Maker Pro
Maker Pro

Current Surge

Lord_grezington

May 3, 2013
124
Joined
May 3, 2013
Messages
124
Hello All

On one of my previous threads (https://www.electronicspoint.com/battery-charging-problems-t260360.html) I used a Buck converter to convert a PWM into DC.

However, if you look at the attached PDF you can see from the voltage on the current sensing resistor that I am now getting a large current surge (caused by the current going through the Cap) for around a ms during start up.

Is there any way I can get rid of the current surge? - I spent ages finding the correct cap/inductor values to cover an acceptable range.

Thanks
Graham
 

Attachments

  • Current Sense Resistor Voltage Spike.pdf
    61.4 KB · Views: 125

Lord_grezington

May 3, 2013
124
Joined
May 3, 2013
Messages
124
Sorry, now attached

Graham
 

Attachments

  • Current Sense Resistor Voltage Spike circuit.pdf
    67.4 KB · Views: 89

Lord_grezington

May 3, 2013
124
Joined
May 3, 2013
Messages
124
Note - the simulation taken was at full 100% duty cycle in attempt to get the surge as large as possible.
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
1uH is way too low unless the PWM freq is int the 10s of MegaHertz. And the 2200uF cap is way too high, try about 10uF.

What is the PWM freq? Try something like 100uH to 1mH.

Since the duty cycle is 100%, you are simply turning the MOSFET on. The surge is the battery trying to charge the 2200uF cap.
 

Lord_grezington

May 3, 2013
124
Joined
May 3, 2013
Messages
124
If I use a larger inductor I tend to get much higher resonance. With the Low value inductor it takes a while to get up to the required voltage, but no resonance.

I have chosen large Cap because I have found that with the larger cap I can achieve much lower voltage fluctuation on the output. I dont get a linear relationship between the duty cycle and the output, but it looks clean.

I am going @ 100Khz

I take what your saying about the cap charging up - I will need to do some thinking (and trialing) with this.

I have attached the 10% and 50% duty cycles (so you can see the result without just switching the MOSFET on.

Graham
 

Attachments

  • 10% & 50% duty.pdf
    66.5 KB · Views: 93

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
The capacitor is way too high for 100KHz. I assume the plot is the voltage acorss R1. If that is correct we are talking about 2A. A 2200uF cap would only lose 0.009 V in the PWM period. There is no reason to control the ripple to that level. I stick to my reccomenation of 1 to 10uF.

And what are the cap and inductor across R1 supposed to do?

Bob
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
As I said before, the inductor is way to small as well. At 48V stepping down to 12V, there will be 36V across the inductor. The formula for current in an inductor is:

dI/dt = V / L

For a 1 uH inductor with 36V across it

dI/dt = 36,000,000 A / sec

Or 360A / PWM period. You want to limit this to more like 1A per PWM period, so that would suggest an indcutor of about 360uH.

Bob
 

Lord_grezington

May 3, 2013
124
Joined
May 3, 2013
Messages
124
I have changed the Cap tp 10uF and the Inductor to 100uH. The surge has indeed gone,

I have attached the ripple, It has now gone up to around 140mV. (I was at around 15mV using the 2200uF & 1uH).

The voltage required is across R3, The previous plots were across R1.

Voltage across R3 is required Voltage, Votlage across R1 is for current sense. As R1 is bottom side I still need to clean the PWM up so I can put it through a OpAmp and then into a pic.
 

Attachments

  • 10uF & 100uH.pdf
    68.7 KB · Views: 108

Lord_grezington

May 3, 2013
124
Joined
May 3, 2013
Messages
124
Hi Bob

I bow down to your experience... Thanks for your help

Schematic - I put the N Channel Mosfet on the high side, I will be driving it via a Mosfet driver with a bootstrap (I cant get this on the simulation, on simulation I just Idealized the driving of the FET). I wanted to keep the N-Channel to try and keep efficiency. This way I got rid of the Extra Cap and Inductor and also kept a common ground (Useful).

50% Duty Cycle - I ended up going for the 50uF Cap and 200uH Inductor. This gives me a great responsive output and more importantly no current surge.

60% duty Cycle Fluctuation - this shows that the fluctuation on the settled voltage of only 13.8mV, this is better than the large Cap.

Graham
 

Attachments

  • Schematic.jpg
    Schematic.jpg
    9.2 KB · Views: 134
  • 50% Duty Cycle.jpg
    50% Duty Cycle.jpg
    26.8 KB · Views: 125
  • 50% Duty Cycle Fluctuation.jpg
    50% Duty Cycle Fluctuation.jpg
    59.6 KB · Views: 119

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Looking good, I am glad I could help. My very first project after years neglect to my electronics hobby, was a micro controller buck converter NiCad charger.

Bob
 
Top