EEVblog® Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: erdemtr55 on October 02, 2026, 06:53:46 pm

Title: Uc3854 problem
Post by: erdemtr55 on October 02, 2026, 06:53:46 pm

Hi Guys,
I tried to simulate a PFC circuit based on the schematic in the attached file, but for some reason, the gate signal is at a very low level (in the mV range). What could the problem be?

Note: 100 kHz, 250 W, Vin 220 VAC, Vout 400 VDC.

Title: Re: Uc3854 problem
Post by: moffy on October 02, 2026, 11:03:31 pm

If you post the LTSpice *.asc file and the UC3854 model we could try them and possibly find the problem.

Title: Re: Uc3854 problem
Post by: erdemtr55 on October 04, 2026, 08:56:00 am

I sent you my file , maybe you can solve my problem
thanks

Title: Re: Uc3854 problem
Post by: moffy on October 04, 2026, 09:03:38 am

Quote from: erdemtr55 on October 04, 2026, 08:56:00 am

I sent you my file , maybe you can solve my problem
thanks

I can't find what you sent, but you are better off just posting the files on the forum. There are more knowledgeable people than I am.

Title: Re: Uc3854 problem
Post by: erdemtr55 on October 04, 2026, 01:50:29 pm

I am  sharing it at now,
Sorry previously

Title: Re: Uc3854 problem
Post by: moffy on October 05, 2026, 03:36:20 am

From the testing of your simulation file it looks like there might be a problem with the UC3854 model, especially since it has been developed by an individual rather than the manufacturer.
The following article: https://dspace.upt.ro/jspui/bitstream/123456789/1802/3/BUPT_ART_Șchiop_f.pdf (https://dspace.upt.ro/jspui/bitstream/123456789/1802/3/BUPT_ART_Șchiop_f.pdf)
goes into detail of how the authors developed a pspice model for the UC3854, and seems to have enough detail to reconstruct it but it would involve a bit of work, unfortunately there doesn't appear to be a link for their model.

Title: Re: Uc3854 problem
Post by: moffy on October 05, 2026, 05:23:48 am

I have to revise my last statement, I modified the LTSpice file to be like the example given in the datasheet and it output full level gate pulses.

Title: Re: Uc3854 problem
Post by: erdemtr55 on October 05, 2026, 01:42:38 pm

I tested mod2 and firstly voltage boost 410 v but then voltage buck 320-310v
I think that it is not normal.What if  how ı solve it?

Title: Re: Uc3854 problem
Post by: moffy on October 06, 2026, 12:32:51 am

Quote from: erdemtr55 on October 05, 2026, 01:42:38 pm

I tested mod2 and firstly voltage boost 410 v but then voltage buck 320-310v
I think that it is not normal.What if  how ı solve it?

The model for the UC3854 is unverified, it cannot really be trusted until proven, the question is does the model behave as the real device?
Here is an application note: https://www.ti.com/lit/an/slua144/slua144.pdf?ts=1724033606595 (https://www.ti.com/lit/an/slua144/slua144.pdf?ts=1724033606595)

Title: Re: Uc3854 problem
Post by: MariuszD on October 06, 2026, 05:18:15 am

Quote from: erdemtr55 on October 05, 2026, 01:42:38 pm

I tested mod2 and firstly voltage boost 410 v but then voltage buck 320-310v
I think that it is not normal.What if  how ı solve it?

The 410V does not result from the controller's operation; even if you remove the switching transistor, the phenomenon will remain. You need to limit the charging current of the capacitor through the NTC.
320V is simply the rectified mains voltage, the boost converter is not working.