dropping voltage on retract channel

[unclecrash]

Hi guys I recently bought an airtronics 2.4 system along with 2 new 5cell packs2700mah putting in a giant scale plane which needs nose weight thats why I went with the 5cell packs. Problem Im having I was testing everything on a older 4cell pack that I dont want to use in the plane.Everthing worked fine. Till I put the dual 5cell packs in then the retracts won't even work. They are not burned up because I tryed the 4cell again and they work. Now Airtronics says 5cell will not hurt there recievers but I dont know why the retracts wont work on there reciever with a 5 cell. I tried the retract on another system and they worked fine with 5cell pack. Now others have said the JR791 will not handle 6volt so Im wondering if there is an easy way to drop the voltage on the gear channel only by using resistor or something cheap. Really dont want to blow another 60 bucks for batteries. Thanks in advance for any help Dave

[Ron W3FJW]

Since you're doing giant scale, you shouldn't be depending upon the Rx providing power to the servos. Signal only. Run the servos directly off the battery (+ & - leads) and only run the signal & ground leads to the servos from the Rx. Takes a lot of stress off of the Rx and provides a greater margin of safety. Even better is to use two packs. One for the Rx and one for the servos.

[jeffs555]

A power diode in series with the power to the servo would drop somewhat under 1 volt. http://www.radioshack.com/product/index.jsp?productId=2062577

[Ron van Sommeren]

But that voltage drop would only occur under load Jeff. At no load servo would still feel 5 cells.

@unclecrash
Same goes for a resistor.

Vriendelijke groeten ;) Ron

[unclecrash]

Since you're doing giant scale, you shouldn't be depending upon the Rx providing power to the servos. Signal only. Run the servos directly off the battery (+ & - leads) and only run the signal & ground leads to the servos from the Rx. Takes a lot of stress off of the Rx and provides a greater margin of safety. Even better is to use two packs. One for the Rx and one for the servos. Ron I like the idea but don't understand. Could you draw me up a scematic of the setup. Im running dual batteries into the reciever now on different channels.incase of a switch or battery failure. I also went back to 4 cell packs. Thanks Dave

[JimDrew]

The problem is not overvoltage, the problem is the retract servo *requires* a 5v PWM signal to drive it. Our customers have run into the same problem. Most modern receivers output 3.3v signal levels and that will not work with this particular servo. Your solution would be to get FMA's servo buffer. It will convert the 3.3v input signal to a 5v signal that will properly drive the servo.

The reason the 4 cell appears to work (and it won't when fully charged), is the differential voltage. The circuitry of the servo is looking for a threshold and too high of an input voltage (the reference voltage) prevents the signal level from reaching .6v or less (TTL low).

[unclecrash]

The problem is not overvoltage, the problem is the retract servo *requires* a 5v PWM signal to drive it. Our customers have run into the same problem. Most modern receivers output 3.3v signal levels and that will not work with this particular servo. Your solution would be to get FMA's servo buffer. It will convert the 3.3v input signal to a 5v signal that will properly drive the servo.

The reason the 4 cell appears to work (and it won't when fully charged), is the differential voltage. The circuitry of the servo is looking for a threshold and too high of an input voltage (the reference voltage) prevents the signal level from reaching .6v or less (TTL low). they work fine with the 4cell fully charged iits when I went to the 5cell the totally quit wound not even jitter or move

[JimDrew]

Then your 4 cell is not fully charging. Trust me, we went through this exact same problem with this servo. It's the only servo on the market (so far) that has this problem. It won't work at all on a 4 cell pack with any of the 2.4GHz Futaba receivers because their servo outputs are 2.7v.

[dmccormick001]

If you search some of the build threads for giant-scale warbirds, you'll find several instances where guys have run into this same problem with this particular servo. Strange thing is, not all of the 791's seem to exhibit the same problem, but a significant number of them simply will not work on a 5-cell/6 volt system. I have to disagree with Jim Drew, however, that the servos that act this way don't work on a 4-cell/4.8 volt pack. In my experience, they do. I have a pair in a Hangar 9 P-51 that will work just fine on a 4.8 volt pack, but plug them into a 6 volt pack and they don't do anything. There are lots of guys who've had this same experience, so I know it's not just mine, and yes, my pack is fully charged.

There's an easy workaround, though, and it's actually a much better way to power any servo that draws a lot of current, as it prevents the current from being routed through the receiver at all. Take an ordinary servo wye, and clip the red wire that runs down the center of the harness, somewhere between the female end (that's the one that plugs into the receiver) and the center junction where all three wires are connected. I usually clip the wire and remove just a little of it, about 1/4", just to make sure there's no way it can re-connect accidentally. Now plug the female lead into the receiver, and the battery into one of the male leads, and the servo into the other male lead. If you have two retract servos, you'll need another wye to connect them together before you plug them into the wye you've altered.

This arrangement will allow the receiver to control the servos via the white/yellow signal wire, but the current won't go through the receiver.

[JimDrew]

The Hangar 9 P-51's is where we found this problem ourselves. Works great with a 4 cell pack that has a load (other servos connected), but would not respond with a 5 cell Nimh or 2 cell A123. If you unplug the other servos and use a fully charged 4 cell pack, the same problem occurs. Like I said, we went through this already and I spent a lot of time trying to figure out what was going on. Since we manufacture radio systems, I was trying to determine if our receivers had some type of a problem. I examined the electronics in the servo and determined without a doubt that they use the input voltage for the reference voltage instead of using a regulated voltage. This means that if your PWM input from the receiver is not "high" enough, it won't count the pulse. It's a simple as that. So, if you are using a modern computer operated receiver, the solution is to use a servo buffer (like from FMA) to change the receiver output pulse from 3.3v to 5v.

