish you luck!

My Kerbal Kontroller progress... Nada... Zip... Zilch... Work has been a killer, and I've been busy at home. The few free days i have had, have been spent tryyyying to decipher the gobbledygook known as C code that is supposed to make my keyboard function... I finished the keyboard in April or May! I can't even remember anymore! I still don't have it up and running!  I'm beginning to dread the thought of coding software for my controller!  I wish I had the kind of money to just hire a coder!  Honestly, I haven't even so much as bought stamps, or even stopped at the local screen printer to ask about flexible ink recommendations for my tape meter. Nor have I stopped at any of the local metal shops to get quotes on the enclosure. Ugh... NO progress at all...

Well... I made SOME progress... Somewhere, on a Post It note, is a scaled set of dimensions for the vertical Instrument face. JUST the vertical face... That's it. 

On a side note, I did manage to snag a valid Windows 8.1 activation code, and got a Windows 10 install on my Hackintosh within hours of the deadline! LOL Nothing like single digit hours remaining to a 1 year count down to light a fire under my procrastinating behind!  Doesn't make much difference as far as KSP is concerned, as I play on Mac anyway, but it does mean I was able to download the Xilinx programming software. It means I can get back to programming CPLDs (Complex Programable Logic Devices). CPLDs allow you to create reconfigurable logic blocks out of standard logic gates. You can either use a specific coding language, or literally draw a schematic using logic gates to program the chip. Not sure how much use I'd have for these types of chips in my controller, but I might be able to use them to create custom drivers for some of my DSKY and alpha readout LEDs. I can't drive those with the MAX7219 chips, as they are common anode. I was gonna use a diode matrix to drive them (It's really simple, but requires manual wiring), but I could create the drivers entirely on the Xilinx CPLD chips. I can certainly live with that! LED drivers are certainly something I've designed from scratch before! Well, that's one thing I can do with my Xilinx chips! 

Edited August 3, 2016 by richfiles

[Freshmeat]

@Pvt. KASA If you use OpenSCAD, you define a single cog and rotate it. You might even get by just defining a block as the cog. Try starting with this, I whipped it up in 15 mins, so it will need a lot of refinement.

max = 12;
for (i = [1:1:max]){
    rotate([0,0,i*360/max]){
        translate([20,0,0]){
            cog(5);
            }    
    }
}

    
module cog(s){
    translate([0,-s/2,0])
    cube([s,s,s]);
    } 

 

To find coordinates of a point on a triangle for a more advanced cog, you will have to use some trigonometry, which might be a bit out of scope for a forum post if you have to start from scratch. The best advise I can give is to draw it out on paper and calculate coordinates from there.