odd: half-bridge, no rail voltage but still drives speaker?

jwagnerhki · 15th April 2006, 07:08 PM

Hi,

I've been wondering about this several times, with different bridges and not only in audio applications. Maybe someone else has seen this? And knows what causes it?

Basically, right now, the schematic and circuit is:
http://users.tkk.fi/~jwagner/electr/...bridge-sch.gif

Prototype:
http://users.tkk.fi/~jwagner/electr/...halfbridge.jpg

It works absolutely fine at e.g. 500kHz and 50V rails, "8 ohm" speaker load. Three supplies: MOSFET driver section takes a +3.3V supply for the logic level input and +12..13V for the medium business end, the 12V is bootstrapped to the high side mosfet via an ultrafast soft recovery diode, and there's the 40..60V rail input for the MOSFET business end. The output LC filter can also seen in the pic.

Now what is odd is that when the 50V rail is completely disconnected, there's still music audible from the speaker.

Of course with the rails missing it's not full blast volume, but audible nevertheless.

It's as if the mosfet gate drivers were somehow driving the speaker/filter, but for the life of it I've never been able to figure out what path that "leak"(?) audio signal might be taking to the speaker. Or if this might be harmful to the drivers of mosfets. Any thoughts?

cheers,
- Jan

mzzj · 15th April 2006, 07:36 PM

Have you measured rail voltages when they are disconnected? I GUESS that gate drivers are capacitively pumping(trough mosfet parasitic capacitances) some small voltage to supply caps.

jwagnerhki · 15th April 2006, 09:11 PM

Thanks for the reply!

Yes that capacitive might actually be the case. I'll have to measure next time our project team is assembled. The mosfet total gate capacitance is something around 3nF (yes it depends on several things, but, roughly speaking

) and PWM 300kHz, but it /could/ be enough to "leak" the drive to the rails...

By the way the same thing also happened with an external bench top adjustable supply connected to the half-bridge rails, and with it turned way down to 5V and less, around which point it did not have any impact on the output audio volume. Of course it's still possible the bench supply isn't capable of sinking current so the rails might have been pumped to something above the bench supply "output voltage".

But I'll have to check your idea.

regards,
- Jan

jwagnerhki · 26th April 2006, 11:06 AM

Ok, finally got around to measure.

Result: the rail has a quite smooth ~500mVDC (MOSFETs have antiparallel 200V 1A schottkys). And there's about 1Vpp going from the half-bridge center tap into the LC lowpass. Does that sound ok?

There's however one other problem, the IR2110 bootstrap supply is DC plus a weak version of the audio signal on top of it. When the rail voltage is high enough, the bootstrap supply can deviate from 12Vdc a lot (into both directions) and undervoltage lock-out kicks in.

Has anyone used the IR2110 in an audio amp? Any example PCB layouts, perhaps?
(and pls don't suggest HIP4080/81 as a replacement, these only go to 80V abs max, that's next to no voltage at all ;-))

- Jan

15th April 2006, 07:08 PM	#1
jwagnerhki diyAudio Member Join Date: Oct 2003 Location: Helsinki	odd: half-bridge, no rail voltage but still drives speaker? Hi, I've been wondering about this several times, with different bridges and not only in audio applications. Maybe someone else has seen this? And knows what causes it? Basically, right now, the schematic and circuit is: http://users.tkk.fi/~jwagner/electr/...bridge-sch.gif Prototype: http://users.tkk.fi/~jwagner/electr/...halfbridge.jpg It works absolutely fine at e.g. 500kHz and 50V rails, "8 ohm" speaker load. Three supplies: MOSFET driver section takes a +3.3V supply for the logic level input and +12..13V for the medium business end, the 12V is bootstrapped to the high side mosfet via an ultrafast soft recovery diode, and there's the 40..60V rail input for the MOSFET business end. The output LC filter can also seen in the pic. Now what is odd is that when the 50V rail is completely disconnected, there's still music audible from the speaker. Of course with the rails missing it's not full blast volume, but audible nevertheless. It's as if the mosfet gate drivers were somehow driving the speaker/filter, but for the life of it I've never been able to figure out what path that "leak"(?) audio signal might be taking to the speaker. Or if this might be harmful to the drivers of mosfets. Any thoughts? cheers, - Jan

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Half bridge smps voltage drop problem	alexclaire	Power Supplies	66	4th October 2012 05:35 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	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

15th April 2006, 07:36 PM	#2
mzzj diyAudio Member Join Date: Jul 2005 Location: 65N 25E	Have you measured rail voltages when they are disconnected? I GUESS that gate drivers are capacitively pumping(trough mosfet parasitic capacitances) some small voltage to supply caps.

15th April 2006, 09:11 PM	#3
jwagnerhki diyAudio Member Join Date: Oct 2003 Location: Helsinki	Thanks for the reply! Yes that capacitive might actually be the case. I'll have to measure next time our project team is assembled. The mosfet total gate capacitance is something around 3nF (yes it depends on several things, but, roughly speaking ) and PWM 300kHz, but it /could/ be enough to "leak" the drive to the rails... By the way the same thing also happened with an external bench top adjustable supply connected to the half-bridge rails, and with it turned way down to 5V and less, around which point it did not have any impact on the output audio volume. Of course it's still possible the bench supply isn't capable of sinking current so the rails might have been pumped to something above the bench supply "output voltage". But I'll have to check your idea. regards, - Jan

26th April 2006, 11:06 AM	#4
jwagnerhki diyAudio Member Join Date: Oct 2003 Location: Helsinki	Ok, finally got around to measure. Result: the rail has a quite smooth ~500mVDC (MOSFETs have antiparallel 200V 1A schottkys). And there's about 1Vpp going from the half-bridge center tap into the LC lowpass. Does that sound ok? There's however one other problem, the IR2110 bootstrap supply is DC plus a weak version of the audio signal on top of it. When the rail voltage is high enough, the bootstrap supply can deviate from 12Vdc a lot (into both directions) and undervoltage lock-out kicks in. Has anyone used the IR2110 in an audio amp? Any example PCB layouts, perhaps? (and pls don't suggest HIP4080/81 as a replacement, these only go to 80V abs max, that's next to no voltage at all ;-)) - Jan