I know this is a little off topic, but people in this forum seem quite knowledgeable about the construction of DC-DC converters, so please, please don't roast me =)
I need to build a dc-dc converter that takes an input that constantly varies from ~18-24v and outputs 12vdc at ~12 amps continuous. The voltage may go transiently go to 36v or it may drop to 0v at times. Below an input of 14v or so it should shut off and it should also have over current/thermal protection if the current gets too high. The input voltage will be 18-24v most of the time. Precise regulation of the output voltage is not important.
Efficiency of this circuit is the most important to me, followed by low part count, then cost. It also needs to practically design itself since I'm pretty unskilled when it comes to circuit design=)
My current plan is to use a few LM2677's in parallel with a 555 timer as suggested by "n-channel": here:
http://www.diyaudio.com/forums/showthread.php?postid=167812#post167812
Does anyone have any better suggestions for IC's? Something more efficient or less complex? How about whole circuits that are ready to go?
I need to build a dc-dc converter that takes an input that constantly varies from ~18-24v and outputs 12vdc at ~12 amps continuous. The voltage may go transiently go to 36v or it may drop to 0v at times. Below an input of 14v or so it should shut off and it should also have over current/thermal protection if the current gets too high. The input voltage will be 18-24v most of the time. Precise regulation of the output voltage is not important.
Efficiency of this circuit is the most important to me, followed by low part count, then cost. It also needs to practically design itself since I'm pretty unskilled when it comes to circuit design=)
My current plan is to use a few LM2677's in parallel with a 555 timer as suggested by "n-channel": here:
http://www.diyaudio.com/forums/showthread.php?postid=167812#post167812
Does anyone have any better suggestions for IC's? Something more efficient or less complex? How about whole circuits that are ready to go?
Hello
I did a google search: http://www.google.com/search?hl=da&q="Synchronous+Buck+Controller"++40V&lr=
I'm sure you can find something that will work.
\Jens
I did a google search: http://www.google.com/search?hl=da&q="Synchronous+Buck+Controller"++40V&lr=
I'm sure you can find something that will work.
\Jens
Hi,
This looks like it could be ok in your application:
http://focus.ti.com/lit/ds/symlink/tps40057.pdf
When you do the layout, be sure to seperate the noisy part of the circuit from the feedback part.
That will secure proper circuit operation.
Best of luck
\Jens
This looks like it could be ok in your application:
http://focus.ti.com/lit/ds/symlink/tps40057.pdf
When you do the layout, be sure to seperate the noisy part of the circuit from the feedback part.
That will secure proper circuit operation.
Best of luck
\Jens
Thanks Guys,
stef1777:
These converters and others I've seen like them have a minimum voltage of 36v and 48v is needed to power up. I need a lower minimum like 14v.
JensRasmussen:
The tps40057 looks good, even though it's more complicated than I had hoped.
Any suggestions for an efficient switching mosfet to pair with this chip?
stef1777:
These converters and others I've seen like them have a minimum voltage of 36v and 48v is needed to power up. I need a lower minimum like 14v.
JensRasmussen:
The tps40057 looks good, even though it's more complicated than I had hoped.
Any suggestions for an efficient switching mosfet to pair with this chip?
pulexirritans said:Any suggestions for an efficient switching mosfet to pair with this chip?
Hi,
I would go for something you can hand solder so Dpack or D2pack type mosfets is a place to start.
The TOP mosfet (switches from Vin) may be a single mosfet for this converter, but for the bottom mosfet (Active diode) it might be a good idea to parallel two devices to get the conduction losses down.
I would look at 30V-40V N-channel mosfets from:
Onsemi
Vishay
IR
Fairchild
ST
Low Qg and Low Rds is what you want.
What about inductor?
Switching frequency?
Remember that the ripple current in the inductor only is related to L, Vin, Vout and Fswitch.
In and output capacitor(s) (Ceramics are great here)
Some of the things to consider
There is no easy way to convert 100+ W with little loss
\\\Jens
JensRasmussen said:
Some of the things to consider
\\\Jens
And one of the most important -> PCB layout.
But the best and simple idea is to use 3 integrated step-down regulator as the LM2679 in parallel. This kind of chip need less external components and some can be shared.
.
stef1777 said:
But the best and simple idea is to use 3 integrated step-down regulator as the LM2679 in parallel. This kind of chip need less external components and some can be shared.
.
Not the cheapest or best way IMO, but I guess it's something everyone can get to work.
Does the LM2679 support ceramic output capacitors?
\\\Jens
JensRasmussen said:
Does the LM2679 support ceramic output capacitors?
\\\Jens
I don't think so. Must use OSCON or very low ESR tantalum.
Maybe others models will work with ceramic but not sure that something exist for the needed 12A...
Paralleling seems critical to design and need to be tested for reliability on a long term use...
.
stef1777 said:May be a solution: 2 x PTN78020W + UCC39002 (or with 2 diodes at output)
See TI application note SLUA270A.
.
Wow. I think this is exactly what I am looking for. Super simple and very efficient.
Thanks Stef1777!
-Nick
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Power Supplies
- Need help w/ efficient DC-DC switching converter