• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 62 subscribers
  • Views 529 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Variable output regulator

Marcel Kica Rosie
Prodigy 20 points

Other Parts Discussed in Thread: LM2679

I trying to build variable output switching regulator, with the folowing specification:

Vin >=38V
Vout (Reg min):Vin*0.8
Iout 0:10A
Output ripple is not a problem
Efficiency is of some concern (>85%)
(*) Starting in pre-biased conditions
(*) Working with infinite output capacity
(*) Short circuit protection

At first i tried the TPS5085 COT PMOSFET controler (limited at 2A), with variable FB divider and fixed external compensation, however i could not get it to perform ok, due to variable frequency vs output voltage and perhaps needing variable compensation.

Second, i tried the LM2679 (limited at 5A), however, during prototyping i blown about several parts, when i found that they do not support mOhm short-circuit, starting in a very high current output situation (battery charging) or working in pre-biased output conditions. -> new requirements (*)

Now, after looking at the parametric table, am i again looking again at the LM267x family and i planing to:
- parallel 2 (to get to 10Amps) by slaving one, with current injection, based on input.
- implement external output current limiting on each one.
- implement external power-down when current limiting does not work.

What other part(s) could i use, and what control scheme do you recommend ?

Thank you,
Marcel.

  • 0
    TI__Guru**** 191445 points
    What is your maximum Vin?
    With Vout minimum = 30.4 V, you can probably expect pretty high efficiceny.
    Most devices will start into pre-bias. From there it depends on what you expect to happen. It is ok for the output to discharge to zero or does it have to start up from the pre-bias voltage?
    Most will have internal current limiting, but many will be externally adjustable.
    Wha do you mean by "infinite output capacity? Generally nothing real is "infinite.
    Short circuit protection is generally part of Outpu current limit. Do you need pulse by pulse, hiccup mode or latched off?
    Generally you cannot parallel devices (some exceptions to this). Best to pick a device rated for your needs.Let me know more details.
  • 0 in reply to JohnTucker
    Prodigy 20 points
    Vin is fixed at nominal 38V +/- 5%.
    The output should be variable between say 1-2V and >30V if possible.
    Efficiency wise, at 19V and 10A out, a top PMOSFET and a bottom diode would not be that efficient, that's why i mentioned it.
    It's ok for the output to discharge to zero.

    The LM part i tested in multiple configurations, at 5A, including pure resistive, and motor output, where it worked flawlessly.
    However, when connected to a generator (as a battery charger), things did not go so great in the end. Infinite capacity at startup in case, means a 200Ah Lead-Acid battery, connected with ~AWG9 wiring. Starting with with Vout = Vbat+50mV (12.6V nominal), ended in smoke in 1 second.
    I think pulse by pulse current limitting would be ok. However, short-circuit protection should be available. For the LM part, which features pulse-by-pulse, i found in the datasheet of the new LM2* parts, that the SOA requires that during a shortcircuit, Vout to be >0.4V, which is not possible in my case.

    I think about paralleling, because i have not yet found a better part.

    Thank you,
    Marcel.