1. Consumer Requirements
2. Transmitters
3. Transmitter Antenna
4. Receiver Antenna
5. Transmission Range
6. Receiver Output
7. Frame Rate
8. Resolution and Latency
9. Servos
10. Waterproof and Water Resistance Ranking
11. Analog v Digital
12. Exact Servo Rotation/Travel
13. Captive Drum Technology
14. BEC — Voltage Regulators
15. Choosing a Servo
16. Worldwide Ranking by real Torque values
17. Product Specification
18. Motor Speed Bands
19. Speed v Load
20. Travel reduction due to load
21. Travel dependant on Tx/Rx
22. Why SailServo
23. Servo Testing
24. Quality Control
Analog v Digital control
As consumers, what do we actually want from our r/c equipment?
Move a joystick to control a servo to rotate a drum for the sails, arm for the rudder or an ESC for a motor with a degree of accuracy and life.
We are not interested in all the electronic waffle only that it works. When we move the joystick a certain distance we want the servo to pull a certain load, over a certain distance, in a certain time, repeated a certain number of operations without failing. Is this too much to ask?
To do this successfully we need to match the transmitter, receiver and servo as one integral unit but unfortunately there is insufficient information to do this.
All the manufacturers do their own thing and there are no standards.
A consumer upgrading their transmitter/receiver to 2.4Ghz will find their servos will not work correctly. In my research I have found a Tx/Rx set which turns my rudder servo only 40 degrees instead of the existing 70 deg. The consumer has no way of finding this out.
When designing and building our boats we spend a lot of time getting rid of the slop in sail mechanisms and rudder linkage. Have you ever thought about the slop in your transmitter – receiver – servo system.
It's called Latency, the time between moving the stick and the servo turning. A customer tells me he can notice the quicker respones with a Spektrum Dx8 over his old Dx7. This is the demonstrable effect of Latency
This article has been written so I can sell my servos with confidence and help my customer in their choice of equipment. In my research I tested 13 out of the 26 drum servos on the Uk market, 4 transmitters and 5 receivers. The Interweb and Forums disgorged hidden gems and my customers tested their receivers and passed on good tips.
Thanks all
Transmitter visual design has not altered for 20 years but user perception and ergonomic design has changed since the introduction of games consoles.
Games consoles perform similar functions but have a totally different design, no neck strap. It may be necessary to include some device to give a visual indication of the sheet/rudder/throttle position. Existing joysticks are large, expensive and not waterproof.
The new transmitter from GWS takes ideas from the X-Box. The size reduction is due to the purpose built joysticks fitted directly onto the PCB. Its antenna could also have been put sideways on the PCB. I will try this later.
HobbyKing has moved its antenna inside the case but at 10 degrees to the centreline. Perhaps it’s a mistake? It is also up 15 degrees which moves the direction of max signal strength nearer to the boat. There is a steel handle directly behind the antenna?
No manufacturer tells its consumers how to position the transmitter antenna for best results other than a few who say “do not point the antenna at the plane.”
2.4Ghz is in the microwave spectrum and the signal is like a beam of light. It travels in a straight line and bounces off solid objects which can cause interference. One forum suggests that the sophisticated error processing power of new receivers eliminates most interference. Glitches in a boat are not that critical unlike in a plane.
Antenna Separation Distance. Body should be more than 5cm from the antenna, except hands, wrists, feet and ankles. (Spektrum Dx6i Manual)
Now for the science bit. Do you remember doing experiments with iron fillings and magnets at school to see the magnetic field?
The antenna’s signal pattern is similar to the magnet and radiates a field of energy in the same pattern. The signal radiates out from the antenna like a fat donut with the maximum signal at right angles along the antenna. There is no signal from the tip of the antenna that is why they say “Do not point the antenna directly at the plane”
Hold up your transmitter. The most comfortable position is at 45 degrees to the ground but this sends the maximum signal at the ground and not your boat.
"Turn the antenna tip point away from the model and ground. Signals transmit strongest from the antenna shaft not tip." Spektrum Dx6i Manual
The signal is always at maximum strength with the antenna positioned horizontally and facing your boat.
2.4Ghz transmitter power can be checked using a GWS Antenna Output Tester
Testing the three transmitters above showed up considerable differences.
The Spektrum light up like a Christmas tree at 30cm with all lights flashing, the HobbyKing had all six LED’s on steady at 10cm but the GWS Park Flyer only lit up two at 4cm from the antenna.
The Spektrum and HobbyKing worked over 100m but the GWS only managed 40m. The GWS transmitter was bought for it’s style but nowhere did the manufacture say it had such a short range. If you bought this product in the UK you have Consumer Laws to protect you against misselling.
velocity = frequency * wavelength
In this case the velocity is C, at say 300,000,000 metres per second. The frequency is 2.4GHz, or 2,400,000,000 Hertz. Dividing the first big number by the second yields a wavelength of 0.125 metres, or 12.5cm. Since we’re using quarter-wave antennae, that works out at just over 31mm.
