Tuesday, December 6, 2016

So lets talk about Batteries. Vol 1.

So let’s talk about batteries.
Many things about batteries depend on each other. Most people flying race quads are going to be flying 1300-1800mah 3 or 4s battery packs. I’m going to assume you know what mah are, and what 3s and 4s mean.
Chemistry is what really limits our batteries. Reaction area, and speed of the atoms and molecules. Reaction area, for LiPo’s works out to “physical size”. Speed of atoms is “literally” the definition of temperature. The warmer something is, the faster that reaction can happen.
LiPo batteries are built of thin layers deposited on big sheets of polyester. (I think it’s polyester) Some Lithium, Some carbon, a membrane that’s impregnated with an electrolyte. Those sheets get rolled up or folded up into the shapes we’re accustomed to seeing. If you’ve ever seen a bunt up battery, you can see those layers peeled apart. Durable batteries have thicker layers. Durable batteries have higher C ratings. We’ll come back to C ratings in a few moments. Batteries with more capacity, have bigger sheets of battery. These directly relate to the size of a battery pack.
For my 9xr ratio, I had a 2c 2200mah 3s pack, that was smaller than 1500mah 20c 3s pack. A 45c 1500mah 3s pack, is about the same size as a 2200mah 20c pack. This is due to the tradeoff of cell area, and cell durability. “Stronger cells” are thicker.
Now we need to talk about abuse. And abuse is something that we do ~very well~ in the multirotor community. Tiny batteries, steep pitch, and multi blade props, four 20-30-40 amp ESCs and 2000kv+ motors. (High Kv essentially means “low turn” for the r/c car guys.)
Battery packs beyond their capacity and number of cells, are rated on the “C” rating. C is “the capacity of the battery” The “C” rating is a multiplier of the capacity of the battery. This lets you do some math to figure out how to treat, and what is considered abuse of the battery. There’s going to be a separate charge rating for both charge, and discharge. Typical limits for charging are in the 2-4C range. That’s 2-4 times the capacity of the battery.
For instance, if we have a 1300mah battery pack, we could safely charge it at 2.6 to 5.2 amps. Slower charging is always better for a battery. To a limit… I don’t have any research handy, so I’ll say you don’t want to charge LiPo at less than 1/10C. This relates back to our bag of chemicals, the slower we charge, the more time atoms have to do their thing. The slower we ~need~ them to move, the less likely they are to damage their surroundings. If you exceed the speed that those chemicals can react, ~things~ start happening. For instance, electrolysis.
Electrolysis is the splitting of chemicals via electricity. Generally, it’s water, and that’s how we get puffed packs. When you push a pack to hard, and you get water splitting into H2, and the O2 then oxidizing other parts of the pack, you end up with a puffy pack, and less reaction area due to oxidation.
Let's stick to abuse. Charging, or discharging quickly, causes heating. Moderate heat can be a good thing. Batteries can “have things done to them” faster, if they’re warmer. To a limit, either by generating vapors, or, let's say, fire.
When charging, or more importantly, discharging, heating comes from within the pack itself. While we’re used to the idea of batteries providing power, they also consume some power. All parts of an electrical circuit, have some resistance. That includes LiPo batteries. This resistance is called “internal resistance”. For LiPo batteries used in commercial situations, which frequently see ~very~ cold temperatures, the procedure for starting involves a two step process. First, is “knowing” the cart/truck/plane won’t start, and applying the starter for a few seconds. This puts a high load on the battery, warming the pack. Then, you wait a few seconds, and try again. The engine would start on the second attempt, because the warmer battery could provide more current without the voltage sagging under load. Some r/c car people have taken to preheating their LiPo packs with some very high C charging. We’ll come back to voltage sag.
Now i’m making it sound like batteries being warm, is good. Unless you’re having peak current issues, it isn’t. Batteries age in relation to their temperature. The warmer they are, they faster they lose capacity and C rating. It’s better to keep batteries cool. Cold even. But not frozen, as that can cause the electrolyte to freeze and directly damage the cells from the inside.
In the past, the internal resistance of batteries were there to save us. You could short out most NiCad, or NiMh batteries and nothing would happen. They’d get warm, and finish discharging. No ~real~ drama, excepting some of the very latest NiMh cells, funny, that was discovered by the r/c car guys too. Maybe we should stay away from r/c car people. Do the same with LeadAcid, and you might get some steam… Their discharge would be self limited by chemistry and internal resistance. This, is not the case with LiPo. The internal resistance of LiPo packs are often in the miliohm range. This means we need to do current limiting somewhere else. Sometimes we don’t….
And when we don’t, because a lot of us use OSD’s, we’re aware of “voltage sag”. Voltage sag is what happens when we get ahead of the chemical reactions that restore the voltage of our batteries. Voltage matters to us, because our motors are a long series of wire coils. And the speed at which we can energize those coils, is directly related to the available voltage.
Back to cell abuse, deep discharge of LiPo batteries is a problem. When the voltage is ~very low~ in a LiPo cell, some of the copper used in the battery can dissolve in the electrolyte. This makes it more conductive, and makes it more prone to runaway. It’s also irreversible.
So how do we stop our packs from dieing? Keep them somewhat charged, so they don’t dissolve into themselves. Keeping a battery charged also has it’s downsides, when you charge a battery, you’re forcing the battery into a state it does NOT like being in. That is, it holding the positive and negative poles as far apart, electrically, as it can. So the lower voltage you can store it at, the better. But batteries also self discharge, so you need to have “some” buffer. The generally accepted voltage is 3.6v, for storage. Finally keep them cool when stored. Higher temperatures increase self discharge rate, and the damage the cells do to themselves while stored.
I ~think~ I have all of that right. I didn't cover balancing or cell choice. But that gets into an even longer discussion.

Thursday, September 1, 2016

A little more Doink(er)

I made some more progress on the Doinker.  In fact, short of the props, it's complete.

There were a few steps leading up to that though.  I had to populate the PicoBlox.  The first thing I added was the power stub for the camera.

I really didn't get any pictures of modifying the receiver tray, and the spektrum satellite.  It's stuffed under the flight controller.  I have thick spacers under the FC to provide clearance for the satellite wire.

It's not the most pretty solution, but the "wedge them in" ESC mounts is quite elegant.

16ga wire is a little big.  Here I'm trying to get the receiver bound.  

Happily, I only got one motor wired backwards.  That's a first.  I've had as many as three of four spinning the wrong way.  The wires on the 1104 motors have varnish on them, so when you shorten, and strip them to attach to the ESCs, sand the wires a little, or else your flux and solder just ball up and fall off.

In the end, I swapped the 16ga lead for 20ga going to a JST plug.  Most of my multi cell lipos are either e-flite 3pin, or T-connector.  So I built two adapters to I can use all three kinds of battery without any real strain.

Next is to order a few batteries, and cut down some props.  

Adding FPV to a normal Micro Radian

E-Flite released a UMX Radian with a FPV camera on it.  I thought this was a great idea, but I really didn't like the idea of spending $99 on top of the cost of the glider for a $40 camera and a $7 mount.  

So the first step to adding FPV to a micro plane, is to get power for the camera.  That means cracking open your bird.  The canopy is held on with "just" tape.  The tape to use for putting it back together is Scotch brand 1/2" clear tape.  The glossy stuff.

The UMX radian has a additional speed control board.  There's three pins that feed it, and my pencil is pointing at it.  

The center pin is battery positive, and the pin nearest the sockets is ground.

I bought a bunch of eflite mini JST plug extension wires.  So I cut one in half and slipped it through the canopy.  To put a hole in the canopy, use a normal 3/32" drill bit and spin it in your fingers.

I taped the wires to the inside of the canopy.  Then I soldered the wires to the reciever.

And there we go, a plug on the top of my glider to power a FPV camera.  The factory FPV camera mount is $7 from E-Flite.  Or you can just hotglue your camera on.  

Thursday, August 25, 2016

Another duck day, and an even bigger project.

It's a giant steel box.  That's the bigger project.  Two weeks ago TheSwitchElectrician came by and we built the platform that the shed stands on.  The platform took us about 4 hours start to finish.

The next week, Dan came by, and he helped build the shed.  Actually building the shed took about five hours.  Shockingly long for a bunch of tin.  

It's pretty roomy inside, and looks to be quite weathertight.  I'm going to need to do something about covering the small gaps at the top, without impeding airflow.  Perhaps stuffing them with steel wool?  

Both hulls are really in the stage for sealing.  With our truncated build day, Dan and I spent it doing epoxy work.  Every screw hole needs to be filled, and due to the quality of wood we've got some oddball gaps at coerners and edges.  

I work at a telecom company.  One of the smocks Dan brought had a fun brand name on it.  

The hulls look really good after being sealed.  The lack of screws somehow makes me a lot more confident in the seaworthiness of the hull.  

One last shot.  I think these are just so pretty to look at.  

Next time, we're doing keel trunks, and perhaps mast steps.  

The desk is clear. That can't stand. A Doinker build.

My desk is a place of projects.  When it's clear I start seeking for something to do.  Two nights ago, it was clear..

