• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 64 subscribers
  • Views 2203 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Help to find PWM controllers and other parts for push/pull DC/DC converter

Adrien Zancan
Prodigy 105 points
Other Parts Discussed in Thread: UC1856, UC3856, UC2856, UC2823A, UC2823B, UC3823B, UC3823A, UC1823A, UC2825A, UC1825A, UC2825B, UC3825B, UC3825A, TL598, SG2524, SG3524, UC3846, UC2846, UC3847, UC1846, UC1526, UC3526, UC2526, UC1525A, UC3527A, UC2525A, UC1527A, UC2527A, UC3525A, UC2524, UC3524, UC2827-2, UC3827-2, UC3827-1, UC2827-1, UCC28085

Hello,

My project is a 1kW high efficiency, 12 V DC to 230 V AC converter.
My present question only concerns the 1st stage which is a step-up stage with following characteristics :

  • transformer used in push-pull topology
  • 10 V DC < Vin < 18 V DC
  • I in max = 110 A
  • Vout = 350 V DC
  • I out max = 3.2 A
  • frequ. = 100 kHz
  • High efficiency
  • not overpriced

I would like to find the right PWM controller but there's a lot of part nums ! Here are the ones which satisfy my Vin and frequency constraints :

  • UC1856, UC2856, UC3856
  • UC1823A, UC2823A, UC2823B, UC3823A, UC3823B,
  • UC1825A, UC2825A, UC2825B, UC3825A, UC3825B
  • UC186* UC286* UC386* ( *= 1 to 8 )
  • TL598
  • SG2524, SG3524
  • UC1846, UC1847, UC2846, UC2847, UC3846, UC3847
  • UC1526, UC2526, UC3526
  • UC1525A, UC1527A, UC2525A, UC2527A, UC3525A, UC3527A
  • UC1524, UC2524, UC3524
  • UC2827-1, UC2827-2, UC3827-1, UC3827-2

The controller must tune the ratio according to Vout and primary current.
For current sensing, I still don't know if I will use a resistive shunt or a hall effect sensor. I'm open to suggestions.

To switch primary coils, I think to use 2 parallels MOS FET IRFP7430 (Qg=300nC , Rdson=1mOhm) on each branch (also open to suggestions).
If I use a 2A gate driver, it gives me a 0.6µs theorical opening/closing time. Any idea if it is enough (considering the 10µs period) ?
If it is good, I can choose a PWM controller with 2A drivers, if not, I can choose small ones and add external drivers.

Also perhaps 1st prototype will use standard diode rectifier, but for future efficiency improvements, I would prefer to use mosfet rectifier ready components (i.e. a pwm controller with sync output pin).

So, the main questions are : Which PWM controller do you recommend ? And is there useful surrounding components ?

Any advice is welcome.

Thank you.

Adrien

  • 0
    TI__Mastermind 25570 points
    Hello Adrien
    The first thing is to try to decide on the correct topology to use. Normally at this power level and for these low input voltages we would suggest a Push-Pull circuit. Diode rectification is actually a good choice for a 350Vdc output. The Vf of the diodes would be about 1.5V (if you use Sic) and if you use a full bridge rectifier you will lose about 3V across them - this is less than 1% of the output power and at 3.2A each diode would dissipate about 1.5*3.2*0.5 or 2.4W - which is manageable. If you decide to use SRs you would need to use 600V parts with an Rds on of less than 240mOhms just to break even and neglecting gate drive losses. I'm not saying that HV SR is not worthwhile but the gains are less than they would be at 12V out for example.

    The controller will use the feedback network to regulate Vout at its setpoint of 350V. You will need to choose the transformer turns ratio of course to give you reasonable duty cycles.

    I can't really comment on your choice of a particular MOSFET - there are a number of very good MOSFET manufacturers in the market. TI offers a range of low voltage MOSFETs with low gate charge - which will help speed up the turn-on and turn-off process,. The product range is at www.ti.com/.../products.html The best approach is to design the power stage first then start looking for a MOSFET that meets the requirements.

    Many controllers have relatively weak outputs which are intended as input signals to external MOSFET drivers. This has some advantages because you can easily find 4A drivers and you can place the drivers nearer to the MOSFETs. If you are paralleling MOSFETs it's vital to use separate gate drive resistors - this will be recommended by the MOSFET manufacturer to prevent destructive high frequency gate oscillations. TI UCC27524A1 is a MOSFET driver with 5A peak current for example.

    You may find our reference design at www.ti.com/.../tiduai7.pdf of interest. It is a bidirectional converter at 300W. The controller operates it as a Phase Shifted Full bridge in one direction (400V to 12V) and as a Push-Pull in the other direction (12V to 400V). Control is via a microcontroller which isn't your first choice but the power stage may be of interest.

    As regards controllers - you need to decide whether to use current mode control or voltage mode control. Normally current mode control is a better choice because you get cycle-by cycle current limiting protection. Then the choice would be to use newer devices rather than older ones.

    As you discovered, TI has a very large product portfolio and it's sometimes difficult to know where to start looking.
    Anyhow, I did a quick search using our product selector www.ti.com/.../products.html which returned 38 devices. The lowest cost 8 pin SOIC part was UCC28085 - so that is where I would start. If you needed some added functionality the UC3525A is in a 16 pin package and appears at the top of the list and is also on your list.

    Hope this helps, please let us know if you need any more information.
    Regards
    Colin
  • 0 in reply to Colin Gillmor
    Prodigy 105 points

    Hello Colin,

    Thanks a lot for your detail answers.

    I have ever calculated these advantages and drawbacks using diode or FET rectifier with these high voltages. Thanks, you confirm my result. Therefore, I will start with diode and perhaps try use MOSFET when I will improve efficiencies.

    I didn't know separate gate resistors was mandatory for parallels MOSFET. Thanks.

    I would like to use current mode control.
    Question, I don't know how to determine the maximum current to set, is it only the continuous input current (i.e. 110 A) ? Or is it higher ?

    Best regards,

    Adrien

  • 0 in reply to Adrien Zancan
    TI__Mastermind 25570 points
    Hello Adrien

    Current mode control usually controls the peak current in the primary side switches. In your case that would be 110A plus the output inductor current ripple (reflected through the transformer).

    As regards current sensing. The best options are a Current Transformer or a Hall Effect sensor. A resistive sensor will dissipate a lot of energy.

    You can find a lot of material at www.ti.com/.../login.shtml which covers all aspects of PSU design. the topic at www.ti.com/.../slup114.pdf covers current sensing in some detail, including current transformers and hall effect sensors.
    Regards
    Colin
  • 0 in reply to Colin Gillmor
    Prodigy 105 points

    Thanks a lot for your perfect helpful answers !