[dmccormick001]

Here's a simple diagram for the "electronically challenged". Probably none of those here, but just in case............

JimDrew:

I don't doubt your expertise or knowledge in this area at all. I wasn't trying to be argumentative or anything, just letting everyone know what I've experienced in actual use. I actually have four of these servos now, they just won't operate on a 5-cell pack, but I have been able to use them with no problems on a 4-cell pack. Perhaps the Rx itself pulled the voltage down enough, perhaps it was something else, but I know they did function just fine with the 4.8 volt pack. I (and a lot of others) decided I didn't like the idea of a pair of really current-hungry servos operating on the same battery pack that my Rx is on, so I use a wye like the one in the drawing to make sure the servos don't run the pack down and cost me a plane.

[JimDrew]

Dave, using a separate battery for the retract servo is actually how the problem was discovered in the first place. A 4 or 5 cell pack worked fine when all of the other servos were plugged in due to the load. Our customer wanted to isolate the retract servo power from the other servos because of the known issues with these servos having the ability to rob virtually all of the battery power when stalled (not good for 2.4GHz radio systems). Using just a fully charged 4 cells pack would exhibit the problem. If the pack was not fully charged, it worked. A 5 cell Nimh or A123 pack would also not work. After buying a set of these servos and doing various experiments I finally ripped the electronics apart to figure out what the true problem is. I just wanted people to be aware of this problem, I would hate for the gear to go up and never come down because of a voltage issue. Like I stated, the servo buffer fixes the actual cause of the problem.

[dmccormick001]

Jim:

Do you know if JR/Horizon has ever addressed this issue? It seems like it would be a really simple one to solve, and would make the 791's a lot more appealing to some builders. I know that when this behavior was first "discovered" their reputation took a real beating, and some guys still won't use them in a giant-scale planes because they don't want to work around the problem. It's a shame, because they're one of the most powerful servos I've ever seen.

Also, do you see a problem in using a modified wye like the one I posted? I've been doing it now for about 18 months no problem, and plan to do it again on a P-47 I'm building now. With the wye, if a servo stalls, goes bad, etc., the most it can do is run down the seperate pack used just for the servos. The RX stays powered by it's own pack.

[Ron W3FJW]

With the wye, if a servo stalls, goes bad, etc., the most it can do is run down the seperate pack used just for the servos. The RX stays powered by it's own pack.
That's what I was saying in my last post. However, best to have at least the throttle & rudder servo off the Rx pack so in the worst case, if the servo pack goes bad, you still have some control over the bird. Hopefully, worst case wouldn't be a failure with full aileron deflection.

[jeffs555]

Should be not problem with that "Y". It will connect the signal and ground wire to the receiver, and power the servo off its own separate battery pack.

What Jim says about the signal levels is most likely correct. Servo City also sells what they call a boosted servo extension and describe the problem the same way. http://www.servocity.com/html/boosted_servo_extensions.html

Or for really cheap you could order from Hobby City.
http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=8356

[unclecrash]

Thanks guys great info. Now I understand from the diagram what Ron was trying to get across to me. Im not the best with electrical.Now mechanical I can solve issues where others will be scratching there heads. Since yesterday have cut my brand new 5cell packs down to 4 cell 2700 mah each pack. One on battery channel 8 with a switch other on channel 7 with a switch so Both will have to fail to bring me down. And am thinking I will also run a bat to the servo on a y with the red wire cut to power just the retracts. What do you think will this get the job done :eek:

[JimDrew]

I doubt this issue was ever addressed. I know that Horizon recommends the servo booster when using a Spektrum system, and I believe that they actually sell one. It's a simple matter of changing what the servo needs to see. You have to remember that when this servo was designed (long before computer radios were common), the voltage output of a standard receiver PWM pulse to a servo was always +B voltage, with "fancy" receivers clamping the voltage to 5v.

The Y cable is a great (and recommended) solution for preventing a stuck gear from completely killing the power to the system. Using a 4 cell pack is plenty enough voltage for operation and low enough apparently with two plugged into the pack to not cause a problem with the pulse. I would still recommend using the servo booster to make ABSOLUTELY sure that the gear go up AND down!

[dmccormick001]

That's what I was saying in my last post. However, best to have at least the throttle & rudder servo off the Rx pack so in the worst case, if the servo pack goes bad, you still have some control over the bird. Hopefully, worst case wouldn't be a failure with full aileron deflection.
Ron, here's what I do...

The Rx has it's own battery pack, 4-cell, whatever size you feel you need. The Rx pack must be a 4-cell 4.8 volt for this "workaround" to work. This pack will supply ailerons, rudder, elevator, and any other servo or device plugged directly into the Rx. For the retract servos, (or flap servos maybe), use a wye modified like the drawing and a seperate battery pack. The Rx will still control the servos connected to the modified wye, but all the current for them will come from the seperate pack and will not flow through the Rx's circuitry. The second battery pack can be 4 or 5 cell, doesn't matter. If something happens to the retract servos, they may not come down (worst case) but you should still have control of the plane.

[Acetronics]

Hi,

+1 for JimDrew analysis ...

This kind of problem has been successfully solved by using Level adaptators between the new 2.4 Ghz Rx ( mainly Futaba ...) and other brands servos or custom Pic modules using Schmitt trigger inputs for signal.

a couple of complementary Lo power transistors like BC 547/557 or 2N 3904/3906 is fine. ;)

Alain

PS ... Flown aboard 3 - 4 meter wingspan "liners" with 2 to 4 ... jet engines. :D