9303 Manual...SERVO SYNC...What is it

[SURREYRCAV8R]

Gents,
new to the 9303 radio and see in the features listings it takes about ServoSync, anyone have any idea as to what that is and how to use it, I have searched the manual and cannot find it anywhere.

Cheers

[dmcquinn]

I'm pretty sure that it refers to a feature whereby the transmitter sends "paired" channels (such as ch 2 and ch 6 for ailerons) back-to back. In this way, the servos respond more closely to each other. Without servosync, the transmitter sends the servo position for channel 1 (throttle), followed by channel 2, etc. It does this many times per second. But theoretically, the command to channel 6 comes a few microseconds after the channel 2 command. By sorting them together, the servos should respond more closely.

I think it is more marketing hype than anything. I doubt anyone can tell the difference.

[Zeeb]

	Quote: Originally Posted by dmcquinn
	I think it is more marketing hype than anything. I doubt anyone can tell the difference.

Maybe I look for stuff like this after having experienced the really bad dual elevator servo lag on a Futaba 9C Super where it was really obvious that pairing channel two and eight created significant lag between the two surfaces?????

Hard to say if it's just marketing hype on the planks as the 2.4GHz system is so much faster than 72MHz, but where you find multiple servos on a surface like ailerons, or the three servos on most heli swashplates, it's evident....

[ULTIMATE1]

ServoSync™ is an ingenious DSM2 feature that automatically re-sequences the bits of data being transmitted based on the type of mixing you select.* For instance, if you select dual elevator mixing or 3-servo CCPM, ServoSync will see to it that those servos working together receive their pulses together. The result is more synchronized servo movement, which translates into an undeniable, locked-in feel when you fly.

[804]

Maybe I'm not doing it right. I originally set up my plane (dual elevator servos) on my 72 9303 and had lag between the two halves. Now it is on 2.4 9303 and I can't see much, if any, difference.
At neutral and full throw it is fine. In between, one lags.
I can set it up so they both travel together on low rates, OR high rates, but not both. So, I have it set up for low rates, since I use that the most.

[Troy Newman]

804,

If you are still having some lag issues and you are on the dedicated 2.4 9303 system. Then it is the geometry of the linkages. What are you subtrim and end points for each elevator servo ?

If they are far off say more than about 5-10% of each other then you linkage geometry needs to be addressed. If the subtrims and end points are within about 5-10% of each other and the throws are the exact same for each side then you should not see any lag between the sides.

The radio is used to fine tune those percentages. The majority for the correct geometry and matching the surfaces is done in the linkages. I personally use 5% as my max difference in the radio. if the difference needs to be more than 5% I adjust the mechanical aspects to get it closer than 5% max different. Sometimes its a new servo arm, or wheel. Often its adjusting the length of a control horn to increase or decrease the throw. I also measure extremely accurately that the surface move identical amounts. If both surfaces are moving the same deflection and the end points and subtrims are close to each other (within 5%) then you will get what you are looking for.

The problem comes in often with some of the aftermarket aluminum servo arms. Sometimes you are left with no choice but to use a large amount of subtrim to get the servo arm squared up with the linkage. JR and Futaba Servos have an ODD number of teeth on the splines. SO if you turn the servo wheel or arm 180 degs to the other side it will now be closer to aligned where you need it. This gets you with 1/2 of a tooth of the servo spine.

Unfortunately the aluminum one sided servo arms don't allow you to do that because you need the servo arm on a given side.

The key to having the elevators in sync is to have both sides with zero subtrim, and the end points matched perfectly. And get the control movement of the elevator to match perfectly with those subtrims and end points.

On the older 72mhz systems even though you got the servos all matched up perfectly the fact that radio ordered the channels for you...meant you would get a lag on the second elevator servo if it was on channel 8 or so. Servo sync will send the data to the two servos essentially together and this will eliminate the lag from the rx signals but the linkage needs to be corrected so you are not creating mechanical lag.

check the end points and subtrims

Troy Newman
Team JR / Horizon Hobby.

[804]

