class d design - diyAudio
Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

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 6th December 2005, 04:26 PM   #1
diyAudio Member
 
Join Date: Nov 2005
Location: brasil
Unhappy class d design

Here's a simulation of a work in progress class-d design i'm working on...
It's pretty rough at the moment & takes an age to simulate, i'm keen to drive the thing directly from rectified mains to loose the big heavy transformer (or separate psu), so i've come up wth this full-bridge design that is at least interesting!
Some of the components are driven out of spec & some are perilously close to spec (i chose components, from the libs shipped with LTspice, for their timings rather than their absolute voltage & current ratings)
Click the image to open in full size.
here's the simulation LTspice simulation

the lack of discrete capacitors has me worried....
Be as critical as possible

cheers
Rob
Attached Files
File Type: zip rttclassd.zip (2.6 KB, 122 views)
  Reply With Quote
Old 6th December 2005, 06:17 PM   #2
BWRX is offline BWRX  United States
diyAudio Moderator Emeritus
 
Join Date: Jan 2005
Location: Pennsylvania
Well, for starters you'll have much higher than 120Vdc if you rectify 120Vac mains. Your supply rail would be more like 168Vdc and would fluctuate as mains voltage isn't always stable. I now see you simulated the mains voltage but have it labeled 110 and 120. You should change the labels and be sure to spec the parts for the higher voltages.

Other than that, you'll need freewheeling diodes across the output mosfets to conduct the inductor current during the dead time or you'll be replacing a lot of output mosfets. I don't know enough about the rest of the circuit to make any other suggestions, except to maybe use small resistors in series with the gates of the pnp driver bjt's to ensure current sharing?
__________________
Brian
  Reply With Quote
Old 6th December 2005, 06:35 PM   #3
Tim__x is offline Tim__x  Canada
diyAudio Member
 
Join Date: Jun 2004
Location: Edmonton area, Alberta
How much current is going through D3 and D7 on M2 and M4 turn-on?
  Reply With Quote
Old 6th December 2005, 08:14 PM   #4
Pano is offline Pano  United States
diyAudio Moderator
 
Pano's Avatar
 
Join Date: Oct 2004
Location: Milliways
Blog Entries: 4
Interesting. Why the bridged design? Is it so you can use a single 120V rail? As BWRX says, you'll get a lot more than 120V after retification. And you'll get some pretty big voltage swings if you don't regulate or use massive filter caps.

Nice simple design. How well do you think the oscillator circuit will work?
__________________
Take the Speaker Voltage Test!
  Reply With Quote
Old 6th December 2005, 11:46 PM   #5
kartino is offline kartino  Indonesia
diyAudio Member
 
kartino's Avatar
 
Join Date: Feb 2005
Location: Gunungkidul Indonesia
Hi buddy,

Be careful with directly rectified from mains.

Please pay attention that the speaker is not designed for connected directly to mains. Why? If you get your speaker blown or melted, the coil is very easy to meet the body of your speaker. Then you have speaker body connnected to mains. And if your speaker box is not so dry enough.... very dangerous!

Please use 35mA MCB with ground fault protection when you will play with mains. But it is still dangerous too..! You still have electric current flow accross your body....ie hand to hand with your hearts in between.

Please make sure what you are doing with mains....

br/tn
  Reply With Quote
Old 7th December 2005, 05:38 PM   #6
diyAudio Member
 
Join Date: Nov 2005
Location: brasil
@BWRX: Good point about it not being the ac voltage, doh! The circuit here is very badly labelled & specced, more a proof of concept than something to be commited to reality! I'll add the diodes, essential balancing resisitors too (& other niceties like decoupling the input, for the final design) I would like to be able to parallel the mosfets too, so that will require more balancing too, probably requiring separate drive circuits for each mosfet?

@tim_x: ltspice reckons nothing untoward at power-up assuming the power supply ramps up smoothly, the biggest worry is the cross conduction thru both mosfets at crossover, the simulation gets peaks of >200A, i hope cabling resistances & the PSU will control this. Maybe i could add some x-over delays too? The mosfets chosen are way underspecced but seemed an OK choice to get the simulation up & running.

@panomaniac: The full bridge design is specifically so that i can use a single rail, however i'm getting more tempted to add a PSU to produce a balanced regulated supply.. It might be that the extra (and fairly high spec) components needed for a mains rail PSU (and other problems) will justify a separate PSU, especially if i use an SMPS.
The oscillator is pretty triangular (using these opamps, i chose them as they seem to be pretty generic ultra-high speed types) I almost treat op-amps as discrete components In ltspice it looks OK at 1/2Mhz, with the RC's here it's only running at 50Khz, enough for bass usage.

