Maker Pro
Maker Pro

Switching Regulator IC with Voltage Controlled Output

Eugbug

Nov 20, 2012
7
Joined
Nov 20, 2012
Messages
7
Does anyone know off hand of a DC to DC converter IC whose output voltage can be varied via a control voltage input?
Any of the devices I looked at rely on varying the value of a resistor to change the voltage.
The application uses a microcontroller to control the output voltage and the other alternative would be to use a JFET or MOSFET in resistor mode or a digital potentiometer to change the resistance.
 

CocaCola

Apr 7, 2012
3,635
Joined
Apr 7, 2012
Messages
3,635
The most cost effective straight forward is to simply swap the analog pot on the existing DC to DC converter (dirt cheap from China) with a digital pot that is controlled by the micro using a serial connection...
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
What output voltage range do you need? If you don't need a very wide range, you can use this trick.

attachment.php


This diagram is part of a battery charger I designed. The LM2676S (U10) is used as a buck regulator with its output voltage controlled by the MCU via Q11 and Q10. This circuit has an output voltage range of about 9~19V DC. In this application, the output voltage was monitored by the MCU, to compensate for semiconductor forward voltage variations due to temperature. This (or some other) method of "closing the loop" is needed if you want accurate output voltage (very important for a battery charger).

The output voltage is controlled by a 0V/3.3V PWM signal from the MCU, which comes in on the wire just below the text "OUTPUT VOLTAGE ADJUSTMENT" and is smoothed by R15 and C15 for the CCV (charger control voltage) node. You could also use a DAC output if one is available on your MCU, connected straight to CCV.

The regulator's output voltage is on the COV node (charger output voltage). R12 provides the main voltage feedback to the CFN node (charger feedback node), and five diode junctions in series provide a (fairly) constant voltage drop of about 3.1V from CFN down to the regulator's feedback input, which the regulator maintains at 1.21V.

If R14 wasn't present, the regulator would operate with a fixed output voltage, determined by the ratio between R12 and R13 with the 3.1V diode drop effectively in series with R12.

Any other current drawn from CFN will cause the CFN voltage, and thus the regulator feedback voltage, to drop, and the regulator will increase its output to compensate. R14 provides the path for this current, which is sinked by Q10 and Q11 operating as a Darlington. The current drawn through R14 depends on the voltage difference between CFN (which has a constant voltage) and Q10's emitter. Therefore, pulling Q10's emitter towards ground sinks more current from CFN and causes the regulator to increase the output voltage, to counteract the current being drawn through R14, and keep the CFN voltage constant. (The regulator remains in regulation at all times.)

A control voltage of 0~3.3V at CCV produces a Q10 emitter voltage of around 1.2~4.5V, so actually when CCV is more than about 3.1V, current is actually sourced, not sinked, through R14. So the CCV voltage controls the output voltage. At CCV=3.3V the output voltage COV is at minimum; as CCV is reduced towards 0V the output voltage increases.

If anything's unclear feel free to ask.

Edit: Make sure that the buck regulator is disabled on startup, and won't be enableld until the PWM is active and the voltage on CCV has been given time to settle. If the regulator starts up immediately, it will start at maximum output voltage.
 

Attachments

  • 7003.gif
    7003.gif
    8 KB · Views: 774
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Kris, are you missing a diagram?
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Kris, are you missing a diagram?
I had attached it, and inlined it, and it was showing up for me. But there was no thumbnail at the bottom of the post. So I re-uploaded it. I hope it's visible now.
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
If you have a microcontroller in the loop already, why use a buck regulator chip? The micro could perform it's function with a MOSFET switch added.

Bob
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Bob, that wouldn't be feasible for several reasons. The micro is busy doing serial comms and other things using several interrupt sources and DMA, and the charger output voltage needs to be fairly stable with some ripple on the input, which would require a fairly fast response. The National Semiconductor "Simple Switcher" regulators are pretty cheap. But the killer is that the regulator runs at around 260 kHz (to keep the inductor size small) and the MCU runs at 4 MHz so using PWM would mean a PWM resolution of about 3 bits which would be far too coarse!
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Okay, I get it. I have done a couple of buck regulators with uControllers, but I did need to use a much bigger inductor that I would have with the dedicated chip running at that kind of frequency, and they were not doing much else.

Bob
 

Eugbug

Nov 20, 2012
7
Joined
Nov 20, 2012
Messages
7
Thanks everyone for the replies! (which I have only just discovered) For some reason I don't seem to be getting notification of replies or they are ending up in my junk mail folder.
 
Top