Avioncis Bending: Differential Resolver (Analysis Part 2)

As mentioned in my last posting, I created a little program with the ATmega32 that indicates the sample point. The controller pulls an interrupt after zero crossing and will initiate the hold and sample ADC conversion.

Here are two different visual representations of the ADC trigger. The long line indicates a alternating state between two samples (this lets me measure the correct 400Hz on the scope easier). The two waveforms are the SIN and COS output from the Differential Resolver.




The image with the dots indicates when the ATmega will pull the ADC during run-time. ADCtrigger (dot) happens 2125 microseconds after zero crossing of the 400Hz reference.



Below is a code snippet that shows the INT port config and the interrupt handler with the fixed time offset.

/* Configure Zero Crossing Counter */
MCUCSR |= !_BV(ISC2);
GICR |= _BV(INT2);
GIFR |= _BV(INTF2);

/*
* INTERRUPT HANDLING
*/
ISR(INT2_vect) {

freqCounter++;

_delay_us( 2125 );

/* Toggle port PA0 .. which is the LSB on PORTA */
PORTA ^= 0x01;
_delay_us( 30 );
PORTA ^= 0x01;

}

Avionics Bending: Differential Resolver (Analysis Part 1)

Finally, things at work are starting to quiet down a little and I have time again to look at the L-1011 project. Over the next few days/weeks I am doing a deep dive into resolvers and other synchro devices. Today, I took the scope to a standard differential resolver that is very common in L-1011 avionics.

Here is a schematic of the most typical Differential Resolver used with project Avionics (most of them are made by Clifton).


So for this test, I put a 16Vrms 400Hz source to the R1 winding and pulled R2 and R3 to GND. I used the S1/S2 winding for the SIN signal and the S2/S3 winding for the COS signal. Here is what that it looks like on the scope:




The signals are stacked as follows:

• Top .... SIN signal from resolver
• Middle ... COS signal from resolver
• Bottom ... 400Hz reference excitation
  

By the way, email me if you need some pointers on how to build a basic 400Hz reference oscillator for less than $5.

Resolver a Zero Point



Resolver at 45º point:


Resolver at 90º point:



Here are some design thoughts:

Sampling at zero crossing does not lead to anything because all waveforms cross zero at the same point and the same time. We are looking for different amplitudes not time delay.



Sample at zero crossing plus 1/2 the 400Hz frequency ... or at zero crossing plus roughly 1250 microseconds: