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

[richfiles]

So, a huge development for me...
I was watching Lord of the Rings: Return of the King, and the end credits started to (finally) roll.
I had the crazy thought... What could I get done during the time it takes to roll these credits?

It was apparently enough time that I fixed my mill's jammed Z-axis! 
Not even joking. I fixed my mill. During the end credits of LotR. It's the most time I've ever dedicated to working on it! 

Since I'm in an apartment, it's not good enough that I just fix it and run it. It's in a closet right now, but I bought some rolls of long rubber mat (about a meter wide, and several long). My plan is to make a simple box frame with some cheap lumber and screw the rubber mat around the box (and maybe some sound dampening too), to shield the walls and floor from chips and cutting fluid. I have tomorrow off, so maybe I can get it done then. Having the mill up and running gives me the chance to more easily cut my panels... I hope... It is a small mill, so I hope I can fit it all in! I also hope my neighbors work during the day. 

**EDIT**

Man, It's been forever since I even thought about using that thing! I cleaned out the closet, so now I have a 3x5.5 foot (0.91x1.6 meter) machine shop! 

I measured up the table I got for it ages ago, and figured out the hardware I need to mount it safely and securely. I think I have a few rolls of LED light strips as well, so I can mount those for lighting in the closet. It's only a 500 watt motor on it (it is a rather tiny mill), and I have power run to the room via one of those old style computer power distribution boxes (they would typically sit under your big old CRT monitor, and have several pushbutton power switches to turn on your computer and all it's various peripherals. Basically, I'll be able to plug in the Mill, LED strip power source, and my DRO (when I get the scales installed), and I'll be able to just push the button on the front of the box to power things up. My milling table even has a proper drain, so if I do a decent enough job in my rubber mat "box" around the mill, I could presumably even use proper cutting fluids and such! The milling table has a door in the front, and you can run a hose from eh drain to the bucket. I could even get a pump and set up a cutting fluid recycler! Maybe...

I probably will take a while to get everything properly set up, unfortunately. The main instrument panel is also rather huge. Trying to mill it would be difficult. If I had room, it'd be possible to mill by doing it in sections, but I have only 5 extra inches of clearance on either side, when the mill is extended to one extreme or the other, so I am quite limited by the small space of the closet it's in. What ought to be doable though, is the smaller DSKY face panel. What I would like to do is machine the display openings from  a sheet of alodine treated aluminum. I have a nice piece that'll work, though if I can get my hands on some solution, I'll refinish it, so the machined surfaces are also treated. Doing that will give the DSKY it's own distinctive feel, as if it were a module, in and of itself. If I can find a reasonable optical filter to use, I'll mask the annunciators and the DSKY LED readouts and block them as two whole areas (like the real DSKY was arranged). I'd like to find a grey or grey-gold filter, but will settle on anything that'll genuinely work.


It's a tiny shop, but it'll do for now! 

The top panel of my controller, I want to paint a similar "hammered" green.
All the various brands of these Chinese mini mills are all actually the same model, made by  the same source. Different retailers just order them with their colors and nameplates. Jet is white, Harbor Freight and Seig are red... Knowing they are all the same, I honestly chose the Grizzly, cause it was green, the koolest kolor! 

Edited August 30, 2016 by richfiles

[richfiles]

Not so much Kerbal controller related, as just simply setting up the tools to be able to actually make stuff for it... I did more work on the mill yesterday and today. Made up the initial frame to suspend the rubber mat from. I got all the mounting hardware to secure the mill to the table, and got some light inside the closet. One of the first things I'll so with the mill, once I'm completely set up, is to mill out the large side view window for the tape meter (it's currently only has a single digit at a time end viewport). Milling out the side will allow a NASA style tape meter to be shown. I'm also going to (reluctantly) mill off the side mount for the rotary switch mechanism. It's such a cool switch setup, but it takes too much width. I'm thinking I'll remount the gear to something rear mounted, or swap the gear for a belt pulley. I can drive that with a stepper with an encoder. I'll punch a tiny hole in the tape, at the 0 meters position, and use an optical interruptor to make that into my index. That ought to work well, and be a relatively narrow profile. The beauty of having a mill, is nothing is really out of my reach, from custom brackets to mount my enclosure pieces together, to custom facia, like what I'll make for the DSKY!


