Page 1 of 1
An efficient way to "roughly" regulate power?
Posted: Mon Oct 20, 2008 12:03 am
I am looking for a small/reliable/efficient way to reduce input voltages
I have an application with a 52v battery (14 series lipo 51.8V) and I already have some regulators that give me the outputs that I need (5v 12v, etc), but I first need to reduce the input to at least around 30 volts...
Short of building yet another high voltage, fairly complex regulator, is there another efficient way to like reduce/divide the voltage down?
Posted: Mon Oct 20, 2008 3:19 am
I think you'll have to go with some kind of switching regulator. It could be a simple homebuilt PWM buck converter if you like.
One thing I would advise though is to have whatever first stage regulator you use drop the voltage as much as possible, maybe down to 15V, rather than 30V which I assume is the theoretical upper limit of your second stage regulators' inputs. I was trying to use some chips from Linear to buck-regulate down from 26V to 5V and I had no end of trouble with burning FETs and smoking regulators, even though it was supposedly in spec. When I used an input voltage of 15V instead, everything worked much better.
Posted: Mon Oct 20, 2008 8:46 am
Try heading to National.com, punch in your desired values in their simulator tool, and they'll give you a National buck converter plus all the associated parts around it.
This is simply too much of a voltage drop for linear regulators to handle due to power dissipation.
Posted: Mon Oct 20, 2008 9:14 am
Don't forget that when fully charged, your LiPos will be closer to 60V than 52.
Posted: Mon Oct 20, 2008 3:38 pm
Thanks for the replies.
I've got a Linar Technologies switching regulator that I'm using to power electronics on various projects. I'm trying to power all the electronics from one primary battery and the power supply could be anything from 7V to 60V. The Linear chip handles the lower ranges great, up to 36V but anything higher requires a pre-voltage reduction.
The efficiency concern is the biggest issue for me because I am pulling 7.5A (max) through this system. If I were to go through any linear regs, they would be smoking to say the least... and highly inefficient. The Linear chip wins hands down on the efficiency, so I'm trying to figure out a pre regulation system to drop the high voltages on some of these batteries without wasting too much power before handinng it off the the primary regulator.
Posted: Mon Oct 20, 2008 4:04 pm
whoa - 7.5A is going to be a real challenge. Here's a NatSemi appnote for a 60-7V input to 5V/7A switcher. Worth talking a look.
Posted: Tue Oct 21, 2008 9:16 am
Have you looked for a small off-the-shelf DC-DC converter that can do what you need? As philba said, designing a 7A efficient switcher isn't trivial and it sounds like this is peripheral to your project, so you probably have better things to do with your time.
Posted: Tue Oct 21, 2008 10:46 am
Why place the load on the full bank of batteries. Just take a tap off at the voltage you need(connect in between the batteries). Just make sure your monitoring the voltage for cutoff at this point as these batteries will discharge faster than the rest. You will also need to charge the batteries individually anyway. I would also plan on rotating the batteries every other charge to distribute the ware.
Posted: Tue Oct 21, 2008 12:37 pm
Maybe it would be worthwhile to tell us about the various loads you plan on having. This feels like we don't have enough information.
By the way, I don't think tapping from the middle of the battery stack would be a good idea as you want the LIPOs to discharge equally. Wouldn't that lead to overcharging or undercharging some of the batteries? Unless you are individually charging each battery.
Posted: Tue Oct 21, 2008 5:23 pm
Tapping the middle is actually pretty common in RC. Castle Creations makes a BEC that tells you to tap the battery in the middle. I think this is ok for applications that have minimal power draw. In RC planes/helis, its the motor that drains the battery, the lipos would power the reciever/servos for a month.
However if you did this as the primary load, you would very quickly end up with unbalanced lipo... not safe to say the least.
I have a two stage power supply that I am putting together for robotics use. The power supply a essentially a DC-DC version of a computer power supply providing 12v and 5v power rails for a slew of electronics at about 90 watts.
One of the applications is to use this on a robotic system that has a 14s lipo pack, and I cannot tap into the middle of the packs due to the high current draw.
My two stage dc-dc system takes an input up to 36v, so I'm needing to build a 3rd stage that drops HV systems down to 30v or so as efficiently as I can.
Looks like I need to just make another switching stage which is fine. Just was hoping to find a crude PWM based system or something but clearly with a 7.5A load, the cost of the filter caps (even at a high frequency) would be more than that of just putting together another high voltage stage.
I saw two solutions from Linear, LTC3703 and the LTC3810. I guess I'll just have to go back to looking at those options.