audio amp h-bridge...

J

John Larkin

Guest
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.
 
On 10/21/2020 1:56 PM, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.
Click to expand...

The Si8275 dual isolated gate driver is a great chip, extremely fast for
an isolated driver, and beefy output. As I did a few months back I\'ve
pushed several watts into a resistive load at 5 MHz just using the
drivers and they had bandwidth to spare.

<https://www.mouser.com/ProductDetail/Silicon-Labs/SI8275ABD-IS1?qs=r5DSvlrkXmLJKA%2F8fwt1Lw%3D%3D>

Very nice for $2 in quantity. Downside is the chips seem to be very
moisture-sensitive, moreso than the Analog Devices isolated drivers that
are also moisture sensitive. so have to be soldered straight out of the
bag or else if that card starts changing color they\'re done for without
baking.
 
On 10/21/2020 3:20 PM, bitrex wrote:
On 10/21/2020 1:56 PM, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.


The Si8275 dual isolated gate driver is a great chip, extremely fast for
an isolated driver, and beefy output. As I did a few months back I\'ve
pushed several watts into a resistive load at 5 MHz just  using the
drivers and they had bandwidth to spare.

https://www.mouser.com/ProductDetail/Silicon-Labs/SI8275ABD-IS1?qs=r5DSvlrkXmLJKA%2F8fwt1Lw%3D%3D


Very nice for $2 in quantity. Downside is the chips seem to be very
moisture-sensitive, moreso than the Analog Devices isolated drivers that
are also moisture sensitive. so have to be soldered straight out of the
bag or else if that card starts changing color they\'re done for without
baking.
Click to expand...

The Si8274 is the PWM-input variant.
 
On Wed, 21 Oct 2020 15:20:52 -0400, bitrex <user@example.net> wrote:

On 10/21/2020 1:56 PM, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.


The Si8275 dual isolated gate driver is a great chip, extremely fast for
an isolated driver, and beefy output. As I did a few months back I\'ve
pushed several watts into a resistive load at 5 MHz just using the
drivers and they had bandwidth to spare.

https://www.mouser.com/ProductDetail/Silicon-Labs/SI8275ABD-IS1?qs=r5DSvlrkXmLJKA%2F8fwt1Lw%3D%3D

Very nice for $2 in quantity. Downside is the chips seem to be very
moisture-sensitive, moreso than the Analog Devices isolated drivers that
are also moisture sensitive. so have to be soldered straight out of the
bag or else if that card starts changing color they\'re done for without
baking.
Click to expand...

The nice thing about the TI chip is that it does everything, audio
sine wave in to power out, including thermal and current limiting. If
we just use mosfet drivers, we have to make the PWM ourselves, and
make sure the bootstraps are properly fed, and probably current limit
somehow. I\'d rather buy all that if I can.

There must be chip for all that.

I don\'t see a good way to parallel two of the TI parts.
 
onsdag den 21. oktober 2020 kl. 21.46.53 UTC+2 skrev John Larkin:
On Wed, 21 Oct 2020 15:20:52 -0400, bitrex <user@example.net> wrote:

On 10/21/2020 1:56 PM, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.


The Si8275 dual isolated gate driver is a great chip, extremely fast for
an isolated driver, and beefy output. As I did a few months back I\'ve
pushed several watts into a resistive load at 5 MHz just using the
drivers and they had bandwidth to spare.

https://www.mouser.com/ProductDetail/Silicon-Labs/SI8275ABD-IS1?qs=r5DSvlrkXmLJKA%2F8fwt1Lw%3D%3D

Very nice for $2 in quantity. Downside is the chips seem to be very
moisture-sensitive, moreso than the Analog Devices isolated drivers that
are also moisture sensitive. so have to be soldered straight out of the
bag or else if that card starts changing color they\'re done for without
baking.

The nice thing about the TI chip is that it does everything, audio
sine wave in to power out, including thermal and current limiting. If
we just use mosfet drivers, we have to make the PWM ourselves, and
make sure the bootstraps are properly fed, and probably current limit
somehow. I\'d rather buy all that if I can.

There must be chip for all that.

I don\'t see a good way to parallel two of the TI parts.
Click to expand...

TI (in parallel with) H-brigde via gatedrivers controlled by the TI output?
 
On 10/21/2020 3:46 PM, John Larkin wrote:
On Wed, 21 Oct 2020 15:20:52 -0400, bitrex <user@example.net> wrote:

On 10/21/2020 1:56 PM, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.


The Si8275 dual isolated gate driver is a great chip, extremely fast for
an isolated driver, and beefy output. As I did a few months back I\'ve
pushed several watts into a resistive load at 5 MHz just using the
drivers and they had bandwidth to spare.

https://www.mouser.com/ProductDetail/Silicon-Labs/SI8275ABD-IS1?qs=r5DSvlrkXmLJKA%2F8fwt1Lw%3D%3D

Very nice for $2 in quantity. Downside is the chips seem to be very
moisture-sensitive, moreso than the Analog Devices isolated drivers that
are also moisture sensitive. so have to be soldered straight out of the
bag or else if that card starts changing color they\'re done for without
baking.

The nice thing about the TI chip is that it does everything, audio
sine wave in to power out, including thermal and current limiting. If
we just use mosfet drivers, we have to make the PWM ourselves, and
make sure the bootstraps are properly fed, and probably current limit
somehow. I\'d rather buy all that if I can.

There must be chip for all that.

I don\'t see a good way to parallel two of the TI parts.
Click to expand...

Well it\'s actually for a bidirectional buck/boost PWM controller for
moving battery power back and forth from a bus, but might be able to
press this one into service as an audio frequency PWM modulator by
feeding the COMP input, it includes overcurrent detect/protection
circuitry. Just add driver. And most modern gate drivers like the one I
mentioned have thermal limiting.

<https://www.mouser.com/new/analog-devices/adi-adp1974-controllers/>
 
On 10/21/2020 3:46 PM, John Larkin wrote:
On Wed, 21 Oct 2020 15:20:52 -0400, bitrex <user@example.net> wrote:

On 10/21/2020 1:56 PM, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.


The Si8275 dual isolated gate driver is a great chip, extremely fast for
an isolated driver, and beefy output. As I did a few months back I\'ve
pushed several watts into a resistive load at 5 MHz just using the
drivers and they had bandwidth to spare.

https://www.mouser.com/ProductDetail/Silicon-Labs/SI8275ABD-IS1?qs=r5DSvlrkXmLJKA%2F8fwt1Lw%3D%3D

Very nice for $2 in quantity. Downside is the chips seem to be very
moisture-sensitive, moreso than the Analog Devices isolated drivers that
are also moisture sensitive. so have to be soldered straight out of the
bag or else if that card starts changing color they\'re done for without
baking.

The nice thing about the TI chip is that it does everything, audio
sine wave in to power out, including thermal and current limiting. If
we just use mosfet drivers, we have to make the PWM ourselves, and
make sure the bootstraps are properly fed, and probably current limit
somehow. I\'d rather buy all that if I can.

There must be chip for all that.

I don\'t see a good way to parallel two of the TI parts.
Click to expand...

Infineon makes an analog-input driver \"for 500W applications\", is 1000W
in a bridge configuration enough?

<https://www.infineon.com/cms/en/product/power/class-d-audio-amplifier-ic/discrete-class-d-audio-amplifier-ics/irs2092spbf/>
 
On 10/21/2020 4:59 PM, bitrex wrote:
On 10/21/2020 3:46 PM, John Larkin wrote:
On Wed, 21 Oct 2020 15:20:52 -0400, bitrex <user@example.net> wrote:

On 10/21/2020 1:56 PM, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.


The Si8275 dual isolated gate driver is a great chip, extremely fast for
an isolated driver, and beefy output. As I did a few months back I\'ve
pushed several watts into a resistive load at 5 MHz just  using the
drivers and they had bandwidth to spare.

https://www.mouser.com/ProductDetail/Silicon-Labs/SI8275ABD-IS1?qs=r5DSvlrkXmLJKA%2F8fwt1Lw%3D%3D


Very nice for $2 in quantity. Downside is the chips seem to be very
moisture-sensitive, moreso than the Analog Devices isolated drivers that
are also moisture sensitive. so have to be soldered straight out of the
bag or else if that card starts changing color they\'re done for without
baking.

The nice thing about the TI chip is that it does everything, audio
sine wave in to power out, including thermal and current limiting. If
we just use mosfet drivers, we have to make the PWM ourselves, and
make sure the bootstraps are properly fed, and probably current limit
somehow. I\'d rather buy all that if I can.

There must be chip for all that.

I don\'t see a good way to parallel two of the TI parts.



Infineon makes an analog-input driver \"for 500W applications\", is 1000W
in a bridge configuration enough?

https://www.infineon.com/cms/en/product/power/class-d-audio-amplifier-ic/discrete-class-d-audio-amplifier-ics/irs2092spbf/
Click to expand...

Here is the frequency response test signal:

<https://youtu.be/4jd-nyBddMo?t=21>
 
On 22/10/20 4:56 am, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.
Click to expand...

HIP4081?
 
On 22/10/2020 04:56, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.
Click to expand...

I know you asked for a chip not a module, but in case you haven\'t seen
these, they have quite impressive specs due to the unusual design -
especially Zout, bandwidth, distortion, noise. I haven\'t used them but I
have no reason to doubt the specs.

https://www.hypex.nl/ncore-family/
https://www.hypex.nl/ucd-family/

Depending on your production volume and engineering costs, it might make
sense. The continuous power ratings of their products are much lower
than the \"audio\" ratings so you\'d need a bigger one than the product
names would suggest.
 
Clifford Heath wrote:

On 22/10/20 4:56 am, John Larkin wrote:
Hi,

I\'m using a TI integrated class-D audio amp in my alternator
simulator, a TPA3255, but it shuts down at 17 amps peak current and I
have a customer that may need more.

So I might want to make an amp out of four hunky mosfets and a
controller chip. The main supply is +48 volts.

Does anyone have driver chips that they like? Ideally, the input would
be baseband audio and it would drive the mosfets PWM. A controller
with separate hi/lo gate drivers is an option.


HIP4081?
Click to expand...
Auggh! A hideous monstrocity. As long as you only want to run at 30-40 V
it actually isn\'t so horrible, but the specsmanship on that chip is awful,
with the best circuit layout one can only get to about 60 % of the
(over)rated voltage.

There must be newer stuff that works much better, that part is over 20 years
old.

Jon
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top