This is the wood box frame. I have not had time to cut and suspend the mats from it yet. Tis a tight fit, Aye tells ya! I forgot about a shelf in the closet, so I may just drape the mat partly over the mill, at light level. I'd rather not block that shelf. I suppose I could consider getting some foam and a few adjustable shower/coat hanger rods (the kind that has a spring or adjustment so it stays up by squeeing between two walls). I'd hold the foam in place with he rods, and could mount it way high, near the ceiling. Maybe I can even get that kind of foam with the... pattern of hills and valleys. I think that's good for sound dampening. I might just line the walls under the mat with that foam too. I wanna reduce noise to as quiet as possible. I do have neighbors.


I saw the light, and realized how important it was to get this mill set up for the first time in 4 years. 

Also, before you ask about the board the light is mounted to: yes, no, no, and oh yeah... 
Is it crooked?
Is it screwed down?
Do I care?
Is it really wedged between he two walls that hard?! 

Edited September 3, 2016 by richfiles

[richfiles]

Man... I had a 3 day weekend. First in a LONG time... I was SO looking forward to it! 
... Strained my shoulder at work, and then got sick, on top of the injury... Wasted the WHOLE thing! Spent most of the time in bed! 

No setting up the mill. No assembling the enclosure and cutting out openings for the instruments... Not even soldering. Shame RL can't have a revert to VAB... 

I did get some nice foam from work though. It's the kind with the "egg crate" pattern on it. I plan to suspend it from the ceiling of the closet with the mill, to act as an acoustic dampener. I figure, the more sound deadening I can achieve int he closet, the better. A mill is useless if the neighbors complain every time I use it.

Edited September 9, 2016 by richfiles

[ryan00793]

I'm looking to start to get into making a panel but i want to make sure it works on windows 10 first since I have windows 10. Also a known good example .ino for the arduino wouldn't go amiss.

[Freshmeat]

If you are on windows 10, you should probably look to the plugin kRPC instead of KSPSerialIO. Unfortunately, I cannot help you with further specifics.

[ryan00793]

4 hours ago, Freshmeat said:

If you are on windows 10, you should probably look to the plugin kRPC instead of KSPSerialIO. Unfortunately, I cannot help you with further specifics.

Why would kRPC be different that say serialIO, it just looks like a c# api to read / write data. Unless you're saying that I would need to write my own communication library to talk with an arduino...

[Freshmeat]

45 minutes ago, ryan00793 said:

Why would kRPC be different that say serialIO, it just looks like a c# api to read / write data. Unless you're saying that I would need to write my own communication library to talk with an arduino...

KSPSerialIO is bugged with Windows 10, and you cannot send data back to KSP. I know that @Antipaten and @PeteWasEre both use kRPC, and can send signals both ways. So if you are using Windows 10, you are best of diving straight into the deep end. Personally, I am still contemplating if a complete rewrite of my code is the trouble worth, kRPC is a lot more powerful than KSPSerialIO but seems far more complicated to use.

Edited September 16, 2016 by Freshmeat
spelling

[PeteWasEre]

I haven't used KRPSerialIO so i can't offer a comparison, but I can say that as a noob to this style of programming (I have a real time C on UNIX background) it was pretty easy. The KRPC documentation is good, the author is always happy to help answer questions, and it gives you the freedom to choose a language you like. Heaps of examples exist in the docs and online.  There is a lot of information available through the interface but so long as you plan out and manage your data transfers performance is ok. Minimise calls and use streams for repetitive data transfers.

The only bit that did my head in was the reference frames, but this is a maths and geometry issue more than a KRPC issue!

@ryan00793 if you want to see the KRPC python code or the Arduino ino you are welcome to take a look on github, or hit me with questions.

[Antipaten]

16 hours ago, ryan00793 said:

Why would kRPC be different that say serialIO, it just looks like a c# api to read / write data. Unless you're saying that I would need to write my own communication library to talk with an arduino...

If you want to use a serial connection with kRPC you need to make a server side script to manage serial communication. It should be possible, but I haven't tried it, to communicate directly using a Protocol Buffers library on the Arduino and add either a WiFi or an Ethernet adapter to the Arduino, but I haven't tried that as my goal is to be able to connect my control panel with a single USB cable to the computer for both data and power.

[Mad_Scientist]

On 9/17/2016 at 1:21 PM, Antipaten said:

If you want to use a serial connection with kRPC you need to make a server side script to manage serial communication. It should be possible, but I haven't tried it, to communicate directly using a Protocol Buffers library on the Arduino and add either a WiFi or an Ethernet adapter to the Arduino, but I haven't tried that as my goal is to be able to connect my control panel with a single USB cable to the computer for both data and power.

I did exactly this. It is not so hard guys, you run a kRPC server in the game, using Python (or your language of preference) on the computer to read the data from the game, send the packages through serial port. On arduino you just listen to serial port, the package contains the address of a device and a new value. If you have input from arduino it works the same, just other way around. I made a prototype for altimeter in 3 days (I am new in arduino, more experienced in programming)

[richfiles]

Consider getting an Arduino clone that uses the CH340 chip for communication. Since they are invariably cheap Chinese clones, they tend to be dirt cheap. Even if it ends up being little more than a hub to distribute packet data between other Arduinos and the PC, or if you actually use it as your primary controller, that's probably all you really need. It uses a different driver than the native Atmel 16U2 or FTDI based communications. I remember ages ago, someone on the Arduino hardware thread had mentioned KSP Serial IO was working for him in Windows 10. That was suspected as the reason, but just the other day, @Mattew has actually successfully run KSP Serial I/O on Windows 10 and transmitted data, both ways, for long durations of time.

So, it is confirmed that using CH340 based Arduino clones solves the communication issues immediately, and it narrows the Windows 10 issue specifically to the driver that the Atmel 16U2 relies on for communication. Hopefully, this will lead to a solution for all Arduinos soon, regardless of whether they have a 16U2 or a CH340 equipped.

To summarize: Windows 10 appears to WORK with CH340 equipped Arduino clones!

Edited September 19, 2016 by richfiles

[Stoney3K]

Sounds like it's worth the effort to hack KSPSerialIO to work with sending/receiving of UDP packets, so you can use Arduinos with Ethernet modules. Ever since those became really cheep, I moved away from serial I/O because it can be quite error-prone when not done right.

[ytmytm]

22 hours ago, Mad_Scientist said:

I did exactly this. It is not so hard guys, you run a kRPC server in the game, using Python (or your language of preference) on the computer to read the data from the game, send the packages through serial port. On arduino you just listen to serial port, the package contains the address of a device and a new value. If you have input from arduino it works the same, just other way around. I made a prototype for altimeter in 3 days (I am new in arduino, more experienced in programming)

I agree. Once you get it running is quite easy and very powerful. If anyone is looking for a starting point, please consider my project: https://github.com/ytmytm/ksp-gegi/

There is Python script (gegi.py) running on PC that acts as an interface between kRPC server and serial I/O. It translates messages from the game into simple commands understood by Arduino code - to turn LEDs on/off, move analog meters or show something on LCD. It also translates messages from Arduino side (when knob is moved or a switch changes state) into kRPC requests - to change state of RCS or throttle.

The Arduino side is completely independent from kRPC, it's just a driver for buttons and displays, so I could easily debug it using serial monitor by simply typing commands.

[Kotagi]

[ryan00793]

So I've started my project since I've started to receive parts however I've not decided on a full design yet my current thinking was the pic below but there isn't enough room for me to label each display. So what do you guys think I should do?

[Freshmeat]

If you want that amount of displays, I think you will need labels. More so if you are able to select what to display on each individual display. So I would cut some displays to make room for labels.

[richfiles]

I recall the LED library for those Max7219 boards supports 8 devices per SPI bus. You can set up multiple SPI busses though, so your configuration would require 2 SPI busses to be set up. Is there any reason you're not just making the panel a little bit taller to giving yourself a little more space vertically? Remember, if you make your enclosure to mount your panel at an angle, then you can stretch the height a little, and not have it go passed the edge of the monitors.

[ryan00793]

5 hours ago, Freshmeat said:

If you want that amount of displays, I think you will need labels. More so if you are able to select what to display on each individual display. So I would cut some displays to make room for labels.

I would rather stretch it out and keep all the displays over removing some for the labels (size isn't set in stone it's just there to give an idea of what I was shooting for) just looking for creative solutions to the problem.

3 hours ago, richfiles said:

I recall the LED library for those Max7219 boards supports 8 devices per SPI bus. You can set up multiple SPI busses though, so your configuration would require 2 SPI busses to be set up. Is there any reason you're not just making the panel a little bit taller to giving yourself a little more space vertically? Remember, if you make your enclosure to mount your panel at an angle, then you can stretch the height a little, and not have it go passed the edge of the monitors.

Mine don't have an spi bus so they can use w.e. (ish (In practice I've noticed that they tend to fall out of sync and lock or stop displaying numbers when CS isn't on pwm)). I can indeed make it taller and just space it out a bit more but I was looking to see if there was any more creative ways to do it. For example moving the scale leds or just using one rgb with different colors or even different lay out of the screens where the leds are on one side and labels on the other. Also good thinking about monitors since I wasn't, I also plan to have a switch / button panel on the bottom with fun stuff like a slider and switches and ofc key locks with big red buttons.

 

Basically what do you guys think of that layout? Should I change it to look different should I nix the annunciator box and go with something like those tft screens (real ones)? Idk it's all up in the air and I'm doing this for fun so I don't mind having to get a better tool for the job if I feel that it's worth the money. For example these 7-seg displays i think were just north of $1.50 each so it's not the end of the world if someones like hey you don't need that many and should do XYZ instead.

PS 3 of the 9 7 seg displays I ordered were faulty so I would need to order 3 more anyway to get back upto 9, so atm I have 6 and I could just order more since the guy is gonna refund the costs of the 3. Just poking around to see what you guys think design wise.

Edited October 10, 2016 by ryan00793

[richfiles]

On 10/10/2016 at 11:36 AM, ryan00793 said:

PS 3 of the 9 7 seg displays I ordered were faulty so I would need to order 3 more anyway to get back upto 9, so atm I have 6 and I could just order more since the guy is gonna refund the costs of the 3. Just poking around to see what you guys think design wise.

I know this feel man... I know this pain... Look at page 10 of this thread... I know your pain... OVER 50% failure rate on the LED modules I got, and they kept on failing after using them for a while. 



If I recall, it's using the SPI serial protocol to send data to the chips, but it's on arbitrarily defined pins that the library initializes. One issue with this, is if you have a lot of long runs of wire, you can introduce noise or timing glitches that cause the signals to fail. One method people have found to help with this is to twist your wires. I've heard some people twist CLK with GND and the Data line with VCC. There is a problem with this that could be troublesome, which I'll explain below. The best bet is to minimize cable run. Consider the total length of wire that needs to run for the Clock and Data lines... You need to include the PC board of the LED drivers, the wires between them, and the wires to your controller. Smart layout will help you a LOT! You'll have a lower overall data transmission length by connecting the LEDs vertically, going down one column, up the next, and down the last column. That's the least amount of criss-crossing of wire runs to connect modules. 

So, I mentioned a potential problem involving VCC before. I noticed with the particular boards I bought, that if you connect power serially (VCC and GND Out from one board to VCC and GND In of the next), they were passing the output through the reverse voltage protection diode (my boards had it)... The problem is this introduces a voltage drop of about 0.6 volts with each additional board. By the end of the chain, the boards barely have sufficient power to even turn on, if at all! The solution is to connect all the VCC and GND directly to your power. I'm not saying your boards are like that, but check your voltages, cause mine were, and they look VERY similar to yours. The way to test this is to use a multimeter with a diode check and a continuity check. Test continuity between he two VCC pins. If you don't get continuity, then switch to diode check, and test if you get OverFlow in one direction and 0.4-0.7 in the other direction. That means your reverse voltage protection diode is wired inline, and you can not daisy chain power! 

Notice how all my power connections are all connected directly between the power source (on the Arduino) and the input power connectors. I never pull power from the VCC output, cause on my boards, the design is actually wrong. I ONLY take the three signal wires from the output end of one module and bring it to input end of the next module. If you measure a diode between VCC In and VCC Out, this will be necessary.

***I have NO IDEA why I marked option 2 with an X. Option 2 is fine. Option one is the one with extra long wire runs to avoid.***


The second one is actually okay, if you really wanna keep consistent order... I mean, it adds a little extra wire length, but unless you actually have problems, probably not the end of the world... Jut too lazy to change the red X to a... green check. First one has long wire runs that some people have had problems with noise or timing issues. I think most people would be fine, but hey, just minimizing extra bus length. The first three all have one basic fault... Max number of devices per bus is exceeded. No problem though, as you can assign additional pins as a second bus. The example below explains it...

