Basic (and less basic) stuff about servos, radios, pulse width and frame rates - HeliFreak
Fun, Learning, Friendship and Mutual Respect START  HERE


Unregistered
Go Back   HeliFreak > R/C Helicopter Support > Antique Out of Production and Vintage Helicopters > E-Sky Helicopters > Belt CP


Belt CP E-Sky Belt CP


Reply
 
LinkBack Thread Tools Display Modes
Old 12-14-2010, 01:01 PM   #1 (permalink)
Registered Users
 
Posts: 4,108
 
My HF Map location
Join Date: Mar 2009
Location: Belgium
Lightbulb Basic (and less basic) stuff about servos, radios, pulse width and frame rates

Ever wondered about some of this stuff, but never bothered to look into it in depth ?

Ř Analog and digital servos ?
Ř High voltage servos ?
Ř Brushless and coreless servo motors ?
Ř 50 Hertz and up to 333 Hertz frame rate capable servos ?
Ř 1520 and 760 microsecond pulse width servos ?
Ř Radio frame rates like Futaba’s 7 - 14 ms specs, or Spektrum’s 11 - 22 ms ?

Might be a good idea to know about it anyway, as this technology is rapidly spreading. “Yes, but sounds too complicated, I don’t want to know.” OK, fair enough, but would you still think the same after you burned a high end 760 µs tailservo because you didn’t program the right frame rate in your gyro before connecting it ? Or just bought the wrong servo for your setup ? Or discover that you don’t use your actual gear at its full possibilities ?

Well, good news, it’s not that hard to understand, and it’s all somewhat related to each other. But it’s not that easy to find a single document that tells the whole story from the beginning to the end in a rather simple way, so here’s my attempt...


1) Servo basics:

Typically, a 3-stranded cable links the servo to the receiver.

The two most popular wiring color schemes are Futaba and JR which are basically compatible but color-coded differently as follows:
JR - Signal=Orange, Positive=Red, Negative=Brown
Futaba - Signal=White, Positive=Red, Negative=Black

Here are most of the color schemes:



There is nothing mysterious about the + and – connections; they simply connect, via the receiver, to the power source and don’t carry any signals. Most servos are rated at either 4.8 or 6.0 volts, even up to 8 volts sometimes nowadays, and you should take care to ensure that your servos are suitable for the battery or BEC voltage in your model.

The third ‘signal’ wire (white or orange) carries the information from the receiver telling the servo what position it should be in. It does this using a system called Pulse Width Modulation (more about PWM later).

At the end of the cable is a ‘Z-connector’ which plugs into the receiver. Futaba-style versions of these connectors have a small plastic blade on one side which ensures it is inserted the right way into the receiver. For some reason JR/Spektrum don’t have this feature which means that if you want to plug a Futaba-style servo into a JR-style receiver you will need to clip the blade off. A useful characteristic of the -/+/S order of the wires is that if you accidentally plug a connector in the wrong way round it shouldn’t cause any damage. On standard radio control servos the range of movement is usually a little over 90 degrees.


2) Pulse Width Modulation:

It’s not really necessary to understand the internal workings of the servo but it is useful to understand how the signals from the receiver control its position. As the name implies, the positional information coming from the receiver is in the form of electric pulses which vary in length. Not in height, as the height of the pulses is simply always the same as the voltage of the power supply.

Put in very simple terms, the signal wire is normally connected to the negative voltage supply but for a fraction of a second the receiver connects it to the positive. This change in voltage is known as a 'pulse'. The servo senses the duration (or length if you wish) of the pulse and using some electronic wizardry the output shaft is placed in the appropriate position.

The length of a standard pulse for a common servo varies between 1.0 and 2.0 milliseconds (ms) where, for example, 1.0 ms might indicate a required position fully to the right, 2.0 ms fully to the left and with a pulse of 1.5 ms the servo would be centered. However, servos will actually allow a greater range of movement than this and the programming of transmitters will often allow the range of pulse widths to be extended from 0.7 to 2.3 ms or more.


3) Pulse Frequency:

The pulses come out of a common receiver at a rate of 50 pulses per second (in other words one every 20 ms). This interval is known as the ‘frame rate’ or ‘pulse repetition rate’ and this defines the rate at which the servo can be given different positional commands.

Don’t confuse this with the length of a single pulse, as described above. Remember, each 20 ms frame contains one pulse with a length of 1 to max 2 ms (in the most common case). BTW, the frame rate is not critical at all for the servo, but the pulse width is, as it accurately defines the position of the servo arm.

It is like someone sitting next to you in your car, and repeatedly telling you to keep driving at 50 km/h. It is not how many times each minute that he repeats this sentence that will make for the best accuracy, but more the fact that he indicates exactly “50”. Similar with the servo, the pulse width is his position reference, not the frames that carry the pulses, and which come in a specific, but not so important rate. I say “not so”, because there are some benefits: if a servo is told more times per second to hold its position, it will have less time to be pulled away from that position by external forces, and will keep that position more accurately when large forces are involved.

The electric signals and servo arm positions look like this, for a typical and common servo:


You notice that we are always using a 5 Volt power supply here, and that the pulse width is changed from 1 to 2 ms for the most extreme positions of the arms, while the frame rate remains always unchanged, 20 ms in this case. Being 50 frames (and short pulses) each second. Clear ?

Just for the record, here is the diagram of a servo:



(the position sensor is in fact just a potentiometer, a known cause of problems)


An example:

The receiver in a radio controlled plane is sending a constant series of pulse to a rudder servo. The pilot's rudder stick is centered which means that every 1/50th of a second the receiver sends a 1.5ms pulse to the rudder servo. Just after it received its last pulse the pilot suddenly moves the rudder to the left which means that the next pulse will have a duration of, say, 1.25ms. However, the servo won't 'see' this change until 1/50th second later because it has to wait for that next pulse.

Sending pulses every 1/50th second (i.e. every 20ms) may sound fast but it’s not fast enough for some applications. For example high-speed tail servos on radio controlled helicopters which, instead of being connected to the receiver, plug into the gyro. These special ‘super servos’ (e.g. the Futaba S9251) are updated by special gyros (like a GY-601/611) at a much faster rate, e.g. 333 times per second (i.e. every 3 ms).

Because these signals are coming from the gyro so frequently there isn’t time for standard pulses (they would still be 'high' when it was time for the next one to start!) so manufacturers reduce them to between 0.5 ms to 1.0 ms. (Note: To prevent damage to them, servos intended for a standard frame rate should not be used with ‘super servo’ devices and visa versa.) So, a faster frame rate improves ‘latency’ which is the time it takes a system to react to a control input.

In fact there are two main characteristics that affect the speed with which a servo can move:

a) The first is the speed that the motor can move the servo horn. Generally manufacturers measure this speed as the time it takes for the servo to travel through 60 degrees.

b) However, this doesn’t take account of how long it takes for the servo to recognise the fact that it needs to move - as we have seen, this is limited by the frequency of those pulses (frame rate !). It may seem incredible that the required servo position needs to be updated every 3 ms but it can make a noticeable difference in demanding situations. It’s also worth noting that latency can be caused by the transmitter/receiver system but that’s another story…

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.


4) Types of Servo:

We’ve talked about their speed but there are many other characteristics that will affect the performance and price of a servo.

For example there is the torque that a servo can deliver. This is the twisting force that it can provide before ‘stalling’ which is measured in various (sometimes rather unscientific) units such as ounce-inches, kilogramme-centimetres or Newton-centimetres.

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.

Some servos have high-performance motors to achieve better speeds or reliability. The coreless motor is an example in which the weight of the armature (the bit that moves) is vastly reduced by eliminating the heavy steel core. The end result is a motor that is smoother, more constant, and stronger action. Futaba introduced the first servo that uses a brushless motor – a technology that promises to prove far more reliable than older designs.


5) The bottom line of the radio-servo story in a nutshell:

Let’s take a familiar example, like the Spektrum DX6 or DX7 radio. This one has a 22 ms frame rate. Note: The DX7SE and DX8 have 11 ms, typical 2.4 Ghz Futabas have 14 and 7 ms frame rates.

