Post by thread:[PIC] r/c receiver interfacing

Bob J. <rocketbob <at> gmail.com> writes: > I am looking to build what is essentially an a huge radio-controlled servo > with a stepper motor. In order to drive the stepper motor, I have to > convert the r/c receiver's pulses into step/direction pulses. To my great > surprise I googled for something off-the-shelf to do this, but have come up > empty-handed. Has anyone done this sort of thing? It is not so simple to do it. There are a lot of things to handle for the stepper drive (including open loop and how to deal with it if the link goes away and comes back). Also with the voltage available for RC driving a stepper will not go down so well (too few volts, too much voltage drop in the driver, too much quiescent current etc). Are you absolutely sure you want to do it with a stepper ? Code for a PIC to do it is really straightforward especially if it's open loop and fixed speed (preferrably the RC frame speed - i.e. 50Hz). Peter P.

Hi Bob, Radio controlled modellers use off-the-shelf stuff, you can buy analogue FM radio system (PPM) or digital FM (PCM, QPCM) or even 2.4GHz ones (DX6 and DX7 for example). Same stuff used by robotic maniacs who even use special servos that rotates infinite (servo is a stepper motor anyway). You can buy many different size and torque of servos - or those not what you want? http://www.futaba-rc.com http://www.hitecrcd.com http://www.graupner.de http://www.spektrumrc.com/ These are manufacturers but you can buy their products at your local modeller shop or e-bay. Tamas On 5/5/07, Peter P. <spam_OUTplpeter2006TakeThisOuTyahoo.com> wrote: {Quote hidden}> > Bob J. <rocketbob <at> gmail.com> writes: > > > I am looking to build what is essentially an a huge radio-controlled > servo > > with a stepper motor. In order to drive the stepper motor, I have to > > convert the r/c receiver's pulses into step/direction pulses. To my > great > > surprise I googled for something off-the-shelf to do this, but have come > up > > empty-handed. Has anyone done this sort of thing? > > It is not so simple to do it. There are a lot of things to handle for the > stepper drive (including open loop and how to deal with it if the link > goes away > and comes back). Also with the voltage available for RC driving a stepper > will > not go down so well (too few volts, too much voltage drop in the driver, > too > much quiescent current etc). Are you absolutely sure you want to do it > with a > stepper ? Code for a PIC to do it is really straightforward especially if > it's > open loop and fixed speed (preferrably the RC frame speed - i.e. 50Hz). > > Peter P. > > > -

The radio output of an RC receiver is a pulse. The width of that pulse determines the servo position. The accuracy and repeatability as well as the output power are reflected in the price of the servo. You can either work from that pulse OR use a high quality servo of your choice to turn a knob. Most of that kind of setups balance a bridge. The chips in the servos have all been unobtainable by me! Good luck... John Ferrell W8CCW "Life is easier if you learn to plow around the stumps" http://DixieNC.US {Original Message removed}

Minor correction: PPM for RC is pulse "position" (as opposed to width in PWM) modulation. 2007/5/6, John Ferrell <.....johnferrellKILLspam@spam@earthlink.net>: {Quote hidden}> > The radio output of an RC receiver is a pulse. The width of that pulse > determines the servo position. The accuracy and repeatability as well as > the > output power are reflected in the price of the servo. You can either work > from that pulse OR use a high quality servo of your choice to turn a knob. > Most of that kind of setups balance a bridge. > The chips in the servos have all been unobtainable by me! > Good luck... > > John Ferrell W8CCW > "Life is easier if you learn to plow > around the stumps" > http://DixieNC.US > > {Original Message removed}

Yes, that's correct John. Probably the easiest to understand description is here: http://www.mh.ttu.ee/risto/rc/electronics/radio/signal.htm Tamas On 5/6/07, Ariel Rocholl <arochollKILLspamgmail.com> wrote: {Quote hidden}> > Minor correction: PPM for RC is pulse "position" (as opposed to width in > PWM) modulation. > > 2007/5/6, John Ferrell <.....johnferrellKILLspam.....earthlink.net>: > > > > The radio output of an RC receiver is a pulse. The width of that pulse > > determines the servo position. The accuracy and repeatability as well as > > the > > output power are reflected in the price of the servo. You can either > work > > from that pulse OR use a high quality servo of your choice to turn a > knob. > > Most of that kind of setups balance a bridge. > > The chips in the servos have all been unobtainable by me! > > Good luck... > > > > John Ferrell W8CCW > > "Life is easier if you learn to plow > > around the stumps" > > http://DixieNC.US > > > > {Original Message removed}

