Author Topic: Self starting SMPS working?  (Read 851 times)

0 Members and 1 Guest are viewing this topic.

Offline AchuTopic starter

  • Contributor
  • Posts: 49
  • Country: in
Self starting SMPS working?
« on: October 17, 2019, 04:53:24 pm »
Hi,

Could any body explain it to me how this self starting SPMS PSU work?

I have seen only those with auxilliary power supply.
 

Offline Miyuki

  • Frequent Contributor
  • **
  • Posts: 908
  • Country: cz
    • Me on youtube
Re: Self starting SMPS working?
« Reply #1 on: October 17, 2019, 05:03:48 pm »
It is same as for example small two transistor astabile oscilator
Or other self oscilatin smps, commonly flybacks
Transistors have positive feedback from that small transformer and without controll signat it wil oscilate itself from saturation to saturation
It starts because it have no stable state in middle
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Self starting SMPS working?
« Reply #2 on: October 17, 2019, 07:04:28 pm »
Wow, that's a nearly perfect circuit.  I am impressed!

The only things I can complain about are perhaps the aux supply being cap-input filtered, where maybe some series resistance would be wise to take the edge off charging C27; and it would be nicer to have cascaded current and voltage error loops, but that does require one more op-amp to pull off.  Of course these days, PFC would be nice as well, but for the time, this is absolutely fine.

Anyway, starting has to do with the high value resistors on the base drive circuits.  I've never built an inverter like this before (with the bias resistors), but I believe the way it goes is, the BJTs are biased slightly on (10s mA?), which gives gain, and pretty soon due to positive feedback (through the drive transformer -- note the load current flows through it) it saturates in one direction.  Then probably due to transformer saturation (but if not, then probably the drive capacitors charging up), drive abruptly stops, and reverses; the system is symmetrical so the process repeats.

Once aux supply comes up, the TL494 acts to stop the self-oscillating inverter.  This is why the drive is inverted by two extra transistors -- the default "idle" state is with both transistors on, holding the drive transformer voltage at zero.

It's a nice way to drive BJTs, as, for merely the cost of an -- admittedly rather complicated -- transformer, you don't have to worry about base current or voltage, or drive power, or hFE at all -- that's all handled by the transformer, and the transistors just latch on and run themselves.  It's effectively making each transistor into a dynamic SCR -- one that stays latched on for a limited time.

I have however made a half-wave version, here:
https://www.seventransistorlabs.com/Images/LED_Light2.png
where the switch operates as a one-shot timer.  The control circuit triggers the switch by pulsing the drive transformer; the switch stays saturated for about 5us, at a forced hFE of about 15t / 3t = 5.  After that, the ferrite core saturates, shorting out base current and rapidly switching off the transistor.  After a brief reset pulse (not handled by anything shown here; I later put a P6KE15CA across the 60t primary to limit peak voltage), the switch is ready to go again.  The control just acts as a frequency modulator, generating pulses to maintain constant output current.  This is also known as a constant on-time control.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: sorin


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf