High resolution dicital potentiometer or PC-controlled. How to?

[neazoi]

Hello, I would like to make an oscillator using the LTC1799 and I was wondering if a means exist to replace the conventional potentiometer/resistor for frequency setting, using a digital one or a PC-controlled.
What I need is a PC/micro controlled resistor, but the resolution should be high, so I can step throughout all frequencies.

An Idea I had, was to use an LDR and a LED pointed towards it. Depended on the voltage at the led, it's brightness should change, so that the LDR value is changing.

Any other ideas for a voltage controlled or digitally controlled resistor?

 

[kam1787]

i don't see how a discrete method would give you 'all' frequencies.

is there a reason why a pot/multi-turn pot/ and 'trimmer' pot are not desired?

 

[esp1]

hi,
I would agree with 'kam' how many discrete steps would meet your your requirement.?

Looking at the d/s I would consider the current control method, say using a P type JFET as a controllable current source.

IMO a 10 or 20 turn precision potentiometer would be the easiest solution.

E

 

[neazoi]

hi,
I would agree with 'kam' how many discrete steps would meet your your requirement.?

Looking at the d/s I would consider the current control method, say using a P type JFET as a controllable current source.

IMO a 10 or 20 turn precision potentiometer would be the easiest solution.

E

A multi-turn potentiometer, or better miniature trimmer would be ideal since it is of pure resistance. However the frequency setting reristor is critical and it has to be of good quality and as close to the chip as possible. To eliminate any mechanical effects and provide some means of electronic tuning, after your replies, I believe the greatest resolution resistor can be provided by an LDR and a light source pointed to it. The voltage of the light source can then be controlled by a multi turn pot, which then can be conveniently placed in a front panel?

 

[FvM]

neazoi already stated that he wants a digital controlled oscillator. The specification is vague, but that's not unusual at a "design idea" project state.

LTC1799 is approximately generating a current proportional frequency (respectively resitance inverse proportional). You could also used a VCO with linear voltage dependency, e.g. the CD4046 oscillator section, probably easier to control. In any case, the frequency resolution will be determined by the resolution of your control signal (voltage, current, whatsoever). For a wide frequency range, you may consider an exponential circuit to get a constant relative frequency step size. Quite clearly, generating a high resolution control signal is the major effort.

An elegant recent method to generate a high resolution variable frequency is a DDS chip. If you utililze the sine generator/low pass/comparator option, you can get up to high MHz low jitter signals with effectively infinite frqeuency resolution. Review respective datasheets and application notes from Analog Devices.

 

[neazoi]

An elegant recent method to generate a high resolution variable frequency is a DDS chip. If you utililze the sine generator/low pass/comparator option, you can get up to high MHz low jitter signals with effectively infinite frqeuency resolution. Review respective datasheets and application notes from Analog Devices.

It is only a matter of simplicity, else a DDS or si570 is ideal (well mostly, if we ignore the noise). DDS is difficult to program and si570 is even more difficult. Not to mention that all DDS chips are difficult to be soldered.
I find the solution of the 5-pin resistor set oscillator very elegant(no expensive and bulky tuning elements, no programming, easy soldering).
My primary goal is to use it with a 74F74 to take 4-phases with 90deg difference, divided down by 4 in frequency (which will hopefully improve frequency stability as well). This could be used in a wideband quadrature switching mixer then.

 

[FvM]

Quite obviously dividing a clock keeps the relative frequency stability. The relative jitter can be improved.

LTC1799 jitter specification is quite poor and may disqualify it for any communication application, because the absolute jitter is copied during down-mixing.