Class D IRS2092, No power up

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all-

I have been working on a project for school for a few months now. In that time I have become very familiar with the IRAUDAMP7S, and it is what my board is based on. I recently had a PCB manufactured for my final version. This version includes a crossover network. The main difference is, I used the stock IRAUDAMP7S originally and I created my own PCB for this part.

It would appear I overlooked the part of the application note about pcb layout when I made the board, and I did not use separate grounding planes. I figured I would get a noisy output but I tried building it anyway. The IRS2092 does nothing when power is applied, no pop, smoke, or ringing. Just nothing.

I checked +VAA (+5.6v), - VSS (-5.6v), VCC (+15v), -B(-50v), and +B(+50v) and they are all what I would expect them to be so it is receiving power but it does not turn on. Is the problem in my reference planes? Could that cause it to do nothing? I have not tried replacing the chip yet as I am trying to find the root cause in case it is dead.

I've attached what the top and bottom layers look like, as you can see there is one very large grounding plane. I am feeling like this is probably it, but I'd like to get some input so I can better figure out what the problem is. The attached pdf can provide a better look at the bottom.

Thanks in advance!!
 

Attachments

  • bottom.pdf
    199.4 KB · Views: 451
  • top.jpg
    top.jpg
    271.6 KB · Views: 1,504
  • bottom.jpg
    bottom.jpg
    217.9 KB · Views: 1,422
Okay, I will take a look at those two points when I am in the lab on Monday. Thanks for the ideas! Do you think my lack of separated analog, power, and reference planes could be the problem? Or is it more likely to be something else when the chip does nothing.

I really mean nothing too, lol I would've expected at least a pop or ringing.
 
ok guys, I am in the lab currently and your suspicions were correct Nigel, the csd pin is resetting.

These are the voltages I am measuring at my pins.

-5.4 csd
-48 oc
-45 vref
-4.3 comp
+5.9 vaa
-5.9 vss
3.9 csh
11.8 vb
-36 vcc
-50 com
-36 dt

Right now I am looking through the datasheets to see how each one of these is supposed to be set to see if I am missing something.

Anyone that is familiar with the irs2092, are these values okay?

Thanks!
 
I dont want to steal this topic, but I have the same problem: irs2092 not starting up, or stuck in reset.
Maybe we have a similar mistake ?

Here is my version:

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


TOP:
An externally hosted image should be here but it was not working when we last tested it.


BOTTOM:
An externally hosted image should be here but it was not working when we last tested it.



Maybe one of you can spot an error directly, that would be nice :)
All my voltages seem to be ok.
 
Kaip:
Nice compact layout!

I think decoupling caps should be ceramic type and as close as possible to pins of 2092. Also maybe add some 1uf X7R at pins of MOSFETS.

May also help to add damping network at MOSFETS.

I am curious about your aux +12 supply. Why not just use a transistor and zener diode? It looks like it may not be properly referenced, but I did not study too closely yet.

Again nice work on layout.
 
spare_parts:
Thanks for the compliment.
The 12V is because the 7812 only allows for a 37V input. I use the 24V to raise that a bit. Nw it should be able to regulate about 35-50V to 12V
Works well.
I just tried a transistor and zener , as a I am becoming paranoid by now, but that didnt help.
Which decoupling caps do you mean ?
The PCB is small so just about everything is close to the 2092 :)
And all traces are minimal.

nigelwright:
Wouldnt the 2092 start regardless of used mosfets ?
If not can I disconnect the mosfets to see if the chip will then start up ?

It is still resetting every second or so.
 
nigelwright:
Wouldnt the 2092 start regardless of used mosfets ?
If not can I disconnect the mosfets to see if the chip will then start up ?

It is still resetting every second or so.

What happens is the 2092 sends out a pulse to the lower mosfet.
Due to high capacitance the mosfet doesn't switch on soon enough and the 2092 thinks there is a current overload so immediately resets itself.

Without gate driver buffers you can just about get away with a pair of irfb4019's
 
NigelWright is correct, try MOSFETS with lower gate capacitance.

Regarding decoupling caps, they work most effectively when connected directly to the pins on the chip because of PCB trace resistance/ inductance, but that is probably not causing your resetting.

just for testing, you could try some IRF540N (usually easy to find and low C) at up to about +/- 45V supply rails.
 
I tried with some IRF520s. Good enough for testing and with low input capacitance.
Now it starts up. Ok the Mosfets get quite hot but that could be fixed.

I will try some IRF540's later but plan a revision to the circuit with some totem drivers
An externally hosted image should be here but it was not working when we last tested it.

Will this be ok , or do you have suggestions ?
I am limited to through hole components and the size of this board so have to keep component count low.
 
I suggest using the IRFB4019 for 4 ohm loads, and the "4020 for 8 ohm loads with higher voltage rails.

The "2092 can drive these fine, see datasheet for info and a power table.

If you add totem pole drivers, I think you lose some of the functionality and attractive features of the "2092, like current sensing of the drivers' and rise/ fall time probably suffers.

Seems like going to totem pole drivers gets you closer to going fully discrete, which in my opinion is more attractive anyway. (like phillips/ hypex/ Putzey's UCD) which is post- filter feedback, another attractive feature.
 
Thanks guys

Now will have to try IRFB4019/4020
The IRF540s were an improvement over the IRF520s.
They run much cooler. You can just feel them warming up from room temperature.

I agree about the totem poles.
And less is more.
My ClassD knowledge is too limited to attempt anything discrete now.
But you have to start somewhere...

One more "problem" : when switching off you hear a load thud.
I remember reading something about that but cannot find it anymore.

Any idea ?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.