LM3450A 340us Off-Timer

Thomas Huber

Other Parts Discussed in Thread: LM3450

Hello!

I built a Flyback Power Supply (24V, 26W) with orientation towards http://www.national.com/an/AN/AN-2150.pdf
I used the LM3450A:
http://www.national.com/ds/LM/LM3450.pdf

Now my problem: The Power Supply stays always in the mode, where every 340us a short impulse comes over the GATE-Pin, then another 340us the MOSFET is off and then trys to switch again and so on. Can you please tell me all cases where the LM3450 changes to this "max. off-time" mode, so I can validate why my circuit stays in this mode all the time?

The 340us are the max. Off-Time (normal operation) of the LM3450A-Off-Timer (datasheet).

Has anyone similar problems using the LM3450A?

Thank you very much!
Kind regards!
Thomas Huber

over 14 years ago

0 James Patterson over 14 years ago

TI__Intellectual 2590 points

Thomas,

Your LM3450 application must be in burst mode when you see the 340us off-times, meaning COMP voltage is below 1.4V.

This could mean that the output power is so low and/or the switching frequency is so high that the minimum duty cycle (caused by leading edge blanking) is providing too much energy to the output. COMP continues to fall far enough (<1.4V) to force burst mode where one on time occurs followed by the max off-time. This continues until COMP rises back above 1.4V nominal (+ hysteresis).

The other possibility is that the ZCD (zero crossing detection) is not wprlomg properly. If the voltage at the ZCD pin does not fall below ~1.24, then the correct off-time will not occur and eventually a maximum off-time will occur. I'm not sure why this would be happening but you can probe the ZCD voltage to check.

I can look at the circuit in detail if you can post schematics, relevant operating points (input voltage, LED stack voltage, LED current, etc..), and (if possible) operating waveforms showing the burst mode.

James Patterson

Staff Applications Engineer

Texas Instruments

0 Thomas Huber over 13 years ago in reply to James Patterson

Prodigy 75 points

Hello James,

thank you very much for your reply!

I put the Scematic in the appendix and I will provide the waveforms later the day, when I can measure again.

I also put the data of transformer construction in the appendix.

The operating points are: Vin = 230V, 50Hz; Vout = 24V; Pout = 26W. I use only the first stage, so there is no LED Driver at the Moment.

Thank you very much!

Best regards!

Thomas Huber

5280.Schematic.pdf

1680.transformer construction.pdf

Note: in Scematic, I put more MosFETs etc. in it to get flexible footprints. As output capacitence I only use one 680uF.
For Clamp Circuit i used the left path in scematic and for Q1 and Q3 i used the suggested ones from the app note (STD4NK80ZT4 (Q1) respectively SPD06N80C3 (Q3)).
Also I use only one transformer. Instead of two parallel 1R resistors, I use three parallel 3R resistors for "R30||R31" (other current limitation setting, higher power disipation possible).