@kartino: I'm living in country where it's a good day if you get less than 3 phases thru u when working near electrical equipment! Almost everywhere lacks a ground lead so having mutually grounded equipment can be more hazardous as leakage in one piece of equipment will electrify everything! I've learnt to wear rubber gloves & disconnect absolutely *everything* before doing work now!
I installed ELCB's on alot of the audio equipment here & couldn't get the damn things to stop cutting out! Fortunately the speakers themselves are all hidden out of reach inside wooden boxes as they are all horn loaded! I think the dangers of live speakers are evident even without a direct mains supply as most PA amplifiers are capable of producing well over 100v P-P.

cheers,
Rob
  Reply With Quote
Old 9th December 2005, 08:03 AM   #7
kartino is offline kartino  Indonesia
diyAudio Member
 
kartino's Avatar
 
Join Date: Feb 2005
Location: Gunungkidul Indonesia
Hi,

From simulation it looked that you need to improve your gate driver. It looked that your gate driver need faster and need delay. You see that there are hundreds amps crossing yous end fets.

Your gate also already start to ON before the opposite became OFF. That is beause you don't have charging delay or your gate too slow to discharge.

Your gate driver supply for low side still 120V, wow... are you sure???

Another, you need are floating ground like felix's for your input stage, isn't?

Also looked that you have problem with communication between input stage and gate driver, level shifter problem. In my opinion it shall have difference circuit between high and low side gate driver to make good connection with input stage.

Please see UCD based in this forum, on eof Chris's (classd4sure) is full bridge. But still need groud reference.

Good luck,
kartino
Attached Images
File Type: jpg simulation_oldroborg.jpg (71.0 KB, 591 views)
  Reply With Quote
Old 9th December 2005, 09:57 AM   #8
djk is offline djk
diyAudio Member
 
djk's Avatar
 
Join Date: Feb 2001
Location: USA
Hi Rob,

From 120VAC with just a bridge and a single 250F cap feeding a 100V switching regulator you can get 600W/8R from a handful of inexpensive parts.

T.I. makes a 1.7W PWM amplifier with THD rated at 0.0018% @1Khz that runs on 5V (TPA6204A1).

Using two pair of IRFP9140/140 and the above T.I. chip as a driver you're in business.

The T.I. part is about $0.50 in 1K lot, the IRF parts about $1 each or so.

With paralleled outputs for the P-channel parts, power can be up to 2KW or so at 2R. There are N-channel parts available that are heavy enough to not need paralleling.

I can surface mail someone a conceptual schematic (amp only, not the regulator) for scan and posting purposes.
  Reply With Quote
Old 9th December 2005, 02:09 PM   #9
diyAudio Member
 
Join Date: Nov 2005
Location: brasil
@kartino: yep i do need to look at the high current at crossover, i think a delay is the best thing here. The low-side driver was simply copied from the high-side for the sake of it! It shouldn't require a higher rail to switch, i made the high & low side drivers identical to (hopefully) extract identical performance from all the mosfets.
The level shifter is a bit simple too, but maybe it'll be OK!
The input stage requires a balanced supply too. The extra low power DC supplies will already be available from other circuitry that would be bundled inside the same case as the amp.
I think i need to look at alot more designs before i carry on with mine!

@djk: I've been looking for suitable integrated pwm amplifiers (there are too many damn IC's nowadays!), this one looks good, do u think it would work well down to 20hz? It seems to have been designed with mobile use in mind so maybe it has a fairly high low-frequency roll-off (i guess this would depend mainly on the circuitry to drive the chip.)
I feel like i'm reinventing the wheel designing my own class-d & my wheel is not getting particularly circular!

cheers
Rob
  Reply With Quote
Old 9th December 2005, 08:51 PM   #10
Pano is offline Pano  United States
diyAudio Moderator
 
Pano's Avatar
 
Join Date: Oct 2004
Location: Milliways
Blog Entries: 4
Smart thinking there, DJK!
I remember my first experiances with power switching, I was boosting the power handling of remote control cars to use much larger motors. Mostly I just burned up MOSFETs.

Getting it right is important, the output devices are easy to kill. (I never did get it right, but there was no www to help, back then.)
__________________
Take the Speaker Voltage Test!
  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
Class D amp design Thys Class D 1 20th July 2007 09:45 PM
Hybrid Class A / Class B design. GK Solid State 27 1st January 2007 02:30 AM
class H design csl113 Solid State 5 23rd June 2006 05:53 PM
My 50W class A design! PanzerLord Solid State 90 11th August 2004 06:17 PM


New To Site? Need Help?

All times are GMT. The time now is 08:39 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