nsg's blog View Details
Posted by nsg | Jul 25, 2020 @ 05:24 PM | 4,546 Views
These motors don't seem to have a very consistent kV. The two motors I have have unloaded kV of 380 and 387. The attached data is from a 387 kV unloaded. Castle Edge HV60, Xoar 17x6, Hyperion G8 HV 2x4s2800. Castle firmware 4.25.

Peak 1950 wt, settles to 1700wt or so, 9000rpm at roughly 80% efficiency.

I use that setup in my 60" EF Raven and 60" EF Laser (slightly less power due to slightly lower kV). It's.. a bit too much power, but certainly a lot of fun.
Posted by nsg | Jul 07, 2020 @ 07:45 AM | 5,255 Views
https://www.aircraft-world.com/aw/fi...r_eos_chargers . Need a Windows computer.

Update instructions attached.

The latest firmware supports HV lipos. The UI, two lines of text controlled/edited by buttons, look archaic compared to modern touchscreens, but the chargers still work well.
Posted by nsg | Jul 04, 2020 @ 12:10 PM | 7,966 Views
EF 69" Raven

MotrolFly 4315 custom wind - measures at 387 kV, was supposed to be 367
Xoar 17x6
Castle HV60
2x4s2600 mah HVLI Hyperion G8
Approximately 1900 wt max, settles at 1700.

EF Laser 60"
MotrolFly 4315 custom wind - measures at 380 kV, was supposed to be 367
Xoar 17.6
Castle HV60
2x4s2600 mah
Approximately 1700 wt max, settles at 1500-1600. That's on older batteries. Newer HV batteries add about 200wt.

3DHS original Slick 70
Hyperon Z4035-299, 299 kV
Castle Phoenix HV85
Xoar 18x10
Approximately 2kwt max
9.5lb AUW

3DHS original Velox 68"
Hyperion ZS4025-16, 320kV
Castle ICE 75
Xoar 17x8
About 1600wt settled
8lb AUW

3DHS Vyper 65" (sold, already missing it)
MotrolFly 4315-300
Castle ICE 50
Xoar 17x8
About 50A max, settles to 45-47A
12 minutes on average
7lb AUW
Posted by nsg | Apr 12, 2020 @ 08:08 AM | 5,814 Views
Quick notes from recent conversations, mostly for myself.

Geared setups suddenly make sense, in 3kwt+ space, anyway. A typical direct drive outrunner setup performs at 80%, give or take. I'm too lazy to figure out why, but it is an observable fact for existing outrunners. It is possible to find a high-speed motor that will output the same power at 90-95% efficiency and convert it to a usable form via a timing belt-based gearbox, at a cost of about 2%. Claimed total efficiency about 90% + some weight savings at 4-6kwt, good reviews from the field. See this blog for details.

Design parameters for a 3D setup
- starts to make sense at about 3kwt, give or take.
- gearbox to support gyroscopic forces from a 24-28" prop
- motor optimized for efficiency in 30%-70% power range, with at least 85% at full power, about 20-30K rpm range
- at least 50V, so kV in 500-700 rpm/V range, lower is better.
- package to fit a typical 74" or greater 3D plane. Most of these are designed for either electric or gas, so there's some room for offset gearboxes.

Proper power setup for servos:
- Needs to support two stalled servos for about 30 seconds, all four stalled for a few seconds w/out shutting down the receiver.
- servo extensions do contribute some voltage drop, so 20ga wire where possible
- servo connector is rated for 5A continuous; observed safe for 10A bursts. Above 10A, two servo plugs from a power source to Rx. Above 15A, use a receiver with a separate plug for a flight battery.
- No BEC for 70+ models: use a flight battery. If using BEC, review the current rating carefully. Most BECs are rated for peak power under certain cooling conditions. Using a battery helps you separate the power circuit from the signal circuit (always a good thing)
- for larger models, measure servo current draw at max deflections before the first flight.
Posted by nsg | Apr 04, 2013 @ 08:43 AM | 7,002 Views
I use 5.5 mm PolyMax bullets for my larger batteries. If I were to start over, I'd probably go PowerPoles, but I've too many batteries/planes already set up for bullets.

My problem with the bullet connectors was the negative bullet - it is exposed. I fried a parallel charging board by accident once - touched exposed battery negative bullet with an exposed charger positive. I found a way to fix that.

Amazon sells a 10mm OD/8mm ID plastic tubing. 10' cost me something like $15. A piece about 7/8" long fits over the negative bullet nicely, held in place by a drop of hot glue. No more exposed bullets.
Posted by nsg | Dec 30, 2011 @ 05:03 PM | 7,460 Views
Maidened a 62" 3DHS Osiris today. Bought a slighly damaged kit from Mr. Irg, repaired, moved the motor & ESC from my 3DHS Vyper, servos left over from my Extra SC, had to buy a roll of midnight blue Ultracote, a new receiver, prop and a spinner. Total ended up about $270. Not bad for a very good pattern plane.

Hyperion H4025-12, ICE 100 Light, 5s4000mah, AR6210, Xoar 16x10, 3x Hyperion DS-16FMD, 1x HD DS-120M (rudder)

Exact match for Osiris dark blue covering: Ultracote, Midnight Blue, HANU885


Posted by nsg | Oct 21, 2011 @ 08:07 PM | 8,183 Views
Not only Flying Styro is back (who'd have thought..), but they are offering some pretty nice deals at the moment. Buy three kits and get free shipping + 10% off. I finally took them up on it - Mosquito, An-2, Supermarin S-6. I have their Yak-3 ad love it.

For those who haven't heard of FSK, it's a Czech company (no, not Alfa.. that's another Czech company) known for highly detailed - count the rivets - scale models that happen to fly extremely well. Hobby-Lobby used to carry them.

Posted by nsg | Mar 26, 2011 @ 06:58 PM | 8,916 Views
Another toy..

Mega 16/25/2, Turnigy 85A ESC, turnigy 6x5 prop
2x HS-65HB, AR500
Motor mount by SmallPartsCNC.com (highly recommended)
Hyperion VX 4s2600 mah

About 55A max, 26-27K rpm.

I think Ultra is a nice improvement over the regular FunJet even if you don't go with a high-performance setup. I flew it on a 400 wt setup while waiting for the Mega, and it flew fine.

I need something better than an AR500, and something better than the TGY ESC, too. All in good time.


Posted by nsg | Dec 15, 2010 @ 09:11 AM | 9,047 Views
.. so long and thanks for all the fish.

(sapienti sat, and I don't mean the Hitchhikers Guide)
Posted by nsg | Oct 26, 2010 @ 04:51 PM | 9,247 Views
Maidened the Aspera I bought from Tim at the NJ 3DHS Throwdown. It takes some getting used to - the only pattern plane I've owned before was Brio, and that doesn't really count. I need to get used to it, but the first impression is very positive.

EF Torque 2818/900, APC 11x5.5, 4s2600mah for about 500 wt. ICE-50, 4xHS--65MG, AR6200. AUW about 1300g or 46oz.
Posted by nsg | Oct 20, 2010 @ 09:19 AM | 9,983 Views
Castle is running a clearance on their test motors. I ended up with three, more or less out of curiosity.

Motors seem to be high quality. Castle being, well, Castle, cross mount holes and holes for a bolt-on adapter do not correspond to any motor I know of. The shaft isn't easily reversible, the grub screw is difficult to access. A Turnigy 42-xxx bolt-on adapter almost fits, the bolt holes need to be filed by 0.5 mm or so. The bolt pattern for X mount looks to be 24x30mm.

3412 series weight 155g, 3422 - 250g, 3434 - 300g.

3412-770. This is pretty close to Hacker A30-12XL, except it weights less. With an APC 13x6.5 on 4s, I got 45A, 8300rpm at 14.7V. 30 seconds at WOT didn't overheat the motor, so it is probably good for 14x7 for 3D. Should be a reasonably good fit for something like 3DHS 51" Slick. Of course 3DHS provides a discount on a Hacker motor bought with Slick at the moment, making Hacker a better deal.

3412-450. Looks to be a good motor for 13x6.5, or possibly 14x7 (pushing it) for 3D at 6s. This will probably work on a 48" EF plane, or most 3DHS 48" ones - it's basically the same as a 150g 900 kV motor on 3s. Short-term max current looks to be about 30A.

3434-250. Needs a high voltage setup. Should be good for about 1500 wt at 10s. Max short-term current looks to be about 35A. 3434-450 is probably a better choice for your typical 65-70 plane, and a good value at $50 compared to similar high quality motors.

Castle was going to market these at $150-200, I believe. Not sure they are worth that much to most people, but they'd certainly be competitive in $100-120 range. 3412-770 and 3434-450 look to be very usable and a decent value at clearance prices.
Posted by nsg | Sep 11, 2010 @ 08:54 PM | 9,713 Views
Reichard Sirius. 3m wingspan, about 2200g AUW. Maidened today, with much appreciated help from Mr. RCEverything.