/* If you need to control more than 8 MAX72XX, you can always create another LedControl-variable that uses 3 different pins on your Arduino board. The only thing you have to do is to initialize another LedControl */

/* we have to include the library */
#include "LedControl.h"

// Create a LedControl for 8 devices... 
LedControl lc1=LedControl(12,11,10,8); 

// ... and another control for ANOTHER 8 devices. Now we control 1024 LEDs from an Arduino, not bad!
// Note : the second one must use different pins!
LedControl lc2=LedControl(9,8,7,8); 

With that, you can define two separate SPI busses and each bus can control up to 8 different MAX7219 modules. You ought to even be able to control your annunciator grid with a MAX7219, if you wire it in the correct matrix.

Edited January 3, 2021 by richfiles

[ryan00793]

@richfiles Yea on mine you can see the diode on the back so I knew it was gonna be a problem (about the same drop for me as well), it's why you can see the two vcc and ground  going elsewhere (doesn't have to be done with ground but I ripped two wires off of the bundle so I just went with bringing ground back as well). However I can't chain them like you are they just don't work well after the second display and maybe that's because I'm not using the real SPI bus but idk, I'm still waiting for the mega in the mail so maybe with a mega I might be able to better drive them off of random pins (I'm just using an unoR3 atm (I don't know if the higher clock speed on the mega means higher switching freq on the pins but maybe?)).

Just saw the edit. thanks for that, I have some tlc5940nt's (5) for controlling the rgb leds I'm going to drop in the box if I do it. I'm kinda liking the idea of having the rgb leds near the switches as well so idk if I'm going to do both.

Edited October 11, 2016 by ryan00793
fixed formatting

[richfiles]

On 8/19/2016 at 7:15 PM, Pvt. KASA said:

I cant use it als a chair as it is because you Arne lying in the back and watching up.
I will rethink it.

I had a thought about your chair. I had suggested making it dual use. That might still not work, but I had this thought, and if it's useful to anyone... well, here it is. 

It explains itself. It lies on it's back, and the legs keep it at your desired angle for KSP play. Whatever setup you have your controller on, include your leg rests there. When not playing, you flip it upright, and you can wheel it around and use it as a regular chair, meaning if you have limited space, this does double duty. Better an idea come 2 months later than never! 

Edited October 12, 2016 by richfiles

[JeffSergeant]

I'm making a simple KSP Information display panel powered by the KSP Serial IO Mod.

The latest iteration looks like this:

Features

LED Display (much more atmospheric than an LCD

 a servo-controlled  multi-purpose gauge, which lets me do more interesting things than a digital display, it's also a little noisy which I actually like... 

and some LEDs.  The LEDs act as indicators to tell you what the number on the Display means.  But also as a descent speed indicator.

 It works.. It's a bit ugly, and it requires an Arduino on a development board.  It's also held together with a wire-wrapping technique that is definitely not up to any sort of code.. not even by Kerbal standards.

Here's a mock-up of iteration #3:  (The top is 14cm by 7cm, for scale).  The only added features are a better form-factor..  and the top is a removable piece of card/paper to support re-programming the function of the gauge and display without having to rebuild anything.. Also features a PCB designed by Bill Kerman himself!

Next steps:

Validate the electronic schematics to confirm I can run it from a standalone Microcontroller and USB->Serial converter (I have everything I need bar two £0.10 Capacitors to do this, which is why I'm playing with 3D designs instead) -Basic test done, set up an Atmega328 chip on a breadboard and communicated with it via serial!  And only bricked 20% of my micro controllers in the process, Success!

Once that's done finalise PCB Layout and Order PCB.  It's probably massively overkill for this circuit but A. It's cool. B. It's actually possible to get small runs printed now for tens of dollars (until the last year or two no-one in this country would touch a hobbyist PCB) and C. It's going to be a present for a KSP fan in my life so needs to be relatively sturdy and safe!

Then I just have to nail the box design and send that off to be laser cut.  and then wait with all digits crossed that it all comes back in one piece... and then assemble it.

Costs: PCB ~£15, Laser Cutting ~£15 Components ~$10 

Hopefully all done by Christmas!