Coax is used to stop some parts of the cable from acting like an antenna - the bits covered by the outer sheath. The last 31mm exposed is the active portion. OrangeRx receivers don’t use coax because they’ve got 31mm worth of wire soldered directly to the circuit board, and hence their antenna cannot be relocated within a distance permitted by a length of coax.
From H2SO4 rcgroup forum
The antenna is only the last 31mm of the end of the cable from the receiver. It is this part that has to be positioned correctly. (The wooden rods are used to show 31mm)
The Spektrum and HobbyKing have the antenna at the end of a coax cable making installation easier. GWS antennas must be kept in a straight inline.
The receiver antenna must be positioned above the waterline.
The antenna can be fitted inside the wooden, plastic or fibreglass hull but never carbon fibre or metal. Placing outside the hull will get the best signal.
ESC can give off interference signals and brushed motors will if not fitted with the correct capacitors.
Guidance on Tx antenna use
“2.4Ghz Spektrum is out of range of virtually all model generated motor & ESC noise and conventional radio interference. DSM range 300ft.”
Spektrum
The receiving antenna works exactly the same way as the transmitting antenna with the strongest signal at 90 degrees along the antenna and no signal at the end.
The position of the antenna has to take into account the two different angles of motion. All boats rotate on the water and sail boats heel.
Ideally the antenna should be vertical and high up as practical but away from metal objects such as mast stays. The signal strength will reduce as the boat heels.
(A Spektrum engineer suggested to position the antenna on the long cable vertically upwards and forget about the stub antenna. Place a piece of clear plastic tubing over the receiver antenna, it keeps the antenna straight and won't interfer with the operation.)
31mm vertical antenna inside the hull will be impossible in most hulls so the next position is to have two horizontal antennas at 90 degrees to each other as high as possible.
With only one antenna, the problem of a weak or no signal at the end could be overcome by bending the antenna into a 90 degree curve. (This needs to be checked if it is correct)
Several manufacturers say in their instruction manual to always make a range check before flying. 30m/100ft.
The other term to watch out for is "Park Flyer". What is the standard size of a park, British, American any country. Again another meaningless term used by the RC manufacturers.
"Full Range" The latest flyers gizmo to have is a tilt/swivel colour camera in the cockpit and wear goggles with TV screens. The camera movement is operated by the flyers head position. They are reporting flying several miles.
Nowhere can you find any factual advice on the best way to position your 2.4Ghz antenna in your boat. Ask somebody at the club.
The only thing I have found is “Dont’t point yout transmitter directly at the plane”. A warning is given that water has an adverse effect on 2.4Ghz signals and doesn’t work below water.
The reciever’s antenna works in the reverse of the transmitter so you would expect no or a poor signal if both antenns are pointed at each other
The only way to find out what actually works is to carry out tests to find out how transmission is effected by distance, antenna height above water and its direction.
A barge was built to contain 6 recievers with the antennas in different directions and 6 multi-turn servos to see if the signal is received.
Two antennas were placed above the deck as a check, they should work all the time. Three antennas were placed horizontally on a Datum line and inline with the centreline to check the effect of rotating them through 90 degrees. One antenna was place vertically below the deck to check the effect of reducing the effective length of the antenna above the waterline. (No 1 flag fell off)
The barge was built to have weights added at the ends to be able to reduce the height of the antennas above the waterline.
All the weights were added for the first test to check if the antennas would work on the waterline. This eliminated the GWS Transmitter/Receiver as it would only work upto a range of 40m. All the others worked upto 185m in both directions showing the antenna would pick up a signal from the tip.
When I got home I found that I had made an error. The antenna were at 10mm above the waterline so they were refitted at the correct height.
The second test was again carried out with all the weights putting the antennas on the wateline. I only carried out the rang test at 50m and they all worked. I added further weights putting the antennas 5mm below the water line and again they worked at 50m.
A collegue then asked, “That’s fine but how far can you put an antenna below the water and still work over a distance?” To answer this will mean a redesign.
A new rig was built so the receivers could be put below the waterline. The original barge was kept to house the servos. Good job I sell servo cables.
The receivers were aranged in a plastic pot so the antennas were pointing in the same direction or straight up. One was curved to see what would happen.
The white plate on the pot base is the datum line with horizontal antennas 12mm above the waterline.
Two 6mm U shaped spacers are removed from the top of the copper tube to lower the pot so the antennas are on the waterline. 3mm, 5mm and four 10mm spacers can be removed to put the antennas 48mm below the waterline.
All antennas worked in all directions 50mm below the waterline at 50 and 100m range and 40mm below the waterline at 180m. Transmitter antenna direction did not seem to matter.
These tests are not meant to be accurate or cover all receivers but to bust the myth that 2.4Ghz receivers do not work below the waterline.
The first post below suggests it is good pratice to have the antenna above the water but no comments made on its orientation.
Model Boat Forum comments on 2.4Ghz.
Model Boat Forum comments on fitting antenna.
The accuracy of a map of Watermead Lake in Aylesbury was established by measuring the distance between two fixed points. End of a bridge and pier.
