Flyback controller duty cycle clamp 45-50%

You are after something like the LT1247 that doesn't allow duty cycle to go above 50%.

A fairly simple comparator circuit using a SOT23-5 packaged comparator may be the best way to control the duty cycle if you don't want anything else effected. A simple RC on the "gate" output of the chip can measure the on time. The output of the comparator can pull the "Control" voltage down. Since this goes in at the point where the integration is done integrator wind up is prevented.

Though Duty cycle hits 78% at startup the Power integrations part TOP256 may be of interest. Internal power switch etc. Stocked by Newark.

regards

Louiiv

Though Duty cycle hits 78% at startup the Power integrations part TOP256 may be of interest. Internal power switch etc. Stocked by Newark.

regards

Louiiv

I seem to remember that TI (?) produced what was apparently a 3844 with a TL431 and an opto-isolator all in a special long DIL package with the creepage limits built in.

Graham H

Thanks for the responses everyone.

What I had in mind is to use a MIC4427 (fet driver dual) between the controller gate drive out and use this to drive the low side FET Q1 and to drive a small PULSE transformer that will drive the high side FET Q2. See image.

I was planning on using a small RC to sync the drivers adding a small delay to the low side input so that both FETS would turn on/off simultaneously. The only problem is I only have 3842's and NCP120X's these have no duty cycle clamps. I know the 3842's can be sync with an external OSC but I would prefer to use the NCP120X if at all possible.

MooseFet Could you elaborate on your comparator circuit please.

ck

Switcher Chip

0-Vcc R1 V1 -------- Gate-----/\\/\\--+-----!-\\ R2 ! Drive ! ! >---/\\/--+--!Control C1 --- -!+/ ! ! Voltage --- ! ! Pin ! ---K1 * Vcc GND

The time constant R1*C1 is made so that V1 goes above K1*Vcc when the on time exceeds the desired. It is about equal to the cycle time of the switcher. R2 may not be needed. It just serves to limit how much current the comparator yanks from the Control Voltage of the switcher chip. Any time the gate drive trys to stay on too long the comparator discharges the control voltage a little thus regulating the maximum on time.

You might need a diode if this goes in at the FB node and a divider is hooked up to that directly rather than via optocoupler. Also, make sure the opamp can't pull it beyond and abs max limits. I usually employ transistors to pull it up so nothing can happen.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

le

4 o

back

r

ip

ll

=A0Switcher

=A0Chip

=A0Pin

uch

r

The node I am suggesting is the "control" node. Most switcher chips have a node where you connect the compensation network. Look for the block diagram here and look at the Vc node.

You can also go in there. But careful with hanging an opamp to that, sans diode it loads the output of the error amp.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

ycle

844 7A

pto

edback

er

le

tor

chip

he

e

mall

an

e.

=A0 =A0Switcher

=A0 =A0Chip

-

=A0Pin

f

much

tor

n e

The circuit is using a open collector comparator not an op-amp. I thought I said that somewhere but perhaps I omitted it