Fully Discrete Class D Amplifier (WIP)

[darkamikaze]

Hi I'm currently working on a fully discrete class D amplifier for my undergraduate senior design project as a Computer Engineer.

I have no prior experience to anything that relates to this topic so I thought I would create this thread for continous discussion and opinions on choices I make throughout this process of building the amplifier. A lot of help would be greatly appreciated!

Okay, so far based on researched a Class D amplifier looks similar to this (http://upload.wikimedia.org/wikipedi...wm_amp.svg.png)

I was able to create a small differential amplifier to act as a comparator and a triangle waveform generator with all discrete components (See attachment.) So far my pspice looks promising and to be honest this is as far as i've gotten X_x. (WIP! ). Any opinions or advice? I'll try to somewhat, log my progress and hopefully be able to have fully functional discrete class D further down in time. Thanks!

[sweetperfume]

Try philips

[stewin]

hi dakar in your project are you planning to go ucd or nonucd ?

you can check on sodvak he has many designs and class d 4 sure also try the below links you might learn a thing or two on fully discreet class d versions.


Philips UCD application note

my first ever discrete class d aud amp

diysmps

Pictures and circuits of my DIY class D and other projects.

My DIY UcD amp from the junk box

Class D.fromru.com

UCD Based Fully Differential Full Bridge 450W RMS Amp

Ultra Simple Class D

diysmps

Index of /Audio/Circuits/POWER AMPLIFIERS-CLASS-D

My DIY UCD

visit the above links and have lots of fun

[Soldermizer]

More on humor here, but I recently learned that there are ("were") vacuum tube based operational amplifiers (in the 1950s-60s, I guess for EE labs or testing). Perhaps apply this to your project: why not a vacuum tube Class D. Is that even possible?

[darkamikaze]

@sweetperfume

I actually tried to PSPICE the Philips UCD but my results were a little odd. (See attachment) It seems that my output was always the Vcc/Vdd of the circuit. It may be because of the parts I chose but those were the only ones available to me in the student edition? Or something in my setup is wrong?

Also, I don't understand the Philips one so well X_x I can't tell what part is the comparator or if there's a triangle wave generator I made a circuitlab.com of this as well found here. for reasons I don't know why since it compiles REALLY slow online

@stewin

I'm not sure whether to go UCD or nonUCD as I'm not entirely sure what that means... but I have been snooping for quite sometime and I've looked at some of those threads. Unfortunately some of them hosted schematics or posted links elsewhere that they're no longer available.

@Soldermizer
My professor actually allows vaccum tubes to be used in my project but they're quite outdated aren't they? Here's his specifications and his mention of vaccum tubes for reference (here)

Thanks everyone for the advice!

[sweetperfume]

Originally Posted by darkamikaze

@sweetperfume

I actually tried to PSPICE the Philips UCD but my results were a little odd. (See attachment) It seems that my output was always the Vcc/Vdd of the circuit. It may be because of the parts I chose but those were the only ones available to me in the student edition? Or something in my setup is wrong?

Also, I don't understand the Philips one so well X_x I can't tell what part is the comparator or if there's a triangle wave generator I made a circuitlab.com of this as well found here. for reasons I don't know why since it compiles REALLY slow online

@stewin

I'm not sure whether to go UCD or nonUCD as I'm not entirely sure what that means... but I have been snooping for quite sometime and I've looked at some of those threads. Unfortunately some of them hosted schematics or posted links elsewhere that they're no longer available.

@Soldermizer
My professor actually allows vaccum tubes to be used in my project but they're quite outdated aren't they? Here's his specifications and his mention of vaccum tubes for reference (here)

Thanks everyone for the advice!

I too got same prob with philips amp but no worries buddy.. Its the problem that your negative switches are unable to get the bias 12v supply....just check it out and dont use powerful transistors...use high voltage low power transistors only.....and set the freq and dead time properly...

[ChocoHolic]

Originally Posted by darkamikaze

I actually tried to PSPICE the Philips UCD but my results were a little odd. (See attachment) It seems that my output was always the Vcc/Vdd of the circuit. It may be because of the parts I chose but those were the only ones available to me in the student edition? Or something in my setup is wrong?

I guess it will start up, when you define the initial condition of C8.
Put its innitial condition to 12V (or -12V, this depends on your chosen orientation during placement of C8).

[stinius]

Originally Posted by darkamikaze

I actually tried to PSPICE the Philips UCD but my results were a little odd. (See attachment)

Hi
Looking at your schematic, the first thing I notice is that the bottom left corner should not be connected to ground, but to ground via a cap, like you do in the upper left corner.

[sweetperfume]

What prob you get with philips???

[darkamikaze]

@stinius
Thanks I just noticed that as well but it didn't seem to make a difference in my simulation X_x

@sweetperfume
I'm not sure but on PSPICE my output is always 35V as shown in my attachments in post 5 X_x. It's also hard to understand parts of the design as to what they do :/