This is central to converting the joystick input movement into the servo’s output of rotating a drum or arm over a controlled distance.
Unfortunately only one manufacturer has given any data which can help the consumer making a reasoned purchase but that’s hidden. It takes me 5 minutes to connect up my GWS Pro Servo Tester to take reading from 6 channels.
Customers of my Testers are sending me readings to add to my data bank. With only eight receivers tested, the PWM pulse range from 460 to 1140µs.
This matters as the variation in sail sheet travel is from 160 to 380mm, a difference of 220mm. From 42 to 82 degrees, a difference of 40 degrees.
The control signal sent from the receiver to the servo has two parts, Pulse Width and Frame Rate.
Pulse changes width as you move the joystick and this controls the position of the servo.
Watch the pulse move. Video from Helifreak
Receivers typically output a square wave digital signal of between 1.1 to 1.9ms. My GWS Servo Tester checks from 0.3 to 2.7ms = 2.4ms.
These pulses are repeated at intervals of 22ms (45Hz) and this is called the “Frame Rate”
The majority of Analog servos use a Frame Rate of 20/22ms. The maximum pulse width of 2.4ms is only 10% of the Frame Rate so for most of the time there is no signal. This speed of sending signals is too slow for flying planes so the manufacturers have halved the Frame Rate to 11ms = 90Hz
For high precision flying helicopters even this is too slow so the Frame Rate is even faster at 3ms. This causes a problem as the pulse width is nearly the same as the Frame Rate so the manufactures have a range of “Narrow Band 760µs” servos where the pulse width is restricted to 0.8ms.
Frame Rates found
ms: Spektrum 22/11, FrSky 18/9, Futuba 14/7,
Hz: 45/90, 56/111, 71/143Hz
To confuse everybody manufacturers specification quote 1520us/333Hz.
1520µs is the neutral/centre stick position which has little relevance, Spektrum use 1500µs.
333Hz is the Motor Driver Frequency. Nothing to do with the Receivers Frame Rate. My supplier says “the 300hz in your servo data 1520/300hz is Motor Drive Frequency.
Sailservo's servos all work at frame rates down to 5ms.
RedBird300 @ Helifreak.
Video — Frame rate does not effect servo speed.
Resolution is the speed of sending digital information from transmitter to receiver in Bits/Second. These values range from 512 to 4096.
Funny numbers used are because the digital signals are in Binary Code measured in Bytes. One byte is 256 bits which you may recognise as the number of colours available in Photo Editing programmes, 0 — 255 RGB
Old transmitters used 2 bytes = 512bits and was superseded by 4 bytes (1024bit) and this is now being rapidly superseded by 8 byte (2048bit). I have seen one Tx at 16 byte (4096bit).
The reason for the speed increase is the simply the new breed of transmitters used by planes and helicopters need a large amount of data to be processed rapidly.
With sailing, we have two different outputs to consider, the rudder moving 60 degrees and the sail sheet moving 100mm to over 300mm.
Your input joystick control accuracy cannot be smaller than 2deg. The joystick movement is nearly the same as the rudder movement so the rudder accuracy is the same as your joystick positioning.
The sail travel accuracy is similar, joystick error / joystick travel x sheet travel = 2/60 x 300 = 10mm
Nothing to do with the speed of data transmission.
A customer told me the Spektrum Dx8 gave a noticeable quicker response than the Dx7. The higher resolution has little to do with the increased response, it’s down the higher Frame Rate and Latency.
“
How ? I connected a frequency meter again to the elevator output of my AR8000 receiver (2048 bit / 11 ms), and made the smallest possible elevator stick movements . I repeated this a whole lot of times, to try to get the smallest movement possible, and at the same time I read the difference in pulse width that came with it.
Guess what, I needed about 5 µseconds of pulse width at least for the smallest possible elevator stick movement !
Knowing that pulse width can vary in this case from 1.11 ms to 1.89 ms, for a total of 0.78 ms or 780 µseconds, we can calculate the number of steps (or movements) my hands can do over the whole stick range: 780 / 5 = 156 steps. In other words, I simply can’t put the stick in more positions from the low end to the high end !
It is clear now that having 1024 or 2048 possible positions isn’t of much use, as we humans are too physically limited to enjoy that kind of resolution.
Data from RedBird300 @ Helifreak. Post 33
Latency is the time it takes for the output to respond to the input’s movement. This includes slop in any mechanical linkage, servo response and the time taken from the joystick being moved to when the receiver outputs the signal to the receiver.
Transmitter/Receiver Latency could be an important factor when choosing a Tx/Rx system and a few examples are given.