One frame is 22 ms long in this case, meaning the radio system will send 1/0.022 = about 45 frames each second to any connected servo. As we read above, each 22 ms long frame contains one single pulse, with a length varying between 1 and 2 ms roughly, for a normal servo which centers at 1.52 ms. Yes, that’s the middle position of that servo, right between 1 and 2 ms, that’s why it is called a 1520 µs servo.

Analog and also digital servos are OK with this rather low frame rate, but some digital ones are capable of receiving much more frames each second, up to 333. For example the Futaba BLS251. They still do work fine at a lower frame rate though. Sending the frames as fast as 333 times each second, means that each frame is only 3 ms long instead of 20 ms, leaving barely enough room to accommodate the common 2 ms long pulse ! And that’s why manufacturers are starting to use smaller pulses, like the ones used in the 760 µs servos, these pulses are roughly only between 0.5 and 1 ms long. And yes again, center position of these servo’s is 0.76 ms, right between 0.5 and 1 ms, hence the name 760 µs servo.

Now don’t be fooled, even very fast systems like the 7 ms Futaba transmitters are still only capable of sending 1/0.007 = 143 frames each second, they still can’t use the full frame rate capability (333 Hz) of the high-end digital servos ! That’s why you don’t see cyclic 760 µs servos, there is no real use, the usual 1520 µs servos do just as fine in this situation.

But a high-end gyro CAN send frame rates this fast to a connected servo, as it has total communication with and control over the tailservo. And that’s why we do see these 760µs/333Hz tailservos.





There is much more to say, electronic experts would add a lot of notes, for example about 1024 and 2048 resolution transmitters, and all the commercial hype and small usable benefits of some of this technology in real life situations, but I guess this covers the most essential knowledge. A bit complicated at first sight, I know, but I hope it sheds some light anyway.

Enjoy, or not... It might read easier after a few beers.
And forgive me for copying some lines from other sources this time, time is money.
Attached Thumbnails
Click image for larger version

Name:	servo_wirecolor.gif
Views:	38341
Size:	21.9 KB
ID:	195222   Click image for larger version

Name:	servo_signal.png
Views:	25351
Size:	10.2 KB
ID:	195223   Click image for larger version

Name:	servo_diagram.png
Views:	20118
Size:	7.4 KB
ID:	195224   Click image for larger version

Name:	servo_BLS251.jpg
Views:	19138
Size:	23.9 KB
ID:	195225  
__________________
TDR 14S - TDR 12S - Logo 500SE - Logo 400 - Furion 450 6S mini-V - 300X - 2 x mCPX - Schluter Bell Huey Cobra '74 - Aura 4 hotliner
redbird300 is offline        Reply With Quote Quick reply to this message
Sponsored Links
Advertisement
 
Old 12-14-2010, 02:16 PM   #2 (permalink)
Registered Users
 
Posts: 3,738
 

Join Date: Aug 2009
Location: Norway , from Sweden
Default

nice work as always Raf

wondering one more thing as I´m trying to understand the servo spes:
the torque value , it is the higher the value the better servo ex: two servo that are almost similer but the torque value: (500-size heli)
servo 1: 0.088 sec/60° / 0.068 sec/60°
• Stall Torque (4.8V): 2.26 kg.cm (31.4 oz/in)
• Stall Torque (6.0V): 2.83 kg.cm (38.3 oz/in)
servo 2: 0.12 sec/60° / 0.09 sec/60°
• Stall Torque (4.8V): 3.8 kg.cm (52.8 oz/in)
• Stall Torque (6.0V): 4.7 kg.cm (65.3 oz/in)
__________________
| Atom 6HV : vbar NEO pro rescue | Goblin 700 : vbar 5,3 pro | HD Cell 500 : brain2 |
ragge1 is offline        Reply With Quote Quick reply to this message
Old 12-14-2010, 02:21 PM   #3 (permalink)
Registered Users
 

Join Date: Aug 2010
Location: L.A
Default

Great read Raf, learned a few things myself!
UPGl2AYDD is offline        Reply With Quote Quick reply to this message
Old 12-14-2010, 02:24 PM   #4 (permalink)
Registered Users
 
