DIYHiFi.org

For the sake of audio
It is currently Thu Nov 23, 2017 4:33 pm

All times are UTC




Post new topic Reply to topic  [ 15 posts ] 
Author Message
PostPosted: Mon Apr 13, 2009 12:43 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
In contrast to devolution, I hope, my forum-member friend. I began with idea of taking two N-channel MOSFETs and, using LTspice, coming up with the simplest, most effective discreet component circuit that could drive those two power transistors in class D mode. All because I was a bit bored.

It turns out that Bruno zeroed in on the solution quite well, not surprisingly. It befits his genius, I'd say.

I tried all kinds of configurations including a class AB voltage amplifier stage that I used in my last class D design--an even simpler circuit that was a modified complementary class AB type. I guess I will try to show an image of it as well as the dual N MOSFET circuit that I now think merits my "N-champ" moniker. I know that this circuit isn't completely optimized. I might do more of that later, if I feel it worthwhile to me. I plan to upload one or more images later.

I had intended to take that class AB driver stage and instead of driving one N and one P-channel MOSFET in source follower configuration, drive one N-channel in direct source-follower mode and another by using a PNP transistor in a modified Sziklai pairing. That approach was not fulfilling my desire for simplicity or effectiveness. The Lower MOSFET in the totem pole tends to suffer from switching speed limitations, for one thing.

The circuit was still not decent-looking enough with the
standard differential pair that level-shifted the signal to the upper MOSFET and also fed it to the lower one. I could not get the propagation delay of the stage lower than about 40nS. In class D, PD translates almost directly into distortion. It is a hideous type, too. I remember my first class D project 25 years ago suffering from it in a very severe way. In really bad cases it sounds like severe crossover distortion that might plague class AB.

The problem was the pull-down resistors for the long tail pair driving the differential amplifier/level shifter. They suck away the drive for differential pair and are slow. I guess that is why the current mirror was invented. I just basically ended up with a cross-linked CM.

As another of my member-here friends recalls, I was not far from the original UCD concept several years ago, before I knew of the existence of Bruno's concept.

Enough apparent horn-tooting. I really just want to show enthusiasm, but the personal background information seems to be an unfortunate aspect of expressing satisfaction and airing that dung out right away and getting it out of my system, so to speak. It's probably mostly self-delusion, anyway, since I know what my IQ is. It is almost a curse for it to be too high at times, sadly. :( . Too high mine is not, in my opinion. I really want a world society based on cooperation instead of competition (predation).

I definitely agree that we build upon the work of those before us. Besides, I would be much surprised that given the opportunity, I will be able to get the circuit to work in real life. It won't help that I intend to use P2P wiring on this thing if I do try. I might just stay in the realm of simulations on this project, and possibly advisedly so.

There you have it, the good and the bad. Thanks everybody for being forbearing peers as you have been. I invite comment, but being non-prolific at writing, my responses will probably be pretty anemic, if at all. It much depends on what comments there would be. :wave:

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Mon Apr 13, 2009 2:23 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
I'm providing an image that has some parts for both inverting and non-inverting input included for test and experimental puroposes. My observation is that inverting input is cleaner but gives lower input impedance. Non-inverting tends to be better at producing lower DC offset. I may or may not try to optimize the circuit, though the output stage seems about right as it stands. The self-oscillating frequency is only about 200khz. I wouldn't be confident about going higher if I had wanted it higher. My preference is high power for bass while another low power non-switching amp handles the rest. Actually, I just design audio amplifier circuits because I have spent so much time on them in the past, not because I even listen to music still. :shock:

The image shows the input signal grounded for test purposes.


You do not have the required permissions to view the files attached to this post.

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Mon Apr 13, 2009 2:26 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
:wave: Carlos

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Mon Apr 13, 2009 2:36 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
A tangential thought on IQ. People can get maglomaniacal if they have a high one. I can only surmise that this is very tragic, since knowing what it is should be a tool for goal setting and planning future endeavors. It is not something to feel proud of like a true accomplishment would be since it is mostly a product of circumstances that barely resulted from what a person did.

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Mon Apr 13, 2009 2:50 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
Here is the complementary symmetry class D amplifier that I mentioned. I have updated it to include two transistors for regenerative feedback to sharpen switchings.


You do not have the required permissions to view the files attached to this post.

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Tue Apr 14, 2009 12:35 am 
Offline
Account disabled on owner's request

Joined: Wed Dec 29, 2004 8:31 pm
Posts: 899
Electrone wrote:
A tangential thought on IQ. People can get maglomaniacal if they have a high one. I can only surmise that this is very tragic, since knowing what it is should be a tool for goal setting and planning future endeavors. It is not something to feel proud of like a true accomplishment would be since it is mostly a product of circumstances that barely resulted from what a person did.


Did you also read this today

http://www.ft.com/cms/s/2/4add9230-23d5 ... e2340.html


Top
 Profile  
Reply with quote  
PostPosted: Tue Apr 14, 2009 12:22 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
It explains why I like to absorb myself into the simulator and focus on it instead of going into a broad range of things, I'd say.

I think i have an idea of how to bridge the circuit without duplicating many parts. If I build this thing, I would like it to be as compact as possible. I wonder if i should bridge and use lower voltage MOSFETs with the increased ease of working with lower voltages or use larger bulk capacitors to allow a smaller switching power supply transformer.

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Tue Apr 14, 2009 1:46 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
The deadtime is about 30 nS in both MOSFET switchings. I hope to post some updated circuit images.


You do not have the required permissions to view the files attached to this post.

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Tue Apr 14, 2009 8:28 pm 
Offline
Cow

Joined: Wed Jan 05, 2005 7:18 am
Posts: 118
Hi.

I think that going for a discrete design is a waste of time nowadays. I have observed UCD400 waveforms and they are not without ringing. You canĀ“t do any better with TO220 devices and their parasitics.

I have tried integrated class D output stage STA500 http://www.st.com/stonline/products/literature/od/8167.pdf in a non audio application. It has much cleaner waveforms decoupled only with 5 10uF 1210 ceramic capacitors. It is now obsolete, but TI has a more powerfull chip DRV8402 http://focus.ti.com/lit/ds/symlink/drv8402.pdf.

It is the same chip as their audio amplifiers TAS5162 http://focus.ti.com/lit/ds/symlink/tas5162.pdf and TAS5261 http://focus.ti.com/lit/ds/symlink/tas5261.pdf, but datasheet targets different set of parameters. They are quite impressive: rise and fall times 9ns resistive, dead time 5 ns.

So if I would build me a class D amplifier today, i would use TI chip in a bridge configuration coupled with leapfrog modulator. DRV8402 has provision to stick current sense resistors in all power supply leads.

There is also a matching power supply module from Vicor http://cdn.vicorpower.com/documents/datasheets/VIB0002_F_BCM.pdf provided that you have a suitable PFC front end. It costs 61EUR in small quantities and only 35 US$ in OEM quantities. Try to beat this with discrete design.

Regards, Jaka


Top
 Profile  
Reply with quote  
PostPosted: Wed Apr 15, 2009 10:54 am 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
Thanks for that advice, Jaka. It makes a lot of sense. I am leaning toward the complementary symmetry discrete version still since I already have all the parts on hand to build it if I ever have time to spare, and it is the most doofus-proof concept.

I mainly like tinkering with LTspice, and home made projects. The UCD type circuit is fun for me to toy with and it is easy for me to understand.

But for serious work, I agree with you. I would probably use that new chip from International Rectifier, IRS20124, and might still. Thanks for reminding me. I'm glad you commented. :)

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Wed Apr 15, 2009 2:25 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
Actually, I think it is the IRS2092 that I am interested in. It comes in a 16 pin DIP package, which makes it a viable option for me. I would have to try to get it to work with all the feedback coming after the output filter, though the data sheet shows the pre-filter feedback approach. I wouldn't listen to an amplifier without full active output damping.

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Wed Apr 15, 2009 2:59 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
Apparently, the International Rectifier Class D ICs do permit self-oscillating operation. That enables cleaner reproduction. My impression is that they can also work with 100% post-filter feedback.

http://www.irf.com/technical-info/white ... naudio.pdf

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Sat Jun 06, 2009 7:05 pm 
Offline
Account disabled on owner's request

Joined: Wed Dec 29, 2004 8:31 pm
Posts: 899
It's more like they permit synchronizing to a clock for reduced performance, othwerise they're intended for self oscillating operation.

I don't believe they can work with 100% post filter. Some portions of the loop may be internal, as their patents show a type of quadrature feedback. The IC's may seem impressive but they've yet to truly deliver a great one, that doesn't rope you into limitations. No doubt better than Tripath but it's not a reason to give up on other methods.

_________________
'I am... everything is... changed... they're calling... your face... interwoven... who is...


Top
 Profile  
Reply with quote  
PostPosted: Wed Jun 10, 2009 10:23 pm 
Offline
Sheep

Joined: Sat Dec 23, 2006 8:27 pm
Posts: 23
Location: NAC
I am glad that I'm in no hurry to make a new class D amplifier right now.

_________________
USMPS http://tech.groups.yahoo.com/group/switchmode


Top
 Profile  
Reply with quote  
PostPosted: Fri Oct 29, 2010 11:05 pm 
Offline
Hen

Joined: Thu Oct 28, 2010 9:19 pm
Posts: 1
Location: carlisle, cumbria UK
Electrone wrote:
Actually, I think it is the IRS2092 that I am interested in. It comes in a 16 pin DIP package, which makes it a viable option for me. I would have to try to get it to work with all the feedback coming after the output filter, though the data sheet shows the pre-filter feedback approach. I wouldn't listen to an amplifier without full active output damping.


I have succesfully made a 2092 class d amplifier.
I did have a few problems getting it to run but after tweaking it i got it to run reliably.
I was disappointed that the application note design would not run for me.

I dont think it sounds any better than my own class AB designs but it was an interesting exercise.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 15 posts ] 

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group