UCC38C44: UCC38C44

Hi BHARATH PRABHU,
Using peak current mode control with a wide input voltage/load range can have issues.
If the ton time gets very small, the turn-on current spike can mean at around the peak current you want to turn-off the current sense signal is still falling rather than rising.
It is possible by decreasing the operating frequency you are pushing up the peak current at which to turn-off and moving away from the turn-on current spike.
It would help me greatly to diagnose the problem if you could send scope traces of the COMP signal and the current sense signal when it is working fine (at low input voltage) and when it is not working fine (high input voltage).
It would also help if you could send me a schematic with values. If you do not want to post this on a public forum please send it to my Email address : j-leisten@ti.com.
Thanks
Joe Leisten

Hi BHARATH PRABHU,
From your CS pin waveforms it looks like my first guess was correct.
At high input voltage the TON period is so short that the turn-on current spike has not finished before the required peak current has been exceeded. This is not a stability problem in the classical sense and will not be solved by adjusting your compensation values.
If you are set on peak current mode control the two choices for solving this problem are as follows:
1) Increase the magnetizing inductance of the transformer. This will increase the TON duration.
2) Decrease the switching frequency. This also has the effect of increasing the TON duration.
I hope this resolves your issue.
Thanks
Joe Leisten

I calculate L as L= Vinmin*Dutymax(0.45 From data sheet)^2/(2*Fswitch*Pin*Krf)

Is it proper? What Krf is good for my wide voltage range operation

100-400VDC and

300-900VDC

2) Decrease the frequency :Should i design the SMPS for 30Khz and select RTCT for 20Khz?

3) Can u share simple methods to select compensation components for stability 

As u can share some example conditions of instability and share the solution for the same by increaing or decreasing the compensation values

L= (Vinmin*Dutymax(0.45 From data sheet))^2/(2*Fswitch*Pin*Krf)
Correction

Hi BHARATH PRABHU,
Could you please send your schematic to my Email address.
In your Email you mentioned that you have sent a schematic, but I only found scope pictures.
Also could you please advise which document you have taken this Equation from.
Thanks
Joe Leisten

Hi BHARATH PRABHU,
Sorry for the delay. Could you please supply the transformer turns ratios.
The reflected output voltage is an important parameter in looking at the required magnetizing inductance value.
Thanks
Joe Leisten

Pri turns =90T

AUX turns = 19T

NP/NAUX = 4.7

27V o/p turns = 29T

10V turns = 12T

10VE turns = 12 T           

iS MY SCHEMATICS PROPER PLEASE CONFIRM AT THE EARLIEST                   

CAN U SHARE ME THE EXACT FORMULA FOR INDUCTANCE CALCULATION

Hi BHARATH PRABHU,
I have responded by Email with the inductance calculation. It is much easier to send attachments by Email than on this forum.
Thanks
Joe Leisten

Hi BHARATH PRABHU,
I studied your schematic once more. It may be that you have a different control problem.
Please could you advise how much power is being drawn from each of your outputs.
In particular I need to know how much power is consumed by the 17VSMPS rail vs the 27VE and 8Vx rails.
There can be a problem associated with controlling a lightly loaded rail with the intent that this will control the output on other more heavily loaded rails.
Please note this is not the same as a Primary Side Regulated controller which senses the actual output voltage via an Aux winding as opposed to controlling the voltage on the Aux rail. These are not the same thing at all.
Thanks
Joe Leisten