Relatively easy build, excellent performance. Near vertical climbout with a fairly modest power setup

5x Hyperion Atlas DS-16FMD servos (wings, rudder)
HS-65HB elevator
Hyperion GS-3020-782 motor
Castle ICE-75 Light ESC
Zippy 30C 2200 mah battery
Aeronaut 14x8
About 500 wt max
Used up 900mah in 20 minutes.
Posted by nsg | Apr 10, 2010 @ 07:19 PM | 9,888 Views
WildWing-1.avi (8 min 12 sec)

The famous $12 keyfob camera mounted on my WildWing.
Posted by nsg | Mar 27, 2010 @ 02:12 PM | 9,548 Views
I almost crashed my 3DHS Vyper 65 the other day. This is (more or less) an investigation log (more or less) for myself.

Equipment as flown:
Hyperion H4025-486
APC-E 17x8
Castle ICE-100Light
4x Hyperions DX-20xSMD
Spektrum AR6200
ValueHobby GForce 4000 mah 5s

What happened:
Loss of radio signal for about 5 seconds. Recovered in time for less than graceful landing, slight damage to cowl. Could have been a lot worse.

What do I know at this point:
- All equipment expect for battery has been flown before
- AR6200 mounted reasonably well, satellite receiver far from main one. Rx been used on my HK-450 heli for several months, had three problem-free flights in Vyper prior to accident
- The plane was in first quarter of a large loop, full throttle, going up
- ICE log shows full throttle up to loss of signal. Battery voltage very reasonable.
- Total servo current is about 1A. Stalling a servo doesn't increase it all that much. The Hyperion spec says nominal curren 0.2A, stall current 2A per servo.

Other data:

68 Velox crashed, BEC/ESC failure. Same equipment I'm using now.
Brownout on a similar setup
Apparently, ICE BEC is only good for 3A (peak? not sure) at 20-24V. That would make brownout a possibility.

  • Asked Castle's help to interpret log data. Castle rep did their new 'ripple voltage' song and dance before we settled on faulty battery conenction as one possible cause. Remember Patrick's "you all use crappy motors" song and dance?
  • Switched to 5.5mm bullet connections for ESC/battery
  • Twisted wire extension (it's 30") for the tail servo.
  • Found that Castle BEC won't connect to Castle Link if it's soldered to ESC leads as it is supposed to. Apparently USB current leaks into the ESC and shuts down the USB port. Figured out how to make it work - plug in a 2s battery. That will help out the USB port, now it all works. Heh.

Posted by nsg | Mar 18, 2010 @ 12:15 PM | 9,892 Views
Okay. EF is hereby nominated for the 2010 funniest customer mailing award. Strong runner for the all-time record award.

(Sapienti sat)
Posted by nsg | Mar 17, 2010 @ 09:24 AM | 9,814 Views
(Notes to myself)

CoG 170-180mm off leading edge == between centre and tail side of the wing tube

Low rates per manual
Ailerons 42mm 30% expo
Elevator 17mm 30% expo
Rudder 52mm 30% expo

High rates per manual

Ailerons 83mm 70%expo
Elevator all the way 75%
Rudder 77mm 75%


Ailerons 15 degrees == 42 mm gap at the wing root
Ailerons 30 degrees == 83 mm
Aileron width at root 160mm

Elevator 15 degrees == 17mm at counterbalance
Elevator width at counterbalance 65mm

Rudder width at counterbalance 100mm
Rudder 30 degrees == 52 mm at counterbalance
Rudder 45 degrees == 77 mm
Posted by nsg | Mar 12, 2010 @ 12:01 PM | 9,840 Views
January 2010, shot from my Cularis. Not too bad for a $14 camera.

LMT-from-above-jan-14-2010-cut-0.avi (9 min 16 sec)

Posted by nsg | Jan 15, 2010 @ 05:53 PM | 11,254 Views
Short notes on my MPX Blizzard


Turnigy SK 35-36-1400 motor (a pretty crappy motor, but works well enough here) Replaced with Hacker A30-12M, same rpms for 5A less.
Castle Phoenix 35
Aeronaut 8x5
2xHS-45Hb in the tail, 2xHS-65HB in the wings
AR6250 Rx
Zippy FlightMax 3s 2200 mah

About 300 wt, AUW just under 1kg.

