Drone Parts

I want to document this here mostly for myself, but also because when I was starting out I kept finding build logs from people who made everything sound easy, and I felt like an idiot when it wasn't easy for me. So this is the honest version.

The Parts I Chose

I went with a pretty standard beginner-friendly stack because I didn't want to overcomplicate things. Here's what I ended up with:

• Frame: iFlight Nazgul5 V3 — picked it because it's beefy and replacement arms are cheap
• FC/ESC stack: SpeedyBee F405 V3 35A stack — this combo gets recommended constantly and there's a ton of support for it online
• Motors: XING2 2306 1700kv — probably overkill for a first build but they were on sale
• Camera: Caddx Ratel 2
• VTX: Rush Tank Solo
• Receiver: ExpressLRS 2.4GHz, the BetaFPV Nano RX
• Props: HQProp 5x4.3x3

Total came out to around $280 not counting the radio and goggles, which I already had (Radiomaster Boxer and some used Skyzone SKY04X goggles I got from the classifieds section on RCGroups).

Where It All Went Wrong Immediately

I'm not a complete solder newbie — I've done hobby electronics stuff before — but I had never soldered to an ESC board and oh my god. The solder pads on the SpeedyBee are fine but I kept getting cold joints on the motor wires and I didn't realize it until I had the whole thing buttoned up and motor 3 was just... not spinning consistently in BLHeliSuite.

I had to desolder and redo that motor connection three times. The third time I finally used a bigger chisel tip on my iron (I have a cheap TS100 clone) and actually let the pad heat up properly before adding solder. Every tutorial says this but it didn't click until I was sitting there frustrated at 11pm with flux residue everywhere.

Also — and this is embarrassing — I forgot to route the antenna for the ELRS receiver before I closed up the frame. So I had to take it partially apart again. Please learn from my mistake and do a dry fit of literally everything, including wire routing, before you start soldering.

The Betaflight Setup Was Less Painful Than Expected

I was dreading Betaflight because everyone talks about it like it's this deep dark rabbit hole. And it kind of is, but for just getting it flying, you really don't need to touch that much. I loaded the SpeedyBee F405 V3 preset in the Betaflight configurator, set up my ELRS receiver on UART1 with serial RX enabled, bound it to my Boxer, and mapped my channels. Done.

I used the RPM filter because I have BLHeli_32 ESCs and people swear by it for reducing noise. Left most of the PIDs at default for now because I figured I'd tune after I confirmed everything worked.

Motor direction took me a second — you do it right in the Motors tab now which is way less confusing than it used to be apparently. I had two motors spinning the wrong way and fixed it with motor direction override in BLHeliSuite rather than swapping motor wires, which I've seen people do both ways.

First Arm — Indoors Over Carpet

No props, obviously. Everything spun up, motors all responded to stick input correctly, and the quad orientation in the Betaflight configurator matched real world movement. That little moment where you tilt the quad and the 3D model on screen tilts the same way is genuinely satisfying. It means things are working.

Put props on, went outside to a flat patch of gravel in my backyard, and did a slow hover test. The thing lifted off way more aggressively than I expected even at low throttle. I had throttle expo set to zero and I think that's part of it — I've since added some throttle mid curve in Betaflight and it feels more manageable now.

First flight was maybe 45 seconds before I chickened out and landed. No crashes, nothing broke, I was just nervous. Second battery I flew for almost three minutes and it felt incredible.

Issues I'm Still Dealing With

The VTX setup is a little annoying. The Rush Tank Solo works fine but getting into the menu to change channels requires the button on the VTX itself, which is buried in the frame. I'm probably going to do VTX control through Betaflight once I figure out which UART to put it on. Not a huge deal but mildly irritating.

I'm also getting some video static when I throttle up hard. It's the classic power noise issue. I've got a 35V 1000uF low ESR capacitor on order to solder to the battery leads and hopefully that cleans it up. This is apparently super common and I've seen it solved with caps almost every time.

The Ratel 2 camera has a small bright spot in the top corner of the image that bugs me. Could be a defective sensor, could be something about the way I have it seated in the frame — I'll mess with it more before I call it defective.

What I Learned

A few things I'd tell someone earlier in this process:

Get a good soldering iron. The TS100 clone I have is fine but I'd just buy an actual TS101 or a Pinecil next time. The temperature accuracy matters more than I thought when you're working on small pads.

Watch at least two or three different build videos for any component you're unfamiliar with. Different people show different things and some things only click when you see them from a slightly different angle or with different commentary.

Don't rush the wiring. I rushed it. I paid for it with two full teardowns. Slow is smooth, smooth is fast, whatever the saying is.

And honestly — just start. I put this off for way too long because I was scared of doing something wrong. I did several things wrong. The quad flies anyway.

I'm going to keep posting here as I tune it and eventually try to get into some casual group flying. There's a local group that does a Sunday morning thing at an airfield about 40 minutes from me and I'm trying to work up the nerve to show up. Maybe once I stop overcooking my takeoffs I will.

If you've done a similar first build and have advice on the cap mod or the VTX control setup, drop it in the comments. I could use the help.

