[OT] Calibrating an auto level?

[N_Cook]

This is purely mechanical surveyor's level, for levelling only,
absolutely no electronics.
A tiny mirror or prism , on a very constrained and damped pendulum, set
in the optical path of a x28 telescope set on a spherical mount to a
tripod, to get initial rough levelling by x-y spirit bubble, then the
pendulum system corrects within the last +/-10 angle minutes of level.
It is a Sokisha B2 probably decades old, working with good repeatability
but poor direct levelling. Checking against levels off a lake it is
about -27.4mm over 100m, but consistent on return legs and repeating.
Requiring offsetting , for any non-balanced levelling, ie non-equal legs
either side of the tripod.
I've the user manual and the carry case contains the tools for adjusting
the calibration but no blurb. Cannot find a technical manual for this or
any other similar, different makes, of auto level. But I found where it
looks like you adjust , by unscrewing (not obvious it unscrews) cover
around the eyepiece.
Anyone any ideas on how to do the adjustment, assuming I have a good
long-leg sighting run. There are 4 brass set screws, holed to take the
adjustment tools , arrayed around the eypiece.
So far I've taken and recorded vernier calliper reading of each of the 4
set screw heads, across the diameter to the threaded boss diametrically
opposite, plus 4 dots of paint.
Would top and bottom ones adjust the vertical displacement?
and would the process be -1/8 turn on one and +1/8 turn on the opposite
other, measure the displacement at distance and reverse if going the
wrong way and repeat until nearer 0 displacement on the staff at the
relevent distance.
Would the other pair of adjusters have to be slackened off ,1/8 turn
say, while doing this , then finger-tip torque tightened back 1/8 turn ,
or at least balanced if not 1/8 turn available then. Its not possible to
see how this adjusment mount is constrained, or not. Hopefully no
threadlock on the set-screws or if there is, then overcome wiht a few
seconds of soldering iron.

[peterw...@gmail.com]

Some SWAGs here.

When I came up, the surveyors I was around pretty much hated self-leveling devices. "How do you know?" was their call. Today, it is all data points and GPS loaded into a self-everything transit.

http://sokkisha.jp/level.html Is a link in Janglish, but readable enough. It has an exploded diagram and points out several leveling screws. It would seem to me that the key is getting the base completely level (dial mike?), then adjusting the eyepiece to fixed level points.

Good luck with it. These old devices can still be useful.

Peter Wieck
Melrose Park, PA

[N_Cook]

Thanks for that, I'd not seen that page . This could well be 1963
vintage, the serial number ends 63.
The GPS people have done a great job at convincing people its soooo
accurate, despite 30 to 50 "corrections" routinely included.
Local archaeologists stopped using GPS, including dGPS, for initial site
positioning, using it only on site. For a site fundamental reference
they pay Ordnance Survey 20 GBP for data on a nearby benchmark or
feature on an old/protected building. They had so many problems with
wrong GPS data over the years.

[N_Cook]

From the exploded view there, the 4 set screws (2 shown there) must
adjust the reticule/crosswires, less problematic than some part of the
optics sytsem I assume, makes me a bit more confident.

[N_Cook]

A technical manual for different make but something similar
http://www.redtailrental.com/assets/upload/Level%20-%20Manual.pdf
Diagram M, shows similar removable cover to expose cross-hair adjusters,
4 + 1 for this one.

If weather stays up, off to the lake again today , set up a pair of
staffs wrt to the lake surface and try adjusting to <>0

[N_Cook]

It looks as though tightening torque would be very critical.
About 1/2 turn off opposing set screws changes the image about 70mm at
100m. Decided the way to do this is to measure one side only vertical
and horizontal , the side where the moving brass block is exposed beyond
the mount , so easily measured diametrically. First attempt went approx
same error the other side of zero. So adjusted to half the measurement
and then make final adjustemnt by varying the finger-tightness of torque
to the opposing set screw, before a dot of laquer.
.

[peterw...@gmail.com]

Make sure everything is temperature-stabilized prior to making adjustments (but you knew this already). Again, back in the day, the instruments lived in the back of the truck, not the cab, so they were pretty much always at ambient.

[N_Cook]

For something that is maybe 55 years old, amazing it works as well as it
does, thinking corrossion products/dust getting in the narrow confines
of the damper. The original specification of these things was +/-1mm at
1km, which I find hard to believe. Before the days of laser spots on the
distant target, I'd have thought well beyond the resolution of even the
youngest eyes even with the most contrasty target possible.

[Tim R]

And lasers will bend, if sufficient temperature gradient over the distance is there.

[N_Cook]

It relies on ,seemingly, uniform gravity in the locality , the sort of
thing the GRACE satellite measures deviations of. I tried waving a 4Kg
lump of lead near the pendulum housing, but no image movement seen.
Next time of an outing with it I'll try waving a speaker magnet near it.

[N_Cook]

Waving a magnet around does noticeably affect the image displacement.