[R-390] Collins R 389 PTO

[Mark Huss]

Since I own an R-390A, I was referring to -390A's.  Did not know an 
R-389 had a 50 turn PTO, which would make matters worse.

As for 1024  step encoders, we are talking bucks here.  need to keep the 
cost practical.  Of course, the big expense is R&D, as Flowertime 
pointed out.  More a labor of love than a money-maker.

Although I haven't actually tried it, it seems that several high-quality 
10-turn pots have the necessary resolution to get close to 10,000 steps 
per travel for under $50 each (the 200,000 turn lifetime concerns me, 
though).  may only be good to 20 hz after taking into account the A/D 
resolution.  The key here is to use a precision voltage source to feed 
the pot and the reference for the AD converter.  this way any drift in 
the voltage will not affect the AD output.  Not counting labor, you 
could probably throw together one for under $500, including circuit 
boards.  given a clean reference source, (>100uv ripple), a simple 
program for a PIC or equivalent would convert each 6.5536 steps into the 
proper setup data for the DDS, perhaps through a look-up table on a 
serial EEPROM.

Then install in the R-390A and use it as a precision goniometer to 
establish the corrections for the non-linearity in the pot at operating 
temperature.  Work out a formula for interpolation vs.temperature for 
temperature correction, and you should have the thing tracking well 
within the original specification.  May not even need the temperature 
comp since a few hundred hertz non-linearity would not be critical.

The three biggest issues would be low lifetime of the potentiometer, 
getting the ripple down on the reference voltage, and jitter from the 
ADC giving you sudden ten hz shifts.

Anyway, thats my 40 cents.

Roy Morgan wrote:
>
> On May 3, 2009, at 9:18 AM, Mark Huss wrote:
>
>> There are several DDS (Direct Digital Synth) VCO kits running around,
>
> Mark,
>
> I appreciate your comments, and will store them and those of others 
> for possible future efforts on the problem.
>
>
>> ..
>> The PTO  covers 1.06 MHz in 10.6 turns, or 100 kHz in one turn.
>
> Are you referring to the R-390 and R-390A PTO's?  The R-389 PTO is a 
> different animal.  According to info I have the thing has a frequency 
> range of  .470  to .980  Mc. and operates over 51 turns (probably just 
> 50 turns in use.)
>
> That is 470 KC to 980 KC and FIFTY ONE turns.
>
> I don't know what resolution you'd need.  But maybe if you had a VCO 
> operating from say 4.7 mc to 9.8 mc you could divide and filter the 
> output to get what's needed.
>
>
> Roy
>
>
> Roy Morgan
> k1lky at earthlink.net
> 529 Cobb St.
> Groton NY, 13073
>
>
>
>
>
>