Posts: 4,108
Thread Starter Thread Starter
 
My HF Map location
Join Date: Mar 2009
Location: Belgium
Default

Quote:
Originally Posted by ragge1 View Post
nice work as always Raf

wondering one more thing as I´m trying to understand the servo spes:
the torque value , it is the higher the value the better servo ex: two servo that are almost similer but the torque value: (500-size heli)
servo 1: 0.088 sec/60° / 0.068 sec/60°
• Stall Torque (4.8V): 2.26 kg.cm (31.4 oz/in)
• Stall Torque (6.0V): 2.83 kg.cm (38.3 oz/in)
servo 2: 0.12 sec/60° / 0.09 sec/60°
• Stall Torque (4.8V): 3.8 kg.cm (52.8 oz/in)
• Stall Torque (6.0V): 4.7 kg.cm (65.3 oz/in)
Servo 2 has enough torque for use as a cyclic servo on a 500, while still being fast enough. Servo 1 looks more like a typical tailservo to me: faster but slightly weaker.
__________________
TDR 14S - TDR 12S - Logo 500SE - Logo 400 - Furion 450 6S mini-V - 300X - 2 x mCPX - Schluter Bell Huey Cobra '74 - Aura 4 hotliner
redbird300 is offline        Reply With Quote Quick reply to this message
Old 12-14-2010, 02:27 PM   #5 (permalink)
Registered Users
 
Posts: 5,405
 

Join Date: Mar 2008
Location: Wales, UK
Default

Hey Raf. Another great post. I understood 95% of that. I started to loose concentration near the end of the read because of my surroundings and i'm really tired. Anyway, what i'm saying is that you have taken a complicated subject it made it easy to understand.

Now I feel great because I have knowledge in another department in this hobby (so much stuff to learn in this hobby) Now I will know a lot more about what I am buying next time, and can make the decisions myself if the servos are suitable for the setup I want and not ask "Will this work with this" all the time.

Cheers Raf!

-Jonny
__________________
Spektrum DX7, Align Trex 250se, Align Trex 500cf, Align Trex 600n LE
Jonnyheli is offline        Reply With Quote Quick reply to this message
Old 12-14-2010, 07:01 PM   #6 (permalink)
Registered Users
 
Posts: 10,134
 
My HF Map location
Join Date: Feb 2008
Location: Manchester, England
Default

Hey Raf, Good stuff as always, and has just come in handy as I am about to take receipt of my Beastx flybarless controller and it has helped me understand a little more about my servo type choices.

From your other recent thread on basic electronics, I noted that you wrote that an analogue servo won't cut it for most on an FBL system. From what I have just discovered, my digital tail servo, Futaba S9257, will be fine with the BeastX, if I set it to 1520 us, and 333 Hz, and also according to what I have found in the BeastX manual, so should my analogue cyclic servos, Hitec HS-65MG, if I set it to 1520 us and 65 Hz.

So my question is, am I going to be so disappointed with the performance, using these analogue servos for the cyclic, that I end up on an almost immediate upgrade path to some faster digital version, in the way that I did for my tail on the 450 and the 250? And, if so, where would I need to pitch my choice in terms of torque/speed and frame rate? Again from the manual it seems clear that even the servo choices that are capable of 333 Hz they suggest setting it to 200 Hz for the cyclic. I wonder why they don't suggest the full 333 Hz here?

I know you said won't cut it for most, but I'm only just starting out on 3D, do you think I would have the skills to even notice, or might it be obvious, just like with a tail?

Also, one last question, as an aside, how variable is the width of the pulse? In other words, can it be 1521.525, or can it only be 1521, or 1530, for example. Surely this affects how accurately the servo can be positioned, not mentioning the mechanical restraints of such accuracy. And also why is 1520 the middle, could in not equally well 1500, or is this something to do with the step sizes, or more accurately the pulse width variability here?

Cheers

Sutty
__________________
Regards, Sutty
BeltEXI, mCX, mSR, 4G3, FBL Trex250SE µB, FBL Trex450Pro µB,
FBL Trex550e VX1 Pro, FBL Trex600ESP VX1e, Parkzone P51D, QRX350Pro
sutty is offline        Reply With Quote Quick reply to this message
Old 12-14-2010, 10:14 PM   #7 (permalink)
Registered Users
 
Posts: 5,174
 

Join Date: May 2010
Location: Erie,Colorado
Default

Damn, Raf! They really need to find you something to do at work!
I don't want anybody calling me long winded again.

Seriously though, great stuff, as always, as expected.
__________________
Avant Aurora 90n,Trex 600E Pro, HC500, HDX450sev4, Trex 250se, mCP-X, Mosquito, X/A Hexa/WKM/UHF/FPV, F330 Flame Wheel,
PenguinV2/UHF/FPV, 757-3 ex Ranger/uhf/Ruby/fpv, QAV250 w/Vector , Ritewing Drak w/Vector Crossfire and fpv.
Tomstoy2 is offline        Reply With Quote Quick reply to this message
Old 12-15-2010, 03:15 AM   #8 (permalink)
Registered Users
 
Posts: 988
 
My HF Map location
Join Date: May 2008
Location: Bangor, N. Ireland
Default

Quote:
Originally Posted by ragge1 View Post
nice work as always Raf

wondering one more thing as I´m trying to understand the servo spes:
the torque value , it is the higher the value the better servo ex: two servo that are almost similer but the torque value: (500-size heli)
servo 1: 0.088 sec/60° / 0.068 sec/60°
• Stall Torque (4.8V): 2.26 kg.cm (31.4 oz/in)
• Stall Torque (6.0V): 2.83 kg.cm (38.3 oz/in)
servo 2: 0.12 sec/60° / 0.09 sec/60°
• Stall Torque (4.8V): 3.8 kg.cm (52.8 oz/in)
• Stall Torque (6.0V): 4.7 kg.cm (65.3 oz/in)
When a manufacturer has two almost identical servos, but one has more torque, it will almost certainly be slower, as in the example above. That's because it probably is identical (motor, electronics) except for the gears between the motor shaft and the servo output shaft. Lower gearing gives move torque, but makes the servo slower, higher gearing makes for a faster servo, but with less torque.

Like in your car, you need lower gear to go up a hill, because you need torque, but you're speed is limited. High gear you go faster, but won't make it up the hill (unless you downshift).

So you can't just say that higher torque is better - the real question is, better for what? As Raf said, servo 1 in your example is better as a tail servo, as the tail needs a faster servo but torque is not so important. Servo 2 is better as a cyclic servo, as it has more torque, and speed is not so important on the cyclic.
__________________
Helis: Belt-CP, Mini Titan, EXI 450, HK250-GT, HK500-GT, HK600-GT, mSR (all modified/customised). Quads: Q4, nQX, HK250, Custom built 560 Deadcat camera platform. Some planks. DX6i & DX9. And a boat & some HK surface controller thingy.
CyprusDave is offline        Reply With Quote Quick reply to this message
Old 12-15-2010, 04:06 AM   #9 (permalink)
Registered Users
 

Join Date: Aug 2010
Location: L.A
Default

Quote:
Originally Posted by CyprusDave View Post
When a manufacturer has two almost identical servos, but one has more torque, it will almost certainly be slower, as in the example above. That's because it probably is identical (motor, electronics) except for the gears between the motor shaft and the servo output shaft. Lower gearing gives move torque, but makes the servo slower, higher gearing makes for a faster servo, but with less torque.

Like in your car, you need lower gear to go up a hill, because you need torque, but you're speed is limited. High gear you go faster, but won't make it up the hill (unless you downshift).

So you can't just say that higher torque is better - the real question is, better for what? As Raf said, servo 1 in your example is better as a tail servo, as the tail needs a faster servo but torque is not so important. Servo 2 is better as a cyclic servo, as it has more torque, and speed is not so important on the cyclic.
Perfect way to explain it Dave, an automotive transmission gearing system.
UPGl2AYDD is offline        Reply With Quote Quick reply to this message
Old 12-15-2010, 04:08 AM   #10 (permalink)
Registered Users
 
Posts: 4,108
Thread Starter Thread Starter
 
My HF Map location
Join Date: Mar 2009
Location: Belgium
Default

Quote:
Originally Posted by sutty View Post
Hey Raf, Good stuff as always, and has just come in handy as I am about to take receipt of my Beastx flybarless controller and it has helped me understand a little more about my servo type choices.