Some background: I've been flying a Betaflight-flashed Eachine Wizard for about a year, which is a fine little quad but I bought it pre-built and never really understood what was going on inside it. I wanted to actually build something from scratch so I could stop being helpless every time something went wrong. That goal was achieved. I am no longer helpless. I am just confused in new and more specific ways.

What I Ended Up Buying

I went back and forth on the frame forever. I almost pulled the trigger on a Diatone or a TBS Source One, but I ended up with the IFlight Nazgul Evoque F5 frame because someone in the Discord I'm in said it was beginner-friendly for building and the arms are replaceable without pulling the whole thing apart. That last part turned out to be relevant. More on that in a minute.

Here's the parts list I ended up with after way too many Reddit threads:

• Frame: IFlight Nazgul Evoque F5 (true X layout)
• Flight Controller: SpeedyBee F405 V3 stack (FC + 4-in-1 ESC combo)
• Motors: Emax ECO II 2306 2400KV
• Props: HQProp 5x4.3x3 (came with a bag of four, broke one on day one)
• VTX: Rush Tank Ultimate (I maybe overkilled this but I wanted range)
• Camera: Runcam Phoenix 2
• Receiver: ExpressLRS via a BetaFPV SuperD diversity receiver
• Battery: Two CNHL 1550mAh 6S packs

Total came out to just under $340 not counting the radio and goggles I already owned. That's probably average for a first serious build? I honestly don't know. It felt like a lot when I was clicking buy on everything but it also felt worth it.

The Soldering Situation

This is where things got humbling fast. I have a cheap Hakko clone iron that I've used for hobby stuff before, so I figured I was fine. I was not fine at first. The solder pads on the ESC board are smaller than I expected, and my first attempt at tinning the battery leads looked like something a raccoon did. I watched a video from Joshua Bardwell about soldering drone components specifically and that helped a lot — his tip about using flux paste even when your solder has flux in it already was something I didn't know and made a real difference.

I burned through one of the ESC motor pads on my first attempt. Not all the way through, but I lifted a little bit of it and panicked completely. Ended up using a tiny bit of thin wire soldered to the pad trace as a workaround and it actually works fine, but that 45 minutes of panic was not fun. My iron was running too hot and I was holding it in one spot too long. Once I dropped the temp a little and worked faster it was way cleaner.

The worst part of soldering was the capacitor leads. The legs are so short and the cap is so close to everything else. I got it done but I'm not proud of how it looks. Functional, not pretty.

Wiring and Stacking

The SpeedyBee stack is honestly great for a beginner because it uses plugs for most connections instead of soldering everything direct. The motor phase wires still need to be soldered but everything else — the VTX, camera, receiver — uses JST or little ribbon connectors. That was a relief. I spent probably two hours just figuring out cable routing so nothing was going to get into the props or pinched by the stack when I tightened everything down.

I used a combination of zip ties and some silicone wire loom I had leftover from a car audio project years ago. Looks a little janky in spots but it's secure. I've already seen it take a crash and nothing came loose so I'll take it.

Betaflight Setup

This took longer than the physical build. I'm not joking. Betaflight Configurator is not unfriendly exactly but there's a LOT going on and I spent a whole evening just figuring out which UART my receiver was on and why it wasn't showing up in the Receiver tab. Turns out I had the serial port set right but forgot to actually enable Serial RX on that UART. Classic.

I used a preset from the Betaflight presets tab for my motor config and then just started with stock PIDs. Everyone says don't try to tune until you've flown it and I'm taking that advice. I set up my rates by copying a profile I saw posted by someone whose flying style looked similar to what I want to do, which is just general freestyle, nothing crazy. Adjusted failsafe to drop mode because that feels safer for where I'm flying (open field, no people around).

Motor direction: took me embarrassingly long to realize one of mine was reversed. Props off, motors tab in Betaflight, fixed it. Easy when you know where to look.

First Flight

I did props-off motor checks first, then a quick hover in my backyard just a couple feet off the ground. Everything spun up, the quad leveled, ELRS link was solid. I was shaking a little. Then I went to the field.

First punch out I gave it too much throttle because I forget that 6S hits different from my old 4S Wizard. It shot up and I immediately rolled into a bando wall. Broke an arm. Which is why the replaceable arm thing was important. I had spare arms because I read someone say to order two extra sets and I'm very glad I listened to that advice.

After swapping the arm (which took like ten minutes, super easy on this frame), I flew more cautiously and had a genuinely great time. It's smooth, it's powerful, and I built the thing myself which makes it feel completely different to fly.

What I'd Do Differently

Buy a proper flux pen from the start. Don't cheap out on solder — I bought Kester 63/37 after my first session and it's so much nicer to work with than the random spool I had. Take photos of your wiring at every stage so if something goes wrong later you know what was where. And honestly, don't rush the Betaflight setup. I rushed it and paid for it in time later.

I'm already thinking about what I want to build next. Probably something smaller, like a 3-inch toothpick style for flying in tighter spots. But first I need to actually get good at flying this one. More updates to come.

