0
\$\begingroup\$

I have built a small and very simple flyback SMPS based on NE555N IC. The circuit exactly I did built from is shown below:

NE555N flyback

It works nicely for small loads.

However I have some issues with it:

  • Despite NE555N datasheet says output (pin 3) voltage is only about 2 volts lower than Vcc, I have only +6.3v on pin 3. Is it a digital multimeter misreading?
  • The MOSFET used there is IRF3205, but currents beyond 6-7A force it to heat up, and 10A (with 100-110 Watts on secondary) even more (heatsink is ready to boil water)
  • Some answers here and some sources on the web (and this circuit) recommend using logic level mosfets. Can I avoid using them (probably until I get some of them)? Will logic level mosfets work perfectly compared to regular ones I using now if their properties (Vdss, Id, Rds(on), Qg) will be relatively same?

Impulse transformer was salvaged from old dead 200W ATX SMPS, its sizes are (WxHxD) 36x42x16 mm. Its main problem that it's cracked (but still works without noises). Primary has 8 turns of 1mm wire, secondary has 160 turns of 0.3mm wire; another secondary has 3 turns for 5 volts source (also 1mm wire).

Heatsink sizes are 32x39x14 mm probably aluminum (also from old hardware). Larger heatsink is avoidable because of limited space where circuit will be stored.

Supply is 12V 19Ah Lead-acid battery pack.

I strive to minimize thermal losses on this circuit, so I am seeking ways to minimize power losses on MOSFET.

Does MOSFET overheat due to undervoltage on gate (hence not fully opened) or cracked transformer core? Do I need some other enhancements for this circuit?

\$\endgroup\$
  • \$\begingroup\$ What is the switching frequency, and inductance of the primary? It sounds like you are using one of the secondaries as the primary in your circuit, flipped around. \$\endgroup\$ – Bitrex Jun 3 '14 at 5:34
  • 1
    \$\begingroup\$ Expecting a flyback converter with a random transformer, MOSFET, and switching frequency, that also has no feedback, to work well across a wide range of output currents is at best quite unrealistic. \$\endgroup\$ – Bitrex Jun 3 '14 at 5:51
  • \$\begingroup\$ The site I've taken circuit from recommends 7-8 turns. BTW can just I brute number of turns in primary without secondary? I am not so experienced with flyback circuits, I am still learning how to do it right. \$\endgroup\$ – user43920 Jun 3 '14 at 6:05
2
\$\begingroup\$

Despite NE555N datasheet says output (pin 3) voltage is only about 2 volts lower than Vcc, I have only +6.3v on pin 3. Is it a digital multimeter misreading?

A ordinary digital multimeter will average the pulse train on the gate of the FET into what looks like a DC level.

The MOSFET used there is IRF3205, but currents beyond 6-7A force it to heat up, and 10A (with 100-110 Watts on secondary) even more (heatsink is ready to boil water)

An NE555 is a poor choice for driving a MOSFET (even logic-level) as it cannot sink and source much current. Put a monolithic MOSFET driver IC between the 555 and the MOSFET and your power dissipation will improve.

Does MOSFET overheat due to undervoltage on gate (hence not fully opened) or cracked transformer core? Do I need some other enhancements for this circuit?

Get an oscilloscope (even a relatively old one) and probe the switching nodes. It's the only way to really know what's going on in the circuit.

\$\endgroup\$
  • \$\begingroup\$ Agreed. Almost certainly your turn-on and turn-off times are so slow that you're getting significant transition inefficiencies. \$\endgroup\$ – WhatRoughBeast Jun 3 '14 at 2:27
  • \$\begingroup\$ I agree that a 555 timer alone is not a good choice. What driver IC can you suggest to use which will take signal and amplify current? Can I just use transistor amplification somehow? (for now I just limited to 555 timer and number of mosfets and BJT transistors) Old oscilloscope shows clean square wave signal without mosfet, but I can't measure it's voltage level now. I will check it with mosfet later. \$\endgroup\$ – user43920 Jun 3 '14 at 3:02
  • \$\begingroup\$ Or, maybe it's naive to say "how many current", instead I just need to take voltage from 12v rail and charge MOSFET gate as quickly as possible to completely turn it on. Then reverse, discharge it the same way. Right? \$\endgroup\$ – user43920 Jun 3 '14 at 3:12
  • \$\begingroup\$ Right. And the way to do that is to use a driver chip between the output of the 555 and the gate of the MOSFET. There are all sorts of high current drivers made for this job, an example is the Maxim MAX627 datasheets.maximintegrated.com/en/ds/MAX626-TSC428.pdf \$\endgroup\$ – WhatRoughBeast Jun 3 '14 at 4:00
  • 1
    \$\begingroup\$ I disagree that the 555 is a poor choice for driving a MOSFET, the bipolar 555 can source and sink up to 200mA. I've used them to switch inductors for boost converters and they have worked great, GIVEN the proper choice of inductor, fet, and switching frequency. There is too much we don't know about this circuit to assume that lack of drive current is the problem. 8 turns on the primary seems very low. What is the switching frequency? Is the primary going into saturation? \$\endgroup\$ – Bitrex Jun 3 '14 at 5:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy