Dual-output switching regulator, wide input range

Go To Last Post
5 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

We are looking for a small switching regulator, dual-output.  Hopefully reasonable cost.

 

Input V:  Nominal 24VDC from 2x 12V SLA batteries.  Say 36VDC max Vin in case hooked up to charger.  Min Vin like maybe 15V when dead?

 

Output1:  A few mA to drive a small AVR.  Say, 10mA-20mA at 3V-5V.

Output2:  About 100mA min at 8VDC.

 

I know you sparkies have discussed switching regulators with one fixed output (e.g. 3.3V/5V) and one variable.  But I can't seem to find the thread(s or models...  Do y'all remember?

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Errrmmmm... Linear have a bunch of multi output switchmode regulators, but they aren't known for "reasonable cost". The lt3509 seems to match your specs. An alternative would be a switching regulator with an LDO for the 3.3V/5V rail - lt3645 is a bit cheaper than the lt3509. I'm sure you can find others from the usual suspects (analog, linear, maxim, ti) via their parametric search.

 

- S

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks, Stephen.  Indeed, the current (pun intended) design has two regulators, with a separate "downstream" regulator for the micro supply.

 

The current switcher seems to get upset when there is a jolt in demand level, so power supply is being re-examined.  In addition there is a need for a very small board, and given that I remembered some positive discussion of some dual-output regulators in the past, I an exploring trying to save a chip for board space reasons.

 

I did find several at Linear, including the one you mentioned.  Just askin'...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
The current switcher seems to get upset when there is a jolt in demand level, so power supply is being re-examined.
Might be able to adjust the compensation for the SMPS's error amplifier that's in the unstable control loop.

A common stimulus is a step input or stepped load; appear to already have a good empirical stepped load.

It'll be more complete to measure the gain margins and phase margins via power supply stability tests at expected, and then some, source and load impedances.

Also likely reevaluate if or when replace the SLA battery with a nickel battery.

theusch wrote:
In addition there is a need for a very small board, ...
There are some very small off-the-shelf SMPS DC-DC modules (mini impedance controlled PCB, PWM controller IC, inductor, maybe capacitors) that might patch into the current PCB for a quick test.

A possible challenge is the 3 or 4-to-1 drop (24V-to-8V).

These very small SMPSs appear to have a 2MHz or greater switching frequency that seems to use effective internal compensation.

 

Ref.

Texas Instruments

E2E Community

Blogs

Power House

Testing power supply: Measuring stability

by

Apr 24, 2013

http://e2e.ti.com/blogs_/b/powerhouse/archive/2013/04/23/testing-power-supply-measuring-stability

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

One option is an isolated, dual-output buck converter (Google "flybuck"), basically a buck converter with a transformer instead of the coil.  Several options from TI, Linear, Fairchild, etc.

You can pull the 2 outputs off the same transformer, each one from its own secondary winding, with just a diode and a capacitor as output.  The output voltages are calculated as the transformer turns ratio, with values adjusted to offset the voltage drop at the output diode (usually 0.3V-0.7V)

The only challenge would be ordering a custom transformer.  There are vendors like Midcom ( http://www.we-online.com/web/en/... ) that can source samples from a few specs like inductance of primary, max DC current, turns ratio and #of outputs.

 

 

while(!solution) {patience--;}