	Quote: Originally Posted by Troy Newman
	804, If you are still having some lag issues and you are on the dedicated 2.4 9303 system. Then it is the geometry of the linkages. What are you subtrim and end points for each elevator servo ? If they are far off say more than about 5-10% of each other then you linkage geometry needs to be addressed. If the subtrims and end points are within about 5-10% of each other and the throws are the exact same for each side then you should not see any lag between the sides. The radio is used to fine tune those percentages. The majority for the correct geometry and matching the surfaces is done in the linkages. I personally use 5% as my max difference in the radio. if the difference needs to be more than 5% I adjust the mechanical aspects to get it closer than 5% max different. Sometimes its a new servo arm, or wheel. Often its adjusting the length of a control horn to increase or decrease the throw. I also measure extremely accurately that the surface move identical amounts. If both surfaces are moving the same deflection and the end points and subtrims are close to each other (within 5%) then you will get what you are looking for. The problem comes in often with some of the aftermarket aluminum servo arms. Sometimes you are left with no choice but to use a large amount of subtrim to get the servo arm squared up with the linkage. JR and Futaba Servos have an ODD number of teeth on the splines. SO if you turn the servo wheel or arm 180 degs to the other side it will now be closer to aligned where you need it. This gets you with 1/2 of a tooth of the servo spine. Unfortunately the aluminum one sided servo arms don't allow you to do that because you need the servo arm on a given side. The key to having the elevators in sync is to have both sides with zero subtrim, and the end points matched perfectly. And get the control movement of the elevator to match perfectly with those subtrims and end points. On the older 72mhz systems even though you got the servos all matched up perfectly the fact that radio ordered the channels for you...meant you would get a lag on the second elevator servo if it was on channel 8 or so. Servo sync will send the data to the two servos essentially together and this will eliminate the lag from the rx signals but the linkage needs to be corrected so you are not creating mechanical lag. check the end points and subtrims Troy Newman Team JR / Horizon Hobby.

Troy, thanks for your reply.
I was very careful in setting up the geometry on this, so that I have only 2% subtrim on one ele. half, and 0% on the other.
On endpoints, (travel adjust), I am within 5% on low and high, between both sides. I am, however at max.--145-150. Could this be a problem?
I am using alum. servo arms, but like I said, they matched up well.
I am using older HS 5645 servos,(50cc bird), but put new gear trains in them at install, and they have always centered well.

[Troy Newman]

I assume that you are using the "MATE" function to get the dual elevators and not a programable mix or a Y'Harness?

I rarely will use the 150 travel adjust. Becasue it doesn't give you room to move on a given servo. I always will try for about 130-135% This way the servos are not near their physical limits but you are using almost all of the available travel of the servo. However I don't think this is your issue. As an engineer I alwasy look at the operation limits of the system. You don't want to use a thermometer with a range of 0 to 212degs F to measure the boiling temperature of water, a temperature of 212degs F normal ATM conditions. Why yes it will measure it, but if you need a gnats rearend measurement to 0.01degs F accuracy that would not be your best choice in tools. Likewise I would prefer not to use the pots to its full extreme end points for its accuracy. the difference between 130-135% and 150% is very small.

In our Servos and models the POT is the tool the servo uses to record its position. Its the measuring stick that says the physical location of the servo arm is not where its commanded to be, motor drive me to where I'm suppose to be. The radio tells it where to go and feedback comes from the pot to see if it is there. If it is not then the electronics in the servo drive the motor to get it back to that location. This is one reason that JR servos are extremely accurate and have a "zero deadband" the pot samples the position and the servo is always trying to drive to that zero location("zero" off the commanded position). When aerodynamic loads are applied to the surface, the servo pot says hey control guy I'm off my commanded spot you better drive me back to it. Likelwise a servo with a wider deadband the servo may be 2-3ms off of the actual command position and the servo is ok with that. The pot and the electronics are only so accurate to drive the servo back center. If it thinks that a sweetspot of 2-3ms is good to go then it will not raise the flag and tell the servo motor to drive to the commanded point. This setup we see at center more often, some see it as oscillation at center, this is when the mass of the surface, wear in the gear train, slop in the linkage, or worn servo pot constantly drives the servo back to center, but in reality it applies to any commanded position on the servo. To check this out...place your trim resolution at 1step for each beep of trim. Now watch the control surface very closely, and click off beeps of trim. A more accurate servo will move every single beep of trim. The movement will be really small, like 1/2048 of the range of the servo.

Now take the servo that has that deadband of 2-3ms. The trim might have to move 2-3beeps before the surface moves. When it does move it will jump the 2 beeps at one time to get it back in the sweetspot range. Rememeber the commanded position of the servo has to get oputside the deadband before the pot feeds back to the servo electronics, hey man I'm not where the radio tells me to be. I'm guessing that this may be some of the issue you are seeing.

Next question is did you program any offset or center changes into the hitec servos? This could be causing it. I don't know I have extremely limited experience with the hitec servos and the servo programmer they have for their product.

I'm guessing and don't know for sure that maybe if you changed the center on the servos with the programmer you have the same effects as using subtrim in the radio. Regardless the servo still has its effective range between 1ms and 2ms with the mid point being 1.5ms. Its my understanding that the Hitec programmer tells the servo to convert the TX center point of 1.5ms to a new physical position on the output pot.

I would think but have no specific knowledge to this effect that you would still have some issues with the output pots range if it was offset much from the actual factory center positon. This would be the same as adding in a large amount fo subtrim in the TX. The programmer is just thinking on its own based on the TX commands. The TX says 1.5ms and the servo has been programmed that a certain location is 1.5ms from the TX. If the servo was at factory setting that same position might be 1.65ms output from the TX. But the programmer re-zeroed the servo.

example. a normal setup might have the servo using 1.1ms at low, 1.5ms at nuetral(or midpoint) and 1.9ms at the high value. If I were to offset the TX's input for nuetral position to 1.6ms. The same travel range would be 1.2ms at low, 1.6ms at the midpoint and now 2.0ms at the high value. While the rotational movement of the servo would be the same or extremely close to the same. The 2.0ms is now approaching the end of the servos range of input or motion. The pot is moving toward its end...At the very extreme ends of the scale its not going to be as accurate as if its operation in the middle of its range. Same idea as the thermometer I mention above.

As we approach the ends of the range, in just about all devices the accuracy will become less. Add to this that a servo pot will wear with vibration. As it wears the resistance across the entire pot can change.

Another issue could be the actual speed differences in the servos. Yes I know that they are the same servo, however there are tolerances with all electronics and two devices even though they are the same components can only perform with a given range of being absolutely the same in all aspects.

Like I said I don't have experience with those servos, however I have setup the linkages and servos using JR 3421SA's on a pattern model and the X9303 and the 12X. Using 12" long pointers out the rear of the model I can move the servos fast or slow and can't see much if any difference in the performance of the pair of servos.

I did have on the X9303 where one servo would jump slightly ahead in partial travels, and the other servo would catch up and overtake again in a very small partial travel. I can compare this to a step deal when watching it. The left elevator would be slightly less travel about 0.1degs then the next small increase the left would catch up and go about 0.1degs past the right. The right would then catch up to the left and be even. Then they would stay dead on on for about 4-5degs and then one or the other would jump out ahead slightly again the 0.1deg and the next increase they would catch up. I would attribute this to minor differences in the servos electronics. The difference is extremely extremly small.

I used the 12X on this same model and its balance feature and was able to correct the "minor lead lag" deal. Using this multipoint style of adjustment the balance feature I was able to get them to track together with no visible lag at any point. Again these were brand new servos with no vibration induced pot wear. As the pots wear under vibration things are going to change.

In your case maybe its time the servos get serviced with new pots. This could be a partial cause of what you are seeing. Afterall the vibration of our models does cause the pots to wear.

Can we actually feel this small differences in flying the model? The larger scale aerobatic models are less affected by the minor differences. Pattern models, pylon racers, and competition sailplanes that rely on extremely precise movements and centering will notice the differences first. Of course the cool thing about the sail planes is the lack of vibration to wear the stuff out.

I hope this explains a little and gives you some places to look for the issue. This is not intended to claim that a certain brand is better than another, instead it is the reality in the way the servos work and perform. Each type of product has its advantages and place. If you are happy with your choosen products then great. As always your Mileage may vary.

To sum up if you have changed the servos programming to change centers or speed, put them back to factory settings and try the setup again. If your servos have some time on them the pots are probbaly worn a little. It doesn't take much to make that deadband even wider than the 2-3ms of the factory setting. Try the trim step deal and watch the servo move. If it is not moving the same for every single beep of trim then its not as accurate as it could be. This could be wear related of just the servos design.

Some servos are more accurate than others. Not just between brands JR vs Futaba Vs Hitec, but within a brand some servos are just more accurate than others.

I hope this helps

No flaming guys this was intended as educational on what the systems work like...not a brand war issue. I'm trying to help 804 get his hitec servos working with his JR radio.

[804]

Troy, thanks again, very informative.
I am using the mate function. And I have not programmed the servos.
I think the issue is the servos. I tried your trim step idea, the right half moves with every click from the start, the left half takes several clicks just to get started, then sometime moves with every click, sometimes not.
The lag is really pretty small, but definitely there. When I originally set it up, I used long straight-edges clamped to the elevator halves to set the travels and trims. This obviously exaggerates the problem.

[Troy Newman]

I would say try another servo in that second servo spot. The one you have might have a pretty worn pot. If you are the original owner and have never used the programmer on them then I can't imagine the two servos being programmed differently. Obviosuly they are not responding the same to the same signal input. Something is up with them.

Troy

[wylieruneyjr]

I find that if you do all that is listed above and still have lag, then there is another issue. As close as you can the distance from the hinge line to control horn linkage is critical. Even being off a half turn will cause lag. The control linkage that is even a half thread lower than the other will get to the end point faster. If you agjust the end points the same , you still have one faster and one slower. Control surfaces are different thickness at each end. So you cannot have the linkage in the same hole on control horn or same amount of threads above the control linkage when thread bolt " like Rocket City " type control hardware is used. I see this mistake too often. And a little bit of differance in length from hinge line to control hardware makes big differances in travel speed of control surface, The more leverage you have on control surface the slower the speed. If you move the control linkage on the lagging servo closer to the hinge line it will speed up, then reset your end points. If you are using control horns move the slow servo linkage down one hole on control horn, you increase its speed. The servos not faster, but it moves the control surface faster.