Transmitter | Receiver | Latency ms | Resolution | Frame Rate |
---|---|---|---|---|
Futuba 10C/TM10 | R6008HS R6014FS | 12 18 | 2048 2048 | High Speed |
Spektrum Dx8 | AR8000 AR8000 | 19 25 | 2048 2048 | 11ms 22ms |
Spektrum Dx7 | AR7000 AR6000 | 28 45 | 1024 1024 | 22ms 22ms |
Spektrum Dx6 | AR7000 | 36 | 1024 | 22ms |
Futuba 14MZ/DMS | AR9000 | 47 | 2048 | |
Hitec Aurora | Optima 7 | 55 | ||
JR 9303/XPS | XPS-8ch | 60 | 2048 | |
Data from JKos @ RunRider |
Having a faster Resolution does not necessarily reduce the Latency but reducing the Frame Rate and choice of Receiver does.
Dx7 with two different receivers shows a 38% decrease and Dx7 and Dx6 with the same receiver show a 22% decrease.
The difference between the lowest and highest is 80% but the cost difference is in the £100’s of pounds.
To reduce latency, some modern radio transmitter/receivers operate at a higher internal frame rate and deliver pulses to servos more quickly too, e.g. the Spektrum DX7se which works at 11ms / 90 Hertz - nearly twice the speed of normal receivers. With more and more radio control equipment moving away from the old 50 Hertz standard it is becoming increasingly important to ensure that servos are matched to the radio equipment they are attached to.
Data from RedBird300 @ Helifreak. Posts 1, 28 and 33
Another point to think about when using drum servos is the accuracy of the Feedback Potentiometer which controls the servo positioning.
“As the drum shaft has to turn more than 1 turn, a gearbox is put between the Pot and the output shaft. I took a servo apart and found that the output shaft could turn 11 times for the Pot to turn its full travel. However, the servo only turned 4 times causing the Pot to turn only 4/11th of the full track. This reduces the voltage range by 64% resulting in less positioning accuracy. ‘Digital’ sail drum servos do not increase accuracy. This is one reason why no two sail drum servos have the same travel”.
From "10. Analog v Digital"
I am teaching myself how to write computer progammes to control a PICAXE microprocessor. Perhaps one day I coudl build a rig to check servo Latency but life’s too short.
The servo output is as important as the transmitter and receiver but receive less recognition. They are just servos. Manufacturers given even less information about their servos and most of that is wrong. Do digital servos need a special receiver?
There are 26 different drum servos available in the UK. SailServo has tested 50% and all stall torque figures are inaccurate by as much as 70%. You will see I ignore Stall Torque as it is no indication of servo power.
The number of turns quoted by manufacturers again is ignored as it is meaningless. It completely depends on your receiver’s output signal and I can get a Hitec 785 to continuously rotate in either direction.
Manufacturers quote speed in sec/60deg. Again this is meaningless as speed is dependant on the drum circumference. Should be quoted as sec/100mm.
Another advance that has been made in recent years is the digital servo. Analogue types only check the position of the output shaft once every time a pulse comes from the receiver (i.e. every 20 ms) and send pulses to the motor at the same rate. Digital servos use a microprocessor to check the position and send pulses to the motor much more quickly which results in better resolution, holding power and acceleration. The only downsides are that digital servos cost more and use more power.
Data from RedBird300 @ Helifreak. Post 1 item 4
Trawling the interweb comes up with some odd facts. “ Some servos continue holding the most recent position even if the control signal goes away and all other digital servos turned off. As usual with RC products, this behaviour is not documented and can have important ramifications
Pololu
A meaningful method of showing all Sailservo's servo data.
Servo Motor speed should not be run below 40% speed drop. Between 20 and 40% is reasonable for intermittent usage and less than 10% for heavy continuous usage.
“Waterproof” is an interesting term used by manufacturers, exactly what does it mean?
If I dip a servo in a mug of water for 2 seconds and no water gets inside then this servo is “Waterproof”
A watch needs to be dustproof but not waterproof as you don’t normally put you wrist/watch in water. If you do, then you buy a waterproof watch. Watch manufacturers use the term Resistance not Waterproof and test to certain depths.
Water Resistance — Wikipedia
Water Resistance — Breiting
To simulate a model boat sinking I filled a one meter pipe with water and immersed four “Waterproof” servos for 24 hours.
Make | Weight gm | Water cc | Seals | Water Resistance | ||
---|---|---|---|---|---|---|
Dry | Wet | Gears | Motor/PCB | |||
Hitec 785 | 110 | 117 | 1.0 | 3.5 | Y | Water/Dust Tight |
GWS 125—1T | 62 | 68 | 1.5 | 1.5 | Y | Water/Dust Tight |
HobbyKing VSD-22ymb | 55 | 66 | 0 | 4.5 | Y | |
SailServo Dcd1 | 62 | 71 | Trace | 5.0 | N | Waterproof |
None of the servos were "Waterproof" so it makes maunufacturer’s claims doubtful. But they may be Water Resistant at reduced depths and durations.
“Waterproof” , “Water Tight” and “Splashproof” are misleading terms as they have no definition. You could dunk a servo in a mug of water for a second and call it “Waterproof” if no water leaks in.
This rating system is a simple method to classify the Water Resistance of servos used in R/C boats.
Immerse the servo vertically in water at a certain depth for a certain time.