 

Edited October 15, 2016 by JeffSergeant
Fixing Tyops

[richfiles]

Well... I took the plunge. I ordered the Futura stamp set to do my tape meter. man, that was spendy! I could have spent $5-10 on a cheap office stamper, but instead I went Futura, and it set me back just shy of $60 USD! 

I could have saved $20, but I figured, the fact I have to import this thing means I may as well buy the full set, so i don't have to deal with this again, if I suddenly find myself in need of more than just the numbers.

Can't believe I had to buy three sets of numbers... No duplicates... Ugh... 

Edited October 22, 2017 by richfiles

[richfiles]

Another development... I my be down to having space for only four of my vertical edgewise meters in my main panel instead of five...

I may have some extra meters... 

I mounted the slide guides for the steel mounting rails for the horizontal control panel (so it can be slid out of and detached from the instrument panel's main body), and discovered that due to lack of foresight, that the steel support rails cut into the edges of the meters. Now, I think I'm okay on the right side, by the vertical velocity meters, since the meter is shorter and round. Means I have a hair more clearance. Even if I don't have enough clearance, I can make a SUPER easy compromise... I can put my digital LED altimeter readout below the meter instead of above it, as I'd originally planned. Long story short, right side is okay.

Left side... I have three options:
A: I can give up a meter and have just 4 meters in the panel. One single meter, and three dual meters. I'd rather not do that! 
B: I can CUT INTO THE LEFTMOST METER HOUSING, as I think that the internal mechanism clears the rail... Maybe? I'd... I'd really rather not do that! 
C: I can buy another dual edgewise 1251 meter and switch over to having 4 full meters... Which is actually the same number of readouts as 3 duals and 2 singles! Sounds good! 

Right? Right!?

Well, the catch is the cost! I got INCREDIBLE deals on the meters I have already. I mean... UNREAL deals... The kind of deals that shocked me. Deals I've not seen since... The meters that are for sale at the moment cost at least 3 times more... After negotiating a reduced price! Additionally, I'm torn. I found a pair of meters that are direct Arduino compatible (0-5 volt DC input). Not bad, no adapters needed... Catch is, they are $65 PER METER! Yikes!  Of course, I'm committed. If a second ebay offer falls through (I offered $42 per meter... LOL  ), then I'm gonna buy anyway. The big question, is do I buy both meters available and retire all of my GE 180 single edgewise meters, committing to a unified 1251 Dual edgewise layout across the board... If I do that, it'll set me back $130... Not sure I can justify THAT much! I spent half that amount on all three of the 1251 meters I already own! Again, YIKES! 

That leaves me in an interesting situation... I'll have either 4 dual meters on my main panel and 2 single meters on my overhead panel, or I'll have 4 dual meters on my main panel and 1 dual meter on the overhead panel. That depends on whether I get just one dual meter, or I blow the whole budget on a pair of them. Regardless of whether I get just one meter or if I get the pair of dual meters, I'll have either a pair of the single meters, or possibly even 4 single meters, that I simply won't be using anymore... 

What to do, what to do... 

So...
If anyone is interested in my single meters, feel free to PM me.

I'm not really sure what to do with them other than try to pass them on to another interested party. If no offers, I'll just put them up on ebay. I can make custom scales for them. One meter is already converted to be backlit (and will require custom scales to be made). The other three meters can still be converted, or left as is. They require an adapter to drive them from an Arduino, but those are cheap on ebay. I could order the correct adapters, or even build them, and verify everything works. Hmm... I'll also need to get a 24 volt power supply to include with them. Again, I don't think I'll be needing the single meters anymore, so if anyone's interested... PM me.

**I have been PMed

That being said, someone just dropped 6 of the same style meter (General Electric 180 Edgewise Meter) on ebay for about $25 + shipping... That's not a bad price, actually! Even better if you can get a good combined shipping deal, since it's all the same seller.

Edited October 17, 2016 by richfiles

[richfiles]

**UPDATES**
I managed to snag the better (priced) of the two meter deals! I got the pair of meters for $42 each, $97 shipped. What's more... They look like the rather uncommon factory backlit models! Sweet! That's two meters I won't have to mod! I can also see how they do their light distribution, and maybe mimic it when I install the lighting on my own! Had to turn down the counter offer for the $130 shipped pair. Those would have been the easier meters to use, as they take 0-5 volt DC inputs, aka... Arduino voltage levels. The cheaper meters are 4-20 mA current loop meters, so they will need to be driven by a driver circuit, but that circuit is only about $4 on ebay, and I have the parts I could just simply build it with too. It'll depend on how lazy I get! 

