Tome,

I agree with all you've said. In general the autopliot
amp is a DC circuit and has a slow response time.

However, this system does generate RF signals which
are used to sense the pitch and roll attitude and the postion
of the servos.

This autopilot system uses variable inductors to sense pitch
and bank information from the attitude gyro. Each one consists
of an iron and aluminum metal vane attached to the corresponding
axis of the atitude gyro which moves in front of a iron core coil as a
pick up. I found a test box designed for the system that includes a
jack that provides a dummy AI gyro allowing the substituion of its
inductors in place of the actual gyro for bench testing. It contains
standard looking RF coils with tunable slugs.

The service manual states an operating frequency of 10 MHz, but that is
stated only as an exmaple. I'm not sure it that is the actual freq or not.
The resonant frequency is determined by the inductance in the AI gyro,
air variable caps in the servos and some variable air caps used for trim
and centering adjustments. The circuirt is arranged so that as the airplane
changes pitch or bank, the changes in reactance cause the circuit to change
frequency. A discriminator in the amp detects the frequency change,
produces
a DC voltage which is amplified to drive the servo motor in the appropriate
direction to bring the circuit back into resonnance via the feedback caps
that sense servo postion. This or course moves the control surfaces and
brings the airplane back to level flight attitude.

The components are connected together with specific lengths of low
capacitance cable. The cable and connectors appear to be very similar
to that used in GM automobiles in the 1960s. At least my 1966 Pontiac
had a similar cable and connector from the antenna to the radio. It is a
hollow plastic tube with a very small center conductor that is loosely
coiled
inside the plastic tube. There is a braided shield over the outside of the
plastic
tube, and a outer jacket over that. Some people have pointed out it is very
similar to RG-62, 93 ohm coax without the spirial wound center insulator /
spacer. I measured about 10.5 pF per foot on a 28" piece of this cable with
connectors installed on each end.

This has been quite a project. I basically have seen all the components
work
in the airplane at various times, but the system as a whole is intermittent.
Most time the wing leveler funciton works correctly but once in a while it
wants to turn right. The pitch circuit has been working but is now acting
up. The altitude hold was badly out of adjustment, but I was able to get
that setup properly and working, but the climb and descent airspeeds are
not right yet. I'm not sure if I've got bad connections, intermittent
components
in the amplifier or both. I'm planning to pull the cable assemblies out
along
with the amplifier and get it on the bench for testing.

Therefore, it would sure help to have a basic understadning of the amp
circuit
during that process. Therefore, I thought learning about the transistors
charateristics
and the inductance values would be helpful. I've already measured all the
cap
values and ranges. Next step is to generate a schematic for the amp and
learn how it is supposed to work.

It might also be fun to design and built a replacement for the amplifier
using modern components.

Thanks again for your help.

Ronnie


"K7ITM" wrote in message
oups.com...
From that era, and from what you know so far (that they are apparently
pretty much all germanium parts), it shouldn't be too difficult to
deduce how it worked. It's unlikely any of the active parts are
particularly esoteric. Usually you can figure out the polarity of the
parts from the schematic, and assuming about 0.3V for forward junction
bias and current gains typical of the power and signal transistors then
in use should get you close to what you need. I suppose nothing in an
autopilot needs to be very fast, so it's very unlikely any are RF
transistors. It's possible that the circuit requires close tolerance
on some feature (like gain), but that's poor circuit design and I'd
hope for better of them. Once you have a circuit diagram, maybe you
could try "spicing" it, if you're not comfortable with a simple paper
analysis.

Good luck with it.

Cheers,
Tom