Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

SMPS with controller IC too inefficient

Status
Not open for further replies.
potcore

potcore

Electrical
Feb 26, 2007
28
Dear Engineers,

I am interested in re-designing an existing SMPS to make it more efficient and hopefully simpler to produce.

The existing SMPS runs off a battery and converts 8-30V DC to 5.6V DC. It is a non-isolated buck converter. There is no controller IC. -The control circuitry is all discrete and this is allegedly because using an SMPS controller IC would be less efficient than the discrete design.

I can't say too much but the power switch is a PNP transistor -again this is said to produce more efficiency than using a MOSFET. I haven't been shown too much but there was certainly a schmitt trigger in there using BJT's and a short circuit protection network using BJT's.

It certainly did not appear to be a quasi-resonant design and was reasonably simple.
Does any reader know of a simple Buck converter SMPS configuration thats very efficient and preferably fairly simple.
Output current is not huge...it just seems to be for some microcontroller chips etc.
 
Replies continue below

Recommended for you

Sort by date Sort by votes
Your question is academic at best. With nothing to compare to, how do you expect a cogent answer?


Moreover, efficiency at the component level is only marginally relevant. What is important is the OVERALL efficiency. Therefore, having some components being allegedly inefficient may not matter.

Note that student posting is not allowed.

TTFN

FAQ731-376


 
Upvote 0 Downvote
The NS is a good place to start. It allows you to explore different switcher topologies, to simulate and adjust component values to get best performance. It even helps with the magnetics and the BOM. Go there. Use it!

Gunnar Englund
--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...
 
Upvote 0 Downvote
Or Linear.com

You can download their Switcher Cad that has all their switching controllers and run simulations.

You can also buy demonstration boards with most of their switching controllers on them to test or see how they would work in your application.

They have 22 that would probably work.

Keith Cress
Flamin Systems, Inc.-
 
Upvote 0 Downvote
As I recently learned at my current job... some companies list a massive number of chips for any particular PS function, but when it comes to what can be readily sourced (i.e., Digi-Key or Mouser), Linear was always the clear winner by a mile. The selection isn't as vast, but they always have something to do what we need and they're in stock!

Dan - Owner
Footwell%20Animation%20Tiny.gif
 
Upvote 0 Downvote
Hi Dan.

Linear parts ALWAYS cost more than National parts but Yep, they work really well and are most often in stock.

I recently ordered one of the demo boards from them, (in fact for that battery charger). They were adamant that they check stock since their in-house(linear) sales department has no stock, that they check with the manufacturing stock,(across the street). They had 250 demo boards in stock!! I was impressed.

I think Digikey also really hammers you on low count Linear products.

Keith Cress
Flamin Systems, Inc.-
 
Upvote 0 Downvote
Yep, the cost is probably 25-40% more for a comparable Linear part, but in our business model that means squat... getting the product prototyped and ready for production is what counts for us. That was (is?) a tough thing for me to get used to as I've always been down the penny-pinching road of choosing a part that's $0.012 cheaper as it adds up over 10 million units. Now I'm supposed to let it go as it doesn't matter if a unit end up costing $20 more than it could, we're only making 50-100. Granted, we'll rent units out for $2,500/day that cost us $100 in parts rather than $80, so I can see the reasoning :)

Dan - Owner
Footwell%20Animation%20Tiny.gif
 
Upvote 0 Downvote
You want to improve the efficiency of your SMPS? Have you measured it? What is it? How does it compare with integrated solutions?
 
Upvote 0 Downvote
I recently undertook a project similar to this.
I wanted a white led driver to run from batteries and be as effecient as possible.
I build it and got a solid 90% across wide variation in battery voltage.
That said it isn't cheap.
Use a mosfet instead of bipolar switch. At battery voltages the Vce is a large power loss. The FET with a low Rds on will have significantly lower switch loss.
Then use schottky diodes for free wheeling purposes. And then use an oversize inductor and switch at a low frequency. This reduces power sw loss and magnetic loss.
Just watch the primary loss areas. Power Switch and Magnetics.

 
Upvote 0 Downvote
Hi, Thanks for your replies. There is just a suggestion in the company that they may want to re-design. I dont want to make any rash suggestions as its not really been in my department but it would be interesting if it could be made more efficient/more reliable/quicker & easier to build and test.

Cost is not the main issue here....volumes are small and reliability in its (possibly) relatively short life is more important, and of course efficiency.

The temperature in which this SMPS works can rise to above 150 degrees Centigrade. Due to its working environment, there is relatively little that can be done to cool it. I believe 150 degrees C is above most SMPS controller IC manufacturers' test data and so maybe thats another reason for using the discrete controller circuitry. -and sticking year after year with a design that has worked (i believe this SMPS was designed some 20 years ago). -It would be very expensive indeed and inconvenient if this SMPS failed in service.
 
Upvote 0 Downvote
Is also high acceleration, vibration or shock involved?

Gunnar Englund
--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...
 
Upvote 0 Downvote
Well discreet does make more sense now.

That will certainly limit options.
You have never stated the important requirement: Current???

The LT1936H is probably your friend.

Buy a reference design and give it a whirl. You may need to replace a few parts for higher temp operation. Or just try it and see if it works. You may need to change some of the other components for the higher temps for your fielded design.


Keith Cress
Flamin Systems, Inc.-
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

sbkenn
  • Locked
  • Question
Stepper controller
Replies
2
Views
89
sbkenn
sbkenn
jmacion
  • Locked
  • Question
relay wont close consistently with open collector input
Replies
16
Views
692
hacksaw
hacksaw
Total Novice
  • Locked
  • Question
How to calculate the EMF in a circuit with 2 opposing batteries? 2
2 3
Replies
41
Views
2K
Total Novice
Total Novice
RyreInc
  • Locked
  • Question
Three single-phase SSRs in place of one three-phase, with zero-cross
Replies
1
Views
204
k75ks
k75ks
waross
  • Locked
  • Question
A problem with a motion detector.
Replies
14
Views
145
waross
waross

Part and Inventory Search

Sponsor

Back
Top Bottom