解决方案搜寻 技术支持

Dual output smps design with top249 and myrra 74010 voltage drop problem

Posted by: amardeep on

Hello
I am trying to develop dual output smps for motor control +5v and +12v
Top249 and myrra 74010 flyback transformer 10w. +5v 330ohm preload and +12v 1k ohm preloaded.
I am following suggested circuit of top249 as per application notes.
Everything is ok voltage is stable as per expectation.

My problem is that if I put some load as resistor then +5v goes low then +12v goes high.

Here is some test report 1.
1)+5v side 68ohm as load = 4.98v & +12v side no load become =12.53v
2)+5v side 33ohm as load = 4.96v & +12v side no load become =12.82v
3)+5v side 10ohm as load = 4.86v & +12v side no load become =12.90v

Test report 2.
1) +5v side no load = 5v & +12v side 68ohm as load become =11v
2) +5v side no load = 5v & +12v side 33ohm as load become = 10.42v
3) +5v side no load = 5v & +12v side 10ohm as load become = 9.84v

Test report 3
1) +5v side 68ohm as load 4.98v & +12v side 150ohm as load = 12v
2) +5v side 33ohm as load 4.96v & +12v side 150ohm as load = 12.2v
3) +5v side 10ohm as load 4.86v & +12v side 150ohm as load = 12.7v

I do know why +5v is getting low with minimum increase in load even +12v has preload also +12v is getting low with minimum increase in load even +5v has only pre load

Please help me to diagnosis the problem

评论

Submitted by PI-Wrench on 03/11/2021

Please post a schematicm For the supply in question.. I especially need to see the circuitry around the TL431.

Submitted by amardeep on 03/11/2021

Thanks for response
Please check the complete circuit. Flyback transformer is rated for 10w

Submitted by PI-Wrench on 03/12/2021

Several things come to mind upon viewing the schematic. First, I am assuming that the 12V and 5V outputs each have a different ground system and float with respect to one another (correct me if I'm wrong). Unfortunately, this is not the optimum situation for good cross regulation between outputs.

Also, there might be oscillation or excessive ripple on the 5V output. Please check the output ripple on the 5V output with an oscilloscope as the load is increased to see if there is any trace of oscillation.. Also, make sure that the ground and output connections on the TL431 control pin voltage divider are connected directly at the 5V output so that any voltage drop in the output traces is corrected by the TL431.
If you find oscillation on the 5V output you might want to try placing a resistor in series with the 100nF cap that is conncted between the TL431 cathode and reference 3.3k-4.7 k is a good starting value. It may also be necessary to increase the voltage divider resistors form 4.7k to 10k, as these resistors work against the 100nF capacitor to determine the control loop frequency response.

Finally, what are the actual requirements for load regulation on the 5V and 12V outputs?

Submitted by amardeep on 03/14/2021

#Yes both of output have different grounds.
# yes I have taken tl431 divider from output capacitor.
I have tried many resistor values values with series capacitor. Like 1.8k, 2.2k,3.9k,4.7k
But the result remains same

For confirmation of my hardware.
I change the circuit switching frequency to 66khz and take smps flyback transformer from old smps(66kHz oscillator) only output 12v.

Replace it with myraa 74010 transformer
I found circuit behaves fine upto 1amps @12v I have tested and no voltage drop etc.

Now I am confident for my hardware.

My requirement is I need max .5amp @5v and .5amp @12v but in that configuration 5v is not regulated as I have post my problem.

Submitted by PI-Wrench on 03/15/2021

You need to position your voltage divider such that it reads the voltage across the output terminals of the supply, not the filter capacitor, especially if there is significant trace length between the filter capacitor and the supply output. The TL431 anode should also be positioned at the supply output so that the TL431 doesn't see a ground shift due to voltage drops in the output PCB traces.

BTW , the device your are using (TOP249) is way too large for your output requirement. This and the value you have chosen for the X-pin resistor will cause transformer saturation during startup. One of the devices from the TinySwitch series would be more appropriate for your input voltage and output power requirement. You should also be able to reduce your input bulk capacitor to ~10uF, unless you have a requirement for extreme holdup time.

Using a Schottky rectifier for your 5V output (1N5819 or IN5822 should be OK) may help with your output cross regulation due to lower losses in the 5V output.

Submitted by amardeep on 03/16/2021

Thanks for valueable suggestion.
I have change the tl431 position and wide the pcb track size and I found more stability in output regulation.
I don't have experience of dual output smps. I presume if only one side is regulated under high load condition pwm duty cycle become wide as I have seen on cro. Thus another output 2 goes high.
One more thing if output 2 goes under high load condition like short circuit or defective ic etc. Then it become none effective on output 1.
Therefore output1 is main concern if its ok then output 2 voltage become almost ok.

My question is under mentioned above conditions does output 2 drop low with 50mA load ? Even output 2 winding is mentioned for 200mA.

Please support me how dual output smps works ?

Submitted by PI-Wrench on 03/17/2021

If output 2 (12V) is short circuited or overloaded, it will engage the primary overcurrent shutdown on the IC and cause the entire power supply to go into autorestart protection mode until the short/overload is removed.
Unfortunately, the transformer you are using is not well documented - in fact, the design is quite old, as list the compatible PI IC as TOP222 (old TOPSwitch-II series).. This means I can't offer much insight as to whether the two secondaries are optimally coupled for good cross-regulation. You may want to inquire to the manufacturer as to the guaranteed isolation voltage between secondaries in case there is a big voltage offset between the 5V and 12V supplies.

Whatever you can do to reduce losses in the 5V output will help limit voltage excursion on the 12V output as a function of 5V load, hence my suggestion for using a Schottky diode for the 5V output.

The biggest concern will be voltage droop on the 12V output when the 5V output is lightly loaded. If you can accept the extra power consumption, a possible solution would be to run a resistor from the 5V output directly to the cathode of the TL431. This will act as an active preload, and force the TL431 to sink more current from the 5V output as it is unloaded. This is an old trick used with the first series of TOPSwitch devices to enable them to regulate down to zero load. You will need to play around with the resistor value for optimum results - start with maybe 100 ohms. If you decide to use this scheme, you can omit the 1k resistor across the optocoupler LED.