Hobbyking H4 - Tuning for better flight & videos

Hi

My name is Sebastian and this is not my own blog. But I asked Peter if he would like to publish a guest entry. Kindly he agreed and on the following lines I want to support quad flying enthusiasts who own a Hobbyking H4 quad. I wan't to show you all some tweeks to tune your own flying H4 a little bit. Just like Peter does so continuously in a great way since more than a year (Thanks Peter!). Most of the tips can be transferred to other multirotor frames I guess. So maybe it is also interesting for different frames.

I have divided the article in two sections. The direct following section is a little bit of my personal history, trying to explain why I can tell you all a little bit about the things I learned and why I'm writing these tuning tips. If you don't like to readd about (my) history, just skip the next passage and got directly to the tuning tips below.

My experience with the H4 frame origins from one year with flying, upgrading, servicing and of course crashing my HK H4. I used my quad at the beginning for gathering experience with just flying. I entered the quad tribe from flying a Blade450 3D helicopter. So I had some 2.2Ah 3s batteries in my inventory. This resulted in initially equipping my H4 frame with 4 NTM2826s 1200kV motors, impressively soft 10"x4.5" 'plastic spoons' from HK, 30A blue seriesESCs and an HK Red MultWii Pro flight controller. Existing batteries slapped together with freshly bought components and voila:

Picture one: Testing the motors on the first version of my new H4

And the best of all: the H was flying!

Of course I was proud like having taught an ostrich how to fly.

But guess which of these components bought at the beginning are left after one year flying and learning:

• The Frame (in most parts)
• The batteries (still flying my heli)
• The ESCs (now with new firmware).

But none of these components are near to the original state they were at the first flight. Further modifications until now include:

• Added two axis gimbal
• Exchanged landing gear
• New motors & props (best decision!),
• New flight controller (second best decision)
• And added/modified many other parts.

But now let's start with the interesting stuff.

After the "Sebastian's history" class , it's time to talk about the main improvements for the quad I did until today.

This specifically means the following topics:

• Getting the quite heavy flying aluminium clump a little lighter without losing too much frame stability
• Reducing EMC (= Electro Magnetic Compatibility) for components mounted on the main frame (i.e. for flight controller, gimbals, sensors, telemtry, radio, …
• Upgrading and tuning propulsion components for squeezing out a little more flight time

So let's start with the first topic. Getting the flying metal hog a little lighter.

I started to slim down my H4, because it felt too heavy when carrying it around. I began to ask myself: where can I  save some weight without weakening the frame integrity too much?

The frame consists of some aluminium bars, sheets of fiberglass, screws and nuts.

Obviously we need the screws and the nuts. OK, we could exchange the steal screws against some made of nylon, but until now I feel a bit safer with the metal ones. I would further use self-locking metal nuts with new nylon if I exchange the screws some time so the nuts still sit safe on top of the screws.

Next option for saving weight would be exchanging the fiberglass decks against some made from carbon fibre. This would be great because CF (carbon fibre) is much lighter and more rigid. But CF is price and I don't have the tools at home to craft this material properly. So this is something to save for a later project.

Last option: the aluminium bars. So how to process these? Easiest option at hand: drill press!

Picture two: drill press vs. aluminium bar

So I first drilled a row of smaller holes, because you don't start with a 14mm drill at metal works. Even when it is only aluminium. When drilling new wholes, take care of the existing ones for the screws keeping the frame together. I have marked the original holes and the positions for the new ones with a scriber. So at the end there are 4 existing holes and 18 new ones.

Picture three: finished with drilling all the small new holes

Sadly I have missed to take a picture from the finished bar with the big holes, but I hope the following picture with the rebuilt frame gives an impression of the finished component. 

Picture four: Drilled bars built into the frame

Picture five: Drilled bars built into the frame

It's a good advise to deburr the sharp rims and chips of the drilled wholes before you mount them back into the frame.

I haven't calculated the weight savings, but I can tell you one thing: my frame feels (with the other savings) quite a bit lighter and it still survived a crash from 13 meter over ground with this aluminium bars just slightly bent in "u" shape. OK there were some other damages, but the frame didn't suffer much.

So after the frame components are discussed what other options are viable to save some weight?

Simple answer: get rid of all stuff you can.

For me this meant moving the ESCs from within the frame to the outer arms of the H. What's the benefit? Removing the heaviest stuff in weight to volume ratio you are flying through the sky: copper!

When you move the ESCs to the outer arms:

Picture six: Escs within the Frame

Pictures seven: Mooving the ESCs to the outer arms

What's the benefit of this modification?

1. Moving a source from very bad EMC interferences which were present along the long 3 phased motor wires through the whole frame to the outer arms of the multirotor.
2. Achieving a way better cooling for the ESCs. This improves the lifespan of the electronics.
3. Saving quite an amount of weight in copper! You don't have to run 3 wires from the frame to the motors, but only 2 from the power distribution board to the ESCs. This means 4 very long copper wires less flying around. Additionally in my case: Saving many many connectors! Because I had every "high power" connection in my H4 equipped with small banana plugs. Simply unnecessary!

And in the end only the wires and the plugs are saving my 73 Gramm. That's the complete weight of my GoPro clone action camera!

Picture nine: saving a lot of weight in copper

So summed up I estimate that I have saved about 100 Gramm in weight. This may not sound very much, but for don't spending any money for new parts I think it is quite OK. And considering that the H4 now has a total weight of 1,6 Kg all sums up to a saving of about 7 percent. That sounds a little bit better than 'only about 100 Gramm'.

So enough weight saved for the moment. Let's talk shortly about the topic EMC.

With removing one of the 3 phased wires running from the centre of the main part of the frame to the outer arms and instead running the 2 permanent DC current wire from the inner frame to the ESCs at the arms, I managed to get rid of some massive twitching at my camera gimbal. You can see some of the twitching in one of my videos (i.e. during the first 20 seconds). And this twitching is a very mild example. I have some videos were the gimbal nearly all the time is twitching around pitch and aileron axis. As the gimbal was influenced that massively I am surprised that the RC receiver and flight controller were working without any hiccups. Maybe I'm a bit too anxious, but it feels saver to fly now that the gimbal is not twitching permanently any more. At the current state of my setup, the gimbal is only twitching during the more extreme manoeuvres (like full stop from fast forward or hard right to left aileron movements). Also for this I have an example video. Twitching can be watched during the last minute of the play time were I tested manoeuvrability in pitch and aileron axis of the H4. So this was the second benefit from slimming down my multirotor.

Now let's get to the last topic of improving the HK H4. Upgrading the propulsion system to get some more flight time with the same batteries. As already mentioned I am using my 2.2Ah 3s 35C helicopter batteries. This is definitely not the best choice for this quad. 3s is OK, but I think 4s batteries may be a better choice. The more worse attribute of my current flight batteries are the 2.2Ah and 35C per battery. This is definitely a too low capacity and too high C-rate. Three batteries get me in sum 6.6Ah and about 600g total weight of batteries. The C-rate should in theory also sum up to 3 times 35, so in total 105C. Even with being conservative with calculating current needs 105C * 2.2Ah are about 200 Ampere which are in theory available from 3 * 3s batteries in parallel. So with 4 motors that leaves about 50A for every motor. Way too much backup with assuming every motor will draw about 20A to 25A max.

For later improvement I plan to get some new batteries. I would cut the number of batteries down to 2, because you will always save weight in cutting down the total number of batteries (cables, plastic housing). And of course you should go up with the capacity per battery. I would probably get some 5Ah, starting at 10C, better 15 to 20C. This would give 10Ah in total instead of the 6.6Ah I am flying now and a total drawable current per motor of about 25A or more. More than enough current and about 50% more capacity at the moment.

The DC current from the batteries will be transformed to 3 phased rotating current for the motors by my ESCs. Originally these were stock 30A blue series ESCs from HK. 30A rating is more than enough for this quad, but I quickly upgraded the firmware of the ESCs to SimonK for better throttle response. Haven't regret that decision until now.

Now we get to the more interesting parts for the thrust. Motors and propellers. You always need to consider both parts together. A too small motor with a lot of KV and a big prop simple would work (too less torque for thrust). Also a big motor, low KV and small prop don't work (too less RPM for thrust). For getting an impression of what motor to prop combination gives you what flight times and performance there are some online calculators available. As mentioned in the history lessons above I started the H4 with Hobbyking NTM2826 1200KV and 10x4.5 slowfly props. Fast RPMs, good torque, props quite matching. But here come the disadvantages: massive power consume and props with a softness like pudding. Flight times even with 3 batteries were about 6 to 7 minutes. Motors were vibrating awfully quite fast over time (bad bearings). The soft props amplified that vibrations due to their softness. The result are videos with a lot of yellow effect (wobble in the video). Couldn't get rid of the yellow for half a year because of the cheap motors and props. The exchange for the bad equipment were SunnySky X2212 980KV motors with 12x5 Aeronaut CAM Carbon light props. The motors were already discussed by Peter some time ago. You have to be careful to get genuine parts. I ordered mine from a vendor in Germany over the big bay. They came in a box for A2212 motors instead of X2212 motors (A version are less durable and capable than X version), but bearings and windings seem to be genuine and they perform very well. So I guess I got genuine ones. In combination with the new props the power is more than sufficient and my flight times went up from 6 minutes to about 8 or even 9 minutes! So plus 30 to 50 percent just with different propulsion! And with a bit balancing of the props, I nearly got rid of the yellow effect in my videos.  That was worth the money I guess :)

Here are some photographs of the actual state of my multirotor:

Picture ten: lightened and upgraded quad

Picture eleven: Closeup of rear with BT telemetry, UBEC and APM power sensing module

Picture twelve: Closeup of camera gimbal at the front of quad

Picture thirteen: H4 on the bench at the flight field

Picture fifteen: Quad on bench at flight flied

So it seems I have finished my guest entry for improving weight, EMC and motor/prop-combination of the H4 HK quadrocopter. Hope it was a bit entertaining for you all and at the end I want to thank Peter again for the opportunity to publish a guest entry on his great and informative blog.