Hi Ariel, It IS actually PWM, just PWM where the maximum "on" time is not 100% of the period. The frequency is usually around 50Hz, and the "on" time varies between 1 and 2 milliseconds (which would be 5 and 10% duty cycle). The signal received by the receiver is PPM encoded, but the receiver converts this into PWM for each servo. Note that unlike regular PWM, this is not used to directly drive a motor but rather to send a command. Sean On 5/6/07, Ariel Rocholl <EraseMEarochollspam_OUTTakeThisOuTgmail.com> wrote: {Quote hidden}> Minor correction: PPM for RC is pulse "position" (as opposed to width in > PWM) modulation. > > 2007/5/6, John Ferrell <johnferrellspam_OUTearthlink.net>: > > > > The radio output of an RC receiver is a pulse. The width of that pulse > > determines the servo position. The accuracy and repeatability as well as > > the > > output power are reflected in the price of the servo. You can either work > > from that pulse OR use a high quality servo of your choice to turn a knob. > > Most of that kind of setups balance a bridge. > > The chips in the servos have all been unobtainable by me! > > Good luck... > > > > John Ferrell W8CCW > > "Life is easier if you learn to plow > > around the stumps" > > http://DixieNC.US > > > > {Original Message removed}

Sean Breheny wrote: > Hi Ariel, > > It IS actually PWM, just PWM where the maximum "on" time is not 100% > of the period. The frequency is usually around 50Hz, and the "on" time > varies between 1 and 2 milliseconds (which would be 5 and 10% duty > cycle). > > The signal received by the receiver is PPM encoded, but the receiver > converts this into PWM for each servo. Note that unlike regular PWM, > this is not used to directly drive a motor but rather to send a > command. Actually, in most early (non digital) R/C setups the pulse DOES drive the motors directly. The leading edge triggers a monostable whose period is set by the current servo position. An exclusive-or of the incoming and position pulses go to an H bridge, whose direction is set by which edge fell first. The result is pulses to the motor (at the frame rate) that are portional to the position error. I know this because when we were developing C-Leg (prosthetic knee) I raised the pulse rep rate to 300 Hz to get a faster response from the R/C servo we were using to position the hydraulic valve controling the knee. The receiver demultiplexes the PWM (not PPM) train using a simple counter and monostable to detect end of frame (and reset the counter/mux). See figure 2 for a schematic of a decoder. http://www.seattlerobotics.org/encoder/200009/Servos.html Robert

Hi Robert, I think I worded this confusingly. When I said "drives the motor directly", I meant "provide the actual motor ON time". Of course, a motor whose PWM "on" time only varied between 5 and 10% wouldn't have much torque. The prosthetic limb project sounds really interesting! Sean On 5/6/07, Robert Rolf <@spam@Robert.RolfKILLspamualberta.ca> wrote: {Quote hidden}> > The signal received by the receiver is PPM encoded, but the receiver > > converts this into PWM for each servo. Note that unlike regular PWM, > > this is not used to directly drive a motor but rather to send a > > command. > > Actually, in most early (non digital) R/C setups the pulse DOES drive > the motors directly. The leading edge triggers a monostable whose period > is set by the current servo position. An exclusive-or of the incoming and position > pulses go to an H bridge, whose direction is set by which edge fell first. > The result is pulses to the motor (at the frame rate) that are portional > to the position error. >

On May 6, 2007, at 8:25 PM, Robert Rolf wrote: > Actually, in most early (non digital) R/C setups the pulse > DOES drive the motors directly. The leading edge triggers > a monostable whose period is set by the current servo position. > an exclusive-or of the incoming and position pulses go to an > H bridge... That's what you call "directly" ?! I think the poster meant that PWM of the motor power (traditional for some speed control sorts of things) is not driven off of the "servo output" PWM channels... > > The receiver demultiplexes the PWM (not PPM) train using a simple > counter and monostable to detect end of frame I see AM, FM, and "PCM" radios advertised, so it's not entirely obvious that they all encode the bitstream the same way... The output of the RADIO is some sort of bitstream that includes the position of each of (12? many?) possible servos and some synchronization scheme that allows you to tell which bits go with which servo. This goes into a decoder of some kind (these days frequently a microcontroller, though SSI logic of the sort described here used to be popular) that demultiplexes the bit stream and provides a signal to each of the servos supported Radio output: SSSS___XX___XXX__X____XXXX______ Servo 1 out: _______XX_______________________ Servo 2 out: ____________XXX_________________ Servo 3 out: _________________X______________ Servo 4 out: ______________________XXXX______ HOWEVER, My impression was that the original poster was more interested in the problem: "given a servo pulse of 1 to 2ms as input, is there a standard way to drive a stepper that maintains "current position" and needs as output "direction" and "number of steps."" My guess is "no", cause it's not too complicated except for taking into account the variables that exist in range of servo travel and degrees per step for the stepper motors. (It seems an unfortunate side effect of electronics economics that relatively beefy stepper motors are cheaper, more standardized, and easier to find for hobbyists than geared motors. I'm not sure you're accurately estimating the relative torque of big steppers and smaller servos, though.) BillW

Sean, Actually, I should have read better original post as he indicated the signal available on the receiver, which yes means after decoded of course. I agree the signal on the servo is PWM, but the FM transmitted is PPM coded - but this signal is not the one you care about when you use the one already decoded by the receiver. 2007/5/7, Sean Breheny <KILLspamshb7KILLspamcornell.edu>: {Quote hidden}> > Hi Ariel, > > It IS actually PWM, just PWM where the maximum "on" time is not 100% > of the period. The frequency is usually around 50Hz, and the "on" time > varies between 1 and 2 milliseconds (which would be 5 and 10% duty > cycle). > > The signal received by the receiver is PPM encoded, but the receiver > converts this into PWM for each servo. Note that unlike regular PWM, > this is not used to directly drive a motor but rather to send a > command. > > Sean > > > On 5/6/07, Ariel Rocholl <RemoveMEarochollTakeThisOuTgmail.com> wrote: > > Minor correction: PPM for RC is pulse "position" (as opposed to width in > > PWM) modulation. > > > > 2007/5/6, John Ferrell <spamBeGonejohnferrellspamBeGoneearthlink.net>: > > > > > > The radio output of an RC receiver is a pulse. The width of that pulse > > > determines the servo position. The accuracy and repeatability as well > as > > > the > > > output power are reflected in the price of the servo. You can either > work > > > from that pulse OR use a high quality servo of your choice to turn a > knob. > > > Most of that kind of setups balance a bridge. > > > The chips in the servos have all been unobtainable by me! > > > Good luck... > > > > > > John Ferrell W8CCW > > > "Life is easier if you learn to plow > > > around the stumps" > > > http://DixieNC.US > > > > > > {Original Message removed}

The "PCM" term refers to the transmitter-receiver link. It is proprietary information to the various manufacturers. The signal is still FM. By transferring digital information the receiver can send to servo pulse to the servo at the rated speed. Without this strategy the servo cannot achieve its maximum (speed , torquee) ratings. John Ferrell W8CCW "Life is easier if you learn to plow around the stumps" http://DixieNC.US {Original Message removed}

> I know this because when we were developing C-Leg (prosthetic knee) I > raised the pulse rep rate to 300 Hz to get a faster response from the > R/C servo we were using to position the hydraulic valve controling the > knee. > Now there is a product I need! I see knee replacements in my near future and it is not really a pretty picture! John Ferrell W8CCW "Life is easier if you learn to plow around the stumps" http://DixieNC.US ----- Original Message ----- From: "Robert Rolf" <TakeThisOuTRobert.RolfEraseMEspam_OUTualberta.ca> To: "Microcontroller discussion list - Public." <RemoveMEpiclistTakeThisOuTmit.edu> Sent: Sunday, May 06, 2007 11:25 PM Subject: Re: [PIC] r/c receiver interfacing > > > Sean Breheny wrote: >> Hi Ariel,

I considered a project similar to this a while back and felt that my best solution would be to use the electric controlled jack screw intended to position a satellite dish as the actuator. Not really all that expensive for the power. Also, don't overlook the many automotive actuators. My resistance to stepper motors is that finding the initial position and assuring that no steps are missed is not trivial. John Ferrell W8CCW "Life is easier if you learn to plow around the stumps" http://DixieNC.US {Original Message removed}

Ariel Rocholl wrote: > I agree the signal on the servo is PWM, but the FM transmitted is PPM > coded - but this signal is not the one you care about when you use the > one already decoded by the receiver. I'm not sure I understand the distinction between PWM and PPM you guys are making here. AFAIK, PWM stands for "pulse width modulation" and this doesn't say what type of information is encoded ("modulated") in the pulse width. See also <http://en.wikipedia.org/wiki/Pulse-width_modulation>. In that sense, any pulse signal of constant frequency that has the information encoded in the pulse width is a PWM signal -- no matter whether the encoded information is position, voltage, speed, motor current, etc. Gerhard

Hi Gerhard- I think that if you consider the "envelope" of Pulse Position Encoding with each of the channels being Pulse Width Modulation (one PWM per channel), then you might have a clearer idea of what they are talking about. With say, a six channel transmitter/receiver, the "frame" of pulse position has six analog bits of information each one being a seperate PWM pulse. This PWM is then sent to the particular channel servo which translates the width of the pulse to a position. After sending all the channels, the "envelop" has a stop or end of frame to allow for the receiver to resync. Hope this helps! Cheers, Rich S. ---- Original Message ---- From: Gerhard Fiedler Date: Mon 5/7/07 9:37 To: piclistEraseME.....mit.edu Subject: Re: [PIC] r/c receiver interfacing Ariel Rocholl wrote: > I agree the signal on the servo is PWM, but the FM transmitted is PPM > coded - but this signal is not the one you care about when you use the > one already decoded by the receiver. I'm not sure I understand the distinction between PWM and PPM you guys are making here. AFAIK, PWM stands for "pulse width modulation" and this doesn't say what type of information is encoded ("modulated") in the pulse width. See also <http://en.wikipedia.org/wiki/Pulse-width_modulation>. In that sense, any pulse signal of constant frequency that has the information encoded in the pulse width is a PWM signal -- no matter whether the encoded information is position, voltage, speed, motor current, etc. Gerhard

Rich Satterlee wrote: >>> I agree the signal on the servo is PWM, but the FM transmitted is PPM >>> coded - but this signal is not the one you care about when you use the >>> one already decoded by the receiver. >> >> I'm not sure I understand the distinction between PWM and PPM you guys >> are making here. > I think that if you consider the "envelope" of Pulse Position Encoding > with each of the channels being Pulse Width Modulation (one PWM per > channel), then you might have a clearer idea of what they are talking > about. Yes, thanks, makes sense... Yet, I wouldn't call this PPM -- if this is to mean Pulse Position /Modulation/. You're using a term ("Encoding") that makes more sense. Gerhard

Or you can use a Nordic chip with a micro in it and drop the PIC altogther. You just need to download a 8051 compiler & away you go. It is pretty simple - especially if you start out with their sample programs & layouts Richard P On 12/05/07, alan smith <RemoveMEmicro_eng2EraseMEEraseMEyahoo.com> wrote: > make it sound too simple as well. Is the Nordic that *simple* to interface with? just a serial interface or? Microchip has a 2.4GHz chip now as well... > > Dario Greggio <RemoveMEadpm.tospam_OUTKILLspaminwind.it> wrote: Hi Bob, why don't you build your own using a PIC and Nordic 2.4 GHz > transceiver? > > (looks like they're giving me money for this :-) but I assure it's not > like that!) > > -- > Ciao, Dario > -

alan smith wrote: > make it sound too simple as well. Is the Nordic that *simple* to > interface with? just a serial interface or? Microchip has a 2.4GHz > chip now as well... Well, I can tell you my experience: started out testing CC2420 daughterboard from Microchip in 2005 (50EUR each...) then tried Freescale's PCB and chipset (cheaper, but I was probably not ready to face 0402 components... :-( ) and then Easy Radio ER900 (35EUR each). Then landed onto Nordic. They cost 18EUR each, had the same range as all of the others, rather easy to interface (it's SPI, just like CC2420 - ER900 used USART) and small. You have to specify some 40 bits plus some more as Setup, and then you basically send your Packet... and get it at the other side. Of course, you don't get Zigbee (at least with those I used) - but they may be just good for you! -- Ciao, Dario