Startup error (Continue with Dear PI_Crusher)

Hi,
I am testing primary zener feedback. Now there is no opto. Unfortunatelly it is not working. (The equivalent clamp resistor is still 5,5k 8w)
I am attaching the control pin signal photo. There are some noises. The pcb design is not very good. I know it but, i am expecting that, it RUNS with bad results like regulation problem or something like that.
IS THE PROBLEM CAUSED BY THIS CONTROL SIGNAL NOISE? If yes, the real error is pcb design...

What are you thinking?
Regards
Zafer

Hello Zafer,

The clamp circuit is important to limit the voltage over-shoot during reset time (right after ON time, at the beginning of OFF time). For now you can ignore it, after you’re happy with the overall performance you can look into this part. Lower value for this clamp resistor is required to decrease the maximum amplitude spike during reset time. You need to allow the reset voltage to go close to maximum allowed device voltage (equivalent with higher resistor value) but at the same time to limit the spike under a reasonably safe limit (lower resistor value). Also, fine adjustments of this resistor can significantly increase/reduce efficiency and AMI noise.
I do not understand what you mean by “there is no opto”.
I assume the big spikes you see on control signal are caused by noise. Unfortunately noise spikes into control pin can generate all kind of problems. Look for ideas how to arrange the PCB traces in a similar example from here:
http://www.powerint.com/en/design-support/reference-designs/design-examples
Cheers,
PI_Crusher

Hi,
Ok, I understand your words about the clamp circuit.
First of all, I want to use secondary TL431 feedback circuit. Then i was change the feedback circuit to secondary zener type. Because it is not more complex and no more problem.
This morning, i was suspecting from PC817C. Then i was try the primary zener feedback circuit. Because of this, i told you, "there is no opto"....

Now, i am working on PCB design again according the your recommandation. Do you think, the reason of my problem is caused completely from that noise?

For example, This problem is caused completely from noise. Why am i see "D7_Cathode_RTN_Vout_31VDC.jpg". I don't understand this ripple. Can you set a scenario to explain this ripple which is caused by noise.

If i fix it, I will send to you Turkish traditional sweet. :)

Regards,
Zafer

Because the optocoupler is connected to the output trough a 68V zenner diode, when the output voltage is lower than this value the optocoupler is totally disconnected, with no current flowing trough the LED. That means the control loop is opened, there is no output to input feedback. In this case the normal voltage ripple you can see on power rails is not any more rejected by the control loop, and that ripple overlap the duty cycle control voltage (internal to the chip).
Cheers,
PI_Crusher

Hi,
There is a good news. :)
SMPS is running now with secondary zener feedback. There is only one guility, 100nF capacitor, which is putted from 1 cm far away from the control pin. Unbelivable!!!

I draw 1,7A from the SMPS. I have no enough heatsink. Because of this, i didn't load it more. But i think, it will run.
It can't startup with that load at 135V and lower voltages, but when i will add the soft finishing circuit, i think it is ok. I have some questions.

1.I'm drawing a new layout with carefully. Would you please chack my design and tell me the poor points. I'm attaching schematic and layout.

2.At an14.pdf file, i am reading one thing.
Symptom: Power supply does not start.
Cause: Switching currents trigger shutdown latch
What to do: Electrically isolate heatsink from circuit
I don't understant this suggestion. Would you please explain it

3.My output voltage is 75VDC, and TL431 is working 36VDC maksimum. According to your referance design (der97.pdf, Output voltage is 48VDC), i am adding V3 zener diode to new layout. But PIExpert don't tell anything about this zener. Am i put this zener to the circuit. If i am put this zener diode, am i change the feedback resistor, capacitor values? If yes, how can i calcultate the new values? Is 56V Zener diode suitable for this function?

Thank you for your help.
Best regards,
Zafer Dedeoglu

Hi,
I make the new PCB. It is running secondary TL431 feedback circuit with 56V Zener instead of V3 zener diode.

I want to add a few more questions to above questions. I'm glad, if you are answer them.

1. Now, new design is not startup with 45R load (1,7A) at lower supply voltage than 170VAC. New design has got soft finishing circuit, but it can't start up. What am i do?

2. Now, i am using EE42/21/15 horizantal core. Am i use vertical EE42/21/15 core? I think there is no transformer construction difference (I have no space). Is it true? And can i change the pin number of windings.

3. I want to add a outer copper foil layer to the transformer. I think, it helps to reduce radiated emission. Am i connect that copper foil to the bulk capacitor (-). (so the source of TOP261)

4. Would you please explain the task of Y class capasitor which is coded to C11 on my design. It is connected to (+) rail to RTN. So, isn't it break the isolation?? And, i saw some schematics on the "tophx_family_datasheet.pdf" file page page 21, figure 41, RTN is connected to earth. Why?
And also page 24, figure 44 there is extra C3 and L2. What is te function is that components. Aren't they drive the RCCB's (Residual Current Circuit Breaker) to cut off.

5.When try'in a new design on the PIExpert, There is a column which name is "Fit estimate" what is the meaning of this column.

Thank you very much and i:'m sorry my long questions.

Best regards,
Zafer.

Hi
I am attaching some photos.
As you can see, i draw current 1,62A at 75,7V and 2,73A at 75,8V.

I have got still some problems.
1.Clamp circuit (Resistor and TVS) temperature are very high (100C)
2.It can't startup with half load lower than 160VDC. (I have got soft finishing circuit but it can't start.)
3.When smps is running, i am loading it %100. Then approximatelly after 10 sec, it is stop and it is try to startup. I'm not sure but, may be, it is stopped according to temperature, (may be the heatsink is still small)

Have you gat any offer?
Regards.

Zafer

Hello Zafer,
Look throughout our application notes and tutorials.
The clamp circuit dissipates significant power, it is normal to be hot. Read our comments and tutorials on clamp circuits, how to optimize it. There are many more design problems you have to deal with.
Cheers,
PI_Crusher