Repair of Gyroscopic Instruments

The context in this case is a KI256 which is a vacuum driven horizon, so the bearings see a ton of sh*it blown in; everything below a certain size according to the filter spec. I have never seen the inside of a KI256 but I bet it is pretty dirty.

Electric gyros should last a lot longer but rumour has it that – in some cases at least – they don’t. They have a DC input and use an inverter driving a 3 phase motor so the frequency could be anything at all…

I do not know of any ball bearing that has NO play in itself. But in cases like that or any other machinery that is required to operate at minimum noise you set a tiny bit of axial load on one bearing with slide fit in the case. So you get zero play in the assembly and that is what I´d do: Put some thin spring washer (wave washer?) under one bearing and it´ll be allright. No clearance in a bearing will normally lead to a blued bearing at high revs. These small bearings have a much higher rev limit than required in your case. With a typical 400 Hz generator you will get 24 000 revs at a three phase motor so I guess that is the kind of range we will expect here. Fitting an open bearing in this instrument is intentionally limiting the life of it – great for the manufacturer !!! Finding a ceramic type would help a lot as this is standard for extra high revs applications elsewhere but that would be a special in the specific size here. Another way for setting the internal clearance in the standard bearing is to experiment with the press fit of the steel sleeve that has to go over the 5/16" bearing o.d. to get the final 1/2" size – a bit tricky but doable. The best idea would be to find that extra space in the instrument to fit a closed 2 Z bearing at 1/8" width but I have no idea about the conditions therein – some more pictures , Peter ? Of course you will never ever fit this modified gear into a certified aircraft , this is certainly only for personal experiments . . . . Vic

OK; yes. The other “legal” aspect is that lubrication (and presumably the use of loctite) is banned per the MM, so the only option is to replace any affected parts. But since this type of servicing would be done, ahem, in the field, none of that is relevant.

I reckon this thread is going to be extremely helpful to a lot of people, especially as HBK move all repair work to themselves in the USA which will render repairs of all this stuff uneconomical to do legally, yet most owners of older hardware will not be able to rip it out and fit “glass” for 20k upwards.

Peter wrote:

In the last KI256 the repairer reported a “worn shaft” which I don’t quite get since surely the only wear should be in the bearings?

We now know from the part number that the radial play is to be in the range of 0.0002" to 0.0004". The bearing will need to be a light press fit into the housing and probably onto the shaft as well. It would only take a tiny amount of wear on the shaft to make the fit slightly sloppy – where “slightly sloppy” is more than the specified play in the bearing.

I reckon you must have found it, Alan. Brilliant work. I ought to buy some and see what they actually look like. It’s hilarious that they get charged out at something like $500.

No such thing as an unhealthy interest in anything to do with engineering… just don’t put it in your Tinder profile

The other thing that is peculiar about the bearing, and which I hadn’t appreciated when I made my initial comments is that it is of odd proportions – that is a large OD given the size of the ID. The technical term is that it is a “modified dimension” bearing (really!) The alternative part number I quoted does appear now to be good:

SSRI-6632ZA0205MCRA7P24LB15P

The basic bearing is SSRI-6632ZA0205, which you will find here:

The decode for the rest of the part number MCRA7P24LB15P you will find here

In case you are wondering why I have this apparently unhealthy interest in bearings, I did a big project early in my consulting career in the 80s around ultra high precision encoders.

Brilliant work vic!

Now… how does one select a bearing for this application?

That’s a good page LeSving but I can’t see any which meet the requirements. These two are ok but not on the diameter, so as per Vic’s suggestion one would have to make a sleeve

90k rpm is surely enough?

For avoidance of doubt my shelf-spare KI256 has been overhauled by a 145 facility… this thread is just for fun, and of course to help others should the overhaul options go the way HBK want them to go i.e. evaporate

I know only one ball bearing manufacturer, and that’s SKF. http://www.skf.com/group/products/bearings-units-housings/ball-bearings/deep-groove-ball-bearings/deep-groove-ball-bearings/index.html

allright, that was not an easy one: Problem is the width of just 2,76 mm , the type is called R 156. It is mostly 1/8" wide with seals, but also available with 2,76mm open . Bore is 4,76 but only 7,94 mm outer dia. , so that leaves an adapter for 1/2" o.d. to be made up. You may try there:

mini bearing

or here, see the PDF:
mini ball bearings

Vic

and the hole is 4.73-4.75 (I don’t have an accurate way to measure it).