SMPS voltage range problem

Status
Not open for further replies.
X

xyz9915

Member level 4
Joined
May 14, 2007
Messages
74
Helped
2
Reputation
4
Reaction score
2
Trophy points
1,288
Location
Pakistan
Visit site
Activity points
1,817
Hello Everyone,
The attached circuit I find in a forum and it is working satisfactorily with Mains AC 100 to 150 Volt & 175 to 230Volts. But I noticed a problem that from the voltage range 151V to 175VAC this circuit stops working. Why? I do not know... Can someone explain me what happens among this voltage range


I am using variable transformer at mains AC input to observe its working voltage range
 

Attachments

  • 50KHz-900W.pdf
    25.6 KB · Views: 132
Last edited:

If the transformer changes the function same as DC? just guess so.
 

The transformer at mains AC input only changes the voltage. But the smps should work like many smps work from universal voltage range i.e. 110VAC-230VAC
 
Since it's working below and above that particular voltage range, I am left wondering if it has to do with some odd harmonic or resonance encountered only at that voltage range. If that is the case, there must be some sort of noise that is interfering with the PWM or Feedback loop?
 

I think this may be true that there might be some harmonics interfering feedback loop.
 

The input rectifier should produce roughly a DC voltage, and the harmonic content of its ripple shouldn't change too much across that range of Vin...

Can you elaborate more on what you mean by "not working" in that voltage range? Does the SG3525 stop operating completely, or does it the output power capacity just drop?
 

I mean SG3525 stop operating in the mentioned voltage range resulting no voltage at output. If I increase voltage above 175 VAC is again starts and delivers output
 

Okay, so if the controller stops operating, then we should be able to discover why by looking at the status of its IO.
When the SG3525 stops operating, is the 12V auxiliary supply still good? What about its reference voltage? What are the status of the Css and CT pins? And the error amplifier pins?

Also your schematic shows a capacitor from Comp to ground, which I don't think is appropriate... though I don't think that alone explains the problem you're having.
 

Dear mtwieg,
1) 12V auxiliary supply is taken separately from 230VAC and not from variable transformer which i use to test smps working at 100v to 240vac, so this 12V auxiliary supply still provides same voltage

2) Voltage at reference pin remains 5v and 2.5v at pin2

3) I think Css only performs a soft start function which protect from in-rush current.
 

Okay sounds fine so far. What about pins 1 and 9 (inverting input and compensation output)?
3) I think Css only performs a soft start function which protect from in-rush current.
Click to expand...
Yes, but its voltage may tell you whether the chip is being inhibited somehow. If it's held low, then its inhibited, if it's near Vref then the error amp should be enabled.
 

Okay sounds fine so far. What about pins 1 and 9 (inverting input and compensation output)?
Click to expand...
I changed capacitor 0.1uF to 1uF but still same problem. Voltage at error amp pin1 changes quickly so it can not be measured without the help of an oscilloscope, so what should i do?

Yes, but its voltage may tell you whether the chip is being inhibited somehow. If it's held low, then its inhibited, if it's near Vref then the error amp should be enabled.
Click to expand...

I did not check status at Pin8 (Css) will try after weekend on Monday
 

xyz9915 said:
I changed capacitor 0.1uF to 1uF but still same problem.
Click to expand...
You mean from pin 9 to ground? See, there generally shouldn't be any capacitance there in a normal design. The error amplifier is a voltage amp, so it should have an impedance network connected between Comp and Inverting input, not between comp and ground. But I actually see in some SG3525 datasheets that its gain is measured with a RC load, which implies such a loading network may work as a compensation scheme... I don't have any direct experience using that IC, maybe someone else can chime in.

Voltage at error amp pin1 changes quickly so it can not be measured without the help of an oscilloscope, so what should i do?
Click to expand...
If it's changing quickly while the supply appears to be off, then this is definitely related to the problem. Do you not have access to an oscilloscope at all? Debugging the issue will be difficult without it...
 

If it's changing quickly while the supply appears to be off, then this is definitely related to the problem
Click to expand...

Voltage changes quickly at pin1 during working of converter, but remain low when it stops working
 

Okay, if pin 1 is low and pin 2 is at 2.5V then the Comp pin should be high, assuming that it's not inhibited by the soft start or shutdown functions. Can you confirm that?
 

The Problem has been solved by replacing 0.1uF capacitor with 10uF capacitor at Pin9 (Compensation) of SG3525. Now the smps works from 100vac to 240vac.

Thanks to Everyone
 

Thanks Doom2pro but i am surprised that usually a large value capacitor (e.g. 10uF) is not used for compensation. Can you help in this respect?
 

I guess the Datasheet for the Gate driver was underestimating the need for a higher compensation capacitor, or the gate driver was expecting it's own bypass and compensation capacitor instead of sharing one with the PWM?
 

xyz9915 said:
Thanks Doom2pro but i am surprised that usually a large value capacitor (e.g. 10uF) is not used for compensation. Can you help in this respect?
Click to expand...

Doom2pro said:
I guess the Datasheet for the Gate driver was underestimating the need for a higher compensation capacitor, or the gate driver was expecting it's own bypass and compensation capacitor instead of sharing one with the PWM?
Click to expand...
I think you are both misunderstanding the function of the Comp pin. It has nothing to do with gate drive or bypassing. The compensation pin is used, in combination with the error amplifier inverting input, to alter the transfer function of the feedback loop. As I stated before, this is normally done by configuring the error amplifier as an inverting amp, with some RC network connected between pins 9 and 1. Putting a capacitor from Comp to ground, is not a typical configuration for the chip, as far as I know. I would strongly recommend using a proper compensation scheme, and actually analyzing the loop response, rather than just trying random component values in odd places.
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…