| Electronics > Beginners |
| Can I use a potentiometer on FB of switching DC-DC converters? |
| (1/2) > >> |
| T3STY:
Hi, This question is directly related to the LTC1771 IC that I'm planning to use, but I ask it more generally for DC-DC switching converters. Most of the switching converter ICs have a FB (feedback) pin that takes the middle point of a resistive voltage divider between output and GND, in order to adjust the output voltage. Now, all of the schematics I have found use fixed-value resistors (some of them are not even standard values), to get a fixed output voltage. But, can I place a potentiometer to manually adjust the output voltage instead? If so, can I (and is it safe) adjust the voltage while the IC is in use? My guess is that I could do this and it would be safe; after all, it would simulate an output voltage drop (or rise) which will force the PWM circuit to increase (or decrease) the duty cycle - looks nothing really fancy to me... But then, I don't understand why manufacturers don't ever advertise those ICs for "adjustable output" (that is user-adjustable output) instead of "programmable output" (that is, fixed output based on pre-calculated, fixed-value resistors). What am I missing here? I know that one question became many, but thanks in advance to anyone willing to answer me. |
| Doctorandus_P:
The answer to this is not so very simple. SMPS circuits tend to generate a lot of electrical noise, especially around the inductor and the traces between the regulator and the inductor. The Feedback pin reacts to small voltage differentials, and adding long wires to a potentiometer will act as an antenna and pick up noise from nearby circuitry. Another potential problem is that long wires add capacitance and with your potentiometer this can lead to extra phase shifts in the feedback signal, and this may lead to instability. Because of these potential problems it is often advised to keep the PCB traces of the feedback circuit short and put the resistors near the feedback pin of the regulator. But potentiometers are often used in SMPS circuits. Most of the shiny brick size SMPS circuits have a potentiometer with a limited range (For example an 24Vdc power supply can be adjusted between 22V and 28Vdc), and loads of the cheap chinese small SMPS boards have potentiometers. So it can be done, but there are some potential troubles. If you want to put your potentiometer in a front panel then keep the wires short and twist them. Twisting the wires will reduce the noise picked up by magnetic fields (Especially the stray magnetic field of the inductor in your SMPS circuit). |
| T3sl4co1l:
Yes, but: 1. Don't set the minimum ratio to zero. That would command infinite output voltage! Place resistor(s) in series with the ends of the pot, to limit how far the total range covers. 2. Mind the range. For example, a self-powered regulator (like the classic offline UC3842 circuit, and relatives) doesn't have much adjustable range between minimum and maximum operating voltage, so neither can the output (a 2:1 range is about the most that's feasible for these; so, for example, a nominal design of 12V, with an 8 to 16V range). Regulators and controllers that are supplied by the input, or a separate aux supply, should be adjustable down to zero, at least if compensation isn't a problem. 3. Place a resistor in parallel with the top half, i.e. from output to wiper. This protects the output in case the pot goes open circuit -- you've likely heard the effect of a scratchy pot before (loudly crackling audio when turning). What's happening is the wiper loses continuity (usually due to dirt or wear). This way, instead of VFB going open circuit, the output is commanded to minimum. 4. When applicable: mind that the Thevenin equivalent resistance of the pot wiper varies with position. If this ties to a feedback network, you may want to use a relatively small divider (like 1-10k range) and add a series resistor (10-100k?) from wiper to FB to set that resistance. (This doesn't apply to most newer adjustable regulators, which have internal compensation and only read the voltage on the FB pin, no current through it. Older types, and custom designs, may use an inverting error amp configuration, where the compensation network wraps around an op-amp, and so the FB divider's resistance matters.) (The LTC1771 is of the "just looking at FB pin voltage" type, no worries about Thevenin resistance.) Tim |
| T3STY:
Wow, thank you both for your awesome answers and advices! Really appreciated! I'd like to ask one more question involving the inductor used in these circuits. Most datasheets will have formulas and suggestions for calculating the inductor to use - and it is obviously going to be affected by the desired output voltage. Since now the voltage is variable, how do I handle this? Should I calculate and choose the inductor based on the largest output voltage needed? Or maybe average the value? My specific application requires about 1.5V adjustability (from 5V to 6.5V on the output) and the calculated inductor at lowest voltage was 2.9uH while the highest was 4.9uH. My instinct tells me that a larger inductor value will be suitable for lower voltages as well, so I'd choose the 4.9uH, but I may be wrong... |
| Doctorandus_P:
--- Quote from: T3STY on February 13, 2019, 02:22:39 pm --- My specific application requires about 1.5V adjustability (from 5V to 6.5V on the output) --- End quote --- With such a small output voltage range it would not have a noticable effect for your inductor. In the classic form of the buck (step down) converter the inductor does not have any influence on the DC output voltage. DC output voltage is a function of the input voltage and PWM duty cycle. The Inductor is not much more than a part of a filter to make DC out of the Pulsed output voltage. The self inductance of your inductor does have a big effect on the dynamic parts of the SMPS circuits, such as ripple voltage on the output and the capability to react to changing output currents. The value of your inductor is usually not very critical. It is not uncommon that if you put in an inductor of twice your calculated inductance that your circuit still works well. Even a factor of 10 often still works. Also: The inductance of an inductor is not a constant value. It changes with current and with applied frequency and temperature. Power inductors ofthen have a tolerance of 20% or more and inductance can easily change byt 10% within their normal operating range (no saturation). Measuring power inductors with an 0.02% LCR meter is a very silly excersise. |
| Navigation |
| Message Index |
| Next page |