[TowerTalk] Fwd: Telrex rotator with sychro

[Hans Hammarquist]

 Hi Jim,
My offer still stays. I can send you one, unused rigt out of the box. 
-----Original Message-----
From: jimlux <jimlux at earthlink.net>
To: towertalk <towertalk at contesting.com>
Sent: Thu, Nov 22, 2018 1:17 am
Subject: Re: [TowerTalk] Telrex rotator with sychro

>On 11/21/18 8:00 PM, Hans Hammarquist via TowerTalk wrote:
>> It is very easy to generate the V1, V2, V3 voltages if you want a receiver synchro to be set at a certain>> angles.

>I'm thinking about the other direction - you have a synchro transmitter 
>on the tower, and you want to measure it in the shack.
Yes, that's what I am thinking about too. You have a Selsyn in the rotor and you want too, cheap, read out the position of the antenna.


>> You can even us a DC voltage for that. You just use the equations you show and replace the
>> "cos(fline*2*pi*t)" with "1". It gets trickier to go the other way around as "C" not necessary is 
>> constant. You have to calculate that out of the known value of V1, V2, and V3.

> An easy hack is to use the maximum of them (going from synchro to 
> digital) or set one of them to max, and the other two at a fraction of max.
Well, what you do is, I seen this in military applications, to set one input to zero and the vary the other two between min (=  -maximum) and maximum. You basically ground one of the inputs and run the other two in a 60-degree split. That way you only need to deal with two signals. It should be possible to do the same way when you decode a Selsyn in the scenario we want to deal with.

> 
>> I'm sure there is an easy way to do that but it becomes a "little" easier if you first make a Scott-T>> transformation (see Wikipedia "Scott-T transformer) because your equations will be easier to reverse>> and calculation of "C" easier too.
>> 
>> C=(Vr^2*Vi^2)^1/2
>> 
>> theta=arctan(Vr/Vi)
> 
>> When you finally are ready to implement your algorithm you have to make a synchronous rectification>> somewhere to get the value and the sign. You can do it before the A/D converter with a synchronous>> rectifier or probably with the software. (I would love to see a software that can do this as that reduce>> the hardware. It should be>> possible.)

> I was thinking more about just digitizing the AC waveform and 
> calculating the amplitude by fitting a sine wave.

> Or, it might be easier (I've not fooled with it yet) to turn each of the 
> 3 voltages into a complex number by "mixing" with cos/sin at the 
> excitation frequency.
Try to remember and this is easy to forget, you are not dealing with any phase shift between the signals. All the signals are in phase or maybe with a slight, but constant phase shift.

>> There are commercial IC-s around that will do this. 

> Yes and they cost about $2000 for some reason. Probably a limited 
> market. The last time I used them was in the late 90s, early 2000s, and 
> when I checked last week, it hasn't changed much.  They're actually a 
> thick film hybrid in a package that looks like a 24 or 28 pin DIP, but 
> with a stainless steel can.
I thought they would cost something like that. The only markets I know about is the military (who else would pay that much?) and the nuclear industry. The Selsyn is very useful there due to its ability to work in high radiation environments.
On a practical note: You don't have to excite the transmitter with a sinusoidal signal. A squarewave passed through a simple lowpass RC filter will do. I did that in an application and it works. The Selsyn is very tolerant to the excitation frequency and can, therefore be used with a, slightly frequency limited, squarewave.
I think there are many hams waiting for us to get an application/software for this as there are plenty of synchros out there waiting to be implemented somewhere.
73 de,
Hans - N2JFS