Devon Slopes

My layout IS still nose heavy, David, despite my best endeavours. I'm still trimming but expect I will end up with 20 grams in the tail. The previous electric version I built was done by cutting the front off a slope nose, and that was indeed not circular, which as you say looks slightly peculiar, but only slightly so. Given PatMc's suggestion of routing the wires outside the fuselage, the clever way for anyone else thinking of doing this may well be to shorten the nose as David suggests, and route the wires outside the fuselage if needed.

Hi David, 750g including a 40g battery, that's very impressive, much better than my 850g, which also has only 40g of battery. How did you do it? Most of it will be the motor, my motor is nearly 70g heavier than the 2826, which sounds similar to the one I used to have, but I only managed to get 90W out of compared with your 150W. One of the reasons I wanted rid of the out runner was the tight fit, problems which you seem to have solved. Did you use the Blejzyk supplied electric nose, or make your own in some way. I'd be surprised if you could get the 2826 in the standard electric nose, which may be a warning for others thinking of this conversion. Do you have a picture? And to answer your question, if you peer closely at the picture 30/11/2018 I laid out the contents of the fuselage in the order I put them in. Things were driven by my need to change the battery through using the cover hatch, and of course that you want to put the battery as far back as you can. This means it has to slide backwards from the hatch, and clear the servos, which dictates how far back you can put the servos. So the battery ends up under the leading edge. The ESC sits almost directly under the hole in the fuselage the wings fit on top of, and then behind the wing are the receiver and finally (pushed far down the fuselage) is a barometric altitude sensor. And after all that, its still nose heavy!

This is how I opened up the holes in the ribs for the wing joiner tubes from 6mm to 8mm. Because the side of the tube has to just touch the main spar, the centre of the new 8mm hole is offset forward with respect to the old 6mm one. So you can't simply open up the hole evenly all round. So I made two ply templates with 8mm holes made in the right place by comparison with the plan. I then "sandwiched" all the ribs together using the incidence tubes, double-sided sticky taped the templates in place and them opened up the holes using sandpaper wrapped around a pencil. Note quite as elegant as it sounds, as a little more sanding had to be done to get the tubes in once the inner wing panels had been constructed, but more of that later.

Many thanks, Peter. I'm still wondering which way to go with this, magnets, multiplex multilocks or tape. A decision will have to be taken soon....

Props. The biggest issue at the moment is the prop. My local model shop didn’t have the recommended prop, a 6”x4”, so I ended up with a 7”x4” Aeronaut prop. When I run with this on the ground the motor starts off by pulling 16A, but rapidly (seconds) fades to 12A. Of course the prop will be less highly loaded in flight, but I don’t have an ammeter on the telemetry (the FrSky one is too large and hefty). However I do monitor the voltage, and it drops pretty precipitately as you power up. In fact it gets too close for comfort to pulling the battery below 3V per cell. Once you’ve been flying for a bit, the drop is not as severe, so my conclusion is that it’s the 3S 450mAhr battery that’s the limiting factor, and it performed better when a little warm. That was easily tested on the ground by connecting a 2200mAhr battery up instead. Then, as one of my friends would describe it, we were cooking with gas – 18 or 19A draw, no obvious fading with time. Oh, and it pulled like a horse. So, what to do as there is no way that huge battery is fitting in the Mefisto fuselage? The battery is marked 30C continuous, 50C burst, which translates to 13.5A and 22.5A. I’m not sure how long a “burst” is meant to be (does anyone out there know), but I’d hoped it was a typical e-glider motor run (maybe 10 seconds), but I guess this is just the well-known lesson of not pushing batteries to their nominal spec. So, I bought some cheap props to try different sizes, maybe I could limit the current draw by having a smaller or finer pitched prop. So far I’ve only done 6”x4” (the recommended size remember), but that was even worse. It started at 16A and faded rapidly to 10A. When I tried flying it the pull was so poor that I was worried about clearing some low bushes as I took off! That to me is a victory for the efficiency of the Aeronaut prop, its bigger than the cheapo prop, has a similar current draw, but more pull. So, for the moment I’m sticking with the Aeronaut prop and limiting the throttle to about 75%. Now, I recall reading somewhere that ESCs are most efficient at that level, is that right? The summer may improve things – I know my old the old version suffered from large voltage drop at winter temperatures.

Thanks, Peter. Do you have any words of advice on how to fix the magnets for those of us who've not done something similar before?

No, this build hasn’t been abandoned, just that nothing has got done for the last two-and-half months due to other commitments, which included the new fuselage for an electric glider. The wing joiner. Andy Blackburn says that “if you’re planning to play rough with your Canberra” you should go up a size in the wing joiner. Slopes Junior thoroughly enjoys his slope aerobatics, so I suspect I need to do this. The standard joiner is 6mm OD carbon tube with 4.5mm carbon rod inside it, so I’m going to 8mm OD carbon tube, with 6mm carbon rod. This means I will have to open out the holes for the joiner in the wing ribs. Attaching the wings. I’m also wondering how to attach the wings onto the fuselage, which Andy recommends doing by tape, but says you can also do it by using a Multiplex Multilock plug and socket, but if you do so you need to make hard points to put them into. Not quite sure what this involves, but it looks like I need to work out how to do this before I complete planking the fuselage. So, I’ve dry assembled the starboard wing laser cuts parts so I can see what’s going on. It all seems to go together nicely, though you have to make another rib R4, which is easy to do using the sandwich method, with the laser cut R4s as the templates.

How did you hold the wings on in the end, Peter? Earlier in the thread you mentioned some wing retaining plugs, and I was thinking of going a similar way, rather than putting tape along the join. But, Andy B. suggests some strengthening around the plugs if you go that route, and I was wondering what you had ended up doing.

(Does anyone know how to rotate a picture?) And here is the completed model. You can see the hatch, secured in the traditional way using wire epoxied to the inside. The nose area looks empty, as all the gear is as far back as possible, and the battery slides over the servos to sit well behind them under the leading edge of the wing. I’m fairly sure it is nose heavy, but have not yet started seeing how I can improve the performance by adding weight to the tail. The all-up weight is about 830g and will probably go to 850g when its trimmed. That’s a bit of a disappointment, as the old version was 750g. On the other hand its really nice to operate, not having to take the wings off to change the battery (and worry about flexing solder joints on plugs so they break). Furthermore, its easy to change between electric and pure glider, just take the wings and tail off one fuselage, and screw them onto another. The old method of changing the nosecone was difficult, and stressful on a Saturday morning when I’d decided to go flying, but then had to spend 20 minutes changing the Mefisto into the appropriate mode. I’m still fiddling to obtain the right prop, but more about that in a later post. But it has answered the problem about power. I well remember the first time Slopes Junior saw the old version of this model working. I opened the throttle and he said “Is that it?” in a derogatory reference to the power output. Those days are gone.

Here you can see the electronics laid out alongside the fuselage in roughly their intended positions. Starting at the front is the prop and motor, then come the servos placed in the tray I’ve made for them, and the battery which will go behind them. Next is the ESC, hidden beneath which is the battery voltage sensor. This is followed by the mass of wiring for the multiplex connectors which go to the wing. Finally comes the receiver, and variometer. The two small black rectangular bits, one amongst the wiring and one beyond the variometer are the aerials. The whole layout is about getting as much weight aft as possible, though I believe I am still going to end up with tail weights. An unexpected time sink was making up the cables. Its the first time I've crimped up lots of cables for a model, but I think it has really paid off in terms neatness and compactness because all the cables are the right length, no excess lengths being pushed into spare space.

I've finally filled up all the available servo connectors on an FrSky X8R by adding a tow release to "full house" glider (it also has a lost model alarm). Now, whilst I could power the rx by putting a Y-lead on one of the servo sockets, that SBUS (or indeed the RSSI) socket looks very tempting. I opened up the rx to check if it was on the same positive and negative buses as the servo ports, but its hard to tell (I suspect connections are buried in the PC boards somehow). So, my question is, has anyone powered the servos connected in the normal way to the X8R via the SBUS or RSSI ports? Or, does anyone know if the internal connection of the SBUS and/or RSSI ports is similar to the one between the servo ports. Thanks, Devon. Edited By Devon Slopes on 09/11/2018 08:13:37

Many thanks for linking the videos, Peter. Inspiring me to continue my build, I just have to complete the changes I'm making to an electric glider first.

Well done Peter. Encouraging to hear that it flies well in modest winds. I do hope you find some pictures of it flying to post.

So, this is the collection of new bits that I have either made or bought for the electric fuselage. Not shown are the things I have from the old version, an ESC, spinner, prop (though I’ll probably need a different one), and receiver (I use FrSky gear). Captive bolts. Above the fuselage are the structures which hold the captive nuts for the wings and tailplane. I found that the nuts originally supplied were very tight, so on the old version I simply glued metal nuts onto the wood, countersinking them slightly into the wood for the wing bolts, and putting a small piece of wood around the ones for the tailplane, in the hope of holding them in. Problem is this did not work well, and I have occasionally had the nuts come off. Not wanting this to happen on the new fuselage, I have used proper captive nuts for the wings, though they had to be cut down a little for the rear wing bolts. The snakes are by Kavan, their 0016 snake outer, which takes 0.8mm piano wire as an inner. I was really pleased to find these, as the snakes that come with the Mefisto also have 0.8mm piano wire as the inner, which pokes into holes in the tailplane horns. So, I can simply move the tailplane between the two models. The servos are a pair of Corona DS843MGs, which are a close match for the Hyperion DS09AMDs I originally used. The latter seem to be becoming difficult to source, and the Coronas are considerably cheaper. Lets wait and see if they are robust. The motor is the thing which has given me the most pause for thought. As I said in the previous post one of the problems with the old version was a lack of power. I got about 90W out of the Overlander T2826/18 from a 3S battery, so my search was for something more powerful, preferably with a similar low kV (1000) as the Overlander, as (I’m told) slow props are more efficient. But, the problem is that the new fuselage is clearly designed for a 28mm diameter inrunner. Inrunners tend to have fewer magnets and so are faster. In the end the best compromise I could find was the 200W 2100kV Overlander BM400. Should have enough power, though a somewhat less efficient prop, but the potential problem is the weight. The T2826/18 was 50g, the BM400 is 114g, so that’s an extra 64 grams in the nose. How will I make this thing balance? The only way to find out is to start assembling it, and see where the Cof G comes out, and then move things around.

It does look nice, Peter. Inspires me to push on with my build.

Not to worry, Andy, it's a trivial piece to make. So I pinned the supports for the tailplane in place on the fuselage. I then put the tailplane in on top, with a layer of plastic between the two. I then pinned the tailplane supports to the tailplane. Removing the pins from the fuselage then allowed me to remove tailplane and supports as a single assembly like this. I checked the supports were seated properly on the tailplane, then put glue in the appropriate places on the supports, and re-assembled the tail as shown below. The wooden pieces on the tips make sure that the tailplane is held at the right angle (i.e. the equivalent of horizontal). Andy, if you are listening, is there any way of checking I have the decalage right? Should the tailplane be a zero incidence? When it’s all dried you can take the tailplane off, and it looks like this. All I have to do now is block in the bottom part. Peter, meanwhile, is making fantastic progress. We started at the same time. He started on the wings first, but completed those months ago, and has now overtaken me and seems to be almost finished. Oh, and in between he’s managed to complete a Hurricane. How do you manage it Peter? My only excuse at the moment is assembling an electric fuselage for a glider.

The back story. I built a Blejzyk Mefisto a couple years ago as a holiday model. It's a 1.9m glider with a four servo wing for which I have both slope and electric noses, the idea being that it could be flown in a wide range of circumstances. As a model for that holiday it was a failure because I didn’t get it completed until just before I left! The lesson is that if you want a holiday model, get it completed well before the holiday. You will be flying it in unfamiliar conditions and so its good to have a model you are comfortable with. On the other hand in the time since it has become a firm favourite. Its light (under 900 grams in electric mode), which means I’m happy to fly it in a large area of nearby parkland. Coming home from work in the summer and spending an hour thermal hunting is good relaxation. Its equally nice on the slope, when its even lighter, but also has thin wings which means it can hold its own in strong winds. So, why is it back in the workshop? Well, I never managed to get the proper electric nose, so made my own by hacking a spare slope nose. The design I came up with was not very good, and so the electric version is a bit of a pest to operate. For example, to change from electric to slope mode you have to take off the prop, undo the screws which hold the motor in, slide off the nosecone, disconnect the ESC and other electronics… the list just goes on and on. Also, the battery is changed by taking the wings off. Oh, and its underpowered. All-in-all not one of my best design moments. Though see the attached pictures of the fuselage in the two modes. So, my argument went, I fly this model often, but my design failures make it time consuming and difficult to operate, so surely upgrading it would be a good use of time and money. The problem is that the UK suppliers for Blejzyk have gone through various changes, which is one of the reasons I’ve never managed to source an electric nose. But there is now a new UK supplier on the scene called MyHobby who had an electric fuselage in stock, and so I snapped it up. You will instantly spot the creeping cost envelope of this project, electric nose has become electric fuselage. Anyway, before I start detailing the assembly over the next few weeks, I should mention there are a couple of other Mefisto builds on the forum, one from FWAL and another from Phil May. The only electric build I can find is on rcgroups.