backfire

Hitec 5925 programmable digital servos (and other Hitec digitals) have a power on boot up time of approximately 1.75 seconds. If power is lost for only milliseconds because of a marginal electrical connection, voltage drop on a long extension, weak battery, or power down spike caused by several servos moving simultaneously; then the servo may re-boot. Please wait another 1.75 seconds for proper servo operation. Alternative: make absolutely sure none of the above faults can ever occur or, use a servo that has a short start up time.

dirtybird

My Feedback: (5)

Where did you get this information? This may explain why they seem to be sensitive to marginal Y harness operation, if true. This should be verified by Hitec.

backfire

I measured the start up time of the servo from power up to actuation time using a time base oscilloscope. A simple circuit was devised to cut the power for a few milliseconds to simulate a power glitch. The tests were run on four 5925's and the results were the same for all. It appears that these are 'as designed' characteristics. I'm sure that Hitec knows this, albeit, doesn't trumpet it.

XJet

That doesn't sound unreasonable for a programmable digital servo.

I suspect that this window of apparent inactivity is when the servo watches for the signals that will flip it into its "program" mode where various parameters can be set using the Hitec servo programmer.

This is pure speculation, but I'm guessing that by only having the program-mode select activated during the first second or so of power-up, the risk of a glitch or noise received during a flight accidentally triggering the program-mode is eliminated.

Well at least that's how *I'd* design it :-)

MikeMayberry

My Feedback: (1)

Interesting post from a "brand new" contributor!?!??

Have you done a similar test with other brands of digital servos?

This is true, but it is not the problem you are trying to make it out to be.

Mike.

backfire

Hi Mike,

Glad to see that you could join this thread! Why haven't you mentioned the boot up time and the fact that the servo can reboot with only a short term loss of power? And, just what is the problem that I am trying to 'make it out to be'? While a 1.75 second set up time is not unreasonable for a programmable device for memory access and/or programmer access, the fact that this servo can reboot with what can be described as spikes on the power supply certainly is. Have you ever tested this servo yourself for these characteristics? Set one up on the bench, create a power glitch, and contemplate the horror of a 1.75 second reboot delay occurring in your expensive, proudly built airplane. Imagine it occurring in a 200 mph jet. I have also run this test on a Futaba 9451 digital servo. It has about the same start up time as an analog servo; power glitches cause only a short mechanical output glitch. For the record; this "brand new" contributor is an electrical engineer with over 37 years of mechanical design and electronic circuit design involving industrial sensors, signal acquisition and processing, control systems, and actuators.

dirtybird

My Feedback: (5)

Very interesting. Since I am also an EE and have the necessary test equipment I plan to verify this. I also have a Futaba and a JR servo to compare it with.
I have had trouble with Hitec servos glitching caused by a bad Y harness. I replaced it with a JR and had the same trouble. Thats why I traced it to the Y harness.
Have you made any measurements as to how long the loss of power must exist before the servo reboots?

backfire

Welcome fellow servo tester,

I have not measured the power supply glitch required for reboot precisely but believe it to be less than 15-20 milliseconds. If you do this test, insure that the supply voltage goes down below about 2.0 volts for the glitch interval. My 5925 test servos dropped out at about 3.0 volts (one dropped out at about 3.5 volts). I say about because it's difficult to measure the exact value. Spikes imposed on the power supply leads by the demands of the servo output stage add to the downside value. Try this: put the servo on a decent 24" extension and use a regulated power supply to gradually lower the supply voltage to where the servo just drops out. Then, quickly raise the voltage back up. The 1.75 second delay will re-occur. Adjust the supply to a value about 0.25 volts above dropout. Put a mechanical load on the servo (about 1/2 rating). Drop out will occur and a new 1.75 second reboot delay will occur. I look forward to seeing what your tests show; on the other servos you have also.

JoeAirPort

My Feedback: (41)

My 5645's took a long time to respond upon power up. The first time I powered them up I was very surprised.

I guess if you have good connections, there is no problem at all. And if there are problems such as a bad connection or battery, then there will be problems regardless of the servo. These things are like tiny little computers (processors) so a boot time is pretty much inevitable. I'm surprised this never came up before in this forum (or did it?).

dirtybird

My Feedback: (5)

I am thinking that it would be best to remove the connection to the motor and determine the level the amplifier drops out first. Then set up that amplitude in a spike(or a little more) and play with the width. Exactly how I am going to do that I don't know. I have a function generator and can simulate the spike but I will have to figure out a way to control the power supply voltage with it.
Its not a good idea to use a regulated power supply for this type of test because as you discovered the regulator gets in the way of the servo load pulse.

dirtybird

My Feedback: (5)

If the servos reboot on low voltage this can be a serious problem. It has been shown the voltage will go low due to the internal resistance of the battery during peak loads. This can occur when several servos are moving at the same time. If the servos drop out for 1.4 sec. things could get interesting. Others have mentioned receiver drop out due to low voltages during extreme manovers. That means the voltage has dropped to less than 3.0 volts. I think if you want to use Hitec digital servos you should use heavy duty extensions and high capacity Ni-Cad batteries. I might even consider going to 6.0 volts

JoeAirPort

My Feedback: (41)

I can see your point there. I always use 6 volt batteries.

backfire

I should clarify the issue of down spikes on the power supplied to the servo. These spikes are fluctuations that occur on the servo end of the 24" extension used (viewed with a scope). They do not appear at the output terminals of the power supply. The power supply used was a high quality unit capable of keeping up with the demands of the servo. Also, I did not intend to imply that these spikes are unique to the 5925's. They appear with most servos, but to a greater extent of coarse with higher power servos. The point I want to emphasize is that very short duration down spikes in the available supply voltage can cause this servo to reboot. What performance margin exists in an aircraft setup that consists of the following: 6 servos (5 being digitals), some long extensions, a one year old 4-cell 1700 mah battery (half discharged), and 3D maneuvers.