Build mods:

I inserted a piece of 3DHS 18 mm wing tube into the boom. That made tail servos removable - the wires are routed inside the tube to the receiver in the pod - and stiffened up the fuse nicely. I used another wing tube as a drill - taped the fuse halves together, taped the wing tube to a 1/4" rod and drilled through the boom until I hit the servo cavities. One mistake was to aligh the tube to motor shaft - the tail boom actually goes down, so the channel I drilled ended up closer tot he top surface of the boom. Not really a problem.

I also added carbon strips to stiffen up the wings. last but not least, covered the whole thing with coloured tape.

Tx setup is an acro plane with V-tail and suitable mixes. I'm still fine-tuning the CoG, rudder-elev mix and aileron diff, but the first flight is defintely a success. It flew fast, slowed down nicely. HS45s seem to be more than enough for the tail surfaces.

Update. TX setup:
- Flight modes on flap switch (AUX1), via PMx4
- FM0 set up for landing, puts both ailerons up, adds elevator (PMx1)
- FM1 is cruise. flaps up/down 2mm via left lever (PMx2), elev compensation (PMx3).
- FM2 launch
- Rudd->elev mix to provide V-Tail diff when used as rudder. Per Multiplex, there has to be about 2 mm differential, PMx3
PMix 1: AUX1->Elev to compensate for the trim change with flaps full up
PMix 2: LLEV to flaps, 2mm up/down, FM2 only
PMix 3: LLEV to ELEV, works together with PMix 2
PMix 4: AUX1->Flaps, works together with PMix 1 (could have used flap system feature for this one)
Posted by nsg | Nov 16, 2009 @ 05:25 PM | 10,429 Views
Test data, for future reference. Firmware 3.0, 3.20

ExtremeFlight Torque 2818/900, APC 12x6

AirBoss 45 Elite, default settings:
10.17V 35.2A 6800rpm
Phoenix ICE 50, firmware 3.0, 8 KHz pwm, low timing (optimal configuration):
10.17V 37.8A 6990rpm

With that prop, the difference amounts to app. 150g - computed, not measured - more thrust for 2.5A more.

Notes on ICE settings:
* Don't use the outrunner mode, set the PWM rate to 8KHz. This is the same as AirBoss default pwm rate.
* The only official non-beta firmware for ICE is 2.16 . The latest available as of today is 3.00 . I noticed no significant difference for this application - 3.0 seems to be a bit more efficient, but not by much.
* AirBoss does seem to run the motor a bit more smoothly - or maybe this is just lower rpms.

Guideline on PWM rate: (RPM/60)*(# of poles) * 2 < pwm rate


Turns out firmware 3.0 had a bug - increased timing at high rpm. That accounts for higher rpm and power consumption. Corrected in firmware 3.20
Posted by nsg | Nov 07, 2009 @ 08:49 AM | 10,754 Views
The second of two posts on setting up JR9303 to control a Multiplex Cularis glider. Setting up the Tx.

Rx channel assignments (wire->Rx socket):

Left Aileron->Throttle
Right Aileron-> Aileron
Elevator, Rudder->Elevator, Rudder
Right flap ->Flap
Left Flap ->AUX1

Setup, device selection

Flight modes enabled

Speed INH
MOTO Spoiler stick
Flap flap lever
AUX3, AUX4 default

Wing type: dual flaps


ELE->FLP: not set up
AIL->FLP: Pos. 1 30%, Pos0: 0%, activated in all three flight modes
Differential +40% for ailerons, all modes, flaps, cruise and launch.
Flaperon Mix +10% fprl, 0% elev
Aileron->rudder, 30% cruise only
Butterfly mix Spoiler->flaperon Pos0 -50%, spoiler->flap Pos0: +70% spoiler
offset +170
Flap rate 20%, cruise only, that affects camber adjustment
Motor hold: off for launch, on for other flight modes, on for Aux4 (throttle hold switch). The motor is controlled by throttle stick in launch mode, switching FM or hold switch will kill the motor.

MX1, spoiler->elevator mix curve. Reverse default from the manual.
MX2, MOTO->ELEV, Launch only, 0, 2%, 8%. This reduces the tendency to pitch up at full power
MX3, MOTO->AUX2, Pos0 0% to +100%, pos1 +100% to +100%, active in launch mode only. That lets throttle stick control throttle in launch mode.

All control surface deflections per MPX manual.

Maiden flight went ok, very minor trim adjustment was necessary. This was my first glider, so some things may seem ok to me which really aren't, but I saw nothing obviously wrong. Like a sailboat, you need to think ahead a bit, everything takes a second or two.

The power is adequate, it goes straight up at WOT. The built-in BEC may be enough - will measure in-flight current, may get rid of the external BEC based on the results.