Startup error
Hi,
I want to design 75V 2.9A output SMPS. I'm using TOP261EN and EE42/21/15 ferit core. This is my first experience for designing SMPS with PI products and i have got startup and overshoot problem. At first, i was build feedback circuit with TL431and it is the same problem. But now, i changed it to secondary zener feedback system for reducing compexity. I am attaching my schematic. But there is still same problems. I looked a lot of datasheets and application notes which are produced by PI and also i looked to the forums. I try some suggestions but still i don't find the solution.
1.Sometimes TOP261EN try to startup but it can't start. I am hearing a little noise every 2 sec. (I am encountered this problem with no load and a light load (approx. 900R) condition.
2. Sometimes TOP261EN is try to startup and it starts a little bit. (Little bit means the ouput voltage is 31VDC) There is a square wave at the Control pin. I don't understand why am i see a square wave at that pin? I think, When the control pin voltage is high, top is reduce the duty of drain pwm signal. So top i stop. (this condition photos are marked as "ERROR"
3. Please look at the "Way to reach stable 75V.jpg" file. I connect 900R load and power up the smps. First the output is reaches 31V. After then i disconnect the 900R load. At first, the output is increasing a little bit, After then the output is increasing extremelly instantly. At that time the output is reaches to 130VDC, and i reconnect the load. The output voltage is decrease and stable 75V. But if i do it with a bigger load, the system is not closed loop and output is died. You can see the scope photos when the top is running properly with this metod which are marked as "RUN". (I am measuring a clear 17V at the bias winding output (at the C4 capacitor when top is running)
Please help me!
Regards,
Zafer
Comments
Hi again,
Thank you for your reply. I am also design this circuit with the help of PIExpert.
Now, i am attaching that original file. (I created this file with PIExpert 7.1 And i can do optimization with core EE42/21/15. After then i saw v8.0 on your web site then i install new software. Now when i attempt to open this file with 8.0, PIExpert changes core type and etc. And also can't do optimization with new selected core. And i can not change to core manually to EE42/21/15. This is another question. I hope, this design is right. I'm also ask this question to PIExpert Forum.)
My feedback schematic is different from the original design. Because i can not run the smps properly with TL431 feedback circuit and i change to feedback cirucuit from TL431 to secondary zener type for eleminating complexity. Other parts are generally same. May be some of the component values are not completly same, because i am testing different values...
Regards,
Zafer
Please check if PIExpert is latest version (8.0.4.1 as of today) and update it from "Help -> Check Latest Update" menu if necessary.
Also, please attach the design file here (rename it to FileName.uds.txt)
Now My PIExpert version is 8.0.4.1. I'm attaching .uds file which i was created with PIExpert 7.1 about two months ago. If i attempt the open this file with v8.0.4.1, PIExpert change the core type and etc.
Note: I am not using TL431 feedback now because of the complexity. If i fix it, i will change the feedback to TL431
Regards,
Zafer
Hello Zafer,
I suspected this kind of changes with the new version. I do not see any other option for you to use the original design without changing the transformer EE42/21/15 core. If this is not an option, try two things. One is to increase the minimum AC input voltage, if possible. The second is to decrease maximum output current, instead of 2.92A use 2.85A. With latest version of PI Expert it is possible to optimize the design if you can accept this kind of compromise. You will get a warning “Gap length too big” and some options to fix this. Pay attention to the new transformer specifications. If you really must keep this transformer size, check the new values. I would recommend switching to another core.
Cheers,
PI_Crusher
Hi,
I can optimize with my EE42/21/15 core at PIExpert 8.0. I'm attaching the .uds file. Everything is same with my original design and core. Only cable diameter is change, because i decrease the output current. So this design must work.
I think there is another trouble! Help!
Regards,
Zafer
Hello Zaher,
Reduce R10 from you schematic (R14 from new project) from 1k to 0.6k, this is going to increase the gain. Also, please measure the inductance of the transformer primary winding (143uH nominal) this value it is supposed to be no more than 10% off. Also, measure the primary leakage inductance, should be no more than 2.1uH. Let me know if these two steps are OK.
Cheers,
PI_Crusher
Hello Crusher,
I test R14 with 570R but it is not working.
Nominal primary inductance is 143,1 uH. But leakage inductance is 3,4uH. Is this the cause of the problem?
The transformer is come with test report. At that report they measure the nominal inductance as 149uH. But there is no words about leakage inductance at the report.
What am i do now?
Regards,
Zafer
The transformer is not that bad to justify this behavior.
Looking at “RUN_PC817_Led_Anode_Catode.jpg” signal. This is the voltage you measure on the secondary side, Anode to Cathode of PC817 input? If so, check C5 and C6 from your old schematic. Are they disconnected? Or you used 470nF instead of 470uF? (That voltage must be very smooth, with just a little of voltage ripple). Also, please show me the signal on the secondary side, Anode and Cathode of D7, and from primary side anode and cathode of D6 (old schematic).
