[ICOM] IC-775DSP antenna tuner observations

[paul at w8aef.com]

The antenna tuner on my IC-775DSP was kind broke.  Kinda in that it would work OK on some frequencies/bands, but not on others.

During early repair attempts I observed that when I forced the capacitors to a new position they had a rough feel to them as they turned, which I assumed was normal so I ignored it.

But after making several observations in the circuit, and replacing a few components, still no worky.

So in desperation I dissassembled one of the motors.

The construction of the motor consists of a two piece nylon spool that grips the armature and rides on an axle, the nylon spool keeps the armatuer concentric with the axle.  Except one end of the nylon spool was not attached to the rest of the spool and allowed the armature to move around and lose it's concentricity.  Close inspection of the loose spool end looked like that was the way it was built, but the other end of the spool's sleeve was glued to the other spool end.  I applied a dab of epoxy to what I thought was the broken part, and whoopee, the antenna tuner works better.  Still not 100% but better than before.

So I checked the other motor and same problem.  Same cure and now my tuner works 100%.

And when I forced the capacitors to a new position they had a very smooth feel as they turned.

Things I learned during this excersize:

There is not much technical information available on how the tuner circuitry works.

My observations include:

IC1a and b amplifies FORward power and REFlected power so the CPU, IC13, an M38022M2-138FP can calculate SWR, to determine whether or not the tuner requires adjustment.  Gain of IC1a and b during USE is a little over 2, during SET is 10.  When tuned the REF voltage at CP1 will be very close to 0 vdc and the FOR voltage will depend on power output, about +4 vdc when running 200 watts output.

IC2a and c and IC3a and d amplifies the RF sine wave and turns it into a pair of negative going pulses.  One pulse indicates the L phase and the other pulse indicates the C phase.  The pulses are fed into IC5a and b, a pair of J-K flip flops.  The outputs of the flip flops are summed at the junction of R27 and R28.  The sum of the trailing edges will create a negative going spike that, depending on it's width (the sum of L and C) will create a DC voltage that is slightly positive or slightly more positive.  This voltage is amplified by 10 in IC6a with a +2.5 vdc offset applied.  This offset voltage is fed to pin 64 of the CPU, which is an Analog to Digital converter input.  If the DC voltage is above +2.5 volts the computer adds capacitance, if below it subtracts capacitance (or vice versa, I forget which).  The important part here is when the SWR is 1:1 the voltage will be very close to +2.5.

Q1 sums the resistance component.  The C component appears at the Drain and the L component appears at the Gate.  The sum appears at the Source, a voltage at or near 0 vdc when the tuner is 'in tune'.  The Source voltage is fed to IC6b, amplified 6.8 times and a +2.5 vdc offset applied.  The offset voltage is applied to the CPU, another A to D converter input.  Sorry, I don't know what the CPU does with this.

The CPU (computer) observes if additional C (in addition to the variable capacitors) is required and applies same.

The drive motors for the variable capacitors are 4 phase motors that are controlled by the CPU.  The drive signal at RW, RX, RY and RZ are negative going pulses.  (before I found the loose armatures one of my motors had no pulses when less capacitance was called for, a new CPU fixed this).  I believe the capacitors are driven to full capacitance during the IC-775 power-up cycle so the CPU knows where they are.

As you pass a 100KHz point when changing frequency the MAIN UNIT feeds the new frequency information to the CPU, so it can automatically recall the last good adjustment and preset the capacitors.

It appears the switched inductance in the tuner is strictly a band function.  I did not check this but it seems reasonable.

I welcome anyone to add (or subtract) information as Icom uses this circuitry in almost all of their auto tuners.

de Paul, W8AEF