Offline full-bridge SMPS - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 8th October 2010, 04:04 AM   #11
diyAudio Member
 
Sch3mat1c's Avatar
 
Join Date: Jan 2003
Location: Milwaukee, WI
Send a message via ICQ to Sch3mat1c Send a message via AIM to Sch3mat1c
Half bridge all the way. I've got 10kW pumping through a half bridge in my induction heater.

For power supply applications, where the output is constant voltage, bipolar, square wave, AC only, there is no advantage. There are two strong disadvantages: doubled switching AND conduction losses.

For some power supply applications, a full bridge has some advantage. For instance, a constant-current type bridge is easier to build in PP or full bridge, because a two-winding supply inductor is required for half bridge. If the supply voltage changes quickly, a lot of current will be spent slewing the filter capacitors (which includes the coupling capacitor, when caps to +V and -V are used in the half bridge, as I traditionally do).

There is only one strong advantage I see for full bridge, which is useful where a DC component is required, or where the output must be connected open circuit sometimes. Examples include DC welders, class D amplifiers, motors, electric vehicle drives, etc. In these cases, PWM can be fed to just one side of the 'H', while the other remains locked (in the low state, let's say). This costs double conduction losses, but saves switching on one side. Meanwhile, the output can be switched over a continuous range from +V to -V, without requiring a tiresome bipolar power supply. (A half bridge, in principle, needs a bipolar supply, but this can be provided for AC loads simply with a coupling capacitor.) If a "coasting" feature is required (this would be handy for electric vehicles, when regenerative breaking is not immediately necessary), both sides of the bridge can be turned off completely, leaving the load open circuit. (It will still drive power into the supply if the load voltage pushes above +V, because of the protection diodes; this could only possibly occur in an accelerating downhill case, which is unusual.) Likewise, a "braking" feature is available, by turning on both sides in the "low" state (effectively shorting out the load).

For AC type inverters, such as forward converters, resonant and etc., I don't see any advantage to H bridge. The filter caps aren't any different, you need the same total amount in either case. If you absolutely must use extra transistors, you can hook them in parallel, and get half, instead of double, the power loss!

Tim
__________________
Seven Transistor Labs
Projects and Resources

Last edited by Sch3mat1c; 8th October 2010 at 04:06 AM.
  Reply With Quote
Old 8th October 2010, 10:05 PM   #12
edfed is offline edfed  France
diyAudio Member
 
Join Date: Oct 2008
when you will meet HV problems around mosfets, and mainly, insulation problems, start to try this:
Click the image to open in full size.

the winding are only 10 turns, all the windings can be made around a single core. the core should be a ferrite core, the form doesn't matter a lot.
toroids or pots act the same.

the command is simple, use a buffer to drive inputs (A & B), the signals from sg3525 are enough, but better with a buffer like UCC27324.
the resistors are 10k 1/4W
the mosfets are N.

it works in a very large range.
Attached Images
File Type: bmp mosdriver.bmp (146.1 KB, 314 views)
  Reply With Quote
Old 9th October 2010, 12:00 AM   #13
diyAudio Member
 
Sch3mat1c's Avatar
 
Join Date: Jan 2003
Location: Milwaukee, WI
Send a message via ICQ to Sch3mat1c Send a message via AIM to Sch3mat1c
Yuck, no series gate resistors? Good way to dramatically increase switching loss.

The parallel resistors do nothing.

Tim
__________________
Seven Transistor Labs
Projects and Resources
  Reply With Quote
Old 11th October 2010, 03:45 PM   #14
diyAudio Member
 
Join Date: Sep 2010
Nice to read so many experienced people on this complex topic

Does anyone know a good information source for building the first smps?
For example, what's an "easy" controller to start with?
What topology i should use?

Is there a well documentated project anywhere to use as example?

Is +/-40V, 5A (per output) a possible target for the first try? (would fit for my 5xTDA7293-Amp)
  Reply With Quote
Old 12th October 2010, 03:22 AM   #15
diyAudio Member
 
alexberg's Avatar
 
Join Date: Oct 2009
To Sch3mat1c,
what controls the output of HB, particularily when both switches are off? Parasitics...
Really poor regulation at light loads
HB output voltage is half of that of FB.
what is the benefit of double current flowing through the half resistance vs the current flowing through the double resistance?
DC bus capacitor ripples are unequal too...
  Reply With Quote
Old 12th October 2010, 04:07 PM   #16
diyAudio Member
 
Sch3mat1c's Avatar
 
Join Date: Jan 2003
Location: Milwaukee, WI
Send a message via ICQ to Sch3mat1c Send a message via AIM to Sch3mat1c
1. In the off state, the transformer voltage is determined by the output transformer and choke (assuming forward converter topology, continuous current). When off, the transformer is clamped to ~0V, at the secondary side, by the choke current. This causes ringing current at the secondary, and ringing voltage at the primary, due to leakage inductance and parasitic capacitance; this is easily damped with an R+C across the primary. Voltage returns asymptotically to zero.

When current is discontinuous (due to compact design, or at light load), the voltage is defined only by the OPT impedance. This is a finite impedance, so the voltage is well defined.

If, instead, one were to drive the transformer with a constant voltage at all times (like a phase shift H bridge), one would get ringing current at the primary and ringing voltage at the secondary. This is commonly seen in the rising edge of a half-bridge forward converter. I suppose this might be disadvantageous because the ringing causes the recitifer diodes to switch rapidly (at the resonant frequency, ~MHz?), which will cause excessive heating and RFI. Obviously, this is more of a problem for silicon junction diodes than it is for schottky.

2. I get excellent regulation at light loads. I don't know what problems you have had.

3. Output voltage is half, so you simply need half the primary turns. Obivously, you wind the transformer for the circuit.

4. For the same number of transistors, double current through halved resistance is equal to single current through double resistance. Something about "no free lunches".

5. Why would they be unequal?

Tim
__________________
Seven Transistor Labs
Projects and Resources
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Offline full-bridge SMPS....need help Nejc_car Power Supplies 1265 25th May 2014 02:20 PM


New To Site? Need Help?

All times are GMT. The time now is 04:39 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2