From your other recent thread on basic electronics, I noted that you wrote that an analogue servo won't cut it for most on an FBL system. From what I have just discovered, my digital tail servo, Futaba S9257, will be fine with the BeastX, if I set it to 1520 us, and 333 Hz, and also according to what I have found in the BeastX manual, so should my analogue cyclic servos, Hitec HS-65MG, if I set it to 1520 us and 65 Hz.

So my question is, am I going to be so disappointed with the performance, using these analogue servos for the cyclic, that I end up on an almost immediate upgrade path to some faster digital version, in the way that I did for my tail on the 450 and the 250? And, if so, where would I need to pitch my choice in terms of torque/speed and frame rate? Again from the manual it seems clear that even the servo choices that are capable of 333 Hz they suggest setting it to 200 Hz for the cyclic. I wonder why they don't suggest the full 333 Hz here?

I know you said won't cut it for most, but I'm only just starting out on 3D, do you think I would have the skills to even notice, or might it be obvious, just like with a tail?

Also, one last question, as an aside, how variable is the width of the pulse? In other words, can it be 1521.525, or can it only be 1521, or 1530, for example. Surely this affects how accurately the servo can be positioned, not mentioning the mechanical restraints of such accuracy. And also why is 1520 the middle, could in not equally well 1500, or is this something to do with the step sizes, or more accurately the pulse width variability here?

Cheers

Sutty
Andrew, sorry but please understand that I really can’t tell you for sure how a BeastX would work with those analogue servos, better ask these kind of questions in some flybarles forum. I’m a V-Bar fanboy, lol, don’t know those BeastX that well, and you know how I hate guessing. Mikado V-Bars do insist on working with digital servos. What I can tell you is that any typical 450 digital servo would do the job very well, any medium quality brand would be OK. Sounds a shame to me to put the cheapest possible digital servo on a FBL system. The servo movements need to replace a flybar in a FBL setup, and I can hardly imagine that analogue servos would do just as well.

Answering your last question would need another post about twice the size as my last one. I intentionally did not talk about that, it’s a science on its own. There is much more as only the pulse width that matters here, like for example radio resolution (1024 or 2048 on newer models), servo motor precision, servo internal electronics (e.g. the feed back loop), the way the signal is processed in the TX, combined to a radio signal where information for different channels can be multiplexed or not, reprocessing the signal in the RX, correction error bits, servo gear ratio and slop, etc….

If you look up some information about the subject, you ‘ll find an enormous amount of discussions going on. In short, getting 2048 steps from a typical R/C servo in real life conditions is generally considered unlikely, best to stick with the 1024 number for a careful estimation. Divide the total possible range from about 0.7 ms to 2.3 ms by 1024, and you’ll have a rough idea. That’s about 1.5 µs. Don’t consider this as the whole and only truth though ! Some people even think that linkage slops and gear slop has much more influence than the typical, much finer r/c servo resolution, and therefore consider the discussion irrelevant. Another thing, deadband resolution on many (even good) servo’s is around 1 - 2 µseconds or more, making the above calculation pretty irrelevant again, lol. I would need to do some tests with a servo with a very long servo arm, and a decent programmable digital signal pulse generator, where I would look for the smallest possible pulse difference that would still cause a reliable and reproducible movement of the arm. But even I don’t go there, lol. BTW, Futaba servos typically use 1520 µs for center position, where many others use 1500 µs as the center position.
__________________
TDR 14S - TDR 12S - Logo 500SE - Logo 400 - Furion 450 6S mini-V - 300X - 2 x mCPX - Schluter Bell Huey Cobra '74 - Aura 4 hotliner
redbird300 is offline        Reply With Quote Quick reply to this message
Old 12-15-2010, 12:36 PM   #11 (permalink)
Registered Users
 
Posts: 10,134
 
My HF Map location
Join Date: Feb 2008
Location: Manchester, England
Default

Damn it, I have just written a long post twice, and now lost it for the second time because I went off doing something else. I had copied it to the clipboard, but then copied an e-mail address the first time, and then a web URL the second time, whilst not thinking, idiot!

Anyway I can only be bothered to write the first bit this time. Here it is.

Thanks Raf, I appreciate you taking the time to reply.

Now the rest of it I am going to speak to you on Skype about it, because I am really cheesed off at having lost it all again.

