# Dimemr Duty Cycle Problem

Discussion in 'The Projects Forum' started by tlcquest, May 22, 2012.

1. ### tlcquest Thread Starter New Member

Jan 26, 2012
3
0
I am currently working on a Dimmer circuit that will go into a LED driver to run an LED but am facing duty cycle problems. Originally I was using a 50K analogue potentiometer and got a 1-99% duty cycle which I was happy but decided to try the new soft pot potentiometer (by spectrasymbol) which came only in 10K as i didn't want a long strip. The problem was that I could only get 67-99% duty cycle with the 10K which is not obvious to notice the dim effect. Could someone suggest a solution to my circuit diagram as I am totally lost. Is there a right calculation to my circuit to determine at least 33-99% duty cycle. Another problem I faced was having to come up with a sample and hold circuit to hold the resistance when i position the resistance on the soft pot as it floats to the maximum 10K when i lift my finger off. Currently am thinking of using a LF398 amplifier chip for sample and hold. Really would appreciate if someone can help. Thanks!!

• ###### circuit.JPG
File size:
21 KB
Views:
19
Last edited: May 22, 2012
2. ### #12 Expert

Nov 30, 2010
18,093
9,679
Changing the 1k resistor proportionately with the change in pot value would help.
Connecting thresh and cont together seems to be a mistake. I hope that's only a drawing error.
Sample and hold circuits sample and hold voltage, not resistance. You could design a circuit to represent a resistance but this all seems much easier to do by just using the correct potentiometer.

3. ### tlcquest Thread Starter New Member

Jan 26, 2012
3
0
Thanks for the reply, apologies it should be to ground for control voltage. I must have used an old draft file. Apparently on the spectral symbol website, they did state something like having to use a sample and hold circuit to hold the last value of the resistance (although i do know traditionally sample and hold tracks voltage). I am really not sure on this but on the youtube sites I did see videos of soft pots arduino music projects which hold the last position of the resistance. I have not done programmed using arduino so have decided to stick with electronic components.

In choosing R1, R2 and C .. I read that R1 and R2 should be in the range 1kΩ to 1MΩ after deciding on the Capacitance. I had to choose 0.1μF due to my specs sheet on the LED driver stating 200Hz max for linearity of duty cycle. R1 was recommended to be 1/10th of R2 (which i suppose to be my 10k potentiometer.) The formula is R2 = 0.7/ f x C1

Maybe someone can direct me to a better calculation if I am wrong. I can't change my potentiometer as this is the only value spectrasymbol has for a short strip. Please can someone direct me to a solution. I have been trialling and error but to no avail in getting 33-99% duty cycle. I have posted the correct image.

File size:
30.8 KB
Views:
12
4. ### #12 Expert

Nov 30, 2010
18,093
9,679
The formula is: time=1.1 RC
I believe the "softpot" is not going to zero ohms. If it did, the discharge path would only have the impedance of a forward biased diode. That would only be a problem if you are trying to use a very low voltage, like 2 volts, to power the timer chip. I am also baffled by a pot that disconnects when you let go of it.

It all comes down to using this softpot as the source of your problems.

5. ### tlcquest Thread Starter New Member

Jan 26, 2012
3
0
Thanks again. I was going to say that I got 43.4% - 99.7% duty cycle on the waveform. I did get some spikes but easily solved with smoothing capacitors. You are right the softpot is a big problem because the minute i release my finger, the waveform disconnects on my oscilloscope so it does float to infinity but unfortunately I will have to use it. I wonder if anyone on the forum has had experience on using softpots.