Need help with DC-DC conversion

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Post Reply
chessman
Posts: 292
Joined: Tue Jan 14, 2003 1:01 am
Location: Issaquah, WA
Contact:

Need help with DC-DC conversion

Post by chessman » Thu Jul 24, 2003 10:32 am

I need to convert 12VDC to +/-56VDC at 8A.<p>I can't afford an off-the-shelf product, but I have plenty of parts around, and I can buy small stuff like transformers.<p>thanks! :cool:

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: Need help with DC-DC conversion

Post by Chris Smith » Thu Jul 24, 2003 11:22 am

First you need to change the DC into a wave form like square wave pulsed DC using an oscillator like a 555. The right frequency for the right transformer is essential for maximum saturation and conversion. <p>An Adjustable frequency oscillator will ensure you can tweak it to the max performance. <p>Then you send that signal into a Transformer that steps up the voltage and then back through a rectifier into DC. <p>The other method is the buck boost method which uses more ICs and a choke to do pretty much the same thing.

chessman
Posts: 292
Joined: Tue Jan 14, 2003 1:01 am
Location: Issaquah, WA
Contact:

Re: Need help with DC-DC conversion

Post by chessman » Thu Jul 24, 2003 11:47 am

Well I was figuring a 555 timer to generate the pulse signal, and then 2 MOSFETS in parallel (would this work????) to switch the 12VDC through the transformer...<p>I mean, I understand the whole concept of SMPS...but what worries me is the 350W capacity that I need to cover.

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: Need help with DC-DC conversion

Post by Chris Smith » Thu Jul 24, 2003 4:22 pm

The way to handle that is a 4013? [and 555] from memory that switches the fets,... one on, the other immediately off in a synchronized flip flop fashion. <p>I’ll email you a Laser driver circuit that drives a laser tube, not diode, and you can beef it up with larger FETS and Transformer.<p>I have to look through about 5000 jpegs, with the search engine, but I think I know which hard drive its on?<p>[ July 24, 2003: Message edited by: Chris Smith ]</p>

bodgy
Posts: 1044
Joined: Tue Dec 04, 2001 1:01 am
Location: Australia
Contact:

Re: Need help with DC-DC conversion

Post by bodgy » Thu Jul 24, 2003 4:44 pm

Have a look at the application notes for the Maxim MAX1605 this produces 75v @6mA without amplification or the MAX5021<p>Or pop over to the Zetex site where they have a high voltage inverter app note using their low power version of the 555 - designed for luminescant display driving, There is also an app note using some of their transistors for a compact neon inverter.<p>Colin
On a clear disk you can seek forever.

russlk
Posts: 563
Joined: Thu Dec 06, 2001 1:01 am
Location: New Hampshire
Contact:

Re: Need help with DC-DC conversion

Post by russlk » Thu Jul 24, 2003 5:33 pm

The LM3524 is designed for this application and National Semiconductor has good application notes. Check out www.nsc.com.

rshayes
Posts: 1286
Joined: Tue Mar 04, 2003 1:01 am
Contact:

Re: Need help with DC-DC conversion

Post by rshayes » Fri Jul 25, 2003 3:01 am

Your first problem will be the switching devices. If the 8 amps is drawn from only one output at a time, you are talking about 448 watts. Double this to 896 watts if both outputs are simultaneously 8 amps. Assume 90% efficiency, and the input current becomes either 41 or 82 amps. If power FETs are used, and a 5% drop is allowed in Rds, then the on resistance would have to be below .014 or .007 ohms. These may be large and expensive devices.<p>The large devices and high currents may force lower operating frequencies. I don't think that this will be a 100 KHz design.<p>If regulation can be avoided, the old push-pull Royer converters using bipolar transistors might be a usable approach. A saturating drive transformer with a non-saturating power transformer might be the best solution. <p>Winding the transformer might be interesting no matter what design is chosen. It will probably involve paralleling several #12 or #14 wires, since larger wire is very hard to bend, even in copper.