Allow the cable to be straight from the casing for the first 10mm.
If any water leaks in then it fails.
Rating 1 Bottom has been put first as it is the most likely source of leaks due to cable entry design. Hulls fill up from the bottom.
Rating 2 Top is next as it is the less likely possible source of leak as the servo top is protected by the drum and the gap around the output shaft is small.
Avantages and disadvantages of Digital servos are fully explained in Analog v Digital.
The are several ways to get the exact servo rotation angle or sheet travel for your boat. The cheapest is to add a Servo Stretcher between the receiver and servo. The next is to use a Programmable Digital Servo. The most expensive is a Computer Transmitter.
At last I have been able to find a 6v Servo Stretcher that is small and easy to adjust. When I mean easy to adjust, the time you have taken to read this section is the time it will take to adjust one end stop. "It's really as simple as this"
This video show the construction a Captive Drum servo retro fitted into my 1/20 scale Jolie Brise sailing in a strong wind. It is the only solution as there is very little space.
When I first started researching ideas on how to control 4 sails on my 1/10 scale “Jolie Brise”. I thought the methods of running the sail sheets around the hull are against best engineering practice.
Bits of elastic or springs stopping the slack sheet from snaking about the bottom of the hull. There are much better methods.
My original brief was to replicate the main sheet arrangement of a 3 + 2 pully block system giving a 5:1 mechanical advantage. The boom travel is 600mm so the main sheet travel 5 x 600 = 3m. As the main sheet is run both sides of the deck, the required sheet to be wound on each drum will be 1.5m.
No existing sheeting methods could handle this.
If you goto Jolie Brise Winch you can see vidoes of my designs and my thought process.
The problem of the loose cord caused by letting out the sheet without any load on it could be solved by retaining the sheet on the drum. My design parameters needed to cater for different size drums and number of turns
The video shows why I stopped trying to design and make my own winch when this new product does everything I need.
BEC stands for Battery Elliminating Circuit. Most commonly found on ESC Motor Controller Circuits. They provide a 1.5v low current supply for the receiver and other 5v electronic devices such as motor sound.
An alternitive way to supply 5v electronic devices from 6v or higher batteries is to use a LBEC or UBEC. L = Linear. U = Ultimate = Digital.
The name for these products is misleading as they do not provide a fixed output voltage normally 5v. They should be called Voltage Limiters as this is actually what they do, they limit the output voltage to a fixed value (5v). The voltage drifts slightly up and down depending on the current load.
Once the supply voltage drops below 5.5 LBEC or 6.0v UBEC the output voltage is 0.5 or 1.0v lower than the input. The output voltage follows the input voltage.
For full details with graphs, battery voltages, efficiency/heat go to Voltage Regulators
Choosing a servo is a nightmare. Some stores offer hundreds varying in torque, speed, size, colour and price; others just a few. Then there is the concern over the very poor quality of imported products.
Drum servo manufactures give Basic Information such as stall torque, speed at no load, voltage but this turns out to be inaccurate and of little use.
Analog or Digital, what’s that all about?
Now there is the added choice between Captive Drum and Standard Drum.
Captive Drum Technology is a simple solution to the complex problem in controlling the sail sheet in confined spaces. A housing is placed around the drum and this retains the cord and stops it falling off the drum. No more cord loops, pulleys and springs.
The video shows the drum construction, retro fitted into my 1/20 model and sailing in a strong wind.
Online Calculator will work out the servo torque required from your boat and sail dimensions.
Sheet Travel is a product of the number of servo rotations and the drum diameter/circumference. It can be effected by the load.
Drum rotation is a major problem as it is directly dependant on the Receivers PWM pulse output. No two Tx/Rx systems are alike, rotation/travel will be different between channels.
It is possible to work out the receiver's PWM pulse width by using a Servo Testing Meter such as the GWS MT1 or by counting rotation angle with the Graduated Disc and sending me your servo and the angle, I can check it with my meter.
The sheet travel produced by analog or digital servos can be altered with a Servo Stretcher. SailServo digital servos can be Reprogrammed.
Stall Torque is used by manufacturers to signify the power of their servo. DO NOT use this information as all servos checked showed their data to be grossly inaccurate. Exaggerated by as much as 30 to 60%. Hitec 785 report is shown here as an illustration.
Ranking. A worldwide list of all drum sevos has been ranked by actual stall torque divided by speed in sec/100mm. 13 out of 26 have been tested giving real values. The remaining manufacturers were given the option of supplying me a sample to test or carrying out their own.
As manufacturer’s stall torque values have been found to be inaccurate, I have used 49% as the factor for calculating values not supplied.
Motor Speed Bands are indicators of a motor’s efficiency which decreases as loads are applied. These have been split into 10, 20, 30 and 40% bands. RMG recommend that servos are not used below 20% speed drop. Electric motors become inefficient below 50%.
Load v Time Graph is the best way to compare servo characteristics. As servos have different drum diameters, time for pulling in 100mm is used as the standard.
