Sorry this has taken a while to update, had some issues and have been awaiting feedback so I know which direction to head in next.
After choosing the parts (part 1) and completing the build (part 2) I've added some extra parts to make this more suitable for racing. Namely a ufl/ipex to right angle sma female connector and a RHCP pagoda antenna. A circular polarised antenna is compulsary for my local multiGP chapter running on 25mW transmit power. This has bought the dry weight up to 250g with the Dalprop t5045c cyclone propellers.
Initial flights: So responsive
I was hoping to report on my first race but unfortunately bad weather had meant it was cancelled. The weather did clear though and so I got the chance to discharge my batteries the RIGHT way.
Quad: 245g dry weigh as per part 2 build plus pagoda antenna, Dal 5040V2 instead of t5045c cyclone
Battery: 4s 1000mah Turnigy Graphene 65c 145g
Pids were fine stock and no excessive motor heat from the dynamic filtering with all notch filters off. Went for some slower light props because I was flying in a smaller space. Combined with a 2450kV motor this meant I didn't have a massive amount of top end speed but wow the pick up was obscene. Between the light dry weight, lower capacitor batteries (light) and the light props that are very easy to spin I found it super responsive and so very easy to change direction at pace. My goal is agility and manauverability while I become more familiar with racing and I wasn't let down at all.
I'm really enjoying the lightweight for another reason too. I am using a micro FS82 flysky receiver with no diversity as a placeholder until my new frsky setup arrives. This reciever is fine for a close-range micro quad like the leader 120 but I got carried away with the floss 2 and failsafed about 20m above tarseal. The quad went down and I lost video signal. After trying my hardest not to run over I was surprised to see everything intact. I only lost video because the ufl connector popped off (which I then popped straight back on. There was also a small scratch on the prop nut and a kink in the pagoda antenna - it seems the antenna had taken most of the hit. I was obviously lucky on the angle it landed but think the lack of inertia through lower mass had helped too. Not even a bent prop!
I put 7 packs through that day and on the second to last I re-armed after a nothing crash and found motor 1 not idling. I unplugged and re-plugged the battery and got that horrid smell of burnt electrical circuit. I powered down once more, waited 5 minutes and re-powered with a new battery. Got idle this time but when I went to throttle up, it dipped on motor 1 before the flight controller took over and limited power to the other 3 motors in order to maintain attitude. Either the motor or more likely the ESC was sick.
Usually when an ESC goes it is catastrophic and it passes no current whatsoever or if it does, has major desyncs and brownouts. Mine behaved differently. To see that the issue was still present I strapped the quad to the bench with all props free to rotate. I then connected a full battery and betaflight configurator so I could power up 1 motor in isolation at a time. Motors 2, 3, 4 were all able to generate maximum thrust but motor 1 was limited to about 30 or 40%. I then swapped the motors attached to ESC 1 and 3. After repeating the test the issue was ties to ESC1 meaning the fault was with the ESC. Not a great outcome since this is a 4 in 1 esc.
Dealing with HGLTech and looking at other options
My first though was that the ESC is just not capable of managing the power for a 5" build. The 20mm x 20mm stacks were originally for micros and although this one is supposedly capable of 28a per esc, the rating may have been overly generous. This got me thinking of expensive replacements like the Asgard V2 and other ESC/FC combos since this FC cannot easily be reused with different ESCs.
For the sake of research and seeing it though I got in touch with HGLTech, makers of this unit. They responded the following day recommended that added capacitance should help manage dangerous voltage spikes and reiterated that the ESC should be capable of 28a continuous and 35a burst. After sending pictures of the build (including the capacitor) I came straight out and asked if it was a manufacturing fault and if so, could they send a replacement. To my pleasant surprise, they said yes. Really I should have gone through the retailer (Gearbest) here but really wanted to get the technical explanation more than just the free replacement ESC. Credit to HGLTech, they got me both!
Simple really, I'll wait for the replacement ESC to show up, wire in 2 capacitors this time (with 1000 microfarad total capacitance) and try again. If it fails it is more than coincidence and the technology is not quite there yet. If it succeeds then it supports the seachange of miniaturisation in racing quad electronics. I'll make another part to this blog series once I've had a few packs through the replacement escs and hopefully make it racing.
Update Jan 2018
A quick update, since installing the backup 428 esc and external capacitors I have not had another issue. From photos of the damaged ESC it was a FET that had burned out. What I failed to mention here is that the ESC burnt out shortly after I had realised the original capacitor I installed was knocked off in the crash I mentioned. My hypothesis is that due to the small size of this board, there is not a lot of capacitance so voltage spikes likely as a result of regenerative motor breaking are causing the damage - hence why the capacitors help.
I still think the this quad is pushing this small stack to it's limit but there are a lot of success stories here: https://www.rcgroups.com/forums/showthread.php?2971570-20mm-x-20mm-FC-ESC-VTX-stack-in-a-5-inch-racing-quad
In the meantime HGLRC has since releasted the XJB F440 tower based around a 40a BLHeli32 esc. I've not tried one myself but this would looks like it may be a better candidate for 5". Link to the F440 direct from HGLRC is below: