Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] Push-Pull Converter Closed Loop Issue

Status
Not open for further replies.

adnan012

Advanced Member level 1
Joined
Oct 6, 2006
Messages
468
Helped
2
Reputation
4
Reaction score
2
Trophy points
1,298
Activity points
4,923
Hi,

I am working on Push-Pull DC/DC converter which is a part of modified sine wave inverter.

The following specs are desired.

Input voltage range: 10.5 -21VDC

Output 800Watts when input is 10.5-14VDC

Output 350Watts when input is 15-21VDC


I am using TL494 PWM controller configured as push pull.

Transformer is PQ32/30 core. Switching frequency is 40KHz.

There are 3 + 3 Primary turns of 0.25mm thick copper strip and 33 (Litz wire) turns on secondary side.

This transformer is only for testing purpose.

Open Loop Testing:

For testing the input source is high power linear power supply.

Digital load is used for loading the DC/DC converter.

In open loop test the duty cycle is fixed and is set to 43%. The input voltage are varied from 10.5 -18VDC. I get 120VDC when the input voltage is 16VDC and load is around 700 to 800Watts and the efficiency is 85%.

There is no issue with the DC/C section. The input current varies when the load is changed or when the input voltage are changed.

Closed Loop Testing:


I am facing issue in closed loop test. When feedback is present (both primary and secondary grounds are common in this case) I have the following scenario

Digital load is set to 1 amp.

Output voltage is set to 90VDC .

The output voltage are stable when the input is changed from 11VDC to 18DVC (or more) and Duty cycle also become very narrow in case of high input voltage. But the input current always remain same and at high input voltage the efficiency goes to 60%.


I need help to resolve this issue.
 

Attachments

  • ClosedLoop.jpg
    1.9 MB · Views: 139
  • OpenLoop.jpg
    1.8 MB · Views: 106

Why not do the two tests with the same conditions?
Other than that the loop is closed. Too many variables.

I do not see the usual output inductor in your ClosedLoop
schematic. That element makes the output current
more constant, much less current ripple out of the FWB
at load. Maybe the ripple (which I'd expect to be gross
with only rectifiers and filter caps, especially crappy
electrolytics w/ high ESL) is making the average
seen by the feedback, much lower than you'd think
given VIN and turns. Have you looked at the output
voltage with a 'scope, rather than a DMM?
 

Thanks for you reply.

I am using oscilloscope . I can share picture of the different test points. I need some time.

I need design equations for the output capacitor and inductor.

Regards
 

you've got no output choke, hence the peak currents in the fets can be very high - the choke/cap filter is there to average the pwm'd output from the diodes - just o/p caps gives only a peak capture type of thing - an L/C o/p filter introduces delays and the feedback loop must be compensated appropriately ...
 

    adnan012

    Points: 2
    Helpful Answer Positive Rating
Thanks for reply.

I have added an output inductor and the efficiency issue has been resolved.
I need the design equations for design of output filter.

The maximum load current is 8A .

Regards
 

I need the design equations for design of output filter.

The maximum load current is 8A .

To attenuate the pulses, you want to install a coil with the right impedance (reactive impedance).
Suppose amplitude is 180V output at 1A 40kHz. In terms of raw math, an easy example is to attenuate 50 percent.

So 180V divided by 1A calculates as 180 ohms. The formula for inductive impedance is:
2 π f L.

Thus 180=2 π 40000 L
Solving for L, makes it 716 uH. This is a crude example since the formula applies to a sinewave, not pulsed. DC.
 

    adnan012

    Points: 2
    Helpful Answer Positive Rating
Thanks for reply.

I have placed and output filter.

I can regulate the converter output voltage.

i have attached a fixed resistive load at the output side. The output voltage is set to 94VDC. Input is varied from 11 to 18VDC and out put voltage remains same.
The issue is with the input current it always remain constant.

At 11 V input voltage the input current is 10A and a 18V input the current remain same.

There are two ways to control the output voltage, one is using the error amplifier and second one is using the DT (dead time control pin) to control output voltage in open loop mode. I have same issue in both modes. The input current does not reduce when high voltage are applied. Input dissipation increased in this case.

Regards
 

Attachments

  • 12V Input Voltage.jpg
    12V Input Voltage.jpg
    2 MB · Views: 92
  • 14V input Voltage.mp4
    4 MB
  • 18V Input Voltage.jpg
    18V Input Voltage.jpg
    2 MB · Views: 94
  • 18V Input Voltage.mp4
    4.1 MB

Hi,
At 11 V input voltage the input current is 10A and a 18V input the current remain same.
If so, then with 18V input something really gets hot. It´s additional 70W of heat.
Check which devices get hot.

Klaus
 

Your movies can be viewed by downloading each to my computer. (They do not play in my browser window. The capability to insert movies was recently added to the Edaboard interface so we're not certain how it should behave.)

We expect duty cycle at 18V to be about 70 percent of the length at 12V. The scope trace appears to be less.

Perhaps this is due to small Henry value of the primary winding? Therefore you get rapid Ampere buildup?

Do you have adequate snubbing at the primary? It needs to absorb spikes on the order of 800W.
 

Thanks for reply.

The heat sink for primary side MOSFETS is getting very hot.

I have snubber at the primary side.

If the duty is fixed to it's maximum value which is 43% ( in open loop) and input volatge are varied then output voltage increases and high effeciency is achieved .

But if adjust the output voltage manually according to the input voltage level the input current stays at constant level.
 

You didn't tell the output inductor value. The output circuit acts as buck converter, inductance is dimensioned for a specific maximal current ripple. However if output DC current is smaller than 1/2 designed current ripple, inductor current becomes discontinuous and duty cycle must be reduced to keep the output voltage constant.
 

Thanks for reply.

The inductor is not properly designed. For testing purpose ,with 2 amp maximum current , what should be the ripple current ?. Can I use 30% of maximum current value. Can I use same type ferriter core for inductor design, which is used for Transformer?

Regards
 

...then maybe the inductor gets into saturation causing high peak currents and high power dissipation.

--> choose a suitable inductor.


Klaus
 

Can I use same type ferrite core for inductor design, which is used for Transformer?
Same ferrite core yes, but with air gap to allow for dc current. Use Ferroxcube SFDT or TDK Ferrite Magnetic Design Tool for inductor calculation.

 

    adnan012

    Points: 2
    Helpful Answer Positive Rating
I reviewed the schematic and waveforms and got more questions than answers:
1. Schematic in post #1 shows 2SK2608 transistor with maximum continuous drain current of 3A. Which transistor is actually used?
2. Which circuit nodes are probed in post #7 waveforms? We would like to see transistor drain-source voltage.
3. How is VCC related to +12V supply?
 

    adnan012

    Points: 2
    Helpful Answer Positive Rating
Thanks for reply.

The actual MOSFET is IXFH80N25X3. And rectifier diodes are SSP2030.

VCC and 12V are same . I have also tried to provide fixed voltage to TL494.
 

Attachments

  • SSP20300CG.pdf
    755.3 KB · Views: 78
  • Littelfuse_Discrete_MOSFETs_N-Channel_Ultra_Junction_IXF_80N25X3_Datasheet.PDF.PDF
    332.8 KB · Views: 56
Last edited:

How to choose inductance value in the tool?
 

Attachments

  • PQ32-30 Inductor Design.png
    PQ32-30 Inductor Design.png
    22.9 KB · Views: 111

There is something wrong at high input voltage.

The Green represents the Gate signal exactly at MOSFET a gate pin. The yellow is the Drian node of the corresponding MOSFET. The purple is the drain signal at the MOSFET of other leg.

The first image shows drain signal at 12Volt input and 113Vout @ 78 watt.
The second image represents signals 1t 16Volt input and 113Vout @ 78 watt.
 

Attachments

  • Input 12.V Load 78 Watt.JPEG
    Input 12.V Load 78 Watt.JPEG
    106.5 KB · Views: 80
  • Signal at 16 V.JPEG
    Signal at 16 V.JPEG
    114.6 KB · Views: 85
  • Signal at 2nd leg 16V.JPEG
    Signal at 2nd leg 16V.JPEG
    101.5 KB · Views: 97
  • Input 16V load 78 V.JPEG
    Input 16V load 78 V.JPEG
    100.8 KB · Views: 85

This behavior starts at 13.34V.

The following image shows result at 13.78 volt.
 

Attachments

  • signal at 13.78 Volt .JPEG
    signal at 13.78 Volt .JPEG
    124 KB · Views: 94

This is at 21Volt input .

The load is constant and voltage are regulated by closed loop.
 

Attachments

  • at 21V Input.JPEG
    at 21V Input.JPEG
    107.9 KB · Views: 72

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top