''special'' mosfet driver... - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Class D
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

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
Thread Tools Search this Thread
Old 10th August 2007, 02:41 PM   #11
Account Disabled
Join Date: Oct 2004
Originally posted by alfsch
hi fredos,
a year ago i had this in mind: (pic)
you think about something like this?

for small mosfet, as trans i tried: bc516/17 , 1A darling

for bigger gates, komp.fet driver: irf7389 , 30v, 30a peak

monster , fast driver: tc4422
There is a problem with that in that there is nowhere for the inductors to reset when you turn the mosfets off.

Here you go....


  Reply With Quote
Old 10th August 2007, 02:56 PM   #12
diyAudio Moderator Emeritus
lumanauw's Avatar
Join Date: Oct 2002
Location: Bandung
Send a message via Yahoo to lumanauw
Hi, Genomerics,

It's your patent?
  Reply With Quote
Old 10th August 2007, 03:21 PM   #13
alfsch is offline alfsch  Germany
diyAudio Member
Join Date: Jan 2002
Location: augsburg
There is a problem with that in that there is nowhere for the inductors to reset when you turn the mosfets off.
coil "reset" < back-diodes

but for stray-ind. a extra snubber is needed, right, not in my pic...
is this, what you guessed?

btw, in the linked patent, i cannot see, why there should be "reduced ripple"...every d-amp has a coil between switching-stage and supply (and most times also a load), what is different here??
  Reply With Quote
Old 10th August 2007, 04:52 PM   #14
Account Disabled
Join Date: Oct 2004
Originally posted by lumanauw
Hi, Genomerics,

It's your patent?

Hi Alfsch, erm... don't know. Don't tell me I might patent it. The reduced ripple current is demonstrated in the pictures.

Yo Fredos.... Attaching pictures doesn't usually work for me so I don't suppose this time will be any different.

You might try a HC, cmos, octal buffer. Use two gates as an input and parallel the other six to drive the emitters of a pair of grounded base NPN/PNP transistors. Collectors up/down to gates of a pair of PMOS/NMOS jobs and drains connected to get your final output.

If you use a tri-state device then you have a means of disabling the driver under fault conditions....

And..... you might be able to use the input gate threshold as your comparator.

  Reply With Quote
Old 10th August 2007, 05:08 PM   #15
Eva is offline Eva  Spain
diyAudio Member
Eva's Avatar
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
The key for ripple current reduction and a completely different current consumption pattern from the power supplies (in comparison with the classic circuit) seems to be this capacitor across the drains of the MOSFETs. My guess is that it forms a LC filter together with the leakage (not coupled) inductance of the coil and prevents the current from shifting completely from one half of the coil to another after each switching event.

Funny approach. It may even be inmune to supply pumping. Can anybody confirm this?
I use to feel like the small child in The Emperor's New Clothes tale
  Reply With Quote
Old 10th August 2007, 05:28 PM   #16
Account Disabled
Join Date: Oct 2004
It started out as a 'what if' on the basis that in switch mode power supplies nodes in series with the inductor only see the ripple current.

The capacitor ended up in there as a path for inductor reset.....

Even though it might save the main resevoir caps from getting hammered that coupling capacitor gets it anyway.

After I'd done a bit of preliminary research to see if it was 'new' I went and filed the patent.

Afterwards I came across this


So, as things turn out it is a Cuk derived topology. They still granted it even though I pointed out the similarity.

No..... it does not overcome power supply pumping.

  Reply With Quote
Old 10th August 2007, 05:46 PM   #17
Account Disabled
Join Date: Oct 2004
Then again.... although it still suffers from power supply pumping you can use it to correct for the problem.

Since it causes the problem in the first place you can.....

Ground the sources through a small resistor and tie an average current mode control loop around things driven from something that senses the rail offset.

It is added complexity but the implementation might be simpler than other methods.

In the paper by Cuk he mentions a battery charger/discharger. Same thing here.

Have a look at figure 12)

If you rotate it 90 degrees and reconnect the load where we are used to putting it in a Class-D amplifier then you'll see......

  Reply With Quote
Old 16th August 2007, 03:27 PM   #18
diyAudio Member
Join Date: Dec 2003
Location: Munich
Does this really pay back?

- Supply rail pumping would be similar as in a half bridge.
- Even if there is a reduced supply ripple (still do not see it....)... Even then a P-Chanel needs 3 times more silicone than N-Chanel.
-The reduced gate drive costs will be eaten up quite fast by the higher MosFet prices.
-You need two coupled inductors.
-Ripple current is not a big deal for low power designs.

Fredos, you must make your living from your designs.
Don't let me go dead as a dump head, -where is the economical clue?
  Reply With Quote
Old 16th August 2007, 04:50 PM   #19
Account Disabled
Join Date: Oct 2004
I don't know about the economics.....

As you say the GB patent version I posted does not overcome supply rail pumping.

The reduction in supply ripple current does occur as suggested. In a normal half bridge the power supplies are subjected to a discontinuous, on off, current demand at the switching frequency as well as that due to the audio signal. In the modified version the current is continuous/triangular at the level of the inductor current.

You are, of course, right about the relative cost/performance of P-channel mosfets. The circuit could be implemented using a pair of N-channel devices but that would mean introducing level shifting and dead-time control. With a P-N pair level shifting is not required and dead-time is, more or less, eliminated. The possibility of cross-conduction is minimised.

In addition the common source connection rides on the filtered audio output signal. Whilst the patent suggests ground referenced operation it would be just as simple to float the drivers and, for example, a current mode control loop on the output. The rest of the front end control electronics could then interface using the same tequniques as found in an ordinary linear amplifier.

The CUK version requires two coupled inductors. The GB patent does not, it is a single coupled inductor.


  Reply With Quote


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
MOSFET Amp Modules on Special! The Saint Vendor's Bazaar 0 13th November 2008 05:48 AM
Mosfet driver IC's hairylee Class D 92 30th March 2008 04:22 AM
Mosfet audio driver mik Chip Amps 11 28th October 2003 04:14 PM

New To Site? Need Help?

All times are GMT. The time now is 08:41 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2