Author Topic: Trying to build/design a Class D amp from a spec drawing, with a minor tweak.  (Read 1877 times)

0 Members and 1 Guest are viewing this topic.

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Hello,

I've just finished building and ruining my IRS2092 based Class D amp.  It was based on their example circuit I'll link here.  That circuit has one difference from mine, in that I had some 48VDC power supplies for the B+ and B- so I made sure I changed the MOSFET's to support that. 

First the schematic as I got it from Infineon with their recommended parts. 

The only change I made as mentioned was to use +/-48V and I replaced the mosfets with some STP40NF10 that support up to 100VDC and 50A that should handle the 48V just fine. 

Upon power up,  (I sent the 12VDC into the pin 12 via my bench power supply plus those B+/B- voltages) I only heard a periodic slow 'tic' noise and that's it.  I did realize, upon review that although I had my input jack connected to the gnd side of the jack, I didn't have the input side linked to the common GND of the rest of the schematic.  When I did so, I got a slight change in noise and then the magic smoke came out.  I blew the IRS2092. 

On reviewing voltages everything seemed ok, and the pins on the IRS2092 have all high level voltage allowances except for the VREF on pin 7, which seems to want between 4.8 to 5.4 and I did read about 9VDC there. 

Now, I'm thinking per what I read there, that the 5V is an output reference voltage but I'm not 100% sure if the R3/R5 is supposed to be a voltage drop/divider to provide the right voltage and expected the -35VDC into it to give the expected 5V range?  Is that why I smoked it? 

I also seem to have destroyed the mosfets, but I think that may have been a mistake I made after I removed the class d chip from my circuit and tried to power it again to read some static voltages?  Not sure.

At any rate, before I buy the replacement parts I thought I'd hit the hive mind and see what seems off with the changes to the schematic etc?

Edit: thought I'd add a link to mouser where I have the BOM for the parts I used for more detail. https://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=d6e7fc6c1b

Thanks,

~Phil


« Last Edit: August 17, 2020, 06:26:12 pm by pompeiisneaks »
 

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Here's the updated schematic with the 48V and the new mosfets..

 

Online Circlotron

  • Super Contributor
  • ***
  • Posts: 3366
  • Country: au
I also seem to have destroyed the mosfets, but I think that may have been a mistake I made after I removed the class d chip from my circuit and tried to power it again to read some static voltages?  Not sure.
Big mistake to power it up with the controller chip removed and the mosfet gates just floating. The gates may drift high and turn both devices partially on and before you know it they are smoking...
 

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
I also seem to have destroyed the mosfets, but I think that may have been a mistake I made after I removed the class d chip from my circuit and tried to power it again to read some static voltages?  Not sure.
Big mistake to power it up with the controller chip removed and the mosfet gates just floating. The gates may drift high and turn both devices partially on and before you know it they are smoking...

That part makes sense, what I don't understand is why I cooked the IRS2092... the ticking at first seems likely to be a power on, safety shutdown cycle, but when I finally correctly grounded the input side it smoked the IC instantly.

~Phil
 

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Maybe this is the wrong subforum for this?  Does it maybe fall under something about learning etc?  Or is this just an area that's not well documented?  Not many people interested?

Or am I not giving enough information?

~Phil
 

Online Circlotron

  • Super Contributor
  • ***
  • Posts: 3366
  • Country: au
 

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Thanks!  I'm giving that a try. 

~Phil
 

Offline viperidae

  • Frequent Contributor
  • **
  • Posts: 306
  • Country: nz
Did you turn it into a high frequency oscillator by missing the extra capacitor and resistor in the feedback path the reference design has? That to me looks like a low pass filter to cut the gain at high frequencies.

Just a guess by the way.
 

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
I'm attaching the reference diagram I used, but I do see now I missed the 150pF cap to ground on that run, I'll have to look at that, maybe that's what it was, but would an oscillation like that burn up the chip?

I just went and hunted down the 'diagram' I used and it must have come from another doc and has no 150pF cap to ground.  Compare the 'wrong version' to the one that does have it...

Anyone see anything else bad about it?

Thanks!

~Phil
 

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Wow, I never have seen a more elitist group of people than those at diyaudio.

I've gotten insulted because of how horribly I did things, and gotten super vague answers to questions that don't help me understand anything at all.  Is it common on that forum to scare people away from learning? I've done a lot of study, reading and trying to understand it, but always do my best when I know something to not treat newbies like morons. 

Answers like "wait are you serious ? you cannot expect this to work right? you just used a tube layout on a class-d amplifier." are super insightful and really make a lot of sense to me... no wait...

Being told I'm wrong, when I already know I'm wrong isn't helpful and teaches me nothing.  WHY I'm wrong does.  Why does a tube layout not work on a class D amplifier?  Is it something about how the leads are oriented, lengths, stray capacitance etc?  or what?

I took a schematic, tried to build a prototype from it, and with my limited knowledge have built many working circuits.

I'm a bit dismayed that learning class D amps is something only electrical engineers can understand?

Or am I just that thick headed?

I'm asking here since the group seems a lot nicer to help new people learn.

At least here people have tried helping.  Thanks for that.

~Phil
 

Online Circlotron

  • Super Contributor
  • ***
  • Posts: 3366
  • Country: au
Some of those guys were a bit up themselves, that’s for sure. I left my opinion.
Maybe you could get a $20 class D amp from eBay and see how they have done it, pcb layout-wise. They are definitely not the last word in good practice but at least they work, though not necessarily optimally. Use that as a starting point and add and subtract from it to your hearts content. But very short connections from everything down to a pcb ground plane is a good start.
 

Offline pompeiisneaksTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Thanks, I'll give that a look.  I'm trying to do something akin to what we do in tube amp land, use high quality components, and high quality stuff and build it myself to save some money from buying a 300$ class d audio amp somewhere else...

My general idea is still for guitar use, but to use a tube preamp into a nice class D solid state power stage... but i digress :D

~Phil
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf