RunRyder - Side effects of digital servos

rrTV-PHOTO

New HD TV

HOME rrTV-PHOTO GALLERIES MY GALLERY HELP-FAQ

myHOME PM pmRR MEMBERS 706 ONLINE 28 EVENTS SEARCH REGISTER START HERE

1 page

2335 views

POST REPLY

HeliHobby	.	Ron’s HeliProz South	.	Century Helicopter
	.		.

Radio - Servo - Gyro - Gov - Batt > Side effects of digital servos

mash
Heliman
Location: Stratford-on-Avon, UK

I plugged a (brand new) HiTec 5625MG into a (brand new) Futaba 138DF Rx. Straight away all the servos started to twitch intermittently. I took the servo back to the shop; the replacement has the same effect. I haven't had time to do a whole load of swop-outs to completely isolate the servo, but a quick check with a Sanwa and a HiTec Rx did not show the same problem. Next step is to see whether any of my other Futaba Rxs take against the offending servo.
Has anyone else come across this? I know many of you will say "don't mix brands" but hey, these servos (even digital) are supposed to have basically the same electrical spec. aren't they? I've never had a problem with mix & match before.
All bright ideas gratefully welcomed!

02-01-2002 Over year old.

z11355
rrMaster
Location: 10000 is enough time wasted.

I think that's seriously questionable.

I have several of the Hitec 56xx's on a JR FM Rx and do not experience
any twitching. Due to the analog nature of the FM signal, my hunch
is that there is a higher positional resolution in FM compared to most PCM systems. Whether that is worth it (or even noticable) vs. the other PCM advantages is
a personal decision. If you figure a servos working angular range
is 90 degrees (+/-45), then @ 1024 steps, you have a maximum
resolution of 0.09 degrees.

The other piece of data is that the 56xx uses
a multi-pole cored motor instead of a coreless (the 59xx series).

02-01-2002 Over year old.

HOMEPAGE

flierbk
Heliman
Location:

Have not heard of digitals acting strangly when used on non PCM radios, have others experienced this?

02-01-2002 Over year old.

BladeRunner
Elite Veteran
Location: Ontario Canada. Member of "some sort" s

Side effects?

The reason you can see the movements on PCM-Z and not on PCM-S is because the resolution is twice as high on "S"; look at it this way; a machine screw has a finer "pitch" i.e. more threads, 1024 in this example. When you turn it one full turn it dosen't penitrate very far, a wood screw on the other hand has a coarse pitch-less threads, 512 in this example. So when you turn that screw it penitrates much farther.

A finer pitch (S-PCM) gives you much smaller, more accurate adjustments.
So when you move the stick a bit the servo moves only a bit, in "Z" PCM when you move the stick the same amount you get twice as much movement in the servo. Does any of this make any sense?
I maybe didn't explain it very well,but, that's how I understand it.

----------BladeRunner.

02-01-2002 Over year old.

Dr.Ben
Elite Veteran
Location: Richmond, VA, USA

The DF designation on that receiver should indicate that it is a non PCM receiver. A nonencoded (non pcm) PPM signal signal is basically stepless. If the servo is acting up, it is not due to the receiver unless the receiver itself is defective.

The old PCM10 (Z series) had two things going against it. First was the 512 stepping that has been mentioned. The second thing was a pokey processor that had trouble keeping up with complex mixing functions like ccpm.

Ben

02-01-2002 Over year old.

Augusto
rrAdvertiser
Location: San Diego, CA

The PCM-Z and PCM-S differ in the speed and resolution. The PCM-S is similar to the Futaba PCM 1024 in resolution and speed. The PCM-Z simply can't be compared to the other two.

That being said that means that ANY servo connected to a PCM-Z will be slower. It's not related to the servo itself but to the processing speed of a PCM-Z system.

Contrary to most people's perception, FM is much more precise than PCM just because it's a continuous control instead of a discrete (1024 steps) control. The difference is that since in the FM system the stick movements aren't digitized the information sent to the receiver is based on a pulse width (time) and not on a 1024 (10 bit) word. That allows for minute differences to be transmitted whereas in the case of PCM the smallest movement that can be transmitted will be the total angle of deflection divided by 1024 in the PCM-S and futaba PCM cases or 512 in the PCM-Z case.

The 10 bits for each channel are stored in the receiver so that if the receiver stops receiving information from the transmitter then the microprocessor inside the receiver uses the last stored value to recreate the last pulse width sent to the servo. There are additional protocols in each brand's operation that tell the receiver which channel to keep as the last position and which channel to take to a different one such as throttle to idle and so forth. This is the main advantage of PCM albeit some people see it as a disadvantage.

The Digital servos are digital INSIDE. In other words, their connection to the receiver is STILL pulse width based in other words time based or ANALOG. There are no digital servos that talk to the receiver in digitized format yet. So all the digitalization is lost again in the interface between the receiver and servo. That's why a servo has no way to tell if the receiver it's connected to is PCM or FM. The only information that a servo can tell is how frequent those analog width pulses are (frame rate).

The high frame rate is just a measure of how quick a servo can pay attention to a new pulse after processing the last one. If the servo is fast enough it will process the pulse by extracting the width, convert it to a voltage and compare it to the voltage controlled by the potentiometer that's connected to the servo's main gear. If they're different then the servo's circuit sends a voltage to the motor to move the pot in the required direction until they're the same.

In the digital servos instead of a voltage to control the servo's motor a pulse is applied in both forward and reverse polarities at ALL times. What the circuitry controls is the width of that pulse. In other words the motor gets full voltage both forward and reverse at the same time all the time. When it's centered the width of the pulses is very very small. That's why you can hear the high pitch wining sound while the servo is centered which is completely normal and expected. If the servo is off the center then the width of the pulse of the side that is required to correct is increased dramatically so the servo reacts at full power to go back to the center. That's why they have so much holding power.

In the case of the super high frame rate servos such as the GY-601's 9251 the frame rate is so fast that the standard width of the pulse was longer than the space between each pulse so they had to decrease the pulse width to make it fit. Because of this, for that servo the pulse width that tells the servo to go to the center is much smaller than the standard one so if connected directly to a receiver the standard pulse width coming from the receiver will be interpreted by the 9251 servo as a full deflection to one end hence the inability to be used directly on a receiver.

Hope this helps clear a bit the confusion about servo types and receiver types.

Avant Aurora Ultimate

02-01-2002 Over year old.

The_Dave
Veteran
Location: Las Vegas, Nevada

Bravo!

ColeopTer,

Very nicely done. It is refreshing that when someone goes on about something he actually knows what he is talking about!

Mark McAlpine - 2005 - We will never forget.

02-01-2002 Over year old.

HOMEPAGE

Phil Cole
Veteran
Location: Redwood City CA

To attempt an answer to the original question...

Why do all the other servos start twitching when the 5625 was plugged in?

My suspicion is that the 5625 is emitting some RF noise due to the previously mentioned motor switching pulses. The most likely path is via the 5625 servo lead. Either the noise is radiating from the lead to the receiver antenna or is being conducted up the lead directly into the receiver.

The first case might be improved by increasing the distance between the lead and antenna, or trying to get them as perpendicular as possible. You can also try shielding the servo lead with foil, or braid stripped out of coax cable. Ground the shield to the negative servo wire at the receiver end only. As you might expect, long extension leads make the situation worse.

If the interference is being conducted into the receiver, an RF filter on the leads might work. A simple 1000 pF capacitor from each of the positive and signal wires to the ground wire, at the receiver end might be enough. Or you can wrap the entire lead through a ferrite toroid core designed for RF noise suppression. Any one that's just big enough to fit the lead through a couple of times would be good. You can buy such filters already made up from some of those guys who make electronic gizmos for RC with the small ads in MA.

Or, you could go to PCM. PCM is able to ignore a low level of interference and maintain reliable control with full resolution. PPM can't ignore low level noise since it looks like small, random variations in pulse width. The receiver is faithfully sending these variations to the servos.

02-01-2002 Over year old.

The_Dave
Veteran
Location: Las Vegas, Nevada

Complex

All I said was is that he went on and on and in the end it was worth the time to read it. Often, that is not the case.

Don't get hard Peter, I am just playin' with ya!