Anyway, It looks like I'm getting rid of three of my GE 180 style Single Edgewise Meters. I'm keeping the one I modded. I'll likely redo it's scale to be atmospheric pressure, since the single meters have a large scale space, and that meter is mostly visual in KSP. I'm debating how I want to handle this. One of my dual edgewise meters has a broken pointer arm. I repaired it, but the arm moves slowly due to increased mass from the repair. The irony, is that I now have more meters than I actually need now... I'm actually looking for data in the packet that even needs to be shown! 

I can do a couple different things with the meters I have:

A: I can install the single and the repaired dual meters in my overhead instrument panel. (2x extra readouts though)
B: I can install the single and the repaired dual meters in my overhead instrument panel, but remove the broken pointer arm, running the dual meter as a single meter. (1x extra readout)
C: I can run a single dual meter up top for the two atmospheric readouts, and the single meter below for Temp. I would not use the repaired meter at all. (correct number of readouts)

Either way, thanks to the presence of the GS meter in my navball (Glide Slope), which has divisions that match the "Gs" meter in KSP... Yeah... I kinda sorta NEED to be using that one! How could I not!  Well, It actually potentially leaves me with extra analog readouts, based on the data sent in the packet. Personally, I'm liking option B, but also totally cool with option C. I just really don't like the repair on that one meter. 

//6   A0︎⃣ vertical velocity (m/s) - (Round Vertical Velocity meter)

//21 A1︎⃣ Liquid Fuel remaining - (Default readout A on Dual Meter 1)
//23 A1︎⃣ Oxidizer remaining - (Default readout B on Dual Meter 1)
//35*A1︎⃣ Liquid Fuel remaining (stage) - (Alternate readout A, selectable by toggle on Dual Meter 1)
//37*A1︎⃣ Oxidizer remaining (stage) - (Alternate readout B, selectable by toggle on Dual Meter 1)

//25 A2︎⃣ Electric Charge remaining - (Readout A on Dual Meter 2)
//27 A2︎⃣ Mono Propellant remaining - (Readout B on Dual Meter 2)

//31 A3︎⃣ Solid Fuel remaining - (Readout A on Dual Meter 3)
//33 A3︎⃣ Xenon Gas remaining - (Readout B on Dual Meter 3)

//48 A4︎⃣ Max part overheat (% percent) - (Readout A on Dual Meter 4...)
... ... I wanted this to be a dual meter, but what else can I show for readout B? If I swap the single meter and this dual, I get option C, but I mix meter styles.

//29 A5︎⃣ Intake Air remaining - (Readout on overhead converted *Dual Meter 5.) *Broken pointer removed to convert it into Single Meter
... ... If I swap meters 4 and 6, I can remove this meter entirely

//13 A6︎⃣ air density (presumably kg/m^3, 1.225 at sea level) - (Air Density Readout on overhead Single Meter)
... ... I wanted this to be a single meter, but if I swap the dual meter and this single, I get layout option C
... ... With layout option C, Readout A will be Air Density, and Readout B will be Intake Air)

//9   A7︎⃣ acceleration (Gees) - (GS meter on left side of navball)


So that's where I sit... I'm kinda out of 0-100% based packets that seem worth displaying. The remaining values are all numeric values, like velocity and such, or total values. They don't translate to a meter... I honestly think I'm gonna go with option C. It uses every single meter, and keeps all air related readouts on the overhead panel. It also lets me eliminate the meter with the one broken (and poorly repaired) pointer.

Side bonus to arrangement C... If I am only using 4 dual meters and 1 single meter... the two duals in the mail are factory backlit, I have one single and one dual backlight modded... That means I literally only have one meter left to backlight mod, and I'm done with backlighting them all! Woot!  I amy still have to do a minor mod to the two backlit meters, but it's a mod I can likely do with just a side cutter and a file.

3 duals + 4 singles = 10 readouts
4 duals + 1 single + navball GS meter = 10 readouts! No loss in readouts, and no unused meters!

Edited October 17, 2016 by richfiles