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.

Help me design an isolated dc-dc converter of 106mW

Status
Not open for further replies.

devgur

Newbie level 2
Joined
Jul 28, 2002
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
18
how to get more milliamps out of an ltc1771

I have 16V 8mA at my disposal I want to design Isolated dc-dc converter of 106mW with maximum efficiency.

can any body help me?
 

First: Welcome to elektroda! This forum is a nice collection of EEs and other people involved with electronics, and we try to help each others here :) (And please remember to read rules and announcements so you get the full picture of the stuff here. Your first post seems to be correctly placed - not like some other newbies manage) :wink:

About your question:
I do not want to go deeply to any supplier's specific chips and applications, just give an example of one part, which might be close to your needs. I rather want to give some general advice what to look at:

1) The steady state consumtion of regulator chip is of course one essential parameter on so low power budget. So Be careful about the data there - because most of the chips are optimized for much higher power levels.

2)You need to see efficiency on very light loads, and for micropower regulator a bit high input voltage. The performance for your application is not always easily readable on the data sheet - so you should be prepared to test several parts before final desicion.

3) I guess you need a voltage lower than the input, and only one voltage, non-isolated. In that case buck regulators tend to be the best and simplest to implement. If you need step-up, inverted, or isolated topology, you have thougher task to have ultra low power and high efficiency!

4) In case the chip needs an external switch (for example Linear Tech LTC1771), be aware that the gate (or base) drive of the switch transistor needs some power, too! For a MosFet low gate capacitance, and, if possible, reduced operating frequency, helps. LTC1771 data sheet says for example that 80% eficiency should be feasible on a few mA loads and 15V in. But, it would perform much better on lower input voltages, and unfortunately you are stuck with a relatively high fixed operating frequency if selecting that part.

Good luck,

Ted
 

it is not so easy to build such a dc-dc converter.in gernaral,a 150mw isolated only have a efficiency of about 50% or less.you can use a low power consumer controller,because the input voltage is only 16v,so yo can use a internal mosfet switch.select chips thant have a mode of :skip mode or burst mode.and use a high eficiency topology,such as a forward converter with syn rectifier.i think you can get a 80-90% eficiency.
chips from linear.com and maxim-ic.com.they have good chips.
good luck.
 

Well, here we have two conditions:

1. 8 mA max primary power supply current from 16V source (P_in max = 128 mW),
2. Output power P_out = 150 mW.

Hence the expected efficiency should be 117% :wink:

Anyway, the low average current of 8 ma means some have to put quite large capacitor on the input of the SMPS to accomodate for large startup currents. Then the design of the inductive components have to be optimized in term of optimal switching frequency for the given ferite and maximuum inductance for the given power load. Naturally MOSFETS should be used both for the primary swithes and for the secondary rectifiers. Then all the standard tricks of designing the snubber networks, proper sectionning of the primary transformer and usage of low ESR capacitors should be implemented. On the topology side some should think of zerro current crossing to reduce the swithing losses.
Naturally the SMPS controller have to be CMOS type, unless some would like to spend all the 100 mW power just on it's consumption ;) In any case it'll be miracle to expect any efficiency better than say 70% for this low power, unless unregulated power is admissable.
 

I have tried cmos oscillator (50% duty cycle) & transformer . But I am not getting good regulation. Efficiency is 75% I want more . Dose anybody has solution
Devgur
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top