Airplane Gyroscope (Compass) repair

[HighVoltage]

I bought this electrically driven airplane gyroscope in a broken state and hope to be able to repair it.
It seems the motor bearing is bad and that is why the electric motor is not turning at all.
This is the first gyro I am taking apart and I am just amazed by the quality of all parts inside.

Does anyone here know, if these bearings can be exchanged or if they even available.
I would also guess that the system will be out of balance, if I should be able to exchange the bearings.
But, this is just for fun and to play with an awesome piece of equipment.

Also, what is the relative complicated PCB for.
It can not only be to keep the motor at the same speed.

This is a Directional Gyro, made by RC Allen
Supply voltage is 14V.

Here are some first pictures.

[SeanB]

Yes the bearings are likely standard ones, or at least you will find a standard C3 grade one that will fit, though you will have to remove all the grease from it and replace with a very thin oil coat only. Some of them will be repairable by replacing the balls only, but those tend to be older air operated ones.

PC board is an inverter, to provide a 400Hz 3 phase drive for the motor of the gyro, so it can operate off the aircraft battery bus of 13V8.

just note that you have to assemble them back in the correct order, and every screw has to go back in the same hole in that cage frame, and every shim has to go back in the same orientation and in the same position. Then for the loose balls in the system you have to adjust the preload to the right level, often by using balls of a slightly different diameter in there, and polished to a certain finish, along with the raceways they run on.

When running right they are reasonably quiet, just the slight vibration and the whine. I used to repair Avionics instruments, so had a lot of exposure to these types of unit. You really do not want to know the price of some of the spares I used.

Check the resistance of all 3 coils of the motor, they should all be the same resistance ( within around 5% of the others) which will prove if the motor windings are working, or if the driver side is cooked, which is more likely here.  The rotor is hidden, and will very likely be fine, they will still run with badly worn bearings, and the drift on the system will be incredibly high ( a lot more than normal drift, and it will vibrate on the bench in any case) before those bearings wear enough to stop it turning.

[HighVoltage]

SeanB,
Thanks for those very helpful suggestions. I followed your detailed advise
- Marked the position of all screws that I removed
- Took pictures of the location and placement of cable routings

Then I took the motor apart.
The lower bearing was full of dirt (Old grease I guess) and cleaned it all with alcohol.
Blew out the dirt with compressed air
Repeated a few times until all the dirt was out.

Here are some pictures of the bearing.
It is a marvel of engineering
The small protective covers over the bearings can be removed to get access to the steel balls.

 

[HighVoltage]

Then I added some high quality very thing oil to the bearings.
And the rotor started spinning right away very lightly.
So, re-assembled all together and it started spinning right away.

I was not aware of the 3 phase motor but it is obvious with the three cables.
With a differential probe I measured the voltage between two connectors:
406 Hz and 35V

But it seems the gyro has a drift now.
I see a lot of adjustment screws but I have no procedure of how to correct the drift.
Any suggestions for that?

[SeanB]

Do not adjust anything, you need new bearings. The old ones have become pitted or worn, causing the drift. You will need to either get or make a small bearing puller to get the old ones out.

drift is normal, it should drift one full rotation every 24 hours, but more than that is caused by either uneven bearing motion or by an unbalanced rotor. As the bearings were full of gunk they will have worn so they are not smooth. Last overhauled and balanced April 1985, so 31 years will have had their toll on them. Replace with good quality SKF or NTN bearings, of the right size. If you have a choice of grades get the closest tolerance ones.

[HighVoltage]

This must be very difficult to calibrate such a Gyro when it is suppose to rotate once a day.
Thinking about it, it will rotate (drift) once a day a full 360 degree only when on the north or south pole.
It will not have this drift anywhere on the equator.
And everywhere else you need to calculate the drift, based on your location on earth.
How interesting

This would be like calibrating your DMM with a 10V reference that has a different value depending on where you are in the world and what time of the day it would be. How crazy.
I tried to find some calibration procedures for such a mechanical gyro but could not find anything.
May be the manufactures want to keep this a secret.

OK, I will see if I can get new bearings.

One more thing:
How can I measure the speed of the rotor, just for fun.
I am thinking probably best with strobe light and looking at the air vents.
 

[SeanB]

Strobe light, or a signal generator and a blue LED across the output, if it will handle driving a 4VAC square wave ( with as narrow a pulse on time as you can get) into the LED. Or you can use a single BJT to drive the LED at 300mA and make sure you only have a 1% duty cycle. Just use a sharpie to blacken one vane so that you have a reference.

[HighVoltage]

The rotating axis of the motor is screwed in and then fixed with a brass nut.
This rotating axis was originally NOT at the limit, when I took it out.
(After loosening the nut, the axis turned freely in both directions)
But this means I can not really find the original spot anymore.

So, I turned the axis a little away from the end point and now the gyro is not drifting anymore.
I think I will have it run for a few days and watch the drift.

Just for fun, I had the Gyro laying on the dash of my car and drove around in our town for a while.
After I came back to the same spot, it showed the same direction on the dial.
Really cool to play with such a precision mechanical / electrical device.

I am located in northern Germany and the earth rotation is about 12 degree / h here
So, I would expect this to show on the Gyro and being my basis for calibration.

But ....
How does a Pilot know of how to set the Gyro, based on where he is and what heading.
It is like constantly changing.
Only if I go North --> South direction will the gyro not be influenced by flying.
This is not so easy to wrap around the head as it seems on a first look.

[SeanB]

Normally there is a correction component applied from a fluxgate sensor in the tail ,but in your case this is just a gyrocompass, so the correction is simply dialled in, and the compass is used, along with regular reference to the magnetic compass, to enable the aircraft to make precision turns, and to navigate with no visible conditions. You turn on the gyrocompass to a heading, then align to the magnetic compass, to make a final correction in straight level flight, because the compass will have a turn error. Compass will swing and have positional error in a turn because of the banking you do to turn, but the gyrocompass will simply ignore most of this ( there are edge cases and turn limits) enabling a precision turn. Then you simply keep correcting the 12 degrees per hour drift ( there will be an extra component because of the rotation of the earth around the sun, but this is not sensitive or low noise enough to pick that up, nor the rotation of the sun around the Galaxy, or the relative motion of the Local Group) every hour, using the magnetic compass, and it runs from there fine, as your course will be not a straight line but a sector of a Great circle anyway.

Incidentally rotation rate is independent of position on the planet, all places rotate 360 degrees every 24 hours roughly, just that some places like the poles have a lower angular acceleration. The gyro erection system, which keeps the gimbals in a horizontal plane in level flight ( and which take a while to erect it unless you use fast erect by pulling or pushing the adjustment dial) will also apply a torque that tends to keep it pointed in the same direction.


OT I really wanted to have a moving map display like I used to work on, but that was a 50kg box, needed it's own 30kg computer to drive it, and used a pair of gyro blocks ( redundancy), a fluxgate sensor, an air data computer to give physical inputs of speed, altitude and such, and which needed around 5kW of power and cooling to operate. Those were mechanical marvels, like the older analogue autopilots and electromechanical navigation and aiming systems were.

[HighVoltage]

Incidentally rotation rate is independent of position on the planet, all places rotate 360 degrees every 24 hours roughly, just that some places like the poles have a lower angular acceleration. The gyro erection system, which keeps the gimbals in a horizontal plane in level flight ( and which take a while to erect it unless you use fast erect by pulling or pushing the adjustment dial) will also apply a torque that tends to keep it pointed in the same direction.

Your knowledge is very impressive, thank you for so many interesting facts.

I don't understand why the Gyro would turn 360 degree per 24h everywhere on the planet.
For some reason I expected the Gyro to behave like a Foucault Pendulum, meaning
360 / 24h at the poles and 0 degree at the equator.
Why is it different for the Gyro?

[SeanB]

At equator direction of swing is normal to the rotation of the planet, so no shift of the motion. At all other points there is some angle relative to the rotation of the planet so it will precess. Think about it, at the equator the force is directly down, and the rotation has zero component to the side while as you go up in latitude there is a sine and cosine component of rotation, which means it will rotate. they are nice, just you need a few floors of space to make a decent one and no draught to spoil the motion.

The gyro wants to keep pointing at the same point in the far universe, and the frame allows it to turn relative to the earth to keep that position unless you apply a restoring force to precess it to follow the nominal horizontal direction.

[Jr460]

I'll try to answer some of the operational questions about the gyro from a pilot side of things.

First off, we normally call that a DG, or Directional Gyro.  Second in small planes it may be driven by a engine driven vac pump.  The idea is that some things are driven by electrical power, and some things by air.  Failure of one or the the other does killing everything gyro based.

When you start the plane up and start taxi, not all the gyros ay be up to speed.  Your Attitude Indicator and your turn/bank will align just fine.  The DG can be at anything random.  One of the checks just before take off is to look at magnetic compass and set the DG to that same reading.  As you fly for a period of time, while in straight and level flight, check the DG with the mag compass.

In the aviation world, all directions are given based on magnetic headings, not true, thus you don't have to in flight correct for that difference because it changes based on where you are.

The next question you should have is why have a DG if you already are required to have the compass.  The short answer is that the compass will lie to you unless you are in straight level and not accel/de-accel flight.  As you progress towards an instrument rating, you learn to make turns with just the compass, with all of it's lead or lag and other problems.  Depends on the direction you are going, or ending up in, and for some errors you latitude.

The next step up is what is called an HSI, Horizontal Situation Indicator.  It is a combo of an electrical DG, with a VOR, glideslope, and heading but to drive an autopilot.  You can get them in small planes, I've had one, makes instrument approaches soon much easier.  In this case the panel doesn't have the gyro in it.  It is in box in the tail of the plane along with a magnetic flux gate.  In most cases it with align very closely on spin up. Once you tweak it, you don't have to worry about the drift as it is applying corrections from the earth's magnetic flied.

[bitseeker]

HV,

That DG looks great on your car's dashboard. All the best on the restoration!

[FlyingHacker]

Since I fly a plane with one of these I would seriously like one in my car too, it is so nice to have as a reference, and so easy to read.

They also make a vertical card compass that looks like a DG. http://pai700.com

It has significantly less lead and lag, and no dip errors like a liquid filled "whiskey" compass. It also requires to power.

[bitseeker]

That's cool. I've never seen one. The only aviation instrument looking thing I have in my lab is a wall clock with an altimeter face.