Peter Burgess

From a practical point of view, I have been using Hitec Digitals for some time now and have never had a reboot problem when flying. (Fingers crossed)
An observation that I have is that although the Hitecs seem to be as fast as other servos they also seem to be more 'damped' in operation, do you share this opinion?

XJet

This behaviour might explain some of the "problems" quite a few people seem to have mentioned when using Hitec digitals with "less than perfect" extensions.

Perhaps the transients caused by the high current demands these servos produce has been causing this reset condition.

Damn - I just put four 5925s into a 3D funfly and am running 700mAH AAA NiMH cells. Although it's a 5-cell pack I might just have to scope the leads and see just what the voltage dips to under transient servo start/stop loads.

ahughes

I was using Hitec 5925 digitals on my split elevator and I was using a 4.8v NiCd battery, 24 inch HD extension cable and a HD Y cable to the two servos. The first time I checked the setup I was moving the elevators up and down pretty fast and I was very surprised when one of the servos would stop moving and then when I would stop moving the stick, a couple seconds later the other servo would catch up and move to the proper position. My second configuration was to utilize two of the receiver outputs (one to each elevator) and two 24 inch HD extension cables. This configuration produced the same results although not as many occurrences.

Needless to say this was quite disturbing. A fellow flier recommended using a 6.0v battery with a low internal resistance (2400RC NiCd) and also devised a clever isolator for the power to use in conjunction with a servo isolator. I tested this setup and from that point I believe I had no more problems with the servos re-booting and put 4.5 flights on my Giles 202.

ZAGNUT

My Feedback: (1)

huh? as in "crashed half way through the 5th flight"?[]


dave

ahughes

Yes, crashed half way through the 5th flight. After much thought on my part as well as others, I do NOT believe the cause was servo related. The Giles was a Creek Hobbies plane which used a spring to retain the wing panels. I believe they pulled out and the panels became misaligned during a very fast roll on high rates. I have since found threads about the inadequacies of the spring for retaining the wing panels on Creek Hobbies ARFs. As usual hindsight is 20-20.

DrFun

My Feedback: (4)

Intersting topic,

I was also suprised for the the wait time for the digitals when I started using them, and understand the concerns and a good point of view.
I had noticed it in 5475's and 5645.
That is one more thing to worry about.

Also a similar problem related to booting is in Futaba 9CAP, it takes 1-2 seconds for the TX to boot and I remember a fellow RCU member crashing a 40% plane because of it, as while he was trying to hook his neck strap he accidentially turn the tx off as the on/off swith is below the hook, and he did not enough altitude or direction to wait for couple of seconds.

After reading that I don't like unhooking my neck strap during the flight.

cap10b

This is another reason for me to get rid of my hitech digitals.
Love the analogs... and I will buy the micros and minis.
But I've had enough of hard overs, slow boots, lock ups.


jds

XJet

There are a hell of a lot of people using Hitec digitals without any problems so rather than just toss them I'd be far more inclined to simply ensure that when you use them you've got good quality extensions and a decent 5-cell receiver pack.

Knowing that a reboot can cause problems is good -- it allows you to take precautions.

dirtybird

My Feedback: (5)

I decided to do a simple test without instruments. I hooked up ten servos to one receiver with two Hitec 5625 on a Y harness, two JR three pole digitals, two Futaba coreless digitals, two Futaba coreless analogs, and two standard Futaba servos on a Y harness. Then I set my 9C on servo cycle and observed the results. Servo cycle drives all of the servos in one direction to its extreme then reverses them all at the same time. This results in more than starting current draw for the servo as the motor must first stop then start in the other direction. I used a four cell 700mah ni-cd pack for the power source.
I saw no signs of the reboot problem in any of the servos. To make it more interesting I put the 5625's on a 36" ext. There was still no sign of a problem.
As far as I am concerned, the reboot time of the servos is not a problem. I do not plan to use more servos than this at a time.
You may want to run your own test.

ahughes

Dirtybird,

Just a note, I also did not get a servo re-boot when I moved the controls manually in a steady fashion, and also using the servo test (I have a 9C myself). The 9C produces a slow moving wave for the servos and I would not think this would induce spikes in the supply line since the change of direction is relatively slow. For the above mentioned tests, my servos slowly and methodically moved back and forth without a problem as well (5 Hitec digitals on HD extensions and one Hitec analog servo).

It was only when quick movements of the stick were made that the problem occurred. Also, when I observed the problem, the servos were connected to/and fully deflecting the control surfaces thus creating a load on the servos, especially when the surface direction was quickly reversed.

This may not be a huge issue for most people but I would not dismiss this so easily given that much of the flying people do at some point involves rapid movements of the control surfaces (especially 3D). These control surfaces are not only loaded by the movement of their own mass but also the resistance of the air on the surfaces as well.

Thanks.

dirtybird

My Feedback: (5)

The 9C servo test is slow moving except when it reverses. At this time the servo is essentially stalled presenting only the resistance of the armeture on the load. This is the greatest load you can put on the system as all servos reverse at the same time. It is not continuous however.
The servos I have are mounted in aircraft and had the mass of the control surfaces attached to them. They were not loaded by wind loads.
If your tests show a failure you would probably want to do something to fix it. Like going to other servos or increasing the voltage or capacity of your battery.
For me I don't seem to have a problem.