Help with 12/24V input, 14.1V output DC/DC Converter

[Sink0]

Hi, i need to designe a DC/DC Converter that would work either with 12V and 24V, and have an Output of 14.1V and 0.5A. I dont need to cover all the range. Just both extremes. I was tring to designe a SEPIC converter but i cant find any IC with reasonable avaibility here on my country (Brazil). I was trying to designe a step-up/Down using MC34063 and fowllowing this AN: https://www.onsemi.com/pub_link/Collateral/AN920-D.PDF but the big range create too high peak currents. Any one have any idea how i can designe such converter. I am out of options. I was thinking on using both a step up, and a step down on the same board, and a jumper would chose which one would be on operation, but that wouls be very dumb...

Any sugestion please?

 

[memarian]

Hi, i need to designe a DC/DC Converter that would work either with 12V and 24V, and have an Output of 14.1V and 0.5A. I dont need to cover all the range. Just both extremes. I was tring to designe a SEPIC converter but i cant find any IC with reasonable avaibility here on my country (Brazil). I was trying to designe a step-up/Down using MC34063 and fowllowing this AN: https://www.onsemi.com/pub_link/Collateral/AN920-D.PDF but the big range create too high peak currents. Any one have any idea how i can designe such converter. I am out of options. I was thinking on using both a step up, and a step down on the same board, and a jumper would chose which one would be on operation, but that wouls be very dumb...

Any sugestion please?

Hi,

This might be an dumb idea but : Try to convert your 12~24V Range to a voltage Like 10V (Step-down) or 28V (Step-UP) and use the second switching regulator for converting that voltage to 14.1 (step-up or Step-down , depend on your first switching regulator).

 

[Sink0]

Would that work? I think that would have less eficiency (but more pactical).

Any other idea?

Thank you!

 

[tantudaisu]

What about buck-boost?
Try to check Linear Technology parts. i.e. LTM4607

 

[mvs sarma]

the design of buck/boost in AN920 seems OK. Logically you had a good application note, I missed it some how. Entire design details are given. compute , arrive at proper values and have a pcb. bread-boarding, definitely lands you in trouble.

but you need to optimize the inductor values etc correctly. otherwise it wouldn't work. so, perhaps you need a inductance measuring tester.
you might home-brew one if you can't lay hands on such tester from your labs.

 

[Sink0]

But again. I have problems with the peak current. The inductor itself will be manufactured by a company, so i belive it will have the indutance that i will specify. The output dont need to be 14.1V matched, dont need to have a good regulation, and neither a 90% eficiency. I just need a cheap and with a reasonable simple designe solution. This converter is JUST to the board can operate from a 12V and from a 24V power suply with no problems, and nothing more. No special requirements. The output itself will be further regulated.

 

[mvs sarma]

short circuit protection may not be possible. peak current can be looked after by the design. you may give it a try. coil can be tried for prototype using some cheap toroid like the ones used in power line spike suppressors.

MC34063A has become relatively old and i don't know whether it is recommended for new designs, if meant for production. I do agree that some modern regulators have come. But designing around national's LM25xx is also not that easy.

 

[Sink0]

short circuit protection may not be possible. peak current can be looked after by the design. you may give it a try. coil can be tried for prototype using some cheap toroid like the ones used in power line spike suppressors.

MC34063A has become relatively old and i don't know whether it is recommended for new designs, if meant for production. I do agree that some modern regulators have come. But designing around national's LM25xx is also not that easy.

Why do you mean that short circuit protection is not possible if there is the Isen pin for that? I could not understand what you meant about the peak current?

I know it is very old, probably thats the reason i can find it easily here. But if i just follow the National's Workbench design will i have problems? And why do you say it is not that easy?

Thank you!

 

[mvs sarma]

i know of some people who had problems designing with LN25xx
while theoretical designs appear to work, the artwork and component placements need to be tough and we have top follow strictly the plans of national.

As you alreadystarted with MC34063A, and your need is 500mA, please go ahead and let us see. It might warrant another transistor at the output stage also.

 

[Sink0]

Its not about "go ahead". It is about to find a solution. If the company accept to import the IC i will probably be using a LM3488 with the Workbench design. But i was looking for some more simple solution....