Product Testing is carried out on all products sold as quality control is a major concern of my customers and I do not trust manufacturer’s data.
Research to find large winches for my 1.7m 1/10 scale French Pilot Cutter lead me to develop the idea of the “Captive Drum” to avoid all those pulleys and springs. As an engineer I hate levers. One day a Chinese servo manufactuer approached me to buy their servos, hence this website.
Going right back to my opening statement.
As consumers, what do we actually want from our r/c equipment?
Move a joystick over a certain distance to control either a servo to rotate a drum for the sails, arm for the rudder or an ESC for the motor with a degree of accuracy and life.
As consumers, we have no way in buying with confidence. The only good thing was the introduction of 2.4Ghz removing the problem of having to have a box of crystals and changing them beside the pond.
The degree of accuracy is down to the purpose and pocket.
At least you now have some facts to go on, what products are available and what questions to ask.
Click the highlighted servo to see it’s Test Report.
Stall Torque for non tested servos is 49% the median of tested servos.
No load speeds appear to be accurate.
Remember stall torque and no load speeds are not indicators of a servo’s performance.
Make | Model Tests | Rank no T/S | £ | An Di | Stall Kg.cm Man Test % | Speed s/100mm Man Tes | V | Dia | Tur | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
RMG | 380EH | 1 | 45.8 | 175 | D | 29.9 | 14.7 | 49 | 0.3 | 0.3 | 6.0 | 30 | 6 |
RMG | 280EF | 2 | 29.9 | 175 | D | 12.2 | 6.0 | 49 | 0.2 | 0.2 | 6.0 | 25 | 6 |
RMG | 280EL | 3 | 28.7 | 175 | D | 19.3 | 9.5 | 49 | 0.3 | 0.3 | 6.0 | 25 | 6 |
RMG | 280ES | 4 | 25.2 | 175 | D | 15.4 | 7.5 | 49 | 0.3 | 0.3 | 6.0 | 30 | 3.2 |
HobbyKing | 11AYMB | 5 | 15.9 | $20 | A | 50.0 | 14.3 | 29 | 0.8 | 0.9 | 7.2 | 30 | 6 |
Titan | Titan | 6 | 13.3 | 95 | D | 5.3 | 0.4 | 7.2 | 26 | 2-6 | |||
HobbyKing | 11YMB | 6 | 13.3 | $18 | A | 40.0 | 11.6 | 29 | 0.9 | 0.9 | 7.2 | 30 | |
*Graupner | Evolution | 6 | 13.3 | 135 | ? | 40.0 | 11.6 | 29 | 0.9 | 0.9 | 7.2 | 30 | |
Hobbyking | 22YMB | 7 | 11.3 | $19 | D | 11.0 | 6.8 | 62 | 0.6 | 0.6 | 7.2 | 30 | |
*Graupner | ECO II | 7 | 11.3 | 73 | D | 11.0 | 6.8 | 62 | 0.6 | 0.6 | 7.2 | 30 | |
HobbyKing | VSD-10Y | 8 | 9.2 | $15 | D | 12.0 | 8.3 | 69 | 0.8 | 0.9 | 7.2 | 30 | |
*Graupner | Regatta II | 8 | 9.2 | 85 | D | 12.0 | 8.3 | 69 | 0.8 | 0.9 | 7.2 | 30 | |
SailServo | Dcd2 | 9 | 8.6 | 50 | D | 14.5 | 6.9 | 48 | 0.8 | 0.8 | 6.0 | 25 | 6 |
*Graupner | Eco Speed | 9 | 8.6 | 68 | D | 14.5 | 6.9 | 49 | 0.8 | 0.8 | 6.0 | 25 | 6 |
SailServo | Dcd1 | 10 | 6.3 | 50 | D | 8.5 | 2.5 | 29 | 0.2 | 0.4 | 6.0 | 25 | 6 |
Robbe | 28371 | 11 | 6.2 | 99 | A | 10.2 | 5.0 | 49 | 0.8 | 0.8 | 6.0 | 40 | 6 |
GWS | 125s-12td | 12 | 5.8 | 20 | D | 10.2 | 7.5 | 74 | 1.0 | 1.3 | 6.0 | 40 | 12 |
CYS | S960 | 13 | 5.2 | ? | A | 20.0 | 9.8 | 49 | 1.9 | 1.9 | 7.2 | 35 | 5 |
Saturn | 12.Winch | 14 | 5.1 | 25 | ? | 12.0 | 5.9 | 49 | 1.2 | 1.2 | 6.0 | 40? | |
Graupner | Winch 4 | 15 | 4.7 | 114 | ? | 4.6 | 2.3 | 49 | 0.5 | 0.5 | 6.0 | 30 | 8.5 |
GWS | 125s-6td | 16 | 4.7 | 20 | D | 10.2 | 6.5 | 64 | 1.0 | 1.4 | 6.0 | 40 | 6 |
GWS | 125s-3td | 17 | 4.7 | 20 | D | 10.2 | 7.5 | 74 | 1.0 | 1.6 | 6.0 | 40 | 3 |
GWS | 125s-1ta | 18 | 4.7 | 17 | A | 7.6 | 7.0 | 92 | 1.0 | 1.5 | 6.0 | 40 | 1 |
GWS | 125s-1/2ta | 19 | 3.8 | 17 | A | 7.6 | 6.5 | 86 | 1.0 | 1.7 | 6.0 | 40 | 1/2 |
SailServo | Acd1-6t | 20 | 3.8 | 13 | A | 10.6 | 7.5 | 71 | 0.9 | 2.2 | 6.0 | 25 | 6 |
