I'm looking for something other than a potentiometer that will let me adjust the ammount of voltage in hundreds of stages say 1-300.

.1V .2V .3V etc etc etc.

Thanks

edit: End result im basiclly looking to hit one of two buttons to control an object, hit up, increase V by X# and stay in that state until its hit again or hit down , decreases V by X# and stay in that state until its hit again. If that doesnt make sense I can try to explain it better.

Current would be about 20mA

 

Hello Acrux,
Are you looking for fixed voltages per stage? Or would you settle for a more linear adjustment?

What kind of current do you require?

Is this a homework assignment?

 

Well I recall in my Air Force days in the 60s, there were Fluke differintial voltmeters that used 10 position rotary switches to set a voltage value for each decade and then a unknown voltage could be applied and you turned switches until a balance meter showed a matched condition. I think it went down to microvolts decade.

It was big and ugly but very accurate and stable. Anyway I assume what you want could be constructed with BCD switches that had weighted resistors to divide down a reference voltage, etc.

Would probably be a little pricey in component cost and size requirements. I wonder if in this day and age a simple AVR or PIC development board couldn't do this in a more cost effective manner. Keyboard entry of desired voltage and PWM followed by low pass filtering to generate the desired voltage?

Lefty

 

Hello Acrux,
Are you looking for fixed voltages per stage? Or would you settle for a more linear adjustment?

What kind of current do you require?

Is this a homework assignment?

No homework, just getting bored with things and seeing what other things i can dabble in.

Can you explain the difference between the fixed vs linear.

End result im basiclly looking to hit one of two buttons to control an object, hit up, increase V by X# and stay in that state until its hit again or hit down , decreases V by X# and stay in that state until its hit again. If that doesnt make sense I can try to explain it better.

Current would be about 20mA

 

Perhaps something like a thumbwheel switch (several sections) and a D to A converter. You would need a device on the output to buffer the current requirement. Also some bucks - TWS's aren't cheap any more.

 

Perhaps something like a thumbwheel switch (several sections) and a D to A converter. You would need a device on the output to buffer the current requirement. Also some bucks - TWS's aren't cheap any more.

Or an up-down counter and a D/A converter. Or a PIC and a D/A converter. For the D/A, he could also use a digital potentiometer.

 

I agree with the previous posts that either D/A or digital potentiometer is the way to go. D/As are more common and digital pots are harder to find and are expensive.

Just to throw out another idea. You could make your own crude D/A converter. Since you only need 300 steps, you could build an up/down counter that counts up to 300. The outputs of the counter can be fed into an opamp based summing amplifier with the gains configured in powers of 2. You would need a 9 bit counter that limits to 300 and stops going up. Use CMOS logic with precision controlled DC power supply so that the logic output is well defined and goes right to the rails. Also, use precision resistors in the summing amplifier. There are other considerations, but this could be a fun experiment to do.

 

I agree with the previous posts that either D/A or digital potentiometer is the way to go. D/As are more common and digital pots are harder to find and are expensive.

Just to through out another idea. You could make your own crude D/A converter. Since you only need 300 steps, you could build an up/down counter that counts up to 300. The outputs of the counter can be fed into an opamp based summing amplifier with the gains configured in powers of 2. You would need a 9 bit counter that limits to 300 and stops going up. Use CMOS logic with precision controlled DC power supply so that the logic output is well defined and goes right to the rails. Also, use precision resistors in the summing amplifier. There are other considerations, but this could be a fun experiment to do.

An R-2R ladder is a lot easier to find parts for than a binary-weighted summing network, especially with 9 bits.

 

Thanks to all so far, I will look into all of this.. anymore ideas keep throwing them out there!

Thanks again!

 

An R-2R ladder is a lot easier to find parts for than a binary-weighted summing network, especially with 9 bits.

Which parts would be hard to find? Just need opamps, resistors, voltage reference and CMOS logic counter. I'm not suggesting this as a practical solution, but just as an educational experiment. Your suggestion would also provide a good learning experience.

 

Which parts would be hard to find? Just need opamps, resistors, voltage reference and CMOS logic counter. I'm not suggesting this as a practical solution, but just as an educational experiment. Your suggestion would also provide a good learning experience.

To make the DAC, I think two values of resistors (R-2R) are easier to obtain than 9 octaves of resistor values. All the other stuff is the same.

 

To make the DAC, I think two values of resistors (R-2R) are easier to obtain than 9 octaves of resistor values. All the other stuff is the same.

Yes, I agree with that. Trying to achieve 300 steps is on the border of what is practical. He would have three choices I can see.

1. Buy 0.1% resistors which is a little expensive
2. Hand select 1% resistors along with some parallel and series combos.
3. Use 1% resistors with fine tuning variable resistors in series

Clearly, not practical for real products that are to be manufactured, but fine for dabbling as the OP was thinking. Definitely he should look at R-2R approaches too. He would learn a lot with this comparison.

 

Yes, I agree with that. Trying to achieve 300 steps is on the border of what is practical. He would have three choices I can see.

1. Buy 0.1% resistors which is a little expensive
2. Hand select 1% resistors along with some parallel and series combos.
3. Use 1% resistors with fine tuning variable resistors in series

Clearly, not practical for real products that are to be manufactured, but fine for dabbling as the OP was thinking. Definitely he should look at R-2R approaches too. He would learn a lot with this comparison.

With 9 bits (and even fewer), the designer has to be aware of counter output resistance, as it needs to be either insignificant (good) or included in the resistor calculations (not so good).

 

With 9 bits (and even fewer), the designer has to be aware of counter output resistance, as it needs to be either insignificant (good) or included in the resistor calculations (not so good).

Definitely there will be some tricky issues such as this. My recommendation to the OP (assuming he actually decides to go forward with any of the recommended approaches) is to think through a design and post schematics (with specified part numbers) for feedback. There is a fine line here. If we point out every issue, then we rob him of the learning experience. However, without enough guidance it could be frustrating. It's no fun to build something and then find a problem that requires a complete rebuild because of poor planning. I did that several times when I was learning, and I would have loved to have a forum like this available back then. Well, since I'm still learning, I'm glad to have it now.

 

Definitely there will be some tricky issues such as this. My recommendation to the OP (assuming he actually decides to go forward with any of the recommended approaches) is to think through a design and post schematics (with specified part numbers) for feedback. There is a fine line here. If we point out every issue, then we rob him of the learning experience. However, without enough guidance it could be frustrating. It's no fun to build something and then find a problem that requires a complete rebuild because of poor planning. I did that several times when I was learning, and I would have loved to have a forum like this available back then. Well, since I'm still learning, I'm glad to have it now.

Agreed. I'll refrain from posting a schematic until one is requested - and then I might still refrain.
As a postscript, a few weeks ago I coincidentally built a freerunning 8 bit counter (a couple of 74HC161's) and an 8 bit R-2R ladder made from matched resistors hand-selected from 5% stock. That was a somewhat of a pain, but the output was monotonic - to my mild surprise.