Yes, I really do type with my keyboard that far back on the desk.  My wrists are very healthy thankyouverymuch.

I ordered my Doinker kit from 65drones.com.  Here's what came in the package:

In there we have the frame kit, four escs, four 1104 motors, a 40channel micro FPV camera/transmitter, and a flight controller with built in PDB.  Oh, and that giant awesome sticker.

It seems like so much less once it's unpacked.  The frame is 3d printed, and has the usual problems with 3d printed things.  It's also been sprayed with some sort of clear laquer, which makes it nice and shiny, and fills in a lot of the layer artificats from being FDM printed.  

The first thing they have you do, is install the reciever, in the reciever tray.  They expect you to go frsky, and my house is not a taranis house.  I'm going spektrum, so their tray doesn't exactly match what I'm going to install.  We'll cover the reciever install in the next installment.

The build directions on theshortcrayon.com are really pretty good.  But they don't cover where these little spacers go.  

They go under the 1104's, for prop clearance, I think.  But that's where the spacers are on when you look at the pictures on 65drones and the shortcrayon.  

I got "a little" further than this, but we're now to the point of needing to solder everything up.  I'll be doing that this week, and hopefully getting it's first flights in.  

Monday, August 15, 2016

Puddle ducks.

Sometimes, projects need to get big.  Most of my projects are small.  This time.. it's big.  like 4x8 foot.  Big.

There's something magic about that first time you sit "in" your boat.  Here's Dan, sitting IN his boat. I should also note, that this was day two of boatbuilding.  I think.

It doesn't look like much, well, it isn't much.  But it is a puddle duck.  http://www.pdracer.com/

Sadly, this set of blog posts is going to be somewhat out of order.  But here we go anyway.

The Puddle Duck is a class driven sailboat.  They tightly regulate the hull shape, but not it's height.  And as long as you don't have foils or submerged balast, essentially anything else goes.  People have run giant sail rigs.  The boats have gotten on plane.  They have been used as overnight sailing boats.  They've been built 8" tall, and 24" tall.

On that specific note, our hulls are 12" deep.  Which is fairly shallow.  Both Dan and I don't mind getting wet.

We've chosen to go the 3/16" thick underlayment method for the hulls.  The underlayment is made with waterproof glue, so is suitable for what we're doing.  We're also doing chine logs, instead of stitch and glue.  I'm sure this is the cheaper method, but i'm also sure it's more time consuming.

The first day, we cut out the hull sides.  As pre-game, Dan cut up the chine logs for us.

For gluing up the hulls, we're mostly using titebond 3, and a whole lot of 1" long screws.  The screws are removed after the glue is dry, and we're filling the resulting holes with thickened epoxy.  You can seriously bulk up epoxy if you buy some very, very, cheap materials.  I think I spent $40 on sillica and Microbaloons, and ended up with more than a gallon of additives.

Later that day, we got the bottom on the hull.

You can see a fillet we screwed to the aft end of the hull to keep the sides square while we got the bottom glued and screwed in place.

As usual, when you get in that building groove, pictures stop taking precedence.  Two building days later, here's Dan working on gluing the bottom on my hull.

I love how epoxy filler looks.  This is how we're waterproofing the tanks, as titebond isn't a good method for closing big-ish gaps.

Here's Dan's hull.  The left side is filled, and some of the holes on the starboard side are filled.  It's going to take two more epoxy sessions to completely fill all of the holes.  

More will follow as the boats get built.

Tuesday, July 26, 2016

Sticky stuff. The glue on my desk.

Last night, I looked at my desk, and it struck me, I have a ~lot~ of glue on my desk.  From left to right:

  • Standard white glue:  I have this for hardening tissue on models.  But it's got it's other uses.  It takes time to tack up, but it's quite useful for building balsa models.  I wonder how it gets along with carbon and kevlar.  
  • Glue Stick:  This is for tissue covered models.  
  • EvoTite: This is foam save medium CA.  Most of my fixed wing stuff, is parkzone foam.  sometimes.. they need to be repaired.  .. often.. they need to be repaired.  
  • Thin CA: I do most of my gluing with this.  I like how fast it lights off.  I like it's wicking ability.  
  • Thick CA: For those joints I might need extra strength in, or might need to re-position once or twice.
  • Thick CA: Same as above, just an extra bottle.
  • Gorilla Glue: Foaming Polyurethane.  This is what you're "supposed" to use on foam planes.  I've used it on those 4' toy gliders, and have been reasonably satisfied, but I won't use it on nice planes.  I really dislike how it foams up, and keeps foaming for hours.  And if you follow Mathias Wendell, you know that it's weaker than normal glues.
  • CA Accelerator: For locking down joints.  I've still never used it...  
  • Testors Wood Cement: I still have no idea how this works.  I'll be trying it sometime.  But it requires a good building board and pins to keep things lined up, for the time being I'm doing most of my building on my computer desk, so I can't dedicate desk space to a pinboard.  
I've got some other adhesives around, but they weren't on my desk.  There's a hot glue gun under my desk, epoxy in the toolbox, along with JB weld.  I've got TiteBond 3 in the garage for the boat project.  

Guillows RC conversions, and a DLG nano.

Well that's a pretty kit.  Guillows could do with some packaging updating.  That "laser cut" sticker, is a sticker.  The box, and instructions inside still refer to the kit being die cut.  While we're at it... these kits could use a much better manual for construction.  But we'll cover those complaints a little later.  

Plans are important,  these Guillows kits come with decent plans.  Speaking of which:

So, here we run into my first little complaint.  The detail of the stringers leaves something to be desired.  There's several stringers on the turtledeck of the F6F, and one of them ends at the rear fuselage former.  The only clue to where that stringer ends, comes from the photograph, not from the plans.  

Also, the Guillows planes are built with, well I call them keels.  There's a dorsal, ventral, a left side, and a right side keel. If you look at the plans up there, can you tell me where the side keels are?  I can't.  That came to bite me later.  

This sheet of plans, is worth writing home about.  Everything in place, an in scale.  These are the sort of plans I like to see.  There's nothing really to say about these, beyond this is what you should expect to see.  

The package also comes with some balsa parts.  And the bits to make the plane into a rubber powered plane.  

Now, if you look closely, you might be able to see what's missing there.  I got two A sheets, a C and D.  No B sheet.  When I started building, I started with the fuselage.  The fuselage really doesn't use many parts from sheet B.  

Someone needs to bless whomever started laser cutting balsa.  You get parts with labels, crisp corners, and labels.  

This is where I got on the first night.  Sheet B had the wing leading engines, both side keels for the fuselage, one little bit of the tail, and the wing upper supports.  

I think this was 13ish grams.  Without the cowling, it looks a little funny.  

I was building this on a weekend.  I found that Guillows webpage has a contact us form.  I dropped them an e-mail.  They asked me for a picture of the parts I got, and they mailed me a replacement bit of balsa.  I must give it to Guillows, they were quite fast about getting me those parts.

And Viola!  A F6F.  More or less, this is where the project is stalled.  I'm debating on how i'm going to do the RC setup.  I'm thinking an 1104 motor, and some 2.8g servos.  I think I can get the whole flying package for 15 grams or so.  But we'll come back to that in a bit.  

I also bought some other kits.  The one I chose to build is the Skyraider.  It's going to get my AR6400 receiver and aileron servo.  

The wings are using music wire to drive the ailerons.  I put in brass bushings for the root, and surface pivots. 

I even transferred the aileron bellcrank from the T-28.

The elevator and rudder use a spring return, so I can use just a pull string to get control  I stole that technique from the DLG people.  

Speaking of DLGs, this is my latest version of the DLG Nano.  It's 52 grams.  I finally built one right.

Friday, January 29, 2016

Building an ITX laptop - Episode 1

I want a laptop that's future proof.  If someone walked up to me and told me that, I'd tell them it's impossible, and buy something cheap so you can save money for the next cheap laptop.  This is because there's no reasonable upgrade path on laptops.  (sometimes you get one or two upgrades...  but that's not a common thing.)

Now, what if a laptop took a standard motherboard?  If that were the case, you could swap out the motherboard as technology moved along.  There's a very tiny motherboard standard that really suits itself to portable installations.  ITX motherboards are the core of the shoebox size computer market.  Some of those ITX cases are even the size of a small book.

People have been strapping monitors to PC cases for decades. And LCD's are now ubiquitous.

Really, that just leaves the power supply as the big issue.  There's no good off the shelf power solution for a portable ITX machine.  And i'm not just looking for "portable."  I want some decent runtime, and maybe even actually to put this in my lap.

I've been thinking about this project since the early 2000's.  I've even bought parts from time to time to support the plan.  (I've got a folder with Ti's simple switchers in 3.3, 5, 12, -5 and -12v..)

This leaves us needing to make a few decisions.

  1. Electrical:  ATX power supply, battery charging, battery, display power supply, power reporting to the OS.
  2. PC components: keyboard, monitor, motherboard, mouse, wifi.
  3. Case design: Can it all fit?  Can it call be cooled?  Can it all be easily accessed?  Does it look good?
Next time, we'll start in on the PC components.