Mark McAlpine - 2005 - We will never forget.

02-01-2002 Over year old.

HOMEPAGE

DangerousDick
Veteran
Location: Cheshire, England

FWIW

I use two Hitec 5925's on both PCM and FM JR Rx's with no problems at all.
I have three Hitec 5945's on Cyclic/Pitch on PCM again with no problems.

I can't beleive it has anything to do with the make of servo, but may suggest you check the voltage coming from your Rx pack. The digital servo will draw plenty of power and if you are using 4 cells with an 'under performing' bad cell then that may cause similar results?

Cheers
Richard.

02-01-2002 Over year old.

jrice45516
Heliman
Location: SoCal

<< The old PCM10 (Z series) had two things going against it. First was the 512 stepping that has been mentioned. The second thing was a pokey processor that had trouble keeping up with complex mixing functions like ccpm.

Ben >>

Strange. My JR PCM10 (ca1989) is a 512 system & has been handling a JR DS8411 servo, which according to JR Servos is a digital servo, just fine since the servo appeared. No "stepping" problems observable. No problems either with the Futaba GY401 S9253 which is digital per Futaba Servos & comes with the Futaba GY401. Nor any problems with the S9251 which comes with the Futaba GY601. But then, maybe time just flies alot faster in Richmond. ;-) The only "thing going against it" here is that i did not buy a spare because i've damned near worn this poor Tx out...not quite yet tho...it still works alot faster than me.

02-03-2002 Over year old.

mash
Heliman
Location: Stratford-on-Avon, UK

Thanks for all the responses; some very interesting stuff coming through.
Phil Cole seems to be suggesting the most likely cause - I'm a half-baked radio engineer myself - but his comments don't give me much comfort. If the Hi-tec or any digital servo was generating noise (RF or servo-pulse-frequency) then that would have an adverse effect on the whole system, eating into the margins. This would not be immediately apparent, especially with PCM systems, but at the limits it would come up and bite.
I decided not to spend my entire weekend eliminating various scenarios - I'm evidently not cut out to be a research scientist - so all I can add to my previous comments is that the problem only shows itself using a 6V supply on the Futaba Rx. 4.8V works absolutely fine! HiTec and Sanwa (Airtronics) show no problem on 4.8V or 6V, so I've put the HiTec Rx in and I'll give it a good range check before I fly it.
If the weather doesn't improve maybe I'll continue the quest but that's all for now, folks!

02-04-2002 Over year old.

Raffy
Key Veteran
Location: Chicago, Illinois

Jrice45516

Thanks for your information! After reading the above informations, I thought I might had to upgrade my PCM10. Had no problems with it either!

04-08-2002 Over year old.

Location:

I installed a 6volt battery into my 60, I run all Futaba servos, only thing I noticed was the 9204 (new servo just out of package) was choppy and would bounce/judder when move to one end of its throw. With the 4.8 it does not do it or it's not noticeable. I decided to go back to the 4.8 for my own comfort zone untill I hear otherwise.

04-08-2002 Over year old.

WV EDIT

Scott_MT
Heliman
Location: Great Falls MT

interference

Don't forget the rf of flourescent lights in the shop. I have found this to be the case on several planes instead of the gas pot hanging off the nose. Just my 2 cents worth.

Thanks Scott S.

04-08-2002 Over year old.

Itsindilas
Veteran
Location: Greece-Athens

http://runryder.com/showtopic.htm?t...31&topicid=7441

Have a look my friend, we are two persons same problem. Coleopter If you wish have a look, please.

04-21-2002 Over year old.

Safe_Cracker
Key Veteran
Location: Chicago, IL

Yepp.

I have a problem with my 8700g causing feed-back through the gyro to the receiver.It causes all the other servos to jitter but not really bad.If the servo is plugged in without the gyro then it stops but as soon as you plug the gyro back in, forget it.It gets worse if i use the higher frame rates for super servo and/or turn the gain up.I figure it causes some type em or rf pulses due to the frame rate frequency.I used a filter with no affect (1pf cap).I came across a electronics project that might cure the problem look here http://www.uoguelph.ca/~antoon/gadgets/noiserx.htm
It can clean up the noise from the culpret servo.

04-22-2002 Over year old.

Itsindilas
Veteran
Location: Greece-Athens

Well my friend in my case, the same happens even if I don't use extensions.
Today I will proceed in a test flight. I rearranged everything. I grounded the boom with the main frame. I changed the whip antenna's position and I will test it.

If the same problem insists then I will wait up to the new batteries' arrival.

Next step is to check the JR 10SXII (whenever it arrives) with my existing module and RX. And then to change the old RX with the new SPCM in order to verify that finaly ZPCM contributed to this problem.

04-22-2002 Over year old.

jrice45516
Heliman
Location: SoCal

Raffy

<< Jrice45516
Thanks for your information! After reading the above informations, I thought I might had to upgrade my PCM10. Had no problems with it either! >>

Don't thank me too soon. The people on this thread are in large part blowing smoke about how "FM" & PCM systems work.

If I knew, I would explain it. But for one thing, it is not "FM" & PCM. It is PPM & PCM. Both are FM. PPM is analog; PCM is digital. To say that "FM" has better resolution than PCM displays an ignorance of real life in which a) the resolution of an analog system is not infinitely small, a long way from it and, certainly, not necessarily as small as PCM resolution, all else being equal AND b) PCM IS FM. Both PPM & PCM are frequency modulated systems. But people, products of the dumbed-down American education system where no one must be allowed the humiliation of failing, are too lazy or worse to use intelligent terminology. To quote world's most successful investor, Warren Buffet, "Bad terminology is the enemy of good thinking."

Gosh & golly. What do we do now?

What is resolution anyway? Many use the term but few understand it &, of those few, even fewer have a score on the Verbal high enough to be able to explain unambiguously what they think they understand. Resolution is the measure of how accurately the servo arm can be positioned.

JR is currently advertising (see e.g. p.15 of the current [July 2002] issue of Model Aviation) their DS8411/DS8411SA servo as having a resolution of 5,900 steps per 120º. If i have my arithmetic correct, 5,900 steps are more than 1,024. ;-) Like I say, blowing smoke. But then there are three types of people, those that can do arithmetic & those that can't.

Are pilots running PPM able to take advantage of this? I don't know.
Would my JR PCM 10 move the servo thru 5,900/2 = 2,950steps? I don't know.
Could even Curtis, handed a radio operating in a mode unknown to him (PPM or PCM), guess the mode accurately a majority of the time given 10 tries? I doubt it. I doubt he could tell a 512 system from a 1024.

Hmm, let's see. 5,900steps/120º = 49.1667steps/degree. You can't see that with your eyeball. You can't feel it with your finger on the servohorn. You can't feel the servohorn move from step 5,899 to step 5,900. In fact, there is enough lash in the geartrain that there are not 5,900 distinct steps to be seen at all. Some get skipped or washed out. Oh, and by the way, the last thing that happens in a "digital" servo, in any RC servo, is a digital-to-analog conversion. The converter which does the conversion is called a potentiometer & its physical size limits resolution. There is no such thing as a digital servo in the RC world...yet.

As you can see, I can blow smoke as thick as anyone. A professional explanation of how these things really work would be nice.

Dances With Woofs

04-24-2002 Over year old.

Safe_Cracker
Key Veteran
Location: Chicago, IL

Oh really.

By the way there isn't a single person here in this thread pretending to know EXACTLY everything.If you would have read it all the way through you would have noticed that we are using trial and error to help us all get over a particular problem.We all have are preferences and that is all that PPM and PCM is really, unless you are really into racking your brain to figure out the real differences.I know i'm not!I prefer to be in control at all times and can deal with some glitches but I can't handle the idea of a preset servo position during interference so I stick with PPM.There isn't any real reason for hostillity in the forums,We are just pilots sharing problems and experiences hoping to get each other through them.So, RELAX!

04-24-2002 Over year old.

1 page

2335 views

POST REPLY

MTA Hobbies	.	Model Rectifier Corp	.	3D Heli Depot
	.		.

Radio - Servo - Gyro - Gov - Batt > Side effects of digital servos

UPDATE SCREEN PRINT TOPIC

Advertisers

Subscribe to This Topic