bruinbear714
Posts: 26
Joined: Thu Apr 17, 2003 1:01 am
Contact:

Re: Need help with DC-DC conversion

Post by bruinbear714 » Fri Jul 25, 2003 9:32 am

Stephen pretty much hit the nail on the head.<p>To generate the PWM needed to drive the FETS, I would look into a TL494, or SG3525 IC. Easy to setup, easy to use, and easy to interface. Since you will most likely parallel a few FETS or use a big FET, I would look into a gate driver like an IR2110 to drive the FETS so you would have a two chip solution instead of mucking around with additional circuitry.<p>I recently designed a push-pull step up converter with 12V in, +-48V out for a car amplifier capable of supplying 5A continuously using the SG3525. If you need more help, contact me and I'll see what I can do.

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: Need help with DC-DC conversion

Post by Chris Smith » Fri Jul 25, 2003 1:44 pm

Your email doesnt work?<p>Sorry, you have hot mail, and my sketch is twice too big. Ill have to re scan, or shrink the bytes?<p>NOPE, no email going through?<p>[ July 25, 2003: Message edited by: Chris Smith ]<p>[ July 25, 2003: Message edited by: Chris Smith ]</p>

ampedtech
Posts: 22
Joined: Thu Aug 01, 2002 1:01 am
Location: earth (for now)
Contact:

Re: Need help with DC-DC conversion

Post by ampedtech » Sun Aug 24, 2003 10:11 am

Chris Smith writes:
“First you need to change the DC into a wave form like square wave pulsed DC using an oscillator like a 555. The right frequency for the right transformer is essential for maximum saturation and conversion. “<p>I’m kind of rusty here so this is for my clarification. I agree that a xformer needs the right freq. to achieve maximum saturation and conversion.
My question is:
What about slew rate of a square wave?
Would not the rise time of square wave appear to an inductor as a MUCH higher freq. than that of the freq. of the repetitions of the square wave?

User avatar
Joseph
Posts: 681
Joined: Mon Dec 10, 2001 1:01 am
Location: USA,World
Contact:

Re: Need help with DC-DC conversion

Post by Joseph » Sun Aug 24, 2003 3:31 pm

Electronic Goldmine has BUZ110s mosfets for 2/$1.00 They are rated Rds of about .014 ohm. You could convert the universal SMPS design to work on 12 volts. It uses a large ferrite toroid transformer which is easier to wind with many short windings than an e-core. Since the voltage on the mosfet totem pole would only be 12VDC you would not have to worry about high DV/DT shorting the mosfets out. At that low voltage, you should be able to parallel many mosfets so long as you use a 3.9 ohm gate resistor for each one. It is a doable project, but will take some effort and considerable learning. You would have 2 big advantages over the circuit which runs off of 120vac. One is that you do not have to deal with converting the input power to DC, you already have 12v. Two is that you already have 12v DC to power the IR2113 mosfet driver and the control circuitry.
http://groups.yahoo.com/group/switchmode/<p>[ August 24, 2003: Message edited by: Joseph Meisenhelder ]</p>

User avatar
Joseph
Posts: 681
Joined: Mon Dec 10, 2001 1:01 am
Location: USA,World
Contact:

Re: Need help with DC-DC conversion

Post by Joseph » Mon Aug 25, 2003 2:38 am

I have added a 12v to DC convverter to the Universal SMPS files. This type of project is much more in tune with most hobbyists who do not want to work with high voltages. With the IRF7343 buffers on the outputs of the IR2113, I see no practical mosfet gate drive limit for the circuit. Since the low voltage mosfets have a lower gate threshold voltage rating, it may be a good idea to add 10pF capacitors at the inputs of the IR2113 and adjust the values of the turn-on delay resistors. With the present 100K resistors, the dead-time is about 1.5uS. It may need to be increased to 2uS. Alternately, you could just increase the resistors to 150K without the caps. That way is how I would try. CMOS reliably has input capacitances of about 5pF.

Post Reply

Who is online

Users browsing this forum: CeaSaR and 28 guests