Ace | wq2811 | 21 | 3.6 | 22 | ? | 11.0 | 5.4 | 49 | 1.5 | 1.5 | 6.0 | 35 | 2 |
Hitec | HS-785 | 22 | 3.2 | 26 | A | 13.2 | 7.5 | 57 | 1.1 | 2.3 | 6.0 | 40 | 3.5 |
SailServo | Acd2-4t | 23 | 2.1 | 9 | A | 6.1 | 3.0 | 49 | 0.9 | 1.6 | 6.0 | 25 | 4 |
SailServo | Acd3-1.5t | 24 | 1.3 | 9 | A | 6.3 | 2.2 | 36 | 0.9 | 1.7 | 6.0 | 25 | 1.5 |
Futaba | S5801 | Discon | 12.6 | n/a | A | 9.8 | 4.8 | 49 | 0.4 | 0.4 | 7.2 | 40 | 6 |
Graupner | Regatta | Discon | 8.9 | 120 | ? | 10.0 | 4.9 | 49 | 0.6 | 0.6 | 7.2 | 40 | 5.5 |
Graupner | Eco | Discon | 1.5 | 65 | ? | 4.6 | 2.3 | 49 | 1.5 | 1.5 | 6.0 | 25 | 5.5 |
*Graupner, Eurgle and Conrad are rebadged versions of HobbyKing. | |||||||||||||
Ranking by Stall Torque & Speed sec/100mm found from testing or using an average. | |||||||||||||
ACE — — CYS — — Futaba (no longer in production) — — Graupner — — GWS — — Hitec — — HobbyKing — — RMG — — Robbe — — Saturn
Product Specifications.PDF
Hitec 785's data was the most detailed but found by accident. Much of the other data has been taken from various sources and the rest I have measured and calculated. As an ex–marketing manager, I am dissapointed at the lack of information given by suppliers.
The servo motor’s speed drop can be found by timing a number of turns.
Electric motors should never be used below 50% speed.
One feature of these tests is to band results by 10, 20, 30 and 40% motor speed reduction as this allows the consumer to make a choice dependant on their use.
I recommend that you used upto the 20% band and 10% if you are working the servo hard as it will prolong its life. Using below 40% can damage the servo and voids the guarantee.
This is not the complete picture as some servos struggle to pull the sheet fully in. See 19 below.
This is the only way to show data for choosing a servo.
This problem has never been hightlighted before, it is an indication of the servo’s power.
The sheet not being pulled back to zero under load could be important in your application. The drum being rotated further at the out postion occurs at large loads on some servos and should not cause any problems.
Sheet travel is totally dependant on your Tx/Rx system. The receiver output drives the servo’s rotation. This presents a major problem. No manufacturer publish the output PMW signals for their receivers and channels. It must assumed that all channels are the standard 1.0 to 2.0ms.
On contacting Spectrum about their Dx5e/AR500 Tx/Rx system they gave conflicting answers. UK 1.1 to 1.9ms. US 1.0 to 2.0ms.
Testing gives 1.03 to 1.77ms (740µs) which is close to the UK 1.1 to 1.9ms (800µs).
The CD1 servo runs continuously over 2.1ms. A customer complained that his CD1 ran continuously so I swapped it for a CD2 which solved his problems. His Tx/Rx was a Planet T5 and the suppliers response was "The Planet radio uses the nominal standard pulse width of 1.0 to 2.0ms"
The latest batch of CD1’s pulse width are 0.8 to 2.2ms.
Below are results for nine different Tx/Rx systems using my GWS MT1 Pulse Meter and a Hitec 785 drum servo to measure the sheet travel.
The first surprise is the wide difference between channels. 26% for Futuba and 8% for Spectrum. It cannot be accidental as the two Futuba receivers gave identical readings.
The difference of 150mm (26%) sheet travel between the highest and lowest PWM values is worrying. How can customers make a rational choice in replacing an existing servo or buying new.
There is a difference in sheet travel between batches of the same servos. Four servo models gave a 3% difference including two digitals. Two servo models gave 8 and 10%.