Cheers,
PI_crusher
Yes, secondary side, PC817C pin1 and pin2. C5,C6 is right.
I'm attaching signals
Sorry, i am attaching other photos.
Regards,
Zafer
Hello Zafer,
I have to assume what you call “D6_UF4007_Vout_75V.jpg”, “D7_MUR1560_When Vout_75V.jpg”, etc. is the voltage you measure on the anode of D6 and D7. Comparing these signals with “RUN_Drain.jpg” it is obvious the transformer is not connected with right polarity.
Check “Original Transformer Construction.pdf” and “Schematic_2.pdf” against your physical hardware and make sure:
1) pin 1 of the transformer is connected to positive rail
2) pin 3 of the transformer is connected to Drain of IC1
3) pin 7 of transformer is connected to Anode D7
4) pin 5 of transformer is connected to Anode D6
If you have these connections wrong, just switch the pins in the right position and everything must be OK.
If you have these connections right, that means the manufacturer made a mistake and reversed the polarity of the primary winding. In order to correct this mistake you need to disconnect pin 3 and pin 1 from the PCB and reverse them: connect pin 3 of the transformer to the positive rail and pin 1 to Drain of IC1.
After these changes you should see all signals with same polarity on Drain of IC1 and Anode of D7 and D6. As you have them now, the waveforms from D6 and D7 are upside-down.
Cheers,
PI_Crusher
Hi,
I am suspencting myself and i did your offers immediatelly. All the signals are die everywhere.
Unfortunatelly there is a misunderstanding.
D6_UF4007_Vout_75V.jpg is the measurement of D6 CATHODE signal according to D6 ANODE.
D7_MUR1560_When Vout_75V.jpg is the measurement of D7 CATHODE signal according to D7 ANODE.
Run_drain.jpg is the measurement of Drain signal according to source.
I am measuring and attaching the signals again.
At Signal1.jpg file:
***Upper signal is the measurement of drain (pin 3 of transformer) signal according to source
***Lower signal is the measurement of anode of D6 (pin 5 of transformer) signal according to source
At Signal2.jpg file:
***Upper signal is the measurement of drain (pin 3 of transformer) signal according to source
***Lower signal is the measurement of anode of D7 (pin 7 of transformer) signal according to source
So the polarties are ok.
I have got two channel scope. Both channels grounds are same point. So i don't try to measure signals on the primary side and secondary side at the same time. Because the referans points are different. Primary and bias winding referans point is source and secondary referance point is RTN.
At signal2.jpg, i am connecing the scopes ground to source and i am measuring the drain. But at the same time i am measuring to D7 anode (pin7) signal? Why can i see that signal because the grouds are different?
I'm so confused.
NOTE:My 10MB quota was filled. How can i add new attachments or am i delete some old attachments.
Waiting for your reply eagerly.
Regards,
Hello Zafer,
The isolation from primary to secondary is for safety reasons only. On the debugging stage you can connect directly the primary ground (Source of IC1) to secondary ground (RTN). You can see a signal with the probe ground connected to primary and the tip connected to secondary because the impedance of the probe is high and there are parasitic elements (capacitors, isolation resistors, electro-magnetic coupling) connecting primary to secondary. For a reasonable accuracy you should measure any primary or secondary signal with the oscilloscope ground tied to the corresponding schematic ground, or connect them together.
Show me the signal from Anode D7 to secondary ground and Cathode D7 to secondary ground. I think you better start a new tread to post these pictures; I will look for your post.
Cheers,
PICrusher
Hi Crusher,
Thank you for your explanation, it is ok now. I am attaching the following photos.
D7Anode_RTN.jpg -
D7Cathode_RTN_RUN_Vout_75VDC.jpg - I am taking this photo when the output is 75VDC.
D7_Cathode_RTN_Vout_31VDC.jpg - I am taking this photo when the output is 31VDC.
I think this ripple on the output (D7_Cathode_RTN_Vout_31VDC.jpg) is derive from feedback mistake. But i'm not sure.
Regards,
Zafer
Hello Zafer,
Since I do not see the design files from our PI Expert software, I have to assume you calculated by yourself the transformer, the feedback loop, etc. I can not help you without this information, seems like the circuit goes into auto-restart.
You have to download and install our software, type-in you input-output parameters, and see where your calculated values are not a match for the required range. Also, you can post here these design files and we can verify directly your project. Or at least give us full description for input design parameters: AC input range, line frequency, desired output voltage regulation, etc. Only then we can see what it is wrong in your application.
Cheers,
PI_Crusher