TOP264 Default UV Threshold

9 posts / 0 new


I built a design from the attached image. I removed R2 and R3 and shorted V-pin to Source in order to disable UV protection. However, the IC turns too late - at 170-180VDC. I am trying to find out what is the internal threshold of the TOP264 IC when no UV resistors are added.

I installed UV resistors (R2 and R3). Now I observe that the IC doesn't try to turn on before specified value (100VDC or so). After this point, it enters auto-restart mode. Control pin discharges to 4.8V, then charges to above 5V every cycle. This remains until 170-180VDC.

Any advice as to where to look to make the converter turn on successfully at low voltage?


d.jpg98.32 KB

There may be something else wrong with the power supply that prevents it from starting properly. Check and see if the transformer windings are the proper polarity. The anodes of D6 and D7 should go negative when the TOPswitch internal fet is on. If one (or both) of these windings are reversed, it can keep the supply from delivering power, as the supply will try to source current directly through D6 or D7 to C7 or C9, during the TOPSwitch on-time, which will look like a short to U1. Also, make sure that VR1 is properly oriented, as this will clamp the transformer reflected voltage durring primary switch off-time and also prevent the supply from delivering power.

Checked - Polarity is proper and component placement and behavior is correct.

However, it seems that the transformer is wrong - it has 5mH magnetizing inductance, while the one in the schematic has 1.2mH.

I am trying to understand why would large inductance create problems only at low Vin. According to Pi Expert design, converter is supposed to work in both DCM and CCM, so larger inductance would only mean smaller current spike - but it should still work.

Is there any theoretical reason why this design (with transformer with higher Lm) wouldn't work at lower Vin, but works at higher Vin, albeit only at variable frequency mode?


If your transformer has 5mH magnetizing inductance, it probably doesn't have a gap. This means the transformer will saturate at a much lower value of drain current and not allow proper power delivery. Apply the proper gap for your specified inductance, and I'm pretty sure your problems will go away.

Thanks for advice - however, my transformer does have a gap, with AL=100nH/T^2.

On re-reading your first comment, you say: "The anodes of D6 and D7 should go negative when the TOPswitch internal fet is on."

However, I notice that anodes of rectifiers are actually positive - see attached oscillograms. Is this incorrect phasing? Bias and secondary are in phase with each other, and out of phase with primary.

Output.jpg1.18 MB
Bias.jpg1.32 MB

If your transformer measures 5 mH rather than the 1.2 mH, design value that still indicates that the gap is insufficiently large, and the transformer will saturate early. The bias and output windings are the wrong polarity, according to the scope picture you posted. For a positive output flyback, the output and bias rectifiers should go negative when the primary switch is on, isolating the secondary circuits from the primary and allowing the primary magnetizing inductance to store energy, which is released to the secondary and bias outputs when the primary switch turns off. You can temporarily fix the polarity problem by swapping the primary leads on the transformer. You will still have issues with the insufficient transformer gap, but you will be able to deliver at least a fraction of the specified power.
Can you post a PDF of the design file?

I just ran a PIXLS spreadsheet on your design, and the TOP264 is way too big for the intended application. A far more suitable part (and likely less expensive) is a part from the TinySwitch or LinkSwitch series. I ran a spreadsheet using the TNY284P, and the transformer design is much more sensible - still an EE13, with 105t primary, 5t secondary, 29t bias winding, 1.72 mH primary inductance.

TinySwitch-4_3V3_3W3.pdf802.12 KB

Fixing the phasing resolved it - thanks. It is too late to change the device, but I agree that LinkSwitch or TinySwitch are probably more suitable.

You will need to dial down the current limit for your TOPSwitch as low as possible. It is also a good idea to use a larger core and less continuous operation (KP of 1 or greater) so you aren't stuck with a transformer that requires a crazy number of primary turns to keep from saturating.