Capacitor voltage down regulator

Thread Starter

dcbingaman

Joined Jun 30, 2021
498
Why do we not normally see this type of circuit in the real world? Granted the regulation is not the best but it could be good enough for a 'pre-regulator' prior to an LDO maybe. It uses less than 100mW and delivers 1.4W to the load down regulating the 12V to around 6V? Has anyone seen a circuit like this before?
1630880080653.png
 

Attachments

Thread Starter

dcbingaman

Joined Jun 30, 2021
498
Note: 2N3906 PNP transistor need to use alternate part in final design as 200mA max current is being slightly exceeded.
 

Papabravo

Joined Feb 24, 2006
17,292
I think that is quite a few parts for a marginal payoff. A transistor, a resistor, and a Zener diode will do the same job. Ultimately it comes down to cost.
 

Thread Starter

dcbingaman

Joined Jun 30, 2021
498
I think that is quite a few parts for a marginal payoff. A transistor, a resistor, and a Zener diode will do the same job. Ultimately it comes down to cost.
Yes but that would be a linear regulator with 12V in and 6V out it would have a maximum of 50% efficiency. This circuit has an efficiency of greater than 90%. Because the transistors are operating as switches and not in linear mode, power is conserved.
 

Thread Starter

dcbingaman

Joined Jun 30, 2021
498
Q2 is running hot. I don't see that it is 90% efficient.
Actually Q2 is either on or off and dropping no voltage. Did you run the sim? It is being switched on and off, not operating in linear mode.
Here is a plot of the voltage on Q1's collector. It is being turned on and off into saturation and cutoff and so is Q2:
1630886467760.png
 
Last edited:

ronsimpson

Joined Oct 7, 2019
1,657
When I saved the file there was an error. So there are probably problems. I normally make switchers well above 100khz and in this case I would shoot for near 1mhz. The 47mH is because of the slow oscillator you made. I would use 47uH.
Actually Q2 is either on or off and dropping no voltage.
Not what I see. Q2 sits on +12 and +6V and when on it has about 6V across it. So current x 6V = hot.
What you showed I see on Q1.
 

Attachments

Papabravo

Joined Feb 24, 2006
17,292
Once you spring for the inductor, you might as well go for the integrated buck converter. Anything else is for entertainment, amusement, or edification. It is tough to beat 1st order cycle-by-cycle current mode control with slope compensation.
 

Thread Starter

dcbingaman

Joined Jun 30, 2021
498
When I saved the file there was an error. So there are probably problems. I normally make switchers well above 100khz and in this case I would shoot for near 1mhz. The 47mH is because of the slow oscillator you made. I would use 47uH.

Not what I see. Q2 sits on +12 and +6V and when on it has about 6V across it. So current x 6V = hot.
What you showed I see on Q1.
After looking at it again, you are correct. Q2 is going to run hot. Not sure how I missed that. I stand corrected. Nice catch. My goal was to create a capacitor only switch mode converter. Not sure that is possible. Thanks for explaining the need for 47mH and how to overcome that by running at higher frequencies. It appears you have a lot of experience in switch mode converter design. Thanks for the updated circuit.
 
Last edited:

Thread Starter

dcbingaman

Joined Jun 30, 2021
498
Why not just buy a small switching regulator from Amazon or Ebay?
My goal was to build a capacitor only switch mode converter that ran at high efficiency. ronsimpson corrected me in that Q2 does end up running hot and defeats the purpose. I am not sure it is possible to make a switch mode capacitor only converter.
 

BobTPH

Joined Jun 5, 2013
4,048
Without an inductor, the switch that is charging the capacitor will have have a voltage drop of Vin - Vout, making it a linear regulator. The inductor also has this drop, but it due to it’s inductive reactance rather than resistance and therefore dissipates no power.

Take the capacitor out, and it becomes PWM, which is highly efficient, but does not regulate the voltage. If fact you can think of PWM as a poor regulator with very high ripple.

Bob
 

Papabravo

Joined Feb 24, 2006
17,292
My goal was to build a capacitor only switch mode converter that ran at high efficiency. ronsimpson corrected me in that Q2 does end up running hot and defeats the purpose. I am not sure it is possible to make a switch mode capacitor only converter.
I am slightly curious about why you thought this might be possible. Was it based on anything concrete, or was it just a kind of gedankenexperiment.
 
Last edited:

Thread Starter

dcbingaman

Joined Jun 30, 2021
498
I am slightly curious about why you thought this might be possible. Was it based on anything concrete, or was it just a kind of a gedankenexperiment.
Nothing concrete. Just thought if you could get all the power going through energy storage devices, could it be done with capacitors alone? I have yet to figure out if it is possible. I am thinking it is not.
 
Top