Cheers


Sut
__________________
Regards, Sutty
BeltEXI, mCX, mSR, 4G3, FBL Trex250SE µB, FBL Trex450Pro µB,
FBL Trex550e VX1 Pro, FBL Trex600ESP VX1e, Parkzone P51D, QRX350Pro
sutty is offline        Reply With Quote Quick reply to this message
Old 12-15-2010, 12:57 PM   #12 (permalink)
Registered Users
 
Posts: 5,405
 

Join Date: Mar 2008
Location: Wales, UK
Default

Quote:
Originally Posted by sutty View Post
Also, one last question, as an aside, how variable is the width of the pulse? In other words, can it be 1521.525, or can it only be 1521, or 1530, for example. Surely this affects how accurately the servo can be positioned, not mentioning the mechanical restraints of such accuracy. And also why is 1520 the middle, could in not equally well 1500, or is this something to do with the step sizes, or more accurately the pulse width variability here?

Cheers

Sutty
I also had this question in mind. Because if you had a servo with a pusle width range from 0.7 to 2.3 (700 to 2300) and it would only work up in 100s, that only leaves 16 positions on the servo arm. Work up in 10s and the servo can have 160 positions. And so on. I would think 160 is plenty but you never know with electronics...

-Jonny
__________________
Spektrum DX7, Align Trex 250se, Align Trex 500cf, Align Trex 600n LE
Jonnyheli is offline        Reply With Quote Quick reply to this message
Old 12-15-2010, 02:08 PM   #13 (permalink)
Registered Users
 
Posts: 988
 
My HF Map location
Join Date: May 2008
Location: Bangor, N. Ireland
Default

Jonny, as Raf said, that's really not a simple question to answer. There are so many other factors involved than just pulse width - accuracy of the feedback loop, tolerance of the electronic components, gear train slop, motor precision, etc.

But in simple terms, the pulse width, as I understand it, is an analogue value, therefore it is infinitely variable, there are no steps. However, due to all the other factors involved, there are other questions to be asked. For example, does changing the pulse width from 800 to 810 cause exactly the same relative amount of rotation as changing it from 1600 to 1610? Does setting the pusle width to 800 always position the servo arm in exactly the same place? Does changing the pulse width from 800 to 850 always make the arm rotate exactly the same amount every time? Unfortunately the answer to these questions is mostly "no", especially with cheaper servers.
__________________
Helis: Belt-CP, Mini Titan, EXI 450, HK250-GT, HK500-GT, HK600-GT, mSR (all modified/customised). Quads: Q4, nQX, HK250, Custom built 560 Deadcat camera platform. Some planks. DX6i & DX9. And a boat & some HK surface controller thingy.
CyprusDave is offline        Reply With Quote Quick reply to this message
Old 12-15-2010, 06:04 PM   #14 (permalink)
Registered Users
 
Posts: 5,405
 

Join Date: Mar 2008
Location: Wales, UK
Default

Complicated stuff
__________________
Spektrum DX7, Align Trex 250se, Align Trex 500cf, Align Trex 600n LE
Jonnyheli is offline        Reply With Quote Quick reply to this message
Old 12-16-2010, 12:14 AM   #15 (permalink)
Registered Users
 
Posts: 988
 
My HF Map location
Join Date: May 2008
Location: Bangor, N. Ireland
Default

And of course, as we're still largely talking analogue signals here, we have to ask the age-old question: Is 2 plus 2 always equal to 4? The answer being "no", it could be 3.9999 or it could be 4.0001. Which really says it all about the fundamental problems with working with analogue
__________________
Helis: Belt-CP, Mini Titan, EXI 450, HK250-GT, HK500-GT, HK600-GT, mSR (all modified/customised). Quads: Q4, nQX, HK250, Custom built 560 Deadcat camera platform. Some planks. DX6i & DX9. And a boat & some HK surface controller thingy.
CyprusDave is offline        Reply With Quote Quick reply to this message
Old 12-16-2010, 02:26 AM   #16 (permalink)
Registered Users
 

Join Date: Aug 2010
Location: L.A
Default

Well Dave, I think you officially blew Johnny's mind, he may even have a headache. LOL
UPGl2AYDD is offline        Reply With Quote Quick reply to this message
Old 12-16-2010, 03:19 AM   #17 (permalink)
Registered Users
 
Posts: 988
 
My HF Map location
Join Date: May 2008
Location: Bangor, N. Ireland
Default

I feel it's my duty
__________________
Helis: Belt-CP, Mini Titan, EXI 450, HK250-GT, HK500-GT, HK600-GT, mSR (all modified/customised). Quads: Q4, nQX, HK250, Custom built 560 Deadcat camera platform. Some planks. DX6i & DX9. And a boat & some HK surface controller thingy.
CyprusDave is offline        Reply With Quote Quick reply to this message
Old 12-16-2010, 08:56 AM   #18 (permalink)
Registered Users
 
Posts: 5,405
 

Join Date: Mar 2008
Location: Wales, UK
Default

Quote:
Originally Posted by UPGl2AYDD View Post
Well Dave, I think you officially blew Johnny's mind, he may even have a headache. LOL
Correct.
__________________
Spektrum DX7, Align Trex 250se, Align Trex 500cf, Align Trex 600n LE
Jonnyheli is offline        Reply With Quote Quick reply to this message
Old 12-16-2010, 02:25 PM   #19 (permalink)
Registered Users
 

Join Date: Aug 2010
Location: L.A
Default

LOL
UPGl2AYDD is offline        Reply With Quote Quick reply to this message
Old 12-18-2010, 11:58 AM   #20 (permalink)
Registered Users
 
Posts: 4,108
Thread Starter Thread Starter
 
My HF Map location
Join Date: Mar 2009
Location: Belgium
Default

Erwin and myself just tested the output signal of a servotester today, to see how far it would drive a connected servo. We measured min and max pulse width, using an oscilloscope and a counter, both times the results were identical.

For a 1520 µs servo, these values should be around 1 and 2 ms at least, for a 760 µs servo around 0.5 and 1 ms at least. Just like a tx/rx does, with the normal settings. But you already knew that when reading the thread. This will give a bit over 90 degrees of movement, like we are used in our helis, but servos can turn about 180 degrees in fact. Only thing is that this range is not useful in a rc heli. So, we were curious how far the Turnigy servo tester would go. We were expecting something like 0.7 and 2.3 ms for example, for a "normal" 1520 µs servo, which would make for the maximum 180 degrees deflection.
Note that this particular tester has 2 positions, one for normal servos, and one for narrow-band 760 µs servos.

These are the results:

1) 1520 µs position: min = 910 µs max = 2240 µs
2) 760 µs position: min = 410 µs max = 1140 µs

Meaning that this particular servotester does not go very low, but that the high side looks better. You can't expect too much for 8$ though, lol. Works nicely though.




http://www.hobbyking.com/hobbyking/s...dProduct=11780
Attached Thumbnails
Click image for larger version

Name:	T-ST_01.gif
Views:	15622
Size:	129.4 KB
ID:	195839  
__________________
TDR 14S - TDR 12S - Logo 500SE - Logo 400 - Furion 450 6S mini-V - 300X - 2 x mCPX - Schluter Bell Huey Cobra '74 - Aura 4 hotliner
redbird300 is offline        Reply With Quote Quick reply to this message
Reply




Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the HeliFreak forums, you must first register.
Please enter your desired user name, your REAL and WORKING email address and other required details in the form below.
User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself. Use a real email address or you will not be granted access to the site. Thank you.
Email Address:
Location
Where do you live? ie: Country, State, City or General Geographic Location please.
Name and Lastname
Enter name and last name here. (This information is not shown to the general public. Optional)
Helicopter #1
Enter Helicopter #1 type and equipment.
Insurance
Please select your insurance company (Optional)
Helicopter #2
Enter Helicopter #2 type and equipment.
Helicopter #3
Enter Helicopter #3 type and equipment.
Helicopter #4
Enter Helicopter #4 type and equipment.

Log-in

Human Verification

In order to verify that you are a human and not a spam bot, please enter the answer into the following box below based on the instructions contained in the graphic.



Thread Tools
Display Modes

Posting Rules
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On




Copyright © VerticalScope Inc. All rights reserved.
 

SEO by vBSEO 3.6.1