Spektrum 2.4Ghz AR500 Receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1054 | 1794 | 740 | 26% | 160 |
R U – D | 2 | 1165 | 1930 | 765 | 24% | 265 |
L U – D | 3 | 1190 | 1855 | 765 | 24% | 265 |
L L – R | 4 | 1127 | 1872 | 735 | 27% | 254 |
Manu US | 1000 | 2000 | 1000 | 100% | ||
Manu UK | 4 | 900 | 1900 | 800 | 20% | |
Spectrum 2.4Ghz + Orange RX | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1060 | 1812 | 750 | 25% | 260 |
R U – D | 2 | 1104 | 1893 | 790 | 21% | 275 |
L U – D | 3 | 1100 | 1891 | 790 | 21% | 275 |
L L – R | 4 | 1139 | 1887 | 750 | 25% | 260 |
GWS 2.4Ghz R–4S Receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1278 | 1734 | 455 | 54% | 160 |
R U – D | 2 | 1320 | 1765 | 445 | 56% | 155 |
L U – D | 3 | 1103 | 1916 | 815 | 19% | 280 |
L L – R | 4 | 1305 | 1765 | 460 | 54% | 160 |
Manu | 1250 | 1750 | 500 | 50% | ||
Manu | 3 | 1100 | 1900 | 800 | 20% |
Eurgle 3CH 2.4Ghz receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1145 | 1864 | 720 | 28% | 250 |
L U – D | 3 | 1091 | 1844 | 755 | 25% | 260 |
Turnigy TGY9X 2.4Ghz receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1068 | 1903 | 835 | 17% | 290 |
R U – D | 2 | 1063 | 1907 | 845 | 16% | 290 |
L U – D | 3 | 1067 | 1904 | 840 | 17% | 290 |
L L – R | 4 | 1052 | 1905 | 855 | 15% | 295 |
HobbyKing HK6DF 2.4Ghz receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1185 | 1970 | 785 | 22% | 270 |
R U – D | 2 | 1125 | 1890 | 765 | 24% | 265 |
L U – D | 3 | 1170 | 2020 | 850 | 15% | 295 |
L L – R | 4 | 1140 | 1865 | 725 | 28% | 250 |
Multiplex Light 2.4Ghz receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1042 | 2138 | 1100 | +10% | 380 |
R U – D | 2 | 1042 | 2138 | 1100 | +10% | 380 |
L U – D | 3 | 1042 | 2068 | 1025 | +3% | 355 |
L L – R | 4 | 1063 | 2139 | 1075 | +8% | 370 |
Stix—Cs AM 27Mhz receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 945 | 1830 | 885 | 12% | 305 |
L U – D | 3 | 990 | 1900 | 910 | 9% | 315 |
Futuba Skysport 4 FM 35Mhz receiver | ||||||
---|---|---|---|---|---|---|
Input | Ch | Min | Max | Bandwidth µs | Below 1 to 2ms | Sheet Travel mm |
R L – R | 1 | 1111 | 1944 | 835 | 17% | 290 |
R U – D | 2 | 1037 | 1902 | 865 | 14% | 300 |
L U – D | 3 | 1215 | 1870 | 655 | 35% | 230 |
L L – R | 4 | 1182 | 2064 | 880 | 12% | 305 |
My interest in Radio Controlled equipment is for marine use especially sailing boats. This puts it apart from planes, cars and robots and introduces its own set of requirements and limitations.
It all started off when I looked for a servo to use in my 1/10 scale Jolie Brise, whose main boom requires a massive sheet load of 5kg to pull it in.
The first thing to do was to find out how other people controlled their sail sheets; only to be horrified to find that everybody uses either arms on the servo or complex arrangements of sail drum, loops of cord, pulleys and springs. Both methods defy logic and basic physics of mechanisms.
Just because it has been used for years does not make it right. Jolie Brise Models
Converting rotary to linear motion can only be achieved efficiently by using a drum servo. The major problem was to find a method of keeping the sail cord on the drum when the servo was let out with no load on the sheet. My last attempt was to run the sail drum on a threaded shaft. When the servo was rotated with no sail sheet load, the drum remained stationary but moved sideways along the threaded shaft. When the wind picked up, the sheet rotated the drum back along the threaded shaft until it hit the driving dogs. It is as complicated to make as it is to describe.
An email came out of the blue from a Chinese supplier with a ready made solution. They placed housing around the drum. Having spent years in product design, sales and marketing I immediately realised that this was a “must have” product, so SailServo was set up.
There are too many stories of poor quality products, so all servos will be tested and only those meeting the “Quality Control Standards” will be sold.
Having worked in an Inspection Department, it is important to wite a “Test Proceed” that will produce repeatable results. It will contain a list of tests with levels of acceptable “Quality Indicators”. These can then be agreed with the supplier to avoid disputes.
From a well known Hong Kong supplier. Out of 2 batches of 5 servos, Dead on Arrival (1), Running slow (1), Erratic behaviour (1), Bottom seal failure (2).
Two factors I had never considered before testing were that some servos:
The past year has been spent refining a method Quality Control not only to check quality but to give my customer’s the true information about the R/C products they buy from me.
The tiny electric motors in servos cannot be used anywhere near their stall torque. The standard SailServo uses is the sheet speed in sec/100mm for a load which drops the motor’s speed by 20%. Speed for 60 degrees at no load is irrelevant as is stall torque.