I want to write this down while it's still fresh because I made a bunch of small mistakes that nobody really warns you about when you're studying for your Part 107 and watching YouTube tutorials about commercial work. The broad strokes are covered everywhere — get your license, get insurance, have a preflight checklist. Fine. But the actual texture of doing a real job is different from any of that prep work.

How It Came Together

A property management company that a friend's dad works for needed aerial documentation of a flat commercial roof. They'd had some water intrusion and wanted photos to send to their insurance adjuster along with photos from the roofing contractor on the ground. Nothing fancy. No thermal imaging, no 3D model, just clean stills showing the whole roof surface, drains, HVAC units, and the areas around the parapet walls where the flashing looked questionable.

My friend basically vouched for me, which meant the pressure was both lower (they weren't expecting a production company) and higher (I didn't want to embarrass him). We agreed on a flat rate for the job including a basic edited photo package. I'm not going to publish the exact number but it was in the range you'd expect for a solo operator doing a first commercial gig in a mid-sized metro area — not premium pricing, but not embarrassing either.

The Prep Work Took Way Longer Than I Expected

I'm used to flying recreationally where I check the weather, throw my bag in the car, and go. This was completely different. I think I spent four hours on the prep for a job that took maybe 90 minutes on site.

First thing was airspace. The building is near a regional airport, well within the ring I needed to check. I pulled it up on and then cross-referenced in the FAA's DroneZone to confirm whether I needed LAANC authorization. In this case the ceiling at that location was low enough that I did need to go through LAANC. I used Aloft to request it, got auto-approved for the altitude I needed, and made sure I had the authorization documentation on my phone before I showed up.

Then there's the actual shot planning. I flew the mission manually rather than using any automated flight path software, but I still sat down the night before and sketched out roughly where I wanted to position the aircraft for each pass. Flat roofs sound simple but there's a lot to think about — sun angle, shadow from the HVAC equipment, keeping the lens mostly nadir for the documentation shots versus adding some oblique angles to show the parapet condition. I also looked at Google Maps satellite view of the building so I wasn't orienting myself from scratch when I got there.

I checked the weather obsessively. Wind at 10 feet is different from wind at 200 feet, and roof inspection work really does require stable conditions if you want sharp photos. I had a backup date ready to go in case conditions weren't usable.

On Site Reality Check

A few things caught me off guard once I was actually there.

The roof had a lot more HVAC equipment on it than the satellite image suggested. Some of it had been added recently and wasn't in any overhead view I'd seen. That changed my hover positions for a couple of the shots because I had to account for the rotor wash potentially affecting the lighter sheet metal covers on some of the units. I wasn't worried about damage — the drone wasn't that close — but it made me more deliberate about positioning.

There was also a person on the roof for part of the flight. A maintenance guy showed up about 20 minutes in to check something unrelated. I landed immediately, talked to him, made sure he knew I was operating, and waited until he was back inside before I resumed. This is just the right thing to do and honestly it wasn't a big deal, but I hadn't mentally rehearsed that scenario and for a second I kind of froze before the obvious answer (land, talk to him) became clear.

The contact from the property company had asked for "a lot of photos" without giving me a specific number. I ended up with about 85 keepers after culling in Lightroom. I wasn't sure if that was too many or not enough for an insurance documentation package. In the end I delivered 60 organized into folders by roof section and they seemed happy with it. But I wish I'd asked upfront what the adjuster actually wanted to see — file count, resolution requirements, whether they needed any specific metadata embedded. I just assumed and it worked out, but that was luck.

The Gear I Used

I flew my DJI Mini 3 Pro for this one. I know some people will say you should show up to commercial work with a more serious aircraft, and I understand that argument, but the Mini 3 Pro does genuinely excellent image quality and for a roof documentation job the resolution and dynamic range were completely sufficient. My main concern was wind — it was forecast at around 12 mph with gusts to 18 mph at ground level. The Mini 3 Pro handled it fine for the most part, though I could see it working harder on the longer hover positions. If the gusts had been higher I would have rescheduled.

I also brought a spare set of batteries (flew on one and a half of them total) and my tablet setup instead of just the phone. The bigger screen makes a real difference for checking image sharpness in the field before you pack up and drive away.

What I'd Do Differently

Have a proper client brief document ready to go before the job even starts. I've since made a one-page intake form I'll use for future clients that asks about deliverable expectations, required file formats, deadlines, any site access specifics, whether there will be people on or near the site, and any known obstructions. It takes five minutes to fill out during a call and would have saved me some uncertainty on this job.

Also — and this is a small thing — I should have confirmed parking. I ended up circling the area for ten minutes looking for somewhere to put my car that wasn't going to get me towed or block a loading dock. Not the end of the world but it was an annoying way to start a job where I was already nervous.

Overall

It went well. I delivered the photos, they were happy, and I didn't screw anything up badly enough to make my friend regret recommending me. That feels like a decent bar to clear for a first paid gig. The money is almost secondary to the experience of actually doing the workflow start to finish for a real client with real expectations.

I've got another potential job in the pipeline — similar kind of documentation work, different type of structure. I'll write about that one too